Rr^\A M^6 Columbia ^ntben^fti) (gift nf ir. SloBf pij A. llak? Digitized by the Internet Archive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/collectionofpape01mayo William J. Mayo at Graduation, iS William J. Mayo. Charles H. Mayo at Graduation, 1887. Charles H. Mayo. A COLLECTION OF PAPERS Published Previous to 1909 By WILLIAM J. MAYO and CHARLES H. MAYO VOLUME I PHILADELPHIA AND LONDON W. B. SAUNDERS COMPANY 1912 Copyright, 1912, by W. B. Saunders Company PRINTED IN AMERICA FOREWORD III llic |)lH'|);ii;il loll <)l lliis \()|iiiiic, "A ( Ollcrl loii of |',i|)crs," the editor Ims heeii inlliieiiced hy the desire to collcet mikI jtreserve a complete file of the \\riliri and 11)0!), hut only some of the more im|)ortant ones taken from the journal files. This explains why papers of a date later than l!)().j are to he found in the present collection. The fj;eneral phm of chissification used in the hrsl \-oliinie of "Collected Papers" has been retained in the arrangement of the "Old I'apers," with the addition of various suhluNidin^rs to accom- modate the more diverse character of the writin<,'s. The papers under each cla.ssification, or subject headin'M Oi-kuations i i-on tiik (i M.i.-iii,\i)iJi;it and Hii.k-passage», WITH Tabulatkd Rep(jht ok 547 Oi'KKated Cases 407 Some Causes ok Failure ok Operation to Cure Gallstone Disease. 410 Some 0bsf:rvation8 on the Surgery ok the Common Duct ok the Liver ii\ A Review ok 1000 Operations kor Oam^stone Disease, with Special Rekerence to the Mortality 430 The Diagnosis ok Gall-stone Disease 438 Some Observations on Cases Involving Operative Loss ok Continuity OK the Common Bile-duct, with the Report ok a Case ok Anasto- mosis BETWEEN THE HePATIC DuCT AND THE DuODENUM 440 A Review ok 1500 Operations upon the Gall-bladder and HiLE-PASsAGta, WITH Especial Rekerence to the Mortality i5i Pancreas Case ok Acute Pancreatitis with Fat Necrosis — Operation: Recovery 401 Pancreatic Cyst 404 The Surgical Aspect of Pancreatitis 4G0 Bibliographic Index 473 Index ok Subjects 479 ALIMENTARY CANAL ESOPHAGUS VOL. I — 1 FOREIGN BODIES IN THE TRACHEA AND ESOPHAGUS* CHARLES H. MAYO I speak of these organs under the same heading because of their intimate structural relations, which often complicate early diag- nosis of the location of foreign bodies or the remote results of secondary changes. The difficulties in the diagnosis and treat- ment, with the attending responsibility, render these cases very serious to the physician and to all concerned. The foreign bodies most frequently met wdth are seeds and other things of a vegetable nature capable of swelling, bones, and metallic bodies. In children it frequently happens that either the character or the location of the offending body is a matter of conjecture, and the symptoms may be so slight as to mislead the physician. At a later period a dislodgment of the irritant may possibly produce a fatal laryngeal spasm, in the one case, or ulceration of the esopha- gus, in the other. A foreign body high in the esophagus may, through local change and swelling, simulate a foreign body in the trachea, or a tracheal body, by lodgment and temporary amelioration of the symptoms, may lead to the belief that a passing esophageal obstruction is the cause of the trouble; therefore, where the character and location of the foreign body are unknown, it is well to pass the esophageal bougie to make a differential diagnosis. Foreign bodies which have entered the air-passages may lodge at any point between the larynx and smaller bronchi. The symp- toms of foreign bodies in the trachea vary with their size, shape, and location, and they may be intense or extremely mild. The * Reprinted from "Trans. Minn. State Med. Society," 1S96, p. 1-26. 3 4 CHARLES H. MAYO early symptoms are spasm of the larynx, choking, cough, and res- piratory distress. Vomiting, and usually more or less shock, may occur. Bloody mucus may be raised immediately, and fatal dyspnea often speedily follows from the laryngeal spasm, or the large size of the foreign body may almost completely occlude the trachea, as in Case I. Violent efforts on the part of the patient may sometimes dis- lodge and expel the irritant, but more often it will serve only to add to the distress and exhaustion. If the foreign body becomes fixed, the symptoms tend to subside, thereby leading attendants to believe that it has been expelled, until, later, a change of the position causes a recurrence of the urgent symptoms, as in Case 2. The diagnosis is often difiicult in children, but the suddenness of the attack, absence of temperature, and comparative relief to the patient between the attacks of dyspnea, combined with the aid of auscultation and laryngoscqpic examination, will assist in the making of a positive diagnosis. The prognosis is exceedingly grave, and depends to a great ex- tent upon how early the patient is subjected to treatment. The dangers are from the immediate obstruction and the secondary septic complications, which are more serious in the child than in the adult. A foreign body in one of the bronchi, unless removed, will almost certainly destroy life, either suddenly or through lung com- plications, such as bronchitis, septic pneumonia, gangrene, abscess, or perforation of the neighboring structures. We may properly divide these cases into two classes for treat- ment: First, where the foreign body is known to have lodged in the trachea or bronchus, in which case there is no doubt as to the necessity of operative interference; second, those cases in which a positive early diagnosis is not possible and necessitates delay until physical signs and symptoms verify the probable diagnosis. The diagnosis determining the operation should not be delayed, no matter how mild the symptoms. The method often advocated, of shaking a child by the heels and pounding him on the back, is unwise unless the physician is prepared to make immediate trache- P'OUEUiN HODIKS IN TUACIIKA AM) IXM'IIACIH .> otomy for the relief of sj)iisiii of the glottis. Foreign bodies loeated in the larynx or high in the trarhea ean he removed through tlic mouth l)y means of tli(> laryngoscope and projtcr instruments. " These do not, as a rule, form the urgent ca.ses, as the foreign })ody may remain high in the traehea for long periods without exeiting dangerous .symptoms. Tracheotomy is neces- sary for the removal of foreign bodies thai liaxc passed the glottis, whether lodged high or low. The high tracheotomy, or division of the upper two tracheal rings, may be preferred in children with short necks, while in ordinary cases the low tracheotomy, or divi- sion of the second three traelieal rings, will prove more advantage- ous. The performance of tracheotomy with the Paquelin cautery or by the use of some mechanical device is out of i)lace in this class of ca.ses. The points to be emphasized in the [)erformance of trache- otomy after exposing the trachea are: First, the insertion of a tenaculum in the median line of the trachea above the point of expected incision, with the traction u])ward; second, a stab j^unc- ture of the trachea and cutting upward; third, avoidance of in- jury to the esophagus by too deep an incision; fourth, retention of knife in tracheal incision until the opening is secured by other instruments. Chloroform is preferred as an anesthetic, since it relieves laryn- geal spasm. Anesthesia is not necessary in operating during a suffocative attack. The use of a probe to loosen a foreign body or excite cough, as in Case 1, as well as the posture of the patient, may prove of advantage. If the foreign body is removed, the Avound will recpiire little care. Our method is to pack the wound lightly with iodoform gauze. If unable to remove the annoying agent, as frecpiently occurs, the sides of the trachea may be stitched to the skin and the opening maintained for a day or two. The irritant will be ejected usually within a few hours. From a considerable ex])erience in the removal of foreign bodies from the trachea I have selected three illustrative cases for the purpose of eliciting discussion. 6 CHARLES H. INIAYO Case 1.— C. F., male, aged six years. After wearing an O'Dwyer intubation tube No. 3 six days for the relief of so-called membranous croup, the swelling left the larjoix, allowing the tube to pass between the vocal cords, where it lodged in the upper part of the trachea. Upon attempting to remove it through the mouth it shpped down into the bronchus, producing a severe attack of dyspnea. Respiration ceased completely in about fifteen minutes, and just before the incision of a low tracheotomy was made. The trachea was held open by forceps, the child held up by the heels, and a probe inserted into the right bronchus, where it loosened the tube and excited spasmodic coughing, aiding the dislodgment. The tube slipped into the trachea and was withdra^\'n through the tracheal wound. Artificial respiration was maintained for about twenty minutes, and then the breath- ing became natural; a tracheotomy tube was inserted and allowed to remain six days. Recovery was perfect, although delayed by a septic bronchitis. Case 2. — C. C, female, aged two years. This child, while playing with citron seeds, had an attack of suffocation which was so severe that her parents though she was dead. She soon recov- ered, however, only to have slightly less severe attacks at intervals. After three days the child was brought to St. Mary's Hospital, where, from the history and symptoms present, a diagnosis of foreign body, probably citron seed, loose in the trachea, was made. A low tracheotomy enabled me to dislodge a swollen citron seed, which was removed with forceps. The recovery was uneventful. Case 3. — M. M., male, aged one and one-half years, was brought to me with this history: Three days previously, while playing in the yard, he was seized with a severe spell of suffocation. There was nothing about him to indicate the cause of the attack, which was followed by many others. A diagnosis of foreign body of unknown character in the trachea was made. A low trache- otomy was performed, and a swollen kernel of corn was coughed up into the opening, where it was fixed by a curved probe and removed with forceps. While foreign bodies in the esophagus are not primarily so dangerous as in the trachea, the remote results are very serious, and the foreign bodies themselves are as varied, with a predomi- nance of artificial dentures, coins, bones, the larger fruit-stones. FOREIGN IJOniKS IN TUArriKA AM) KSOPHAGUS 7 and temponiry ol)slnicli()ii from food holiis. The impaction of the forci^'ii hody in I he csoplia^Mis gctK-rally takes place at its narrowest portion, at I lie le\'el of the cricoid cartilage, and next in frequency at the cardiac end, !)ut may occur at any point. The symptoms depend upon the size and shajjc of the foreign body. If large and smooth, sudden death may occur from j)ressurc upon the aperture of the huyiix or from spasm of the glottis. If j)ointed and irregular, there will be dyspliagia or complete obstruction. l)y.si)nca and a constant di.scharge of saliva and nmcus, with ten- derness of the neck and soreness behind the sternum, may occur. If left to itself, the foreign body excites inflammation, which often leads to its dislodgmcnt, but more frequently complicates matters by extending into the neighboring tissues, causing perforation into the pleura, trachea, mediastinum, or aorta. When the foreign bod}' is large, the diagnosis is comparatively simple. The exploring finger can reach as far as the arytenoid cartilages. When situated below this point, its presence must be decided by inference, aided by esophageal sounds, stomach-tube, etc. Always bear in mind that symptoms of the presence of a foreign body in the esophagus may be simulated by irritation of the mucous membrane, caused by the passage into the stomach of the foreign body, the sensation existing long after its dislodgraent. When it can be done safely, the foreign body may be extracted through the mouth. If unable to remove it in this manner and it can be easily pushed down into the stomach, this may be done. If neither method avails without the use of undue violence, an external esophagotomy should be performed and the body extracted through the opening. Sometimes during the manii)ulations of removal the foreign body is ejected by the severe ^•omiting and retching l)roduced, but emetics should not be given for this ])urpose, as the esophagus has been known to be ruptured by violent vomiting. Various instruments, such as expanding sponge probang, long forceps, blunt hook or coin-catcher, and soft-rubber stomach- tube, have been employed to dislodge the obstruction. In cases of obstruction by bodies of an irregular character the swallowing of a knotted skein of thread fastened to a strong cord will often 8 CHARLES H. IMAYO entangle the foreign body in its meshes upon its withdrawal. An external esophagotomy is much safer than violent efforts at ex- traction or forced attempts to push it downward. Other attempts at removal having failed, operation should be made not later than twenty-four hours after diagnosis of impacted foreign body. The operation of external esophagotomy is usually made on the left side of the neck; carefully separating the tissues between the great vessels and trachea, and avoiding the inferior laryngeal nerve, the esophagus is forced into the wound, and its opening readily effected. If necessary, a finger can be introduced through the opening for the purpose of exploration, as in Case 2. In case of impaction low down in the esophagus a gastrotomy may be combined with an external esophagotomy and the body loosened from below. If necessary, the hand may be inserted into the stomach to effect its dislodgment, as recommended by Richardson, or gastrotomy will admit a probe passed from below, by the obstruction through the opening of the esophagus, carrying a string -^dth a sponge attached, which, upon its withdrawal, may raise the obstruction sufficiently for its removal — the latter method being similar to the Abbe string method of dividing strictures of the esophagus. The wound is dressed with a light iodoform gauze pack, which serves the purpose of drainage and prevents discharge during the process of deglutition. The healing occurs by granulation, immediate suture not being safe. The following case reports illustrate the three ordinary methods of removing foreign bodies from the esophagus. Case 1. — E. L., male, aged sixty-five years. Entered St. Mary's Hospital to be relieved of the major part of the breast bone of a chicken impacted in his esophagus for six days. He was in bad condition, due to repeated attempts to force the body do'RTLward. It was lodged opposite the cricoid cartilage, and was removed after fracture of one wing by forceps, using the finger as a guide. Case 2. — W. C. K., male, aged fifty-three years. Came to the office ^\nth a portion of an oyster-shell caught in his esophagus. The obstruction was only of a few hours' duration, but the patient FOREIGN BODIES IN TRACHEA AND ESOPHAGUS 9 was suffering severely. The obstruction was readily passed into the stomach by the safe pressure produced by the soft-rubber tube. Case 3.-^C. A., female, aj^'cd three years. Was brought to St. Mary's Hospital with complete obstruction of the esophagus I)roduced by an impacted prune-stone. The esophageal sound located the stone at the middle of the esophagus. .Ml attempts to remove it through the mouth with forcei)s and hooks having failed, a low e.vternal esophagotomy enabled us to get at the foreign body, which was situated three inches below the opening, and remove it. Recovery was uneventful, the wound granulating in about two weeks. 'J'he successful treatment of three cases of cicatricial strictures of the esophagus by external esophagotomy combined with gastrot- omy, the passing of probes, and the use of string dilators, has demonstrated the advantage of this method in giving access to all parts of the lower esophagus. REMOVAL OF AN OPEN BUCKLE IMPACTED IN THE ESOPHAGUS (WITH Z-RAY SKIAGRAPH)* CHARLES H. MAYO The report of the following case, with the accompanying skia- graph, is an indication of the future usefulness of the Rontgen rays in practical surgery. M. F., aged three years, was brought to St. Mary's Hospital January 7, 1897, for the removal of a buckle which had been im- pacted in the esophagus for three days without attempt at removal. A skiagraph was made of the child's chest, which showed the buckle plainly visible, open, its teeth projecting up and to the right, just behind the upper part of the sternum (Fig. 1). A left esophagotomy enabled me to effect its removal by hooking a bent probe into a loop of the buckle and extracting it, blunt end first. The wound was packed with gauze, and the patient made an uneventful recovery, leaving the Hospital January 20, 1897. While it was known that the patient swallowed a buckle and had an obstruction of the esophagus, the points of value to be ascertained were the location of the buckle, and whether or not it was open. The former could easily be ascertained by an eso- phageal bougie, but not the latter. The bougie might be harmful, should the buckle be open, by forcing the points through the wall of the esophagus. The danger of its removal by grasping with forceps through the mouth and drawing up the points impacted in the wall of the esophagus for three days will be readily seen. The child was exposed to the a;-ray twenty minutes while asleep in its mother's arms, without removing any but the outer clothing. The dark center of the radiograph is caused by the spine, and the shadow of the heart renders it larger below. The buttons of the clothing are easily distinguished (Fig. 1). * Reprinted from "Northwestern Lancet," 1897, xvii. 10 Fij:. I.— Skiiisraph showing' opin buckle impactetl in the esophagus. CICATRICIAL STRICTURE OF THE ESOPHAGUS* WILLIAM J. MAYO The inaccessible situation of the gullet, its relation to impor- tant structures, and the difficulty attending manipulations within its narrow lumen, all tend to place esophageal obstruction among the surgical problems which are difficult of solution. The cases are sufficiently rare to render an individual experience incomplete, and yet are frequent enough to stimulate our best endeavor for their relief. Konig makes a verj' practical classification of esophageal ob- structions into: (1) Those located within the esophagus, such as inflammatory spasmodic or cicatricial strictures, foreign bodies, tumors, and diverticula. ("2) Pressure obstructions located with- out the esophagus, especially tumors involving the thyroid body, tracheal and mediastinal glands, or aneurysms of the arch of the aorta. Abscess from Pott's disease may also be the cause of pres- sure obstruction. It is to a variety of the first group that I wish to call your at- tention at this time. Etiology and General Character Cicatricial stenosis of the esophagus is the result of the healing of an ulceration. The latter is produced by a traumatism, such as the swallowing of caustic alkali, acids, or hot fluids; occasionally by a wound, or it may be due to the prolonged lodgment of a * Presented to the Section on Surgery and Anatomy at the Fiftieth Annual Meeting of the American Medical Association, held at Columbus, Ohio, June 6-9, 1899. Reprinted from "Jour. Amer. Med. Assoc.," July ^9. 1899. 11 12 WILLIAM J. :mayo foreign body. The most common cause, especially in children, is the accidental swallowing of concentrated lye. Carbolic acid, ammonia, etc., are frequently taken by adults, and the immediate mortality from the poisonous effects reduces to a small proportion the number who might hve to develop cicatrix. The breaking down of a syphilitic gumma may leave an ulcera- tion; cases of stricture having their origin in this manner have been reported by LubHnski. Senn says that syphilis may cause a fibrous stricture of any portion of the alimentary canal, and states that it is not ulcerative in character. Tuberculous ulceration of the esophagus is usually secondary to swallowing infected sputum. Flexner reports 19 and Cone 28 cases of this variety. None of the ulcers in these cases healed sufficiently to cause stricture. Primary tuberculosis of the gullet is very rare; however, it occurs and may be the cause of stenosis. Zenker re- ports such a case. Poncet cites a case of obstruction due to the ulceration attending actinomycosis; no attempt at healing was noted, the interference being mechanical. Fibrous strictures of the esophagus without ulceration have been variously described as idiopathic, syphilitic, gouty, rheumatic, or due to chronic esopha- gitis. Audry records two such cases supposed to be cancerous, which, after death, was sho"WTi to be due to hyperplasia of the muscular coat. Ingals reports scA^eral cases occurring in one family. Rumpel cited a case of fusiform dilatation of the esopha- gus, due, as he believed, to a spasmodic contraction of the thick- ened muscular coat at the lower end of the gullet and collected 20 cases from medical literature. The pathologic condition found in the reported cases of this description shows a remarkable resemblance to the large fibrous stricture of the pylorus, and to that form of fibrous stricture of the rectum, the etiology of which has been the subject of controversy for years. It is probable that these strictures of the alimentary canal have a common origin, and for convenience fibrous strictures of the gullet are classed with the cicatricial form. Simple strictures of the esophagus, of unknown origin, are not CICATRICIAL STUICTLRE OF THK KSOI'H A(;LS 1 '{ uncommon, Kendall Frank records a imiiiher of this variety. Ewald states that the simple or j)C|)tic ulcer may be the cause of stenosis and cites cases; it is possible that this is the etiolofjy of many of the. so-called "simple strictures." Congenital strictures are described at length by Carey, with a report of cases from the literature of the subject. As the treatment of the simple and congenital forms is essen- tially the same as that of the variety under discussion, they are included in the same class. The locations of strictures of the gullet for anatomic reasons are most common in three localities: (1) At the isthmus of the esophagus opposite the cricoid cartilage; (2) at or near the bifurcation of the trachea, where the gullet is crossed by the left bronchus; (3) at the diaphragmatic opening. According to the researches of von Hacker, this latter locality is mo.st commonly affected by caustics. In the adult the gullet is from nine to ten inches in length, and Richardson has shown that from the incisor teeth to the dia- phragmatic opening it is about 143^ inches, measurements which aid the exact location by the sound. If the stricture can be passed by an olive bougie, Tillman advises that the tip be inserted well beyond the stenosis; on withdrawing, the handle is marked as resistance begins and again as it passes through the stricture. The distance can thus be easily estimated. Not infrequently several strictures are found, or the whole of the esophagus may be obliterated, as in cases reported by Richardson and also by von Hacker. Diagnosis The diagnosis of the obstruction is easj-; in fact, advanced mechanical stricture gives a group of symptoms which are self- evident; of these, dysphagia and regurgitation of food are most prominent, and if the obstruction lies in the cervical portion, dys- phonia and dyspnea can be observed. Sounding with suitable bougies completes the diagnosis. In children with advanced cicatricial stenosis the emaciation, the anxious, hungry expression, the constant desire for liquids and rapid regurgitation, form a clinical picture which hardly needs a 14 WILLIAM J. AL^YO question. In adults the diflFerentiation as to the nature of the stenosis may be more difficult. Spasmodic stricture is not rare, and usually occurs in people of a hysteric tendency. According to Moullin, slight traumatisms are frequently the exciting cause of the hysteric form, and the injury may be grossly exaggerated by the patient. The lack of sufficient cause, the suddenness of the onset, and the neurotic history will usually suffice to clear up the diagnosis, yet cases are recorded by Pepper and others in which it was necessary to resort to sounding under anesthesia positively to exclude organic stric- ture. Esophageal diverticula may cause difficulty in diagnosis. The origin of these pouches is stated by Maylard to be: (1) Congenital ; (2) the result of a strictured condition below; (3) from pressure or traction. These diverticula are most common on the posterior wall, about on a level with the cricoid cartilage. Error in sound- ing is thus liable to occur, the bougie passing into the pocket. The possibility of such occurrence will put one on guard and prevent serious difficulty in differentiation. Butlin has collected a number of cases from the London hos- pital reports, many of which have been diagnosed as stricture. Malignant strictures and pressure tumors suggest at once clinical phenomena which will render exclusion easy. Esophagoscopy cannot be considered an important aid to diag- nosis. Von Hacker, in Billroth's clinic, used a Leiter panelectric light \^dth small benefit, and while numerous devices have been developed since, not much evidence of value has accumulated. Stoerk describes a new esophagoscope, but its usefulness seems uncertain. Einhorn is one of the few authorities who place re- liance upon this clinical method in diagnosis. Prognosis The prognosis of untreated cicatricial stenosis is bad — there is a constant tendency to contract, indeed, there seems to be no limit to this process. The esophagus gradually dilates above the stricture; its muscu- CICATRICIAL STUK TLUK OF THE ESOPHAGUS 15 lar cojit hyporlropliics, and, \>y ii scjuccziiij^ process, some nourisJi- ment is forced through the stenosed porli\\'/n\ ciial)!!!!;^ one to pass through the curved i)haryiix without injuriously impinging on the posterior wall. This elasticity is necessary, as it also permits some j)lay at the point and facilitates search oi" the lace of the stricture for the opening. Fig. 2 — Showing manner of introducing the silk cord with the rubber drainage-tube looped in position ready for introduction. Fig. 3. — Showing the manner In which the tube is drawn out in order to pass through a tight stricture Many cases appear to be impermeable to the probe, which in time may be safely passed. The frightened little sufferer adds to the general discomfort and renders a hasty judgment in this regard only too natural. I once heard an eminent genito-urinary surgeon say that impermeable strictures of the urethra hai)pened largely VOL. I — i 18 WILLIAM J. MAYO in an early experience, and I am sure that this is true of esophageal stricture. With care and gentleness a bougie can usually be in- serted in the opening, although several sittings may be required. In such cases a number of whalebone bougies lubricated with glycerin should be passed into the gullet and against the stricture in the same manner as filiforms are used in the urethra, and by Fig. 4. — Showing a double rubber drainage-tube, looped into one already drawn through. As many tubes as may be necessary to obtain the desired dilatation may be drawn through in this way. alternating probes one will usually slip through. It is best to stop here, and then every other day repeat the process, using per- haps several increasing sizes at one sitting; the first probe intro- duced should have a very flexible handle, and when in place, straightens the throat curve so that the larger and stiffer bougies readily follow, each one being left in the stricture until the next CKATHKIAL STHirTlIfK OK TlIK KSOI'lIAGUft 19 one is ready by its side. In ji few eases it may be neecssary to use an anestlietie the first time or two, but usually the patient will sit on a low stool lacing the operator, and i'roni tlic relief ;ilford<'(l by the ability to retain a little nourishmeni, will soon face the ordeal with an unexijeeled degree of fortitude. The fretjueney of sounding depends on the case, every other day being sufficient, and in sen.sitive patients perhaps too frecpient. Many months are oc- eujned in this gradual dilatation, and after ai)j)arent cure a sound should l)e passed occasionally for years. Case 2, — Sfricfure of EsoplHu/iis: flnuliial Dilafdtioti: Re- covery. — T. C, a male, aged two years, was admitted to St. Mary's Hospital February 24, 1898, with a history of having accidentally swallowed concentrated lye seven weeks ])reviously. Difficulty in swallowing developed at once. At first this was due to trau- matism and resulting ulceration, later to the contraction. At no time since the injury has solid food been taken, and a large part of liquid nourishment swallowed is regurgitated. Before admission to the hospital several attempts at sounding the esophagus had been made, but without success. On examination a stricture was detected in the thoracic esoph- agus. Under anesthesia several whalebone bougies were placed in position, and by using first one and then another, a probe was finally ])assed into the stomach. The stricture dilated easily, and a fair-sized bougie was introduced at this time. Systematic sounding every third day was carried out, and March 8, 1898, the little i)atient was discharged, able to take liquids easily, and chopped meat and softened bread with but little difficulty. Since that time soundings have been made at intervals, the esophagus being of nearly normal caliber. Non-dilatable Strictures Stenosis involving a large extent of the esophagus may jirevent gradual dilatation or even continue to contract while attempts at dilatation are being carried out, and in a few cases the difficulties and dangers attending the sounding of a tight stricture make a resort to some more rapid method desirable. For strictures in the vicinity of the cricoid cartilage external e.sophagotomy after Billroth is the ojjcration of choice. Like the 20 WILLIAM J. MAYO perineal cut for stricture of the urethra, funnel-shaped retraction of the cut portion is caused by adhesion to the external tissues divided, and it lessens future contraction. This operation, first performed by Mitchel and later by Annandale, has stood the test, and should not be long delayed. A most interesting case of this kind is recorded by Fenger. Frank has successfully performed esophagectomy in a case of simple stricture in this portion of the esophagus, the divided ends being sutured over a tube passed through the nose. For dense stricture above the arch of the aorta and below the point which can be directly divided, Gussenbauer's combined esophagotomy is the best operation: through an external incision in the neck a tenotome is introduced and passed downward to the stricture, which is then di\aded. I take the liberty to introduce an illustrative case previously reported : Case 3. — Tico Strictures of Esophagus; Prolonged Dilatation and Combined Esophagotomy; Recovery. — J. H., aged four years, was seen in January, 1893. In the spring of 1892, while living in North Dakota, the child swallowed concentrated lye. After a month's severe illness he gradually recovered, but ^dth an increasing difficulty in swallowing. During the past three months all food has to be strained and he has frequent attacks of regurgitation. The material regurgitated is brought up somewhat slowly. He is a moderately well-nourished boy, of good disposi- tion, which materially aided treatment. On the introduction of a catheter it was arrested at the isthmus of the gullet. After some manipulation a No. 3 urethral catheter was passed three inches farther doTvnward, where it was completely arrested. It was very evident that there was a stricture at the level of the cricoid cartilage, and a second in the intrathoracic portion of the esophagus, undoubtedly produced by the action of the esoph- ageal muscles at the time of the original accident, causing the effect of the lye to be located at these points. Regular sounding was instituted, and continued at intervals of three to ten days, with but slight interruption for one year. At the end of two months the lowest stricture was passed by a fine whalebone bougie. This stricture was very dense, and re- sisted sounding obstinately; it was only at intervals of several nr'ATHKIAI. STUKTT'UE OF THK ESOPIIAfa'S • 21 sittings tliat tiic hoiigic could he iiit roducc*! into I lie stom;ich. At the ('11(1 of a year's perscvcranct; llie upiKT stricture had yielded to the extent of admitting a No. 5 catheter and the lower one a No. .'{. The l)oy was well nourished, did not re^'urj^'itate unless he took solid food, and was ahle to take softened hread and finely chopped meat. During the preceding two months no perceptible improvement was manifest; therefore on PY'hruary 15, 1804, he was admitted to St. Mary's Hospital, and on February IGth left external esophaj^otomy was performed. The ni)[)er stricture was now readily dilated })y forceps. 'J'he hjwer stricture, which wa.s below the level of the uj)per end of the sternum, wa.s carefully nicked with a dull knife on a fjrooved director and dilated with forceps, j)ermittin<,' of easy catheterization of the esopha^Mis. Unfortunately, the majority of dense strictures are in the lower esophagus, or if there is one stricture in the upper portion, another will usually be found lower down. In three cases in the experience of the writer more than one stricture was present, but the lower in each instance was the more dense. How to reach strictures situated below the arch of the aorta in the thoracic esophagus by direct means has been, and is yet, a problem. NasollofT, Quenu, and Hartmann developed an opera- tion for external eso])hagotomy on the cadaver, which Rehn first l)cri'ormcd on the living subject, an incision from the fourth to the eighth rib on the right side of the spinal column being employed. The difficulties and dangers of this plan of attack are so great that indirect measures are to be relied u[)on. For this purpose two methods are available — Abbe's string- saw and Ochsner's operation; in the latter a loop of rubber tubing is used as a dilating medium. The one method supplements the other. Division of the stricture by the string-saw invented by Abbe has l)een jjcrformed a number of times with great success. Through a gastrotomy wound a stout silk cord is passed in a retrograde way through the esoi)hagus and out through the mouth or through an external esophagotomy. The stricture is made tense by engaging bougies into it from below, and by a sawing motion of the cord the tight bands are divided, while the important soft parts are crowded back out of the way. After full dilatation 22 WILLIAM J. MAYO has been secured the incisions can be closed, or a rubber tube is inserted to a point above the stricture and brought out of the gastric incision, the latter being united to the margins of the ab- dominal wound. In two or three days the tube is removed and sounding from above employed in the usual manner. Case 4. — Stricture of Esophagus; Division and Dilatation after Gastrotomy and External Esophagotomy; Testimony as to the Value of Abbe's String Method of Division. — H. W., aged three years, was admitted to St. Mary's Hospital on October 5, 1892, with the folloT\dng history, given by her mother: One year previously the child accidentally swallowed caustic lye. For four weeks she was very sick, then gradually improved for several months, although totally unable to swallow soHd food at any time. During the last six months she has had great diffi- culty in swallowing liquids, and has regurgitated from the esoph- agus more than half of the nourishment taken. She is emaciated to a considerable degree. Upon being given two oimces of milk it was drunk with avidity, but was nearly all regurgitated. The dilated esophagus above the stricture holds nearly four ounces of fluids. Bougies were inserted through the pharynx into the esophagus, but were completely arrested at its lower end, while careful search of the face of the stricture with a number of whalebone bougies, under anesthesia, failed to pass the stenosed portion. At intervals of four days this search was continued, but without result. While undergoing these repeated examinations the child became much better and was able to retain a larger proportion of liquid nourishment, and improved visibly. Proper manipulation was much interfered -^dth by the short, sharp curve of the child's pharynx and its small size; therefore external esophagotomy was urged upon the parents to permit of more direct access to the parts involved. This was decUned, and the little girl was taken home at the end of two weeks. On July 19, 1893, the child was readmitted to the hospital. The slight improvement had been of short duration, a gradual closing of the small channel had taken place, and for the past four weeks it had been nearly, if not quite, impermeable, so that the little patient had been nourished by rectal enemata. As before, careful search failed to pass the stricture. External esophagotomy with a hope of finding a way through the damaged esophagus did not promise the rapid relief that the child's nutri- ricATKKiAL sTRinrHr: of tiik kscjimiaol's 23 lion tleniaiuled; tluM-clorc on July 'iO, IBO.'J, a ^'astroslomy was made alter the iiietliod ot" lH'iij,'('r, tlu" incisitjii l)eirifi made us lii^^li lip and as close to the median line as jxjssihle, to permit of ret ro^radc dilatation. Fen^^er's oj)eration was chosen, as permitting more direct and easy access to the cardiac orifice ot" the stomach, in place of the von Hacker or Witzel methods, which are far superior as a means of preventing leakage from the fistula, but also present greater obstruction to intraj^'astric manipulation. The child was now well nourisiied through the gastric fistula, and twice a week the lower face of the stricture, which was at the diaphragmatic opening, was carefully searched for an outlet. On two occasions the finger was introduced through the fistula into the stomach and used as a guide for the i)robe, the patient being anesthetized. The dila- tation of the fistula necessary to introduce the finger in so small a subject greatly increased the leakage through the gastric open- ing, and the child again began to fail, from the inability of the stoiiiach to retain nourishment. On August 25, 1893, a left external esophagotomy, at a low point in the neck, was performed by C. H. Mayo, and a bougie introduced through this opening and pressed against the stricture, pushing it downward into the stomach and holding it steadily. With a finger in the stomach a long malleable German-silver probe was passed through the stricture and out of the esophageal fistula in the neck, A heavy double strand of braided silk was drawn through the channel, having an end out of the opening in the esophagus and also out of the gastric fistula. The stricture was about three quarters of an inch in length, as nearly as could be ascertained. Using one strand to cut the tissues, after the method recommended by Robert Abbe, knots were tied on the second strand and pulled into the obstruction to keep the stricture tense, and the opening was greatly enlarged. This method was of the greatest value, and had but one disadvantage — that of cutting the edge of the gastric fistula when drawn taut, which was in part obviated by holding the lower end of the cutting thread in the bite forceps, held on a j)lane with the stricture from within the stomach. Knots on the second thread, to render the stricture tense, were equally good and easier of execution than pushing bougies from below into the small opening, as recommended by Abbe. During the succeeding month, at intervals of four days, this process of division was carried on, and perforated shot clamped upon the thread were drawn through to assist in the dilatation, iis they readily followed the tortuous passage. Bougies then 24 WILLIAM J. MAYO became permissible, inserted first from the neck and later through the mouth. In five weeks the threads were removed and dilatation was carried on with olive-tipped whalebone bougies made for the purpose. A sister of the little patient was taught to pass the probe, and the child was discharged from the hospital in two months in a good general condition; able to drink milk readily and to take chopped meat and bread with little effort. The esophageal fistula had spontaneously closed, and the gastric open- ing, after being touched with the cautery, was nearly cicatrized. March 19, 1899, the patient is well. The following operation was first successfully performed by Ochsner in February, 1899. In April, 1899, 1 had an opportunity to employ this method, with satisfactory results. The technic is as follows: The anterior wall of the stomach is dra\\^ out of a left oblique incision through the abdominal coverings ; a small opening is made into the stomach sufficient in size to introduce the finger. A whalebone probe, to the tip of which a silk string guide has been tied, is now passed through the esophagus either from above or retrograde, as in the Abbe method. With this guide a loop of silk is drawn out of the gastric incision in such manner as to leave the guide as a third string. Into this loop a small soft-rubber drainage-tube, three feet or more in length, is caught in the middle; by traction on the ends of the doubled thread through the mouth this loop of rubber tube is drawn through the stomach and made to engage in the stricture. The greater the amount of traction, the smaller the stretched rubber tube, until it is sufficiently reduced in size to enter the stenosed portion; by alternating the direction of the pull the tube is drawn out by its free ends and in by the silk loop. Increasing sizes of tubes can be employed, and, if necessary, the third string can be used as a string saw after the Abbe plan of procedure. Case 5. — Stricture of the Esophagus; Gastrostomy and Dilatation by Ochsner's Method; Recovery. — G. H., female, aged nine years, was admitted to St. Mary's Hospital March 3, 1899, with the following history: Four months before she accidentally swallowed CICATUICIAL STUICTL'llE OF TlIK K.SOI'HA(a:.S '2.'} coMcoiif nitcd lye. Tlic iilccriilioii |)ro(luc('(l was slow in Iicaliii^', and tlicdifliculty in swallowiii^f ^M-adiially Ix'caiiK' iiion" proiKjiiiifcjl. For the past month only htpiid has l)('en attcnij)t('wer esophagus was readily detected by a bougie. After considerable effort a fine whalebone probe was j)assed. There apjjcared to be two points of contraction, about an inch apart, the lower being at the dia- l)hragniatic opening. Systematic probing during the next seven weeks did not yield nnich result; occasionally a larger probe could be passed, and again only the finest could be used. During this time rectal feeding was employed to supplement the very limited nourishment obtained by means of the esophagus. The little pa- tient became so much reduced that on April 20, 1899, the abdomen was opened by a left oblicjue incision. A fine probe armed with a silk thread was passed through the eso})hagus into the stomach. The thread was caught and drawn out of the incision; by this means two threads of heavy silk were drawn upward and the silk guide left in position. A quarter-inch rubber drainage-tube was lubricated and caught l)y the middle in the loop; by traction from above the tube was drawn through the stomach and into the eso])hagus, considerable traction being required to engage it in the strictured area. By first using traction U])ward on the strings and then downward on the free ends of the tube the stricture was rapidly dilated. The dilating tube was removed, and a half-inch rubber tube was introduced into the gastric opening, and through this tube the three threads were drawn, the free ends being tied to the ends projecting from the mouth. A gastric fistula was then formed after the method of Bernays, the rubber tube being inclosed by three superimposed circular purse-string sutures of catgut, one-fourth of an inch apart, intro- duced in the wall of the stomach around the incision, the margin of the fistula being then sutured to the abdominal woimd. This causes a cone or nipple-like lu'ojection of the wall into the lumen of the stomach, and through this tu])e sufficient nourishment was introduced. The general condition im])r()ved ra])idly and sys- tematic sounding was carried on as before, with increasing sizes of bougies. Internal esophagotomy, first performed by Maisonneuve in 1861 , has since been made about 'io times (Richardson), with a death-rale 26 WILLIAM J, IVIAYO of over 25 per cent. Above the arch of the aorta the operation is unnecessary; below it is a "chance shot" and success a lucky acci- dent. Konig believes that permanent gastrostomy is preferable. Sands' instrument shares with ISIackenzie's the doubtful honor of being best for the purpose. Meyer, in an interesting summing up of internal instrumental esophagotomy, says that the danger of accidental injury to important structures is not greater than the introduction of infective material without the gullet, such infec- tions being responsible for more than hah the mortality. He recommends gastrostomy for the purpose of feeding and also to allow preliminary cleansing of the operative field, if internal instrumental esophagotomy is adopted. Internal divulsion is nearly as dangerous as the cutting operation and less effective. Fletcher's esophageal di\Tilsor has been used a few times. Mixter has devoted a great deal of attention to this subject. He recom- mends that Symonds' tube, as originally advocated for malignant strictures, be introduced, and believes that that steady pressure will gradually increase the caliber. It would seem that a stricture of sufficient size to permit of permanent tubage could be overcome by gradual dilatation. The literature of the subject is filled with cuts of more or less ingenious instruments for the purpose of dilat- ing these strictures: the spiral rolled tin sounds of Rosenheim, the laminaria tent of Senator, and a host of others, interesting but of questionable value; of these, linear electrolysis, as introduced by LeFort, has had the greatest reputation. The marvelous nature of the cures MTOUght in many cases leads one to believe that the enthusiastic advocates of electricity failed to exclude the spasmodic variety with sufficient care. Impassable Strictures In a moderate number of strictures a probe cannot be passed through the mouth, yet after an external esophagotomy the bougie can be manipulated to so much better advantage that the opening may be found, and, after a few soundings, conducted through the fistula; the future dilatation is continued through the mouth. CICATRICIAL STUICTLRIO OF TFIK KSOI'HACL'S 27 Graser particularly advises this in children. KaiiimcnT reported a case before tlie New York Surgical Society in which, after faih'ng to pass the stricture by retrograde sounding through a gastric- in- cision, success followed probing from al)ove through an external esopliagotoniy. Konig had a series of fine siKer balls made and threaded; one swallowed at Ix'dtime would usually pass the sten- osed area during the night, and was drawn back through by the thread in the morning. Zeehei.sen records two cases in which an opening passable to a sound was .secured in this manner. Billroth succeeded in several cases with a cylindric cloth bougie partly filled with mercury, the weight and adaptability of the metal carrying it through. Ga.strotomy and retrograde dilatation were recommended by Schede in 1883, and first performed by Trendelenburg. Frank collected 20 cases cured in this manner. Since that time a large number of successful cases have been recorded. Great and unex- pected difficulty is often experienced in attempting retrograde dila- tation through a gastric fistula. It is a surprising fact that under such circumstances it may be almost impossible to find the car- diac opening. For this reason the dilatation should be carried out as the primary operation if possible. Richardson directs that the anterior wall of the stomach be delivered and a small transverse incision made into its cavity near the lesser curvature, in the neigh- borhood of the pylorus; by traction on the stomach just below this incision the lesser curvature forms a sulcus along which the instru- ments glide into the cardiac orifice. Hagenback turned defeat into triumph in a case in which he was unable to find the stricture from below, and made a gastric fistula for feeding purposes. He caused the patient to swallow a small perforated .shot to which a thread was attached; this passed through the stricture and was hooked out of the fistula, acting as a guide for future manipulation. Abbe recommends that a string guide be introduced by which dilating bougies can be drawn upward retrograde, and that this guide should be retained in place until a sound can be introduced from above. Observation has shown that it is possible to pass an 28 AVILLIAM J. MAYO instrument retrograde when impassable from above, but that regu- lar dilatation from below without a guide may prove very difficult. Twenty-eight cases from the literature of retrograde dilatation are analyzed by George Woolsey, and the comparative merits of the various methods of cutting are clearly stated. It may happen that the condition of the patient will not permit prolonged attempts at retrograde dilatation, and a rapid gastros- tomy should be done by the Witzel or the multiple purse-string method of Bernays, as either permits immediate feeding, does not leak, and yet can be readily converted into a direct opening by a dilator when needed for later attempts at retrograde dilatation. With the improvement in nutrition which follows gastrostomy and the absolute rest to the esophagus, in a short time a probe may be passed from above. Case 6. — Cicatricial Stenosis of the Esophagus; Gastrostomy; Gradual Dilatation Through the Mouth.— V. M., aged three years, was admitted to St. Mary's Hospital July 1, 1894. Nine months previous to admission the little fellow had accidentally swallowed concentrated lye. Symptoms of esophageal obstruction becoming more and more pronounced, attempts were made to dilate the strictured gullet without success. At first solids were regurgi- tated, later fluids, and for six weeks nourishment had been main- tained by rectal feeding. The child is emaciated to an extreme degree, and is too feeble to stand. Fluids are eagerly swallowed, only to be regurgitated at once. Attempts to pass whalebone probes through the stricture were unavailing. Gastrostomy after the Witzel method was per- formed and immediate feeding resorted to. Regular feeding through the gastric fistula soon improved the patient's general condition. Systematic search for an opening through the stricture from above was made every other day for several weeks before a small whalebone probe passed. After this success the bougie failed to find the orifice a second time for a number of days. Three months of persistent effort finally developed a moderately sure passage. For five months all the feeding was carried on by the gastric fistula, a good-sized opening through the stricture having developed by this time. On February 22, 1895, the patient was discharged, being able to eat ordinary food, and since that time he had been regularly firATRICIAL STRICTURE OF Till-: ESOPHAGUS '20 sounded and had ri'iiiaiucd in cxccllcnl licaltli until June, 18!)H, wlien he was reachnitted to the hospital with coinj^lete obstruction, hiivin^ three (hi\'s l>efore aceidentally swallowed a mass of ehewing- <,'urn which had ti since remained in good condition. There will yet remain a few cases in which a large part of the esophagus is obliterated, and permanent gastrcstomy after the Frank method is the melancholy outcome. In summing up, the following conclusions may be formulated: 1. Systematic sounding should be commenced in from two to four weeks after the swallowing of a caustic substance. 2. Should the traumatism be severe, immediate gastrostomy will lessen infection and hasten cicatrization, sounding being carried on as before. 3. Non-dilatable strictures in the vicinity of the cricoid carti- lage should be divided by external esophagotomy. 4. Stricture above the arch of the aorta may be safely cut by a combined internal and external esophagotomy. 5. Dense thoracic strictures are best dilated by Ochsner's method, and, if necessary, divided by Abbe's string saw. 6. Impassable strictures should be treated by retrograde dila- tation. 7. A dilated stricture should be occasionally sounded for years, if not for life. 1 STOMACH i SURGERY OF THP: STOMACH* WILLIAM J. MAYO The stomach is now amenable to surgical relief from a number of conditions which until recently have been supposed to be purely medical in character. In many instances the stomach offers un- expected opportunities for operative interference. First, it is an organ more or less fixed within a certain definite part of the upper abdomen. Second, in a large majority of instances it can be irri- gated and rendered fairly clean before operation. Third, bj' rectal alimentation it can be given rest after operation. Fourth, its thick muscular walls afford a good hold for sutures, while its blood- supply is of a nature to favor union after extensive resections, herein differing materially from the thin-walled intestine, nourished by the mesentery, with more or less constant vermicular action. The cardiac extremity of the stomach is the only fixed point, and it is suspended, so to speak, in a way to vary its position, whether contracted or dilated — in the former case lying deep under the liver against the crura of the diaphragm, and in the latter occupying a prominent position in the hypogastrium. The lesser omentum, containing the main blood-vessels, is of great surgical importance, as are also the relations of the mesocolon and other delicate structures in the neighborhood of the pylorus. As pointed out by Tillaux, the cartilage of the ninth rib forms an important guide to the lower border of the stomach. The diagnosis of gastric disease, as a rule, is not difficult. The stomach can be distended by ether, as recommended by Felitzet, or by the use of bicarbonate of soda and an acid, as practised by Jacobi. Usually, tiic introduction of a definite ([uantity of water * Reprinted from tlie "Medical Record," XovemlH-r 10. 181)4. pp. ,'>S(>-.>S^. VOL. 1—3 33 34 WILLIAM J. MAYO will sufficiently mark its outline, and, by forcing air into the colon per rectum, as practised by Senn, the relation of the stomach to the transverse colon can be mapped out. By such means I was able to diagnose a pancreatic cyst, the tumor being shown to be retroperitoneal and lying behind the gastrocoHc omentum. The operation proved the diagnosis. The value of the examination of test -meals and the amount of acid as affecting the diagnosis of cancer as pointed out by von Jaksch in his "Clinical Diagnosis," the presence of free hydrochloric acid being a factor in the diagnosis as against cancer and in favor of ulcer. The effect of the weight of tumors in dis- placing the stomach is ably shown by Osier in a series of papers, beginning in the "New York Medical Journal" of February 3, 1894. As preliminary to operation upon the stomach, irrigation is of great value where it can be practised, and, as pointed out by Kussmaul, much benefit may be derived from lavage where ob- struction exists below, thus relieving the stomach of irritating material, removing the distention and interference with breathing, and putting the patient into better condition for operation. Shall the stomach be distended previous to operation, to facilitate its discovery.'' Distention may assist in finding the stomach, but it complicates the further manipulation and introduces an element of danger of wound contamination from escape of its dilating con- tents. The after-treatment of gastric operations is changing toward the earlier administration of food by the stomach, especially in exhausted patients, although rectal feeding is the rule during the first few days. Before closing these general considerations the writer wishes to call attention to gastric distress produced by traction of ad- herent omentum. Billroth has written upon this subject, es- pecially in reference to the small "buttons" of omentum protrud- ing through little gaps in the upper abdominal wall. Konig re- ports some 20 cases operated upon. We have observed two cases of this description: Case 1, a boy of fifteen, with a small hernia of adherent omentum protruding through a little defect in the SURGKI{Y (M" Till-: STOMACH .S.> nuMlian line ahovc the iiinl)ilicus. Case ^2, a woman of filly, with adherent oinenluni in a femoral hernia. Neither of the patients had any local symptoms at the site of the protrusion, and both were relieved ll)y operation. We have also observed cases of this form of omental adhesion with corresponding]; symptoms in ef)n- nection with post-operative ventral hernias. Wounds and injuries of the stomach are to be treated on sur- gical principles, and where evidence of perforative wounds are found, operative repair is imi)erative. Rose speaks of persistent vomiting of blood as a reliable sign of gastric injury. In case there is injury to the head at the same time, the possibility of blood having been swallowed from the upper passages should be borne in mind. Foreign bodies in the stomach are a not infrequent cause of danger and suffering, and gastrotomy for the purpose of their removal has been greatly perfected. Richardson gives a careful analysis of reported cases, and makes many valuable modifica- tions in the technic, when the foreign body lies in the lower esopha- gus. Bull also reports excellent work in this direction. The value of a potato diet in the non-operative treatment of foreign bodies of the alimentary canal should not be forgotten. Fistulas are, as a rule, the result of operative technic, and rarely come from accidental or pathologic causes. They vary greatly in the difficulty of repair on account of location, those adjacent to the bony framework being more dijQBcult of operation than those more remote. The value of the provisional suture to prevent escape of stomach-contents during manipulation is ap- parent. Chronic dilatation of the stomach, when not due to organic lesion, may be caused by chronic gastric catarrh with sagging of the fundus, rendering evacuation difficult. Taylor has studied this condition carefully, and believes that it produces a kinking at the pylorus, with prominence of the mucous fold, causing a bar-like obstruction to theemptying of the stomach. He practises oi)era- tive dilatation of the pylorus with success. Bircher, followed by WVir, raises the fundus mechanically and reduces its size by putting 36 WILLIAM J. RIAYO a longitudinal plait along the walls, turning a fold into its cavity. This operation is indicated in cases where lavage has for a time failed to relieve the symptoms. Ulcer. — In the diagnosis of ulcer, the age of the patient, the quantity of the vomit, the lack of emaciation, which in ulcer is more often a profound anemia mthout great loss of flesh, the presence of free acid, and the slow course of the disease are all valuable symptoms of ulceration as against carcinoma. Opera- tion may be demanded for acute perforation, provided it can be diagnosed in time. As can be readily perceived, the location of the perforation, whether on the anterior or on the posterior wall, will greatly influence the ease of repair. Ulcer of the stomach being most frequent on the posterior wall, a perforation may oc- casionally take place, and the escaping material, on account of its mechanical surroundings, become encapsulated as a subdiaphrag- matic abscess. Weir points out the frequency of left pneumo- thorax due to the secondary results of perforative ulcer. The following is a report of a case of subdiaphragmatic abscess (No. 1537, St. Mary's Hospital Record) : Scandinavian, fifty years of age, with history of chronic stomach trouble, became suddenly ill, developed pain, and, later, cough and hiccup. A deep swelling became manifest in the sixth inter- space, just to the right of the sternum, for relief of which he entered the hospital. A deep incision was made and pus evacuated; but as a sinus formed and failed to heal, the cartilages of the sixth and seventh ribs and a large portion of the lower end of the sternum were resected. The sinus was followed through the diaphragm to the neighborhood of the lesser omentum, where a small cavity existed; this was packed \\'ith gauze and slow healing followed. The secondary results of ulcer in producing contraction at the pylorus may call for operation for obstruction — either the digital divulsion of Loreta or, better, the Heineke-Mikulicz pyloroplasty. Weir reports a case of this character in which gastro-enterostomy was performed, and in selected cases advocates its expediency. Cancer of the Stomach. — The age of the patient, lack of free acid in the vomit, the emaciation, and the possible presence of SURCEKY OK TIIK S|()MA( II .'37 pt'plont's ;iiiIay 5 to 8. 1S9G. llopriiitetl from "Jour. Amer. Med. .Vssoc.," Juse, lS9(i. 43 44 WILLIAM J. MAYO corroborative value only. Many new methods of examining the stomach have been reported. The use of a small electric light to illuminate the gastric cavity; also complicated apparatus for the purpose of reflecting or measuring the interior of the stomach, etc. These methods are of little value to the practical surgeon at the present time, although they may possibly be of use in the future. The examination of the urine for the finding of certain products supposed to be indicative of cancer is interesting, rather than important. Mechanical obstructions of the cardiac orifice of the stomach or the esophagus are most commonly due to malignant disease or the cicatrization following upon the action of caustics; more rarely to aneurysms or tumor pressure from without. The diagnosis of the location of the obstruction by means of esophageal bougies is easy and needs no comment, while the history and rational signs and symptoms point to the pathologic nature of the obstruction. The treatment of those forms due to stenosis as a result of scar tissue is exceedingly trying. Some of the less resistant cases, when seen early, can be dilated by means of bougies used through the mouth, but, after failure of catheterization in the usual man- ner, the opportunity of more direct manipulations afforded by external esophagotomy has been pointed out by Sonnenberg, and should be borne in mind. In two out of three cases of cicatricial stenosis of the esophagus of my own this was well illustrated — one, a male aged four, in whom a probe could not be passed through the mouth, but was passed through the external esophageal incision. If unable to pass a bougie, retrograde dilatation by means of gas- trotomy is a rational procedure, and with the string method of dividing the stricture introduced by Abbe, we were enabled to deal with otherwise hopeless conditions in two cases. In the gradual dilatation which is so necessary in the successful after- care of these sufferers I have found the olive-tipped whalebone bougies the ones of most value, and after the treatment is well advanced, the use of probes having several graduated bulbs on one stem, as recommended by Solis-Cohen, is of great service. Gastrotomy for the purpose of retrograde dilatation is, per- MF.cnAMCAI. CATSES INTKHI F.IU NC WITH (iAsTHH AfTICJ-V 45 hai)s, host done l)y FiMi^'er's ()Mi((iie left laterul incision tliron^'li the ahdoniinul wall, which l)rin<^s this opening more directly in line with the cardiac orifice. For work on the pylorus or the re- raoval of foreign hodies or gastric exploration generally, the central incision al)o\'c the unihilicus has many advantages. Gastrotoniy for the removal of foreign hodies accidentally or purposely swallowed or slowly collected, such as hair-balls, is an operation of great efficiency. Much credit is due to Richardson for his work in this field. Gastrostomy for the purpose of feeding, necessitated by cardiac or esophageal obstruction, when done by Fenger's method, is subject to great annoyance in the waj^ of leakage. This was particularly marked in a case under my care. Fortunately for the comfort of the little patient, the purpose of the opening was shortly obviated by relief of the obstruction. The necessity of repeated attemjits to penetrate the strictured portion of the lower esoi)hagus, as well as for feeding j)urposes, made Fenger's opera- tion, in this case, the only one available. Other methods of operation, while giving better closure, would equally prevent I'eady access to the cardiac orifice. Both the Witzel method and that of Frank are free from this annoyance, and for temporary purposes the Witzel method is of the greatest benefit, as immediately after removing the tube the fistulous tract closes. Tliis was so well marked in one instance that the accidental slipping out of the tube was followed in a few hours by great diffi- culty in reinsertion. Frank's sprout method, as it requires no tube, is undoubtedly the best for permanent feeding, and as it is an opera- tion which can be speedily performed, it is classed with the Witzel and is far superior to that of Hahn, von Hacker, or Fenger in preventing leakage. Obstructions at the outlet of the stomach are exceedingly common, and are too often pronounced malignant without proper examination. This is especially true of the pyloric stenosis sec- ondary to ulcer, and as even a tumor may be found, in the latter condition due to the peritoneal thickening over the cicatricial area, much care should be given in the diagnosis to determine the nature 46 willia:^! j. zsiayo of the obstruction. The history of prolonged previous ulceration, with the presence of free hydrochloric acid, and its slower course are among the readier means of differentiation, and it should also be borne in mind that a non-malignant pol;sT3oid or valve acting tumor at the pyloric orifice may be the cause of the obstructive symptoms, as well as a cancerous growth. For the rehef of non- malignant stricture at the pylorus Loreta's method of divulsion is open to the objection of strong probability of recontraction, so common in other strictured mucous passages, such as the urethra. The Heineke-Mikulicz pyloroplastic operation is the one of choice, and is wonderfully well adapted to the average case. For example, I ^"ill briefly report the following case : M. M., male, aged forty-six, has suffered from severe and painful gastric sjTnptoms for seven years, for which he has been treated almost constantly, obstructive symptoms gradually becom- ing more prominent. For the past eighteen months he has vomited daily almost the entire amount of food taken, and has lost 80 pounds in weight. Just above the umbilicus a small tumor could be felt, which was more or less movable, and on dilating the stomach with air this mass moved to the right and upward, while the air-dilated stomach filled almost the whole abdominal ca^aty. By making pressure on the dilated stomach and listening with a stethoscope over the enlargement, a hissing of gas passing through a fine orifice could be heard. Test-meals showed the stomach-contents to contain free acid. Celiotomy showed, as expected, a cicatricial obstruction of the pylorus due to former ulceration. Pyloroplasty was readily performed, and as the patient took the anesthetic badly, he was allowed to come out from under its influence and complained of no pain during the introduction of the Czerny-Lembert sutures in the stomach- wall, but required further anesthesia in closing the external incision. This lack of sensitiveness in the visceral peritoneum has been pointed out by Greig Smith. The patient's recovery was prompt, and the gain in weight remarkable. For inoperable obstruction, such as advanced malignant dis- ease, gastro-enterostomy is the operation of choice. Anastomosis by means of bone plates, inaugurated by Senn, first popularized .MKCHAMCAL CAUSES 1.NTEU1-E1U.\(; WITH (lASTlilC ACTKJX 17 lliis ()|)('rati()ii. It iniiy, liowover, be open to tlic oljjcctioii of too small an opoiiinji; with too large a coaptating surface for future contraction. The suture method is slow, with danger of leakage, and since the average j)atient is only too fre(juently at the point of collapse from chronic starvation, the Murphy l)utton, even with the (lisatlvantage of its possible jiassage backward into the stomach rather than onward into the intestine, is a quick and safe method. I have employed the ])utt()n in three gastro-enterostomies — twice successfully and once followed by death; in the latter case the union was perfect, although the patient, on the verge of collapse from starvation due to advanced malignant obstruction, succumbed to exhaustion on the fifth day. Of the two that recovered, one is alive and well, with a gain of 40 pounds in weight at the present time, one and one-half years after the operation. This patient was suffering from acute starvation, and his condition would not war- rant a pylorectomy, which a tumor of unknown nature indicated. In making a gastro-enterostomy the jejunum should be caught to its origin, and a loop formed with the direction of its peristalsis in the same direction as the stomach, to prevent impaction of the upper portion, or, as Kocher recommends, transversely to the axis of the stomach, thus allowing gravity to aid the passage of the food in the proper direction. The cureting of gastric cancer through a gastrotomy wound, as advocated by Bernays, has little to recommend it. Pylorectomy is but infrequently done. Its frightful mortality in malignant disease without reported perma- nent cures is not encouraging. The reason for this great mortality lies in the debilitated and starved condition of the patient at the time operation is resorted to, although Kocher has recently re- ported five primary recoveries out of seven pylorectomies. In the case previously referred to of tumor of doubtful nature obstructing the pylorus, in which the stomach and jejumun were joined with Murphy's device, it was my intention to wait until the patient was in better physical condition as a result of the gastro-enterostomy and then, at a secondary operation, resect the pylorus and com- pletely close the end of the duodenum and the stomach, utilizing permanently the existing fistula. The patient received so much 48 WILLL^M J. iLA.YO relief from the first operation that he decHned to submit to any further interference. I fully believe that preliminary gastro- enterostomy would put the patient in much better condition for the major operation, and, by allowing permanent closure of both duodenum and stomach at the points of section, the mortality would be greatly reduced. In any case it is better, as pointed out by Bull, to carry the gastric opening well away from the dangerous suture line. Brandt has treated dilatation of the stomach caused by chronic catarrh by plaiting its anterior wall with a Lembert suture, and he reports great benefit from this procedure, the object being to re- lieve the dependent portion, allowing gravity to aid the gastric contents to work toward the pylorus. Cases of chronic dilatation of this character are usually due to some form of pyloric obstruc- tion, and the operation is, therefore, rarely indicated. Among the external causes of interference with the stomach are adhesions of the pylorus or duodenum to the gall-bladder, due to the inflammation excited by gall-stone. The most common cause of external interference with the action of the stomach is the fixation of some portion of the omen- tum in a hernial ring, producing traction upon the stomach, and of these incarcerated omental hernias, the least often recognized are the button-like protrusions through little defects in the median line above the umbiHcus. We have operated upon a small number of such cases, with marked relief to gastric symptoms. Irreducible omental hernias of any variety are almost always accompanied by gastric distress, and it is the rule that this dis- appears after the radical cure of the hernia. Not only do omental adhesions cause distress, but they may fix or distort the stomach to an astonishing extent without symptoms directing attention to the hernial protrusions. In one instance a male, fifty-four years of age, had suffered for seventeen years from gastric pain and chronic indigestion. Test-meals showed free acid. On dilating the stomach with air it was found to expand to a remarkable degree downward and to the right; careful examination revealed an old irreducible omental hernia of small size on the right side, which had MIOCIIAMCAI. CAT'SES INTERFFOUI \C; WITH <;A>THI( A( TION 1!» existed lor years. Itadical oper;!! ion on I In- licrnia, w il li lilx-ratioii ol" liic omentum, proniplly relie\-e(l llie s_\iii|)lonis. 'J'liese cases are common. Chronic gastric distress following alxh^niinal o|)era- tion is not infre(|iieiilly dii(> to post -operative adhesions of l)rni■^(•d or ligaled onienlimi, and may necessitate secondary oi;eralioii lor I heir iiheral ion. CICATRICIAL STENOSIS AND VALVE FOR- MATION A CAUSE OF PYLORIC OBSTRUCTION A REPORT OF FIVE CASES RELIEVED BY OPERATION* WILLIAM J. MAYO Non-malignant forms of pyloric obstruction are not infrequent, and with some exceptions have been so commonly confused with cancerous disease that patients have been allowed to die without surgical intervention. When the stenosis results from the cicatrization of an ulcer which existed years previously, the patient may be in middle life or later before the obstruction becomes marked, and the cachexia of chronic starvation closely simulates that of malignant disease. Now and then cases operated upon by masters in the art of surgery, e. g., Senn, Lange, Weir, and others, have been reported in American literature. European surgeons have been more advanced in this respect, and have furnished many important contributions to the subject. Stricture of the pylorus following upon the healing of gastric ulcer is the most common form of non-malignant obstruction, and 4 out of 5 cases upon which I have operated have been of this variety. The site of previous ulceration in these cases was along the lesser curvature and anterior surface of the stomach, and in 2 the cicatrix was very extensive, a prolongation extending downward in the pyloric region and causing great distortion of the canal. In 2 of the remaining cases the strictured area was narrow and *Presented to tlie Section on Surgery and Anatomy of the Amer. Med. Assoc, Philadelphia, June, 1897. Reprinted from "Jour. Amer. Med. Assoc," October If), 1897. 50 SO.MK CAISKS OK rVLOHK" OUSTHITTION' ."51 circular in form, one with its origin in a cicatrix along the anterior wall and the otluT luiving no apparent cause. In these 2 eases the dilatation of the stomach was extreme, although the mechan- ical obstacles to its emptying were not so great as in the 3 cases in which the scar tissue was more extensive. In 1 of the '■I cases the pylorus was held up very high by a short gastrolie|)atic omentum, and in both cases the downward sagging of the stomach produced a well-marked kinking or valve formation at the pylorus, which was evidently obstructive in its nature. In a patient who died from chronic starvation, supposed to be due to pyloric obstruction of a cancerous form, and therefore not subjected to operation, the postmortem showed a large fibrous hypertrophy with obstruction, such as described by Greig Smith. The pylorus was stenosed, yet not to a marked degree, the stric- tured portion easily admitting a lead-pencil. There was no e\i- dence of ])rcvious ulceration. Here again this valve formation was well marked. Some observations of the 2 cases operated upon and cxj^ieri- ments made with the postmortem findings just referred to lead me to believe that a pylorus somewhat rigidly held upward by a short gastrohej)atic omentum, or one in which a certain amount of mechanical obstruction is present, will permit sagging of the stomach and valve formation of the pylorus, with consequent ob- struction. It is possible that this condition may exist in many cases of chronic gastric catarrh and dilatation without producing suflBcient trouble to cause death or require operation. The production of valve formation at the pylorus is similar to the valve formation of the ureter at the pelvis of the kidney as a cause of hydronephrosis and pyonephrosis, so graphically por- trayed by Christian Fenger. Enteroi)tosis favors such obstruction, as the greater the sagging of the intestines and the greater the downward displacement of the stomach, the more perfect are the mechanical conditions which tend to valve formation. One year ago I read a paper before this section on a similar 52 WILLIAM J. MAYO subject, and referred to valve formation as a possible cause of the difficulty at the outlet of the stomach, and I am now still more confirmed in the belief then expressed. It certainly explains many of the changes incident to the causation of obstructive symptoms without previous local disease. Kussmaul has been able to produce on the cadaver a rotation obstruction which is evidently similar in nature. A few cases of congenital stricture have been reported by Landerer and Maier, It is altogether questionable whether movable kidney can produce mechanical pyloric obstruction, although this conclusion has been drawn as to the relief of gastric symptoms supposed to be given by nephrorrhaphy in neurasthenic women. The diagnosis of marked pyloric obstruction and the consequent dilatation of the stomach is not difficult — in fact, the vomiting of large quantities of ingested material or its removal by the stomach- tube is of itself significant. After inflating the stomach with atmospheric air by means of an ordinary valve syringe and a stomach-tube, its outline can be rendered evident to the usual methods of physical diagnosis, inspection, palpation, percussion, and auscultation. The gastroscope and the gastrodiaphanoscope have little practical value to the surgeon. The differentiation between the malignant and the non-malignant forms of obstruc- tion is often difficult and may be impossible without exploratory incision. The examination of the stomach-contents has some value. When the test-meal of Ewald and Boas shows an absence of free hydrochloric acid with Giinzburg's test and at the same time shows lactic acid by Uffelmann's method of examination, the indications are for cancer; yet prolonged non-malignant obstruc- tion may cause such disease of the mucous membrane as to mislead the observer. In a considerable number of such examinations made by Graham the conclusions of Rosenheim have usually ob- tained, but these tests are far from reliable in character and at best are only confirmatory. In 2 of my cases of obstruction from former ulceration the cicatricial mass could be so plainly felt through the abdominal walls as to lead to a fear of malignant tumor. S(hMK CAUSES OF I'VLOKIC OHSTKL( TIO.N 53 Afirr all, llic cliiof faclors in the difrcrcntial diii^iiosis will he obtained i'roin a carclul examination into the personal history of the patient, esjjeeially as to previous uleeration, and from the lenfi;tii of time the obstruction has existed. In this respect the occasional cases wherein the scar tissue of previous ulceration has degenerated into inalif,niancy are of interest. I wish to call attention to the significance of enlarged glands in the greater and lesser omenta as evidence of malignancy. I have on three occasions observed enlarged glands having all the appearance of malignant disease in non-malignant pyloric ob- struction, evidently due to chronic sepsis from the absorption into the lymphatics of decomposing stomach-contents. In his article, "The Floating Stone in the Common Duct," Fenger speaks of this septic glandular enlargement about the head of the pancreas as nowise malignant. I have found it in one case in the axilla in a non-malignant tumor of the breast, and in the deep lymphatics in a case of sloughing uterine fibroids. Bull says that 50 per cent, of pyloric cancer patients die of starvation before glandular infection takes place. In his work on "The Surgery of the Alimentary Canal" May- lard speaks of the large fibrous stricture of the pylorus as having only fibrous structure, but says that if epithelial elements be found in the enlarged glands, malignancy is established. I think that the accuracy of statements can be questioned as to the absolute malignancy of any case in which the proof is enlarged glands which have not been microscopically examined. Operations for the Relief of Pyloric Obstruction The median abdominal incision between the ensiform cartilage and the umbilicus has proved to be the most satisfactory in giving easy access to the field of operation. This can be enlarged by a transverse cut across the rectus muscle, if necessary for free ex- posure. Pyloroplasty, devised by Heineke and Mikulicz, is the opera- tion of choice, and in narrow strictures is easy and certain in its results. As much of the incision as possible should be in the 54 WILLIAM J. MAYO healthy rather than in the scar tissue, as the latter is stiff and does not coapt nicely, and has the disadvantage of all scar tissue that necrosis is apt to be the result of the suture pressure. If the con- tracted portion is very long, the duodenum can be folded or knuckled upon the stomach after the adossement method of the French surgeons. This latter plan worked admirably in the second operation on my fifth case. As pyloroplasty cannot be readily performed in severe cases in which the scar tissue is very extensive Fig. 5. — Showing the maimer in which the first two stitches are applied. The forceps serve as retractors to hold the wound apart, as shown in Fig. 6. These stitches are inserted before the incision is made. or the pylorus bound down by adhesions to a degree which renders the operation hazardous, the mortality should not be great — about 10 per cent. Out of 5 cases I was able to do pyloroplasty in only 2. Gastrorrhaphy, introduced by Bircher, may be a good opera- tion in valve formation, but cases which are suitable to this pro- cedure could be relieved by pyloroplasty. SOME CArSE.s Ol I'VLOItK OUSTUl ( TION O.J r.orcla's ofXTalioii of diviilsion has a inorlality of 40 [xt cciil., and fails in a larj^'c proportion of tlic cases, making a primary recovery, to yield a p( rniaiiciil cnrt-. Greig Smith, however. })elieves that it has a fiekl in the hir^e fihrous form of obstruction. Pylorectomy will seldom he done for non-mali^'nant stricture. I can understand how a surj^eon could conunence a pyloroplasty and, (iridinu this incHiod iinsiiitai)l(', he forci'd to make a pylorec- Fig. 6. — Showing the wound drawn transversely, with a few interrupted sutures in place. tomy as the only way out of the difficulty, hut the mortality of 30 per cent, is too high when contrasted with other methods of relief. Weir, Meyer, and others have done the operation succe.ssfully in this class of ca.ses. Gastro-enterostomy for non-malignant disease is an operation of expediency and not t)f choice, ("ircinustances often compel its ii.se, however, as extensive contracture of the pylorus or the pres- ence of tissue unsuitahle for the securitv of suture union. 56 WILLIAM J. :mayo Inaccessibility of the field of operation, by reason of the pres- ence of dense adhesions to important structures, is also an im- portant indication. For one or more of these reasons 3 of my 5 patients were relieved by gastro-enterostomy. This operation by the suture methods, according to McGill, gives about 50 per cent, mortality, and, with mechanical aids, about 23 per cent. The best mechanical device is, undoubtedly, the Murphy button. The greatest drawback to its use is the liability of the button dropping backward into the stomach upon separation. This happened in 3 out of 8 gastro-enterostomies in which it was employed in this clinic. A year ago I suggested tying a string to the intestinal side of the button with the end in a double bow-knot eight inches away, thinking that intestinal action at that distance from the site of the adhesions would act as a kite-tail guide or tractor. In three cases it acted admirably. In a fourth case the button had not appeared when the patient left the hospital. Great care should be exercised to secure the jejunum at its origin and form a coil as close as possible to the proximal end. Rockwitz says that the union should be made so that the intestinal and gastric peristalsis will be in the same direction, to prevent impaction of the unused portion of the bowel. Braun advises entero-anastomosis to empty the duodenum more perfectly, and Kocher makes a right-angled attachment of the bowel to the stomach, in order that gravity may aid the downward passage of the gastric contents. In all operations in this region the omentum can be readily fastened with a few sutures in such a position as to add greatly to the security of union. The unequal thickness of the stomach and intestine is a source of weakness in this opera- tion. It should be borne in mind that a stomach with a normal capacity of three pints, dilated to two or three times this extent at the time of operation, will certainly contract after relief of the obstruction, that this contraction will effect the opening and correspondingly reduce its size to the same extent, and that such harmful contraction may not necessarily be due to the scar tissue SOME ( AFSES or PYLORIC OBSTRITTION .} i forming the margins of the fistula. In this Hcs tli«- iiiiccrfniiily of gastro-enterostf)my as a permanent opening. Taking all these facts into consideration, the suture method of Wolfler offers the fewest objections, as the opening can be of suf- ficient length to provide for future contraction and, with the im- proving techuic of modern methods of suturing, the luorfalily will steadily diminish. Fig. 7. — Showing the wound as closed by two rows of continuous Lembert sutures by means of the needles and threads used in the original sutures, the unthreaded end of the sutures having served as retractors as well as for the purpose of indicating the extent of the suturing necessary. In operations on the stomach the emptying of this viscus should be carefully attended to previous to operation, to prevent con- tamination of the wound by escaping contents. A preliminary hypodermic of morphin with chloroform anesthesia has, in our hands, best controlled the tendency to retching so annoying in 58 willia:m j. imayo operations of this character. The after-treatment has been simple; morphin, if needed, to prevent peristalsis, and rectal feeding for three or four days, with strychnin hypodermically and stimulants per rectum as needed. Case 1. — Stricture and Valve Formation; Pyloroplasty. — M. M., male, aged forty-six, American, admitted to St. Mary's Hospital January 10, 1895, \n\h. a history of gastric symptoms extending over a period of seven years. For the past two years he has vomited large quantities of partially digested or decompos- ing food once or twice in every twenty-four hours. He has grown weaker, and lost about 100 pounds in weight. Early in his ill- ness he vomited blood a few times and suffered severe pain after taking food. Physical Examination. — Patient very much emaciated, heart's action feeble, mth dilatation of the right ventricle. With the stomach-tube a large quantity of gastric contents was removed. Dilating the stomach \sith air, it was found to extend downward and to the left, to a point on a Hne with the anterior superior spine of the ilium. The pylorus was plainly felt on a level with, and to the right of, the umbilicus. Free hydrochloric acid was present in the test-meal. Operation January 13, 1895. A small cicatrix was found on the anterior wall of the pylorus, and from this a band incircled the pylorus, producing a stenosis of about the caliber of a goose- quill. The sagging of the stomach was very marked, produc- ing a kinking or valve formation just at the stenosed area. En- larged glands were found in the mesentery, having the usual appearance of mahgnant disease. Pyloroplasty was done, and the patient discharged February 15th, cured. When seen six months later, he had gained 70 pounds in weight and was doing his work on a farm. Case 2. — Cicatricial Stenosis; Gastro-enterostomy . — W. G., male, aged fifty-three, German, admitted to St. Mary's Hospital May 26, 1896, with a history of ulcer of the stomach which had caused seA'ere trouble for nine years. During this time he had been unable to work, and was more or less constantly under treatment. For three years he had been unable to take sohd food, and at intervals of from eighteen to thirty hours vomited up great quantities of material from his stomach. Physical E.ramination. — Patient extremely emaciated, and SOMK CAL'SKS OF I'VLOUIC OIJ.STIU ( TKjN 50 SO weak as to lie imaMe to walk willioul assistance. On dilating' the stomach with air, it filled ;ilniost, the uhoie ol" tiie left side ol' the ahdonion. Examinations of stomach-contents were con- tradictory and uncertain. To the right of the umhiHciis a thick- ened area, amounting ahiiost to a tumor, could be felt. Operation May !^Oth. A very extensive, hoard-like \'eiii.s is easily tliilfrciili;it»'<| I'mm tin- \\liili>li scars of foriiicr ulceration, while the exlcnl of the disease and adhesions to im- portant viscera render the radical treatment more or less feasible. We have made seven exploratory o[)erations for supf)osed can- cer of the stomach. In one case the tumor |)roved to he tuhercii- h)Ms omenlniii, aiire- vcnts egress of the food. The condition is similar to that of valve formation of the nn'ter at the pelvis of the kidneys, described by Christian Fenger as intermittent hydronephrosis and pyonephrosis. (Jastroptosis may act in a .similar manner, causing chronic indiges- tion and distress. In the relief of pyloric obstruction gastro-enterostomy plays the most prominent part, and but two methods have won recogni- tion — the simple suture plan and the Murphy button. The ad- vocates are rather equally divided; for example, Kocher, Doyen, and Senn use the suture, Czerny, Kiimmell, and a large share of the German surgeons hold to the button. We have used the button exclusively in the cases operated upon, and either the recovery of the patient or the postmortem has established, the competency of the device. Unfortunate button, experience drawn from cases of acute intestinal obstruc- tion will not hold good in these cases, as at the point of attachment both the stomach and the intestine are healthy, while in the former ease the proximal portion of the bowel is usually congested for a great distance. To set up a sloughing process in damaged tissues is not good surgery, as the chances are that it will not limit itself and perforation often occurs; the suture is safer in these cases. As to whether the anastomosis between the stomach and je- junum shall be made on the anterior or the posterior wall of the stomach, there is also a difference of opinion. All our cases have been anterior operations, and we see no reason to regret it. The posterior method requires a larger in- cision, more manipulation, and also creates an artificial defect in the transverse mesocolon to get into the lesser cavity of the peri- toneum. I do not believe that these disadvantages are counter- balanced by any practical results to the patient. In using the button, grasp the jejunum at its origin, form a coil of about 14. inches in length, and introduce the male half into the intestine first. In placing the female half in the stomach-wall, cut through the muscular and peritoneal coats before introducing the suture, and keep close to the margin of this incision with it; other- 92 WILLIAM J. MAYO wise the thick coats ruffle up and prevent easy approximation. Unite so that intestinal and gastric peristalsis will be in the same line. If the union is nicely done, the anterior wall of the stomach will tip downward and form a funnel at the point of union. Malthe says that if after the approximation the button lies to the right of the spinal column, it will go down — if to the left it will drop back into the stomach; it does not seem to make much differ- ence whether the button passes out or drops back; in the latter event, no harmful results have been reported. Some of our cases of button gastro-enterostomy have already gone from three to six years since operation and remain in perfect health, although in several of these cases the button has never been passed to our knowledge. The pyloroplasty of Heineke-Mikulicz, which consists in mak- ing a longitudinal incision two or three inches in length, with its center at the stricture, and suturing the wound transversely, is an operation which is not so popular as it was formerly, and many surgeons have substituted gastro-enterostomy for it in every case. We had relapse follow the operation in two cases ; in each instance at the secondary operation the pylorus lay very high and was ad- herent, with evident valve formation. Reasoning that gastro-enterostomy succeeded because gravity emptied the stomach through the dependent opening, we have, in three cases, pushed the pylorus downward and fastened it in this pos'tion by a suture to the abdominal wall. The results thus far have been excellent. The gastroplication of Bircher and the gastropexy of Duret have little to recommend them, while the dilatation of Loreta is an obsolete operation. Traumatisms of the stomach are so purely surgical that I will not take up time further than to say that the results of the applica- tion of modern surgical principles to these injuries have been most satisfactory. MALIGNANT DISEASES OF THE STOMACH AND PYLORUS* WILLIAM J. MAYO Carcinoma of the stomach causes about 1 per cent, of the total death-rate (Van Valzah), and is the most common form of mahg- nant disease. In an analysis of 30,000 cases of malignant neo- plasms Welch gives the percentage as 21.4; Virchow, 3o, and Haberlin, as 41.5 per cent. Age has a more important bearing on carcinoma of the stom- ach than on carcinoma of the lower portion of the gastro-intestinal tract. In the former locality it is peculiarly a disease of middle and later life; it is most rare under thirty years of age, although Moore reports a case at the age of thirteen. It is slightly more frequent in males than in females — 5 to 4 (Welch); the dispro- portion, however, is not so great as the older authorities were in- clined to believe. The progress of the disease varies, and is materially aflFected by the age of the patient, the situation of the growth, and its histologic structure. Death may follow in three months from the first symptoms, or it may be delayed to two and one-half years; the majority of patients die within the year. Park is of the opin- ion that all forms of malignant tumors are on the increase; how nuich of this increase is apparent rather than real and due to better methods of diagnosis is open to discussion. Surgery oflfers the only hope of cure: it is important that car- cinoma of the stomach be considered a surgical disease, and a sus- picion of gastric cancer should cause the physician to send the patient to the surgeon for exploratory incision to complete the *Rcprintoil from '"Trans. Amor. Surg. .Vssor.," IHOO. 93 94 WILLIAM J. MAYO examination. The same principles should govern here as in prob- able malignant disease of the breast or uterus. The great difficulty which arises in carrying this proposition into eflFect is the question of early diagnosis. In a general way it can be said that the early medical diagnosis of cancer of the stomach does not depend upon any one specific sign or symptom, but rather on a collection of facts, each one of which, if taken alone, would have but little value; and, again, most of these symptoms are not developed to a char- acteristic extent until late. Osier reports a number of cases of latent cancer of the stomach in which there were no symptoms during life to lead to a suspicion of malignant disease, the true condition being found at postmortem. A person of middle or advanced life, of previous good digestive power, begins to complain of pain in the region of the stomach, difficult digestion, and loss of appetite. There is a continuous loss of flesh; a progressive waste of the albuminous tissues; later there is more or less vomiting, depending on the situation of the growth, perhaps of blood partially digested and likened to coffee- grounds. A tumor may gradually become discernible (115 out of 150 causes. Osier). A dry cachexia develops, and transitory edema, usually affecting the lower limbs and due to a local phlebitis, may occur. This seems to be a clear clinical picture of malignant disease, but the fault is that, sui-gically speaking, the case has been hopeless since the first few weeks. Czerny, Kraske, and many other eminent authorities believe that radical operation is out of the question by the time a tumor can be felt. Examination by means of stomach-tests adds to our diagnostic resources. The absence of free hydrochloric acid is an indication of value, and it occurs in 60 per cent, of all cases, but is not always well marked early. In the forms of malignant disease developing upon ulcer, of which a number had been collected and described by Hemmeter, the free hydrochloric acid may even be increased. Another valuable indication is the presence of lactic acid accom- panied by the Oppler-Boas bacillus. Reduced motor power of the stomach independent of obstruc- tion is a constant symptom, but is also common to non-malig- MALKiNANT DISKASKS OK ST()MA( M AVI) I'VLORUS *.l) iiiint disease, such as gastric myasthenia. Terrier, in his recciif work on surgery of the stomach, says tliat "the examination of a fasting stomach gives the largest amount of information. " The only positive diagnostic indication is the finding of cancer elements in the fragments wa.shed out of the stomach; here again, unfortunately, the spontaneous detachment of such speci- mens in an early stage of the disease must necessarily be a very rare occurrence. The value, however, is absolute, and efforts in this direction are increasing. In the laboratory of Boas the first systematic efforts to detach pieces of the growth for this purpose were made. To locate the point from which the pieces are detached Kuhn invented a spiral sound which could be watched with the j-ray. Hemmeter has systematically developed this method of examination, both in the means of detaching the frag- ments and in the certainty as to the location of their origin, the examination being conducted after the manner of uterine scrapings. He calls especial attention to the diagnostic importance of atypical as well as pathologic mitoses of the cells. One important point is to be determined: Is there danger of increasing the spread of the disease by diagnostic cureting? The removal of pieces of malignant growth for microscopic examina- tion is a doubtful practice. Senn warns against the rough hand- ling of a malignant tumor for the purpose of making a diagnosis, and says that such "diagnostic massage" may result in increased activity. Halsted opposes the removal of portions of a malignant growth for diagnostic purposes on account of the danger of inocu- lation. Whether this objection would hold good in the necessary manipulations for the removal of fragments from a gastric cancer is worthy of consideration. Examination of the blood shows reduced specific gravity, di- minished hemoglobin and number of the red corpuscles, and re- duction in their size; these changes are due to the anemia and have no special characteristic in malignant disease of the stomach. The white corpuscles are often increased by the accompanying inflam- mation and ulceration. The urinary changes are in no way pecu- liar to the cancerous process. The indican, so frequent an accom- 96 WILLIAM J. MAYO paniment, is the result of the retention, and the other changes to the albuminous waste and low state of the nutrition. Van Valzah says that careful and repeated examination of a suspicious case of carcinoma of the stomach should develop a probable diagnosis in two or three weeks. The scientific zeal with which the diagnosis is sought has in itself a distinct danger. The suspicious case is often kept under observation too long in the hope of making a positive diagnosis. The surgeon should not ask the physician for a diagnosis of gastric cancer; if we wait for that we are pretty sure of being too late. It is the suspicious cases which should be explored, and it is the duty of the physician to urge this when in doubt. A few indica- tions, with the history, should suffice, and the matter laid before the patient to decide as to an exploration. There is no reason why the radical cure of cancer of the stomach should not approach that of the uterus or breast. It is true, the difficulties of diagnosis are increased, yet the early symptoms are quite as positive. At an early stage the tumor in the breast is unnoticed or no attention is paid to it, and the early history of cancer of the uterus is notoriously defective, with the result that the number of cures in either breast or uterine cancers is not large — the majority have passed the curable stage before the diagnosis is made. This has been and will be true of the stomach, but the results should be much better than they are. The curability of cancer of the stomach depends upon — (1) The histologic structure of the neoplasm; (2) its location; (3) extension to neighboring structures; (4) lymphatic infection; (5) the general condition of the patient. It is essential that the rela- tive value of the manifestations of the malignant process be studied, for upon a correct appreciation of the conditions present depends the whole question of treatment. 1. Every cancer is at one time a local process, and in that stage curable, but the duration of this stage is usually short. The rapid- ity of progress of carcinoma of the stomach is largely influenced by the relation of the cells to the stroma. If the cellular elements predominate, the growth is soft and its extension rapid. If the MALIGNANT DISEASFIS ()V STO.M VrH AND PYLORUS !) / stroma is in excess of the i)urencliyiua, the tumor mass is harder and slower of growth. Ilemmeter chissifies the mahgnant ejji- theHal tumors of the stomach into — (a) Adenocarcinoma; (h) cyhndric-cell carcinoma; (c) mcdulUiry carcinoma, the common degenerations being coUoid and scirrhus. Tiie colloid type is an unfavorable indication, and if it exists to any extent, indicates a late stage of the disease beyond radical intervention. Secondary nodules, the result of a primary gastric cancer, are especially prone to this form of softening, and may obscure the original focus. In the metastatic deposits in the omentum colloid degeneration may develop a tumor of very large size. Bland-Sutton describes one weighing 10 pounds^ Scirrhus refers to the relatively large pro- portion of connective-tissue stroma which is undergoing more or less contraction, and indicates a slower process. This variety is the most common form of gastric carcinoma (7'-2 per cent., Brinton). Pyloric growths are usually scirrhus, and often follow the lines of the blood-vessels incircling the pyloric end of the stomach, producing an early obstruction, and they may cause death while still an operable growth. In some cases the contraction may not take place, leaving the pylorus open and gaping. More rarely diffuse infiltrating carcinoma of the whole stomach-wall may exist, either as a primary or as a secondary process. This condition is more common than cirrhosis of the stomach (Bristow's water-bottle stomach), and often mistaken for it. Hektoen, in describing two cases of the primary form, calls attention to the necessity of careful search for the carcinoma cells to prevent mis- conception as to the nature of the disease. The cylindric-cell carcinoma is the most favorable form of disease for extirpation. It is slower of growth, and does not develop adhesions to neighbor- ing organs in the early stages. 2, The location of the growth is important, an early diag- nosis depending to some extent on the mechanical features present, and the accessibility to operative procedures is very largely deter- mined by the situation. Approximately, 10 per cent, involve the cardiac area, 30 per cent, the body, and 60 per cent, the pyloric «nd of the stomach (Gussenbauer) . The cardiac orifice is seldom VOL. I — 7 98 WILLIAM J. MAYO the primary seat of disease, but rather an extension upward from the fundus or downward from the esophagus. The symptoms are painful deglutition and, later, obstructive regurgitation. The clinical diagnosis is comparatively easy, but the situation renders radical operation exceedingly difficult and usually impossible. The progress of mahgnant disease of the cardia is rapid. It involves neighboring structures early, and death follows active symptoms within a few months. Carcinoma of the body of the stomach is most difficult of early diagnosis, as a rule presenting few symptoms until it is well advanced, unless by extension it involves one of the orifices and thereby introduces mechanical phenomena. Einhorn thinks the gastrodiaphanoscope is useful in outlining a tumor of the body. The ar-ray has been used for the same purpose; the utility of either is doubtful. Kocher says that under deep anes- thesia a tumor can sometimes be mapped out early. It is fortunate that the pylorus is the part so often affected. The mechanical obstruction favors an early diagnosis, and, as shown by Winiwarter more than twenty years ago, many cases die from starvation pro- duced by a surgically curable growth. Israel states that the pylorus itself is not so often the primarj^ seat of disease as has been believed. The infiltration from any place in the pyloric end of the stomach naturally extends in this direction, and is arrested at that point. In the earlier stages the process cuts off rather sharply at the duodenum. The situation is most favorable for extirpation. Obstruction with resulting dilatation of the stomach has usually been the most im- portant factor in determining operation in the cases submitted to extirpation. Kocher states that if food is regularly found in the stomach fourteen hours after taking, mechanical obstruction ex- ists. Myasthenia with retention and hyperesthetic gastritis with retention produce this symptom, and are not necessarily due to a mechanical obstruction, although valve formation often coexists. In these cases chronic dilatation is found, and often hyperchlor- hydria or supersecretion. While the differential diagnosis is not difficult, surgery offers the best means of cure in all forms of chronic dilatation of the stomach. The finding of a dilated stomach brings the case within surgical limits, and if chronic MALK.'NANT DISEASES OF STOMAfU AM) I'^I.OUI S f)9 rclciilioti cxisls, oilier lliiiij^rs Ix'iii;^ ('(luul, ()pcriili\(; rclit'f is in- dicated. Tlie amoiiiil of dilatation due to i)yIoric cancor depends to some extent upon tlic situation of I lie pylorus. If held high by a short gastroliei)atic oiucntuin, the stomach is emptied by musen- lar eontraetion. In malignant disease compensutory liypertrophy is deficient, and dilatation eomes on rapidly. If the pylorus lies low down, gravity aids the onward course of the ingesta, and a larger amount of obstruction is well borne. 3. Direct extension to surrounding structures is through ad- hesions; along these bands the malignant infiltration takes place. In most instances this is a late phenomenon, and acts as a contra- indication to ojjcration, but not necessarily so if the ca.se is other- wise favorable. The adhesions may be recent and due to a septic process rather than the malignant disease. Advance adhesive inflammation takes place in the vicinity of ulceration, and later fistulas may form between the stomach and duodenum or colon, or, rarely, with the abdominal coverings. Perforation occurs in about 6 per cent, of the cases, causing death at once if in the free peritoneal cavity or subphrenic abscess if protective adhesions exist. Indirect extension through the blood-vessels is rarely the cause of metastatic deposits. The arteries are very resistant, Init the veins are more easily invaded, and from the infected thrombus emboli are carried to the liver, lungs, etc. Peritoneal inoculation is a rare sequel of cancer of the stomach. The most common method of advance is by extension to adjacent structures, and thi.s. is true of recurrences after extirpation, which are usually local. Prom an operative standpoint adhesions are a serious complica- tion, and, as they are usually infiltrated with carcinomatous cells, the extirpation under such circumstances will be unpromising. Haberkant's statistics of extirpation show that where adhesions were extensive 72 per cent, of the patients died. If no adhesions existed, only 27 per cent. died. Gussenbauer and Winiwarter, in 542 cases of cancer of the pylorus, found that in 37 per cent, there were no adhesions; in -41 per cent, there was no metastasis to other organs. 4. Lymphatic infection is a still more serious complication. 100 WILLL\M J. MAYO because, if palpable, glandular malignancy can be established and it is altogether probable that the involvement has passed beyond reach. Enlargement of glands without microscopic proof cannot be considered necessarily malignant. In more than half of the 29 cases of non-malignant disease of the stomach which I have operated upon for the relief of dilatation from benign causes en- larged glands could be palpated. Fenger, in his study of the effects of stone in the common duct of the liver, noted the fact that en- larged glands could sometimes be felt about the head of the pan- creas, leading to the belief that malignant disease existed. Halsted found such non-malignant glandular enlargement occasionally in malignant disease of the breast. It has been frequently noted in the submaxillary glands during the course of lip carcinoma, and in malignant disease accompanied by inflammatory lesion. In due time these glands undoubtedly become malignant, just as inflam- matory adhesions are later infiltrated. In my experience the septic glands are smaller and softer than the carcinomatous. jNIikulicz says the lymphatic glands tributary to the stomach lie in four groups: (a) Along the lesser curvature and cardia; (b) along the greater curvature; (c) in the gastrocolic omentum; {d) about the head of the pancreas. The modern operation for the removal of the glands -^ath carcinoma of the breast suggests a similar gland- ular extirpation in radical operation on the stomach. Lindner does not believe this feasible. In 28 recurrences after extirpation 15 were local, 12 distant, and only 1 involved the glands. Bland-Sutton says that glandular involvements are only found in about one- half of the patients dying with gastric cancer. It is probable that this is too low an estimate, and that at least 65 per cent, dying of cancer of the stomach have well-marked glandular in- volvement. In the earlier stages, however, the percentage is smaller. The usual route of travel is to the deep glands, liver, etc. Binnie, in discussing carcinoma of the umbilicus secondary to cancer of the stomach, says that such unusual routes of travel are indications of inflammatory or mechanical impediments in the normal channels. He believes the infrequent recurrence in the glands after extirpation of gastric cancer is due to the destruc- MALKJXAXT DISEASES OF STOMACH AND PYLORUS 10] tion of the normal lymphatics at the time of operation. Russell has recently pointed out that the cure of cancer of the uterus de- pends rather upon wider local extirpation than upon removal of the glands, and this statement is equally true of the stomach. An investigation leads to the belief that glandular infection in cancer of the stomach is not uniform; many cases die without such in- fection, and a moderate enlargement may be septic. Should they become infected, their situation renders radical removal practically impossible. 5. The condition of the patient is a factor of importance. Ma- lignant disease of the stomach has a peculiar depressing effect. The nutrition is bad, and the healing power is greatly impaired. Under equal conditions a most formidable operation for malig- nant disease about the head or neck would be well borne. A degree of cachexia out of proportion to the demonstrable disease is of bad omen, unless the debility can be shown to be due to a mechanical condition, such as obstruction. Ascites, even in a limited degree, contraindicates a plastic operation, such as gastro- enterostomy, and renders any kind of an operative procedure hazardous, as firm adhesion does not take place. Sarcoma of the stomach is a rare disease. In 1899 Van Valzah and Nisbet found 43 cases recorded. It is most common in males under thirty-five years of age. Finlayson recently reported a case in a child three years of age. The disease may present itself as a smooth, rounded tumor, and by its weight may cause considerable prolapse of the stomach. In some of the cases reported the tumor had a limited origin in the stomach-wall and was easily extirpated. It rarely occurs as a diffuse sarcomatous infiltration. Sarcoma seldom con- tracts, and involvement of the orifices does not produce the me- chanical interference with the progress of the food which is so marked a symptom of scirrhus. Incontinence is the rule. Sarcoma is more often a secondary involvement of the stomach than is carcinoma. The diagnosis was seldom made in the cases operated upon, the operation being commenced usually under a misapprehension as to the nature of the growth. In the majority of instances there were no marked gastric symptoms to suggest 102 WILLIAM J. MAYO that the abdominal tumor had its origin in the stomach. In a few cases ulceration and hemorrhage have been late symptoms. In 1893 Baldy removed almost the whole of the stomach for sar- coma without success. A number of partial gastrectomies have been recorded, with successful outcome. In several of the cases the tumors removed were of considerable size. The exploratory incision for suspected cancer of the stomach, instead of being a last resort, should be one of the first, and promptly undertaken if the condition cannot be proved to be non- malignant. A median incision between the ensiform cartilage and the umbili- cus is most useful for exploration upon the stomach. A small open- ing readily exposes the pyloric portion, which is the most common seat of disease. The writer at times has experienced considerable difficulty after opening the abdomen in ascertaining the real con- dition. What are the local appearances which enable us, by palpation and inspection, to say that a certain thickening of the stomach- wall is malignant? There are no definite signs which occur in all cases, and much depends upon the experience of the surgeon. It is hardly practicable to remove a portion of the growth and have a microscopic examination made before finishing the opera- tion, nor is such practice devoid of danger of inoculation. In all the cases I have explored the growth was hard and had an appear- ance of increased vascularity, and the general "feel" of the dis- eased part was distinctly malignant. At times, with the involved area thoroughly exposed, and in some cases laid open by incision, it is impossible to state from macroscopic examination whether it is benign or malignant. Czerny's experience is very interesting. Four patients in which gastro-enterostomy was done for supposed malignancy lived so long as to preclude the possibility of cancerous disease. Exploration of the neighborhood of the cardiac orifice is dif- ficult and often unsatisfactor3\ A small opening may be made into the stomach cavity, and with a finger introduced through the incision the anterior wall can be invaginated to enable the MALI(;.\A.NT DISKASKS OI' STOMATH AM) l'Vr,OIU S 1 ().'{ exploring digit lo sciirch the opening int(j tlie esophagus. Shouhl the orifice he tightly closed, pressure for a little time may he neces- sary before a dimi)lc can he fell at I he normal opening (Abbe). For inslrumcnlal examination of the <-anliac orifice a small in- cision should be made in the anterior wall and the lesser cur\ature held taut; this forms a sulcus leading directly to the opening (Richardson). The body of the stomach anteriorly can be readily examined; by traction a considerable portion can be drawn forth for in- spection. Hy an incision through the gastrocolic omentum the posterior wall can also be inspected. The pylorus is, fortunately, easily explored, and unless adhesions have formed, it can l)e brought fully out of the abdominal cavity. If it i.s held within the cavity of the abdomen and under the li\ er by a short gastro- hepalic omentum, this latter structure can be i)artly divided v.ith- out injury to the nerve- or blood-supply, readily mobilizing the pylorus and upper portion of the duodenum. The writer has in five instances been unable to make an exact diagnosis, even after exploratory incision. In two cases the pylorus was held high up and under the liver by adhesions, the symptoms being obstructive. If the disease is cancerous, the position and the extension to surrounding structures by adhesions would prevent radical opera- tion. In each a gastro-enterostomy was performed for relief of the mechanical condition present. In one a hard mass on the posterior wall could be felt, with signs of extensive inflammation in the lesser cavity of the peritoneum. There had evidently been a perforation at one time, but even with an incision in the anterior wall of the stomach I was unable to tell whether or not it was malignant. Here, again, gastro-enterostomy was performed on the supposition that, if the perforation was due to simple ulcer, recovery would be hastened, and if carcinomatous, some relief would be experienced. In the fourtli case a small stomach was found deep under the liver; with a finger inside the stomach cavity I could not reach the cardiac orifice. The clinical evidence led to the belief that the obstruction was malignant, and gastrostomy was decided upon without further attempt at diagnosis. In the 104 WILLIAM J. MAYO fifth case the evidences of malignant obstruction at the cardiac opening were marked; the condition of the patient was such that gastrostomy was done as quickly as possible without any attempt to explore freely as to the nature of the obstructive lesion. In 12 cases malignant disease was quickly shown to be too far advanced to extirpate, and not presenting symptoms such as obstruction, requiring palliation, nothing was left to do further than to close the abdominal incision rapidly. In one of my earlier explorations, in which extensive malignant disease was found without symptoms requiring palliation, the incision was closed and the patient put to bed to wait the usual length of time for healing to take place, the debility increased, and, while he lived several weeks, he was unable to return to his home before death. Since that time I have closed the incision under such circumstances with permanent silver-wire sutures after the Halsted plan. The patients get up on the third day and leave the hospital in a week. Silver wire buried in fixed aponeurotic structures, in my ex- perience, does not give rise to the atrophy necrosis sometimes seen when placed in the muscles. The hernia liability in any event is of no great consequence to the victim of an incurable malady. It may be justly said that exploration in cases of ad- vanced malignant disease of this description is not good practice, yet there will always be a certain number of cases in which the diagnosis will be uncertain, although advanced disease exists. The extent to which the exploration will be carried will vary with different operators. Personally, I have not made undue effort to perfect the diagnosis if extirpation did not seem proper, and have been content to give such relief as the circumstances would permit. Complete removal of the stomach has won a foothold, but to what extent only the future can determine. The more radical observers believe that even if the disease has affected only a limited and apparently excisable portion of the stomach, the whole organ should be removed, that pylorectomy and partial gastrectomy are not based on correct principles, and the large percentage of local recurrences after partial operations certainly MALKJNANT DISIOASKS OF STOMACH AM) I'VI.OKI S ]()."> give color lo this view. Bernays calls at lent ion lo llic iinjirove- ment in llic resiills of cancer of llic ulcrns since complete hys- tereeloniy has replaced anipulalion of the cervix, and helieves tliat this will be true of the stomach. On the other hand, the number of successful complete gastrectomies are so few and of such short duration that the problem cannot be solved at this time. It is, however, altogether {)robable that complete gastrectomy is des- tined to become a most valuable operation. In cancer of the body of the stomach nothing else in the way of radical operation offers a reasonable prospect of cure, and all the forms of pyloric cancer with a tendency to infiltrate the body of the stomach can be placed in the same category, reserving pylorectomy and partial gastrec- tomy for the not uncommon cases of ring-like infiltration. These conditions kill by obstruction, and there can be no more reason for a complete gastrectomy in such cases than in removal of the whole of the large bowel for cancerous stricture of the colon. It is in these cases that the partial operation has achieved its triumph. The first complete removal of the stomach was performed by Conner, an operation condemned at the time by reason of its unfortunate termination. It was Schlatter's success — not his originality — that called attention to the possibilities of gastrectomy, and the cases of Richardson, Brigham, Macdonald, and others quickly gave the operation a standing. The ease with which the duodenum has been approximated to the esophagus in many of the reported cases is surprising; it certainly is an important factor in the result. The duodenum, on account of its diameter and func- tion, offers the best prospect of replacing the stomach, and it is yet to be determined whether an opening between the esophagus and jejunum would serve equally as well. Schlatter's case was of this description, and the result was, functionally, as satisfactory as esophagoduodenostomy. In the determination of the advisability of gastrectomy the ability to approximate the duodenum to the esophagus is of im[)ortance. Richardson calls attention to the fact that a permanent duodenostomy might be made should it be impos- sible to make a satisfactory anastomosis. A number of the deaths after partial as well as complete gastrectomy has been caused by 106 WILLIAM J. MAYO including a portion of the mesocolon in the ligatures used to free the greater curvature of the stomach, with resulting gangrene of the transverse colon. If the gastrohepatic omentum is tied off first, the fingers can be slipped underneath the pylorus and act as a safe guide to the ligation and division of the gastrocolic omentum. As to the form of union, the Murphy button and the simple suture plan have both been used. Opinion is divided as to which is of the greater merit. As elsewhere in the gastro-intestinal tract, the results seem to be about the same, the one used depend- ing more on the individual preference of the operator than on any specific indication. Mayo Robson, in the Hunterian Lectures for 1900, gives statistics showing a death-rate of 50 per cent.; the cases from which to draw conclusions are as yet too few. Pylorectomy and Partial Gastrectomy Operations limited to the pyloric end of the stomach and its immediate \dcinity have been largely practised for malignant disease, and the result, taken as a whole, has not been satisfactory A few cures have taken place in exceptionally favorable cases, mainly the malignant strictures where the mechanical effects led to an early diagnosis and operation. Limited extirpations, to be successful, must go wide of the disease on the stomach side. This is particularly true if the infiltration extends laterally along the stomach-wall, and less so if the line of invasion follows the blood- vessels in a circular manner. The difficulties of diagnosis have rendered late operations the rule, and the absolutely unfavorable prognosis has encouraged operations in many cases unfit for so serious a procedure. The results, both immediate and remote, have been correspond- ingly bad. An occasional patient, apparently with a hopeless extent of disease, gets well after extirpation and stays well; this leads the surgeon to an effort in similar cases, with very disap- pointing results. An analysis of the more recent cases of Kocher and Maydl shows a decreasing immediate mortality and a more encouraging percentage of permanent cures, depending on an earlier diagnosis .maij(;n"a.\t diseases of stomach and pyi.oki s 107 and a l)cttcr selection of cases. Maydl lost only IG per cent, of patients, and in i8 per cent, there was early recurrence, the re- maining 50 per cent, being still alive at the time of his report. Mayo Rohson says the evidence shows that operation can !)e rea- sonably expected to cure one patient in three or four. The Kocher operation for pylorectomy and partial gastrectomy has been adopted by surgeons generally, and, while modified in some particulars by various surgeons, can still be considered the best plan of procedure. It avoids the fatal suture angle of the Billroth method by completely closing the stomach end and form- ing an independent anastomosis between the duodenum and the remaining portion of the stomach. In a comparison of these two methods Guinard shows a mortality in 148 cases by the Billroth method amounting to 38 per cent., and in 64 cases by the Kocher method only 16 per cent. In our experience the following plan was found satisfactory: First, tie and divide the necessary amount of the gastrohepatic omentum; this mobilizes the pylorus and lesser curvature and permits easy delivery. The fingers are now passed behind the pylorus from above into the lesser cavity of the peritoneum; this renders ligation of the gastrocolic omentum free from the danger of injury to the vessels in the transverse mesocolon. The stomach should then be clamped above the disease, and a cir- cular cut made with a knife completely around the healthy stomach to the mucous coat; the peritoneal and muscular coats are stripped back a half inch and the few bleeding vessels are caught with forceps and tied; the mucous coat is cut inch by inch from above downward and sutured at once with a continuous cat- gut suture, preventing leakage. After complete separation the tumor is covered and turned to the right out of the way. The muscular coat is then sutured with a continuous catgut suture; finally, the peritoneum is turned in by a good silk Gushing stitch. The stomach is now sutured to the pedicles of the tied omenta, anchoring it to the right, furnishing further protection against leakage, and also preventing undue traction on the duodenum, which, after safe amputation lieyond the disease, is fastened to 108 WILLIAM J. MAYO the anterior wall of the stomach by a Murphy button. The opera- tion is, with some slight modification, the Kocher method. Kocher uses the posterior anastomosis with suture. Czerny has used both the anterior and the posterior method with suture and the button. He thinks it makes little difference in the result. Greig Smith says that any extirpation requiring over one hour for its performance is open to serious objection. Czerny speaks favorably of Tuholske's method of doing the operation in two stages where the patient is much debilitated — first the gastro-enterostomy, to be followed in three weeks by pylorectomy. He has practised this in one case; the interval enables the patient to recuperate. Kiimmell has also advocated this in starvation cases. Czerny says that one objection can be made to this method — the adhesions following the gastro-enteros- tomy interfere with the extirpation. It is a question whether the average patient, after experiencing the relief afforded by successful gastro-enterostomy, would submit to a second operation, and especially to one which promises so little as to permanent cure. Kocher has made the extirpation by his present method 30 times, with 5 deaths. Malthe, in the Christiana clinic, had only 1 death in 11 cases. Kocher, out of a total of 57 extirpations, had 11 pa- tients alive, and only 5 of these alive long enough to be called cured. The immediate mortality of pylorectomy varies from 25 to 55 per cent. In properly selected cases it should have as low or a lower death-rate than gastro-enterostomy. Pylorectomy requires that the patient be in good condition, and the growth in the early stage — gastro-enterostomy for malignant disease has no such limitations. Levy has described an operation for the resection of the cardiac end of the stomach, but I am unable to find that it has been per- formed upon the living subject. Chlumsky says that any death within thirty days should be classed as operative mortality, and I believe this conclusion a just one. The palliative operations depend on the situation of the growth, and are based on mechanical conditions. Malignant obstruction of the cardiac orifice demands gastrostomy, and best at an early date, without waiting for marked starvation symptoms. MMAi.SASr IHSEASES OF STOM-VrH ANIJ I'VLORUS 10!) Relief from the irritation of the i)a.s.sing food markedly delays the progress of the disease, comparing in this respect with colos- tomy for the relief of cancer of the rectum. The original ojx'ra- tions were undertaken with the one view of feeding the patient, the Fenger operation being the type. The writer did his first gastrostomy after this plan, and the distress from the leakage al- most equaled the benefit. In many cases the irritation from the leakage and enlargement of the fistula from ulceration constituted so distressing an after-result that efforts were made to obviate this, and attempts to form a muscular sphincter about the opening were more or less successfully adopted. The von Hacker operation, using the left rectus muscle-fibers for this purpose, is the best of the kind. But not until Witzel published his method of lateral folding of the stomach-wall about a rubber tube, forming an oblique channel and valvular opening, was the problem solved. We have made four gastrostomies by this method, and the result in each instance was most gratifying. The Kader operation, which Curtis says was first described by Stamm, is an adaptation of the \Yitzel method, the only difference being that the tube is introduced directly into the cavity of the stomach and the walls brought up about it by a circular purse-string suture (Stamm) or by lateral interrupted sutures (Kader). The object is to cause a cone or nipple-like projection to present into the gastric cavity. In the one case in which the writer used this method the result was equally as good as in the Witzel plan. It is of special advantage in some cases on account of the ease with which a gas- trostomy can be quickly done on a very small stomach. The Marwedel operation is also a modification of the Witzel, the rubber tube being obliquely buried in the wall of the stomach itself. It is highly commended by Dennis, and undoubtedly is as perfect as the method of either Witzel or Kader. The Ssabanajew- Frank method of gastrostomy is on an entirely different principle, and has the great advantage in not requiring a tube. The fistula formed is a mucocutaneous one, and, therefore, permanent, while by the peculiar displacing upward of the cone of the stomach brought out of the deeper layers of the abdominal wall, a spout- 110 WILLIAM J. MAYO like opening is maintained which is self-closing and does not leak. The one disadvantage is the difficulty of making the operation on a contracted stomach. McCosh gives the mortality of gastrostomy for malignant disease at 30 per cent. The palliative treatment of advanced cancer of the body of the stomach is unsatisfactory; for- tunately, mechanical conditions arising from this form of disease are infrequent. Occasionally, on exploration an inoperable growth of this kind is found, and the question is: Can we prolong life or induce future comfort? If sufficient healthy stomach remains on the cardiac side to permit gastro-enterostomy, this should be done. The cureting of such growths through a gastrotomy wound, advo- cated by Bernays, has little to commend it, nor can the actual cau- tery, so useful in advanced cancer of the uterus, be used to any great advantage. Should the growth involve the orifices of the stomach, starvation may necessitate duodenostomy or jejunostomy for feeding purposes. Edward Martin reports a case relieved by duodenostomy, made in a manner similar to the gastrostomy of Witzel. Jejunostomy has had some degree of prominence, and at one time was a rival of gastro-enterostomy. Maydl performed 2.5 jejunostomies with 4 deaths, and strongly urges the operation in selected cases. Heidenhain reports several very satisfactory re- sults from this operation, and believes that in cases in which the gastro-intestinal fistula must be made at a point within the possible future progress of the disease, jejunostomy is indicated rather than gastro-enterostomy . Obstructions at the pyloric opening are most common, and the gastric retention Avhich results demands relief. Gastro-enteros- tomy is the most generally useful operation performed on the stomach, in suitable cases prolonging life, relieving pain, and pro- moting comfort. It would be an unprofitable undertaking to go into the various methods of making the anastomosis, as the literature of the sub- ject is already enormous. Two methods of making gastro-enter- ostomy have stood the test of time — the simple suture operation and the Murphy button. At the present stage of development MALKINANT DISKASKS OK STOMACH AM) I'VLOItIS 111 the results are about llie sarue, dcpcnditi^ more on the exjXTiciK-e of the operator than on the method employed. In performing the <)[)eration hy either method it is important that the jejunum should l)e grasped at its origin and a eoil from 14 to Hi inehes in length formed. At this point the mesentery is of sufficient length to prevent traction. Care should be taken to have the direction of the peristalsis the same in the stomach and intestine when the anastomosis is effected. There has been considerable discussion as to whether the fistula should be established on the anterior (Woifler) or posterior wall (von Hacker) of the stomach. The latter necessitates an artificial opening into the lesser cavity of the peritoneum, to reach the desired part of the posterior wall, and requires a larger incision with greater exposure. Its supposed ad- vantages are: that gravity will aid in passing the food downward, and if the button has been used, that it is less liable to be retained in the stomach; and, lastly, that regurgitant vomiting of bile and pancreatic juices is less frequent — all of which are important points in favor of this locality' for anastomosis. Carle and Fantino dem- onstrated the superior advantages of the posterior operation. These investigators also showed, by experimental work, that bile was frequently, if not usually, present in the stomach for three months after gastro-enterostomy, and in moderate quantities did not interfere with digestion. On the contrary, at the Breslau clinic the anterior method gave the best results. If the posterior operation is chosen, the suggestion of Meyer, that the edges of the divided mesocolon be sutured to the posterior wall of the stomach, should be carried out to prevent contraction of the mesenteric opening. For the anterior operation, a point on the healthy por- tion of the stomach should be chosen as near the pylorus as will probably remain free from encroachment of the disease. It should be placed near the greater curvature. We endeavor to get the lower border of the anastomotic opening about one inch above the inferior border of the stomach. When completed, the traction weight of the attached bowel draws the stomach over until, at the point of attachment, the anterior wall becomes the inferior, and a funnel-shaped entrance into the bowel is secured. This can b;» 112 WILLIAM J. MAYO readily observed before closing by lifting up the abdominal wall with a retractor, exposing the field of operation. I believe that this mechanical condition in the completed gastro-enterostomy prevents, to a large extent, the vicious circle of the biliary and pancreatic juices, which has proved so prolific a source of danger to the patient. The anterior operation has usually been made too high up on the stomach-wall. The anastomosis should be at a low point, so that gravity will empty the stomach and prevent the entrance of bile. Of 11 anterior gastro-enterostomies which I have made for malignant disease, in only 2 cases was regurgitant vomiting marked. One subsided after a lavage, and the second required lavage once or twice a day for five days; both cases recovered. Of 20 gastro-enterostomies made for non-malignant disease, in one was regurgitant vomiting present. It is a more common com- plication of malignant disease, and perhaps in part due to the changed nutrition affecting the glandular secretions as well as the reduction in the motor power so characteristic of cancer of the stomach. Entero-anastomosis between the proximal and distal limbs of the jejunum, as advocated by Jaboulay-Braun, to prevent this complication, has been considered of value by Weir. The latter uses a small Murphy button for the purpose, and with long forceps introduces each half of the button . from inside the lumen of the intestine, making a small opening through the intestinal wall large enough to admit the cylinders and clamps without sutures. Mikulicz made entero-anastomosis four times after regurgitant vomiting had commenced, as a secondary operation, with success in checking regurgitation. Doyen has recommended that the bowel be divided and the distal end anastomosed to the stomach, the proximal end being joined to the small bowel below, in this way avoiding, as be believes, the danger of the vicious circle, and pre- venting dilatation of the proximal portion of the loop. Rutkowski adds to the gastro-enterostomy a Witzel or Kader gastrostomy, and introduces a rubber tube from the surface through the external fistula and the stomach into the intestine MALIGNANT DISEASES OF STOMACH AM) I'VLORUS 113 hy way oi the anastomotic opening;. \\'it/,el speaks highly of this method, both in i)reventing the establishment of a false route and also as j)rovi(ling a means of early feeding through the tube. It has been recommended that the tube be fastened into the gastro- enterostomy opening by an absorbable suture, to prevent it from slipping upward into the stomach. The necessity for such com- plicated operations for the ])urj)ose of preventing regurgitant vom- iting is oj)en to serious question. Czerny had only 1 fatal case of intestinal regurgitation in 65 gastro-enterostomies, and this case was complicated by free hemorrhage from an ulcerating carci- noma. Czerny believes the button prevents spur formation while in place, and prefers the posterior operation. Mikulicz had 74 gastro-enterostomies in the Breslau clinic, with 24 deaths, spur formation accounting for 6, or 25 per cent, of the mortality. In the fatal cases due to spur formation the proximal loop was enor- mously distended and the intestinal tract empty, the patients dy- ing of starvation. Mikulicz abandoned the posterior operation on this account. At the time of the report he was using the anterior method with the ^Murphy button, and his results were very satis- factory. The suture operation of Wolfler brings the lateral wall of the jejunum to the side of the stomach, and Senn fixes the bowel at several points each side of the opening to prevent angulation, and favors a long visceral incision to prevent contraction. Kocher loops the bowel up in such a manner as to produce a rather marked angle, the anastomosis being at the apex of the knuckle. Fenger has modified the anterior suture operation of Kocher in a most important particular, and theoretically it would appear to be a good method. The suturing is proceeded with as in the Kocher l)lan, but from the middle of each half of the anterior surface of the stomach and bowel an incision from three-quarters of an inch to an inch in length is made — that is, from the center of the un- finished upper half toward the lesser curvature ol the stomach and downward along the convexity of the bowel to the same extent. On closing this wound the anterior portion of the anastomotic opening is greatly lengthened, and spur formation, with its at- tendant evils, prevented. It is not so injuriously affected liy VOL. I — 8 114 WILLIAM J. MAYO future contraction as the Kocher operation. Fenger's modifica- tion is simply "pyloroplasty" upon the anterior surface of the opening. The writer has used the Murphy button in all his operations, and has been satisfied with the results. Death followed 3 out of 11 gastro-enterostomies for malignant disease, and in only 1 out of 20 non-malignant cases. In no case was the button found to be at fault. Two died of pneumonia from aspiration and 2 of exhaustion. A form of collapse after operation for malignant disease in the abdomen, coming on about the fourth day in my experience, has been a rather common fatal termination. The postmortem does not show adequate cause for the exhaustion. In the Breslau clinic collapse was given as the cause of death in 13 out of 24 fatal cases. In using the button the female half should be placed in the intestine first after a preliminary purse-string suture. The stomach-wall should be cut to the mucous coat before placing the suture, and the latter should be close to the margins of the incision, to render approximation easy; otherwise the thick wall of the stomach will rufSe up and expose the suture at some point. This is less true of malignant disease than of non-malignant, as the shorter period of obstruction does not permit of adequate com- pensation, and the ability to develop muscular structure is defi- cient. Klimmell places the stomach half of the button in position and closes the incision to the cylinder by sutures. In a later report Kummell says that he has since had an accident from this cause, and he has now returned to the purse-string suture of Murphy. Kammerer speaks of the button as being especially adapted for the posterior operation and the suture for the anterior. Supple- mentary sutures outside of the button are unnecessary, and may prove a source of danger. I have never used them. In using the button, one should be careful to prevent a hematoma infiltrating the walls of the stomach outside the grasp of the button, as the sloughing process may cause an infection of the clot and perfora- tion beyond the limiting adhesions. The main objection made to .malh;\.\nt diseases of stoniac ii and pylorus 11.> Ihe l)ullon is that it often falls hack into the stomach. This occurred in a iiurnhcr of my cases, and no harm has resulted. Weir and Kiiinmell have each modified the intestinal half of the button, enlarging it in such a way as to prevent its passing into the stomach. I do not know of the value of the modification. Malthc says that if the button in position lies to the right of the spinal column, it will pass downward; if to the left, it will droj) l)ack into the stomach. For some time I have paid no attention to the passing of the button. Meyer speaks of finding the button in the rectum in many of his cases. Colicky pains and symptoms of obstruction may appear while the button is in transit. This readily subsides by prohil)iting food for a short time. Contraction of the anastomotic opening after the button operation does not often occur. Meyer was able to find only one case. That some contraction should take place with the diminish- ing size of a dilated stomach is to be expected, but in none of my cases has failure to empty the stomach occurred after the operation. I have one case of over six years' duration, and several beyond three years. Gastro-enterostomy for malignant disease has a mor- tality of about 38.3 (Robson). Czerny, in 65 recent cases, had a mortality of 38.5. Mikulicz, 74 cases; mortality, 3*2 per cent. Pernian, of Stockholm, 4'2 cases and 15 deaths. The special preparation of the stomach for operation is of im- portance. Under normal conditions bacteria do not flourish in the stomach, although present under ordinary food conditions. With carcinoma, motor insufficiency, retention, and, in the later stages, ulceration, present conditions favoring development of germs, as well as the saprophytes of putrefaction. In Halsted's clinic Cush- ing has been able to secure a high degree of sterilization of the stomach by means of careful antiseptic cleansing of the teeth and mouth and heat sterilization of the food. Lavage as a means of aifling the cleansing process is very necessary. The mechanical re- moval of the gastric contents washes out the unabsorbed food- [)roducts and prevents decomposition. Guillot does not favor lavage immediately prior to operation upon the stomach, believing that it tends to weaken the patient 116 WILLIAM J. MAYO at a critical time. He also opposes purgation before the operation on the same grounds. Among surgeons generally the opposite view is held, lavage and purgation being considered essential as preliminary preparations. The writer does not ordinarily favor marked changes in the diet of patients shortly before operation. The patient of average intelligence can materially aid the surgeon in selecting articles of diet which experience has taught him cause the least harm. If these articles can be sterilized by cooking and the remains be removed by gastric lavage before decomposition occurs, we will have accomplished something in the way of securing a proper wound-site. Undoubtedly a greater amount of good would be accomplished by a diet beginning some days before operation, but the necessary experimentation to secure proper feeding takes valuable time, and the immediate result is often temporarily to disarrange the already enfeebled digestive f>ower. Attempts to add to the patient's strength before operation by rectal feeding to supplement gastric absorption may be objection- able. In many cases the rectum becomes intolerant after a few days, and its value may in this way be seriously impaired for car- rying on nutrition after operation. My own experience has been that the patient does fully as well if either method be employed. Not more than a few days should be spent in preliminary prepa- ration. The stomach should be carefully emptied of its contents just previous to the operation; this is seldom as successful as one could wish, since the water will often return quite clear, and on opening the stomach a few minutes later a quantity of dirty fluid will be encountered. This renders accidental wound soiling pos- sible, and in elevating the stomach out of the abdominal incision gravity may cause the fluids to pass into the esophagus. Pneu- monia under such circumstances is very liable to occur after operation. I had two deaths from this cause. Fifteen out of a total of 20 deaths after operations on the stomach in the Heidel- berg clinic were from pneumonia. Czerny does not think the pneumonia due to the anesthetic, as it occurred twice in five opera- tions under cocain, and as it often came on in the first forty-eight MALIGNANT DISKASKS OF STOMACH AND I'VLOKIS 117 hours, it could not be clue to confinemont to bed. In his cxix'ri- ence the complication was most common in males with a pre- vious bronchitis or emphysema, and due, he believes, to the incision interfering^ with abdominal respiration. In de!)ilitated cases very little anesthetic is needed — a preliminary hypoder- matic of morphin, with just enough ether or chloroform to enable painless division of the abdominal coverings and again to close. No pain is felt during the gastro-intestinal manipu- lations. Local anesthesia by cocain in very debilitated sub- jects is an ideal method, provided the operation is short and does not reciuire traction on the margins of the abdominal incision. Abbe and others have used it to a considerable extent in gastric surgery. The experience of Bloodgood and Gushing in hernia work, and Matas in cocainization of nerve-trunks, suggests a wider field for its employment. The after-care is mainly to counteract shock, which the nearness to the great sympathetic ganglion and direct injury to the terminal filaments of the vagi often renders severe. Morphin, strychnin, and atropin are useful to meet indi- cations, and, if necessary, saline infusions. Rectal enemata of hot saline solution or coffee are valuable adjuvants to prevent col- lapse. After the immediate danger has been overcome, every effort to prevent exhaustion and death at a later stage must be- made. The majority of surgeons prefer rectal feeding for the first few days. Successful rectal feeding requires experience and good judgment; the tendency is to overfeed and to use larger quantities than are well borne. The need of liquids is most ap- parent, and large enemata of saline solution at least once in twenty- four hours meet this indication. The general tendency is to earlier feeding by mouth, and less reliance is placed on rectal alimenta- tion. Rectal feeding carries the patients along, but it is inade- quate, and they do not gain. Guillot begins liquid nourishment by mouth two hours after partial extirpation of the stomach, and Roux gives whatever the patient desires and as soon as called for. This practice shows great coufidence in the methods of suture in preventing leakage, but after an abdominal operation of this magnitude, digestion for 118 WILLIAM J. IVIAYO the first twenty -four hours is nearly at a standstill, and food under such circumstances is liable to do harm. Chlumsky, in his ex- periments as to the strength of union after intestinal anastomosis, demonstrated that from the third to the fifth day the union was weakest. There was little difference between the button and suture in this respect. These conclusions are borne out by clinical experience, and care should be exercised in feeding until union is complete. Elderly people bear confinement to bed badly, and do much better if allowed to be up within the first week. LYMPHATIC INVOLVEMENT IN CANCER OF THE STOMACH* WILLIAM J. MAYO To THE Editor of the "Medical Record": In the issue of the "IVFedical Record," IVIay 5, 1900, is an abstract of a paper on "Malignant Diseases of the Stomach and Pylorus," read before the American Surgical Association. In the abstract I am quoted as saying that — "Lymphatic involvement, so called, is now known to be a simple septic glandular enlargement, the relation between the abdominal glands and the carcinoma being utterly different from that which exists between the breast and axilla; and that abdominal adenitis is, therefore, of little importance." What I did say is, "Lymphatic infection is a still more serious comjilication, because, if palpable, glandular malignancy can be established, and it is altogether probable that the involvement has passed beyond reach." In 28 cases of non-malignant dila- tations of the stomach from various causes which I have operated upon, in more than half the enlarged glands could be palpated, due to an accompanying chronic gastritis. This may be, and frequently is, true of cancer of the stomach, the enlarged glands being the result of septic complications. The modern operation for the removal of the glands with carcinoma of the breast sug- gests a similar glandular extirpation in radical operations on the stomach. Lindner does not believe this feasible. In 28 recurrences after extiri)ation 15 were local, 12 distant, and only 1 involved the glands. An investigation leads us to the belief that glandular infection in cancer of the stomach is not uniform, — many such patients die without such involvement, — and that a moderate en- * Reprinted from the "Med. Record." 1900, Ivii, p. OiS. 119 120 WILLIAM J. AL\TO largement may be septic. Should they become infected, their situation renders radical removal practically impossible. Inasmuch as involvement of the lymphatics takes place in at least two-thirds of the cases of cancer of the stomach, I feel that the statement should be corrected, and more especially in consideration of the large circulation of your journal. SOME OF THE DISEASES COMMON TO THE STOMACH: THEIR SURGICAL TREATMENT* WILLIAM J. MAYO There are a few points in the anatomy of the stomach which are of especial importance in their surgical relations, and, as the first portion of the duodenum in its origin, development, and func- tion partakes more of the stomach than the intestine, it may well be considered at the same time. This part of the duodenum may be said to be the vestibule of the small intestines, for it is not until the biliary and pancreatic juices are poured into the second portion that the real work of the small bowel begins. Up to this point the acid reaction and character of its contents predispose to the same forms of ulceration and kindred phenomena which mark the stom- ach rather than the small bowel. One mentally grasps the anatomy of the stomach by consider- ing it an intestine with a double mesentery, the upper part being formed by the lesser omentum and gastrophrenic and splenic ligaments. This is the most fixed part of the stomach-wall. The lower attachment is formed by the gastrocolic portion of the great omentum, which gives mobility to the organ. The anterior and posterior surfaces are covered by peritoneum. The blood- supply is from three principal sources, and in its freedom of anas- tomosis calls to mind the palmar arch or the circle of Willis. This allows of extensive incisions, and gives a certainty of wound heal- ing which is lacking in the small intestine with its single mesentery. The smaller vessels lie beneath the mucous coat, and the thickness *Read before the Minnesota Academy of Medicine, November 7, 1900. Re- printed from "The St. Paul Medical Journal," January, 1901. 121 122 WILLIAM J. MAYO and looseness of attachment of this latter membrane allow its easy separation, permitting independent suture. The gastro- hepatic portion of the lesser omentum anchors the pylorus behind the ribs, and by division of this structure the distal end of the stomach can be easily mobilized. The position of the normal stomach is a disputed point. Luschka maintains that the direction is nearly vertical, the lesser curvature corresponding with the median line of the body, and there is no doubt but that this is the normal position in infants. The weight of authority gives to the stomach of the adult a small degree of obliquity. The cardiac orifice is fixed at the twelfth dorsal vertebra, and the pylorus lies behind the edge of the right costal border, on a line with the tip of the ensiform cartilage. Doyen says that the vertical position of the lesser curvature gives to the pyloric antrum and first portion of the duodenum the ap- pearance of a fish-hook. When the stomach is distended, it rotates somewhat on its long axis, the pylorus passing downward and to the right. With- out going into the question of the exact position of the normal stomach, it seems certain that the pylorus cannot be easily felt, nor is the lesser curvature within reach of palpation, and if they can be so palpated, they are in a condition of descent. Changes in the size of the stomach within normal limits are effected through the greater curvature. The surgeon is inclined to look at the function of the stomach as being largely mechanical; it acts as a reservoir in which the ingesta are macerated in a weak solution of pepsin and hydro- chloric acid, breaking up the food-masses, equalizing the tem- perature, and, with its powerful muscular apparatus, slowly propelling its contents into the small bowel, the pyloric sphincter acting to prevent intestinal overloading. Bacteria do not thrive in the stomach. The experiments of Gushing go to show that the germs of the upper intestinal tract are of far less virulence than in the lower bowel, and by proper cleansing of the teeth and mouth and cooking of the food a rela- tively high degree of sterilization can be secured. SOME DISEASES COMMON" TO TlIK STOM ATM 1 '^.'5 MelJiixh of Kxainination. — In .sj>iU' of llie (jpliiiiisin of the gastrologist, the value of the examination of test-meals and the various methods of absorption for estimating the motor power of the stomaeji are not ^'reat, and only when taken into consideration with the history and clinical si<;;ns have they any real weifj;ht. The general tendency is to rely too much upon the results of laboratory analysis and too little on the ordinary clinical examination. While not wishing to minimize the importance of such analysis, as con- firmatory evidence I would emphasize the fact that the history and clinical course are of first consideration. The laboratory must go hand in hand with clinical observation. The surgeon is concerned with certain definite conditions largely mechanical, or at least in which there is a demonstrable change in the organ itself, and these changes are usually capable of clinical demonstration. For instance, the salol test for loss of motor power is unreliable, but the finding of food in the fasting stomach regularly seven hours after taking is both practical and reliable, and if found four- teen hours after, indicates stagnation or retention. For making the outlines of the stomach evident neither the gastrodiaphanoscope nor the gastroscope is of real use, nor can the giving of bismuth or the using of metal sounds for the purpose of x-Tny be compared in efficiency with simple dilatation of the stomach with bicarbonate of soda and tartaric acid, or, what is better still, a stomach-tube and a Davidson s\Tinge, with which it is very easy to distend the stomach with air, and if, on distention, the lesser curvature and pylorus remain in a normal position while the greater curvature lies below the umbilicus, dilatation is evi- dent. If the pylorus and les.ser curvature are detected below the costal margin, the stomach has descended and the distance between the lesser and greater curvature marks the extent of the dilatation if it exists. The history of the patient, his present condition, and the chemical and biologic examination of the gastric contents, taken into consideration with the position and size of the stomach, give a basis for diagnosis and, as a rule, indi- cate whether an abdominal incision may be expedient. In the beginning nearly all operations on the stomach are exploratory 124 WILLIAM J. MAYO in character, for in spite of the most elaborate preliminary in- vestigation, the exact condition is seldom known beforehand. A median incision is most convenient for this purpose, and through it the pylorus and anterior wall of the stomach can be explored. The posterior wall can be examined by passing the hand behind the stomach through an opening in the gastrocolic omentum. Inspection of the interior of the gastric cavity is more diffi- cult. Tiffany advises an incision through the anterior wall of the stomach, and with the hand behind, successive portions of its interior are pressed up to the incision and inspected through this opening. Maylard, of Glasgow, in a paper before the last International Surgical Congress, August, 1900, recommends that an incision be made through the anterior wall of the stomach and its contents removed with a siphon tube. He then introduces an old-fashioned Ferguson vaginal speculum. The entire interior can in this way be inspected mth ease. This suggestion of May- lard's is of great importance in the recognition of ulceration and new-growths, and gives a certainty to the diagnostic incision which it did not previously possess. For emptying the stomach at the time of operation of its fluid contents we have found an ordinary Davidson syringe with a piece of gauze tied over its suction end as a strainer to be most valuable. Ulcer of the Stomach. — Syphihtic and tuberculous ulcerations occur so rarely as to be of small interest. Until within the last five years surgery was directed more to the relief of secondary conditions, such as contractions arising during the heahng process, or to the occasional emergency operation, in which the sudden perforation of a gastric ulcer demanded instant action. These were operations of necessity and were life saving, but with the advent of better knowledge of the subject and better technic operations of expedi- ency have become frequent, with the intention of arresting a dis- tressing and dangerous malady and bringing about a speedy and certain recovery. It is still an unsettled question as to when the emancipation and pain of a chronic intractable ulcer of the stomach demand surgical relief. The problem is one in which the experience of SOME DISEASES COMMON' TO THK STOMACH 1 'io the operator is llic most iinporlaiit factor. The results ol" ojxt- ations for pislric ulcer are constantly iuiproviu^'; Rodman says that "in the- l)c<,nnning a new operation must demonstrate the reasons for its existence by saving the otherwise hopeless cases, and the early mortality is high; later, by a judicious selection, the results are greatly improved." It may be said at this time that prolonged and imsucccssful medical treatment is not for the best interest of the patient. Ger- hardt states that 28 per cent, of the cases of gastric ulcer treated medically eventually prove fatal. Leube, of Wiirzl)urg, in the tabulation of 1000 cases of gastric ulcer, gives a direct mortality of 4 per cent, from hemorrhage or perforation and 21 per cent, were not cured. It is probable that about 75 per cent, can be cured by medical and dietetic treatment in from four to five weeks. Eighty- five per cent, of the cases which cannot be cured by medical treat- ment can be cured by operation. The diagnosis of ulcer of the stomach in the average ca.se is easy. The pain, vomiting of food with hyperacid secretions, occasional small hemorrhages, and such objective signs as a sensi- tive epigastric or dorsal point, all aid in the differentiation. Hut it should not be forgotten that ulcer may give rise to but few symp- toms, and obstructive dilatation may be the first evidence that an ulcer has ever existed. Age is not so important a factor as has been supposed; I have watched a number of cases in which active sym])toms were present in the fourth and fifth decades of life, and in two instances have operated to cut short its course after the fortieth year. The presence of a tumor was long considered a differential sign as against ulcer. A thickening over an ulcerated area may sometimes be felt, and occasionally with great ease. For chronic intractable ulceration three methods of ojjcration have been practised: Excision, incision with suture, and gastro- enterostomy. Each has its proper indication and field of useful- ness. Gastro-enterostomy relieves the hyperchlorhydria and puts the stomach at rest, and if more than one ulcer is present, it is the operation of choice. Xicolaysen has gathered 30 cases from the literature successfully treated in this manner. In 5 cases of 126 WILLI.^1 J. MAYO our ovm the results of gastro-enterostomy for the rehef of open ulcer were satisfactory, but not speedy. In each case six months or more elapsed before the symptoms were relieved. The con- tracted stomach seemed to be unable to hold enough nourishment, and the opinion of Malthe that a number of months of painful digestion is to be expected after gastro-enterostomy for open ulcer has been borne out by our experience. Excision has been prac- tised in a small number of cases. To do this successfully requires that the ulcer be single and in an accessible situation. In any operation for intractable ulcer great care should be exercised in handling adlierent portions of the stomach-wall, as under such circumstances partial gastrectomy^ may become necessary to prevent secondary perforation. At least 6 per cent, of all gastric ulcers perforate (Finney), and of 56 cases of perforation collected by Dickinson from the St. George's Hospital reports, about one-half died at once and one-half formed adhesions, with resulting abdominal or subphrenic abscess, fistula, or other secondary complications. The result depends upon the suddenness of the perforation, its location, and the amount of the gastric contents at the time, ulcers on the lesser curvature and posterior surface being more liable to be at least temporarily protected by adhesions. In 793 perforated cases Welch found 62 per cent, on the lesser curvature and the posterior wall. The treatment of acute perforation is immediate suture wdth omental graft, or, if necessary, gauze isolation and drainage. Sec- ondary processes the result of perforations should be cared for on ordinary surgical principles. Tinker has collected 268 cases of perforation, which have been operated upon vnih 48 per cent, mortality. Perforation of ulcer in the first portion of the duo- denum has been sho^Ti by Weir to be not an infrequent accident, and in its course and results not to differ materially from gastric perforation, except that it gives rise to a right-sided peritonitis resembhng a cholecystitis or appendicitis, and, as Richardson has pointed out, the result in either gastric or duodenal perforation depends largely on the quantity of fluid turned loose in the abdom- inal cavity, as well as the speed and thoroughness with which it SOME DISEASES COMMON TO THE STOMAril l'-27 is removed. Finney again (alls attention to the sudden increase in the leukocyte count as an early symptom of perforation of the stomach, and says that in this resjjcct it is similar to typhoid perforation. Hemorrhage the result of gastric ulcer may demand surgical intervention. Repeated small hemorrhages are perhaps a more positive indication for operation than a single large hemorrhage. Such cases have been treated by Kiister and others by primary incision, or with the actual cautery and gastro-enterostoray. Andrews and Eisendrath have elaborated a definite operation for the treatment of hemorrhage from gastric ulcer. Through an incision into the gastric cavity the bleeding point is located and tied en masse from within, protective sutures being introduced from without if possible. Gastro-enterostomy is perhaps most often performed of necessity, and frequently the bleeding point can- not be located even at postmortem, or several ulcers may be present ; fortunately, the hemorrhage is usually checked by this operation. The most common indication for operation following ulcer is obstruction, the result of the healing process, with subsequent dilatation. This usually occurs at the pylorus, but may take place in the body, and an hour-glass distortion of the stomach result. The latter form is not common. Watson found but 29 operated cases in the literature, including two of his own. In hour-glass stomach two general plans of attack have been adopted : first, gastro-enterostomy with attachment to the proximal pouch. Unfortunately, this does not drain the distal cavity. Second, a plastic operation, in which the two pouches are joined; for many reasons the latter is the preferable operation. Pyloric obstruction the result of contraction is a frequent condition, and not so often diagnosticated. The diagnosis is that of dilatation of the stomach with retention, and such com- plications as the accompanying chronic gastritis may give rise to — hyperchlorhydria, supersecretion, and so forth. For the relief of benign obstructions, pyloroplasty and gastro- enterostomy have been rival procedures. The first requires that the amount of scar tissue be small, otherwise a plastic operation 128 WILLIAM J. MAYO is hazardous, and, secondly, that the hypertrophy of the muscular coat be equal to the task of elevating the food from the artificial pouch of the dilated stomach. These conditions are seldom present, and nearly one-half of the pyloroplasties eventually require gastro-enterostomy for relief. The latter operation is more reliable, but has a slightly higher mortality. In 11 pyloroplasties all recovered, but we were later compelled to do a gastro-enteros- tomy in 4. In 41 gastro-enterostomies for various causes the operation was a mechanical success in all, but 5 died. Some Forms of Gastrectasia of Uncertain Origin. — There are a number of cases of dilatation of the stomach in which, if there is an obstruction, it is not discoverable. The symptoms are not unlike obstructive dilatation in the severe cases, while in the milder forms the condition is that of an ordinary dyspepsia. The causa- tion is obscure, although most frequently ascribed to constitu- tional maladies and wasting diseases, yet it is not uncommon in people who are otherwise robust, and in these cases is supposed to be a neurosis. If compensatory hypertrophy of the gastric muscles takes place, the peristaltic waves can be seen as the stomach contracts. In other cases no such hypertrophy exists, and the muscular coat is thinned and stretched without visible signs of its action at any time. In this way, then, I think we can distinguish two general classes: in the first, obstruction to easy passage must exist,^ — the hypertrophy proves it, — and in the second the absence of hypertrophy indicates some other origin. In analyzing the first group of cases the position of the pylorus is important. I have a number of times noticed that, with other conditions equal, if the pylorus was at a low point, dilatation did not come on so early nor so completely as when the outlet was at a high level. In a paper read before the American Medical Association in 1895 I described a form of obstruction of the pylorus called "valve formation," in which a high-lying and fixed pylorus became more and more compressed as the stomach filled, the distention finally relieving itself by copious vomiting of the retained ingesta. In these cases there is the constant need of elevating the food up- SO.MK DISKASK.S COMMON TO THE STOMACH 1 '29 ward, with rcsultjuit liyportro[)liy, but without diminution of the caHber of tlie pyloric opening. Any sHght overloading causes an acute angle at the outlet and interferes with the delivery of the food into the duodenum. I believe this is often a cause of dilata- tion, and have oj)cratcd on four such cases. The so-called "fish- hook" i)ylorus, in which the pyloric opening is directed upward, increasing the muscular efforts, acts in a similar manner by in- creasing the work. Spasm of the {)ylorus has been a popular diagnosis of late, and is supposed to depend upon an ulceration which may be too slight to be seen, yet sufficient to cause a spasmodic closure of the pylorus, interfering more or less with the passage of the gastric contents. I have met with four cases in which this appeared to be the condi- tion. The pyloric sphincter w^as tightly contracted in each, and on the under surface a thickening could be felt. In two there were slight adhesions, and in three there was some glandular enlarge- ment. The pylorus was opened longitudinally in all, and the pyloric region carefully examined; in one a small ulcer could be seen; in two there seemed to be an erosion of the mucous coat; in one nothing could be found. In all these cases there was hj^po- chlorhydria or supersecretion. In each case, on opening the gastric cavity a quantity of clear mucus and gastric secretions were en- countered, but not much evidence of decomposing food, differing in this respect markedly from cicatricial contraction. In each case the diagnostic incision was closed transversely by pyloroplasty. In all these cases there was temporary relief, but in two it was necessary to do a gastro-enterostomy later. The previous history of the cases was similar — a rather sudden commencement, pain a constant symptom, aggravated liy eating, loss of flesh, hj-per- chlorhydria, or supersecretion. Test-meals showed excessive acidity and a somewhat dilated stomach, but without much retention of food. In two the pain was just beneath the ensiform cartilage; in two so far to the right as to suggest gall-stone disease, and they were operated upon with this diagnosis. It is possible that these cases are similar to fissure of the anus in the painful spasm produced. All these patients starved themselves rather VOL. 1—9 130 wiLLL^ii J. :nl\yo than endure the certain pain which followed eating, yet in none was vomiting a prominent symptom. The pyloroplasties in which rehef did not follow the operation were found to be adherent at a high level. In four cases since then I have fastened the pylorus at a low point to the abdominal wall after pyloroplasty. hoping that this might relieve the mechanical condition and enable the stomach to drain to better advantage. Whether this will prove true cannot yet be determined. Pyloric spasm as an explanation of the obstructive phenomena in certain cases is very alluring; it is not necessary that anything should be found on exploration, the inference being that the spasm relaxes under the anesthetic, and the ulcer is too small to be apparent. That this condition does exist cannot be doubted, but we should be careful not to make it a "scapegoat" for other more material causes. Gastroptosis may also cause dilatation, and the two difficulties not infrequently coexist. In such cases the pylorus is prolapsed downward and to the right: the lesser curvature is in e\-idence below the margin of the ribs, but the greater curvature is propor- tionately much lower, rendering the relative position of the pylorus such as to necessitate an increased muscular effort to force the food on its journey into the small bowel. Gastroptosis is usually a part of a general condition associated with a prolapse of the abdominal viscera, and for which no adequate explanation has yet been advanced. There are at least two factors which may be causative: a loss of muscular tone and neurasthenia, although the latter is sometimes considered a result rather than a cause. There are two varieties of cases in which no h^-per- trophj^ exists. The first may be said to be a secondary dila- tation, the obstruction causing a primary h\'pertrophy and finally, increasing beyond the hniits of compensation, the muscular coats of the stomach become thinned and stretched under the increased strain. The second variety is where myasthenia exists without obstruction, seemingly due to a simple atony. The resulting retention causes dilatation and aU its disastrous com- plications. Gastric myasthenia is usually a chronic affection. yet sometimes occurs in a most acute form, the dilatation pro- SOME DISEASES COMMON TO THE STOMACH 1 '5 1 ducing violent symptoms ami f^cnorally irsiilling in death in a few days. Without rej^ard to its origin, dilatation causes a number of symptoms common to all. Tlie stomach is shown to he too large by distention with air, splashing sounds, and other signs, the subjective .symptoms depending on the degree of retention and its effect on the gastric mucosa. Chronic inflammation of the mucosa usually coexists, due to the .stagnation and fermenta- tion; supersecretion or hyperchlorhydria may be present, the result of the hyperasthenic gastritis. The.se .symptoms are of great clinical importance, and w'orthy of more attention than the scope of this pa])er permits. Supersecretion and hyperchlorhydria are conditions closely connected with pyloric spasm, either as cause or effect, and may be sufficiently pronounced and the effects on the individual so unfortunate as to demand surgical intervention. In these cases drainage of the stomach would be indicated whether dilatation exists or not. Many individuals suffering from dilatation eat so carefully and so little as to maintain a balance w'ith the compensa- tion; in other cases drainage of the stomach pouch is desirable. Gastro-enterostomy is most generally useful, and has few con- traindications. The opening should be established at or near the greater curvature, and this is of great importance in securing adequate drainage and preventing regurgitant vomiting. Whether the suture or button is used or the anterior or posterior operation employed makes little difference in the result, provided the anastomotic opening be placed at the bottom of the pouch of the stomach. Pyloroplasty may be of benefit in certain cases; as a result of my own experience I believe that much depends on the position of the pylorus. If the outlet is high, the weakened muscles must still elevate the food, and relapse after pyloro])lasty may occur; on the contrary, if the pylorus lies low, gravity aids its progress. Gastroplication endeavors to overcome this necessary elevation of food by i)laiting the walls of the gastric pouch so as to change the relative position of the pylorus. It may be advantageous in selected cases if the pyloric opening is ample, but personally I have never met with a case in whiih this opera- tion appeared to be indicated. SOME INDICATIONS FOR GASTRO- ENTEROSTOMY* WILLIAM J. MAYO Surgery of the stomach has advanced rapidly and now occupies a secure position. Its Hmitations, however, are not as yet estab- lished, and there is much debatable ground between the internist and the surgeon. This is more especially true of the inflammatory condition of the stomach, of which ulcer is the more common. It would be manifestly impossible for me to cover the entire ground of gastric surgery in the short time at my disposal, and I shall, therefore, confine myself to one phase of the subject — that of gastric drainage by means of gastro-enterostomy. Dilatation of the stomach is not a disease, but a symptom, yet the ease with which it can be diagnosed and the desirability of its relief without regard to cause render it a factor of prime import- ance. The original operation of gastro-enterostomy as performed by Wolfler in 1881 had this object in view. Pyloric obstruction due to cancer not amenable to radical operation gives a clear indi- cation for gastro-enterostomy, and as much can be said of the benign forms of obstruction, such as the healing of an ulceration in the vicinity of the pylorus. The relation between cause and effect is apparent, and the argument is simple and convincing. Gastro- enterostomy for the relief of open ulcer and allied inflammatory conditions without obstruction develops a new line of thought and there is room for discussion. All in all, this operation is by far more valuable than any other operative procedure upon the stomach. In 136 operations on the stomach performed in St. Mary's Hospital, more than half have been gastro-enteros- *Reprinted from "The St. Paul Medical Journal," November, 1901. 132 SOME IN1)1( ATIOXS KOU (JASTRO-ENTEROSTOMY 133 toiiiies, and our experience lias Ik'cii that of o])erator.s in widely different fields. Cancer of the Stomach In somewhat over 100 cases of malignant disease of the stomach which the writer has examined with a view to surgical operation, more than half wore so plainly advanced beyond radical interven- tion that exploration was not necessary to ascertain the fact. In spite of the most elaborate examination of the stomach-con- tents and other methods of diagnosis it has been only those cases in which the location of the growth introduced mechanical features that a diagnosis could be made sufficiently early to make an at- tempt to cure by a radical operation. One of the most pronounced symptoms of gastric carcinoma is the diminished muscular power of the stomach, and if the growth be pyloric, a small amount of obstruction gives early evidence of its presence. Adequate compensatory hypertrophy is not found, al- though present to a marked degree in non-malignant obstruction. The great difficulty in making an early diagnosis in malignant disease renders gastro-enterostomy the operation most often indi- cated. It is to be hoped that in the future an earlier diagnosis may more frequently enable radical operations to be performed. In 53 operated cases of cancer of the stomach we have made gastro-enterostomy 18 times, with 4 deaths. The longest period of life after gastro-enterostomy was nineteen months, with ability to work until a few weeks prior to death. Excision of the pylorus and a greater or less amount of the stomach was undertaken 11 times, with 1 death; 6 patients are still alive, and 1 of these nearly four years, although there arc now evidences of metastasis in the liver in this case. In the remaining cases the extent of the disease or location not giving rise to obstructive phenomena, op- eration was not indicated, and the abdominal incision was closed. For the past five years we have sutured exploratory abdominal in- cisions for inoperable malignant disease with permanent sutures of buried silk or silver wire. This enables the patient to get about the next day and leave the hospital \\'ithin the week. If left in 13i WILLIAM J, MAYO bed the usual length of time, many of these patients become pro- gressively weaker and are unable to return to their homes and friends. The mortality of gastro-enterostomy for malignant disease is 20 per cent, or over — fully as high as after pylorectomy. The first requirement for the performance of a radical operation is that the disease should not be too far advanced, and the patient must be in good condition. Gastro-enterostomy has no such limitations, and the mortality is influenced to a large extent by the condition of the patient. Ulcer Excision or other form of surgical treatment is indicated in a few cases presenting special features, but the common situation of the ulceration, its varying extent, and the reasonable possibility that more than one ulcer exists, make gastro-enterostomy the prac- ticable operation in the majority of cases. Not infrequently the site of the ulcer cannot be discovered, rendering gastro-enterostomy the operation of necessity. The symptoms of ulcer of the stomach depend somewhat upon the situation of the disease; ulcer is most common near the pylorus, a position which may introduce certain mechanical features, and it is in the relief of these secondary phenomena that this operation achieves its triumphs. Gastro-enterostomy relieves the hyper- acidity and allows prompt emptying of the ingesta, preventing irri- tation and aiding nutrition. The ulcerated stomach is often contracted, and among the earlier observers it was supposed to be always small; this is partly true. In acute ulcer the stomach is small, and if the ulcerative proc- ess is not in the vicinity of the outlet, it will probably remain small. On the contrary, it is during the healing process that many ulcers in the pyloric region become most troublesome. Ulcers in this situation are often extensive, and in chronic cases perhaps but partly cicatrized. Enough distortion or narrowing of the pyloric outlet takes place materially to obstruct the opening, the unhealed portion of the ulcer keeping up irregular symptoms of its presence in addition to the dilatation. In such cases symptoms of open ulcer SO.Mi; INDICATIONS FOR GASTRO-ENTEROSTOMY 135 alternate with periods of health, and later signs of ulc-er in a stomach more or less dilated supervene. The majority of cases when once cicatrized remain healed, but a minority occasionally lapse into open ulcer. The capacity of the stomach affected hy ulcer is not greatly changed in the majority of cases, hut if so, it has a surgical signifi- cance. This gives us a good working basis of comparison. First, ulcers in the i)yloric region, with a normal or enlarged stomach, and, second, ulcers in a contracted stomach. In the first group gastro-enterostomy is the operation of choice: it delivers the ingesta at a point sufficiently far from the disease to prevent irritation, and the healing process is not inter- fered with and progresses rapidly. We have made gastro-enteros- tomy 11 times for this condition, with 1 death; the remaining 10 were cured and remain well. At times a small ulcer at the pylorus causes pyloric spasm, and symptoms are produced resembling me- chanical interference. In four cases of pyloric spasm we found di- latation only once, and then not at all marked. In this form of disease pyloroplasty is fairly effective, but does not compare with the benefits derived from gastro-enterostomy in suitable cases, al- though the division of the pyloric sphincter stops the spasm and the enlargement of the opening exerts a healing influence on the ulcer. In the second group of cases gastro-enterostomy does not give immediate relief, as a rule. In four cases of our own the symptoms in a modified degree continued for some months. However, ex- perience goes to show that after gastro-enterostomy the ulcer will eventually heal, but the results are not so good as in the first group of cases. The pylorus being open and the stomach small, it is self- evident that the main function of gastro-enterostomy is already well performed, and it is probable that in some cases the artificial opening will not remain patent. In two of these four cases secon- dary kinking of the small bowel at the anastomotic opening caused symptoms of chronic obstruction, and in both the communication was, at the second operation, found to be much reduced in size. In one case but six weeks elapsed, in the other, thirteen months. Theoretically, we would expect the opening to close in this group of 136 WILLLAJM J, MAYO cases, the pylorus being of good caliber and the stomach properly emptying itseK, the fistula becomes cicatrized from non-use. In the contraction attending closure in these cases the attached in- testine was angulated, causing distress. Entero-anastomosis be- tween the two limbs of the attached intestine afforded prompt relief. Taken alone, this would seem to contraindicate gastro- enterostomy on the small stomach, which would not be just, to the operation, since both of these cases were cured of the ulcera- tion, as shown at the second operation. The temporary arrest of activity in the stomach and prompt drainage of ingesta allow speedy healing of the ulcer. It must also be borne in mind that these cases represent the chronic and intractable variety of ulceration. The majority of dilated stomachs are due to old ulcers which have permanently closed, and only the mechanical interference with the progress of the food remains. Benign Obstruction at the Pylorus Pyloric obstruction of benign origin is usually the result of the healing of an ulceration in the vicinity of the pylorus. In some cases it is due to valve formation. Obstruction from without the stomach, due to pressure of tu- mors or adhesions, may also be an indication for operation. In some cases a dilated stomach is found without apparent obstruc- tion, but as the muscular wall is often greatly increased, we may take it that some interference with prompt emptying of the gastric cavity exists. Non-malignant obstruction of the pylorus is usually accompanied by marked hypertrophy of the stomach-wall, with dilatation in the later stages, which may be aptly compared to the heart with valvular insufficiency. Most of these patients eat little, and unconsciously^ try to accommodate the food quality and quantity to the diminished gastric musculature. Others wash the residue out of the stomach with a stomach-tube. Vomiting is not so common in these cases as in those more acute in character, al- though in the later stages vomiting of large quantities is to be ex- pected. A peculiarity about some of these cases is the formation of a gastric pouch, due to the more rapid stretching of the greater HOME INDICATIONS FOR GASTRO-ENTEROSTOMY Itil curvature. It is for Lliis rcuson lliut pyloroplasty may fail to cure. The pylorus being held high, even if the opening be made of ample size, the fatty degenerated muscle-fiber is unable to elevate the food from the pouch, and the patient is not materially relieved. In 15 j)yloroi)lasties we were compelled to make gastro-enterostomy G times as a secondary operation. It is in these advanced cases that gastro-enterostomy gives the great advantage over pyloroplasty. The opening can be made at the bottom of the pouch, furnishing adequate drainage. Hour-glass stomach, in rare instances, may demand gastro- enterostomy upon the proximal pouch. We had one such case. Gastroplasty is, however, the operation of choice, as shown by Wat- son, We have performed gastro-enterostomy 35 times for benign obstruction of various forms, with 1 death — this in a case of chronic starvation of most pronounced character. Analyzing the 67 gastro-enterostomies upon which the report is based, we find an average mortality of less than 9 per cent. — highest in the malignant cases, as would be expected (23 per cent.) ; for open ulcer, about 6 per cent. The condition of the average patient with an intractable ulceration would lead us to expect this death-rate, although if the operation be performed at an earlier date, it should not exceed 5 per cent. In the benign obstructions which form more than one-half of the cases only 1 death in 35 cases, or less than 3 per cent., occurred. The operation has three points of interest: Shall it be made with the suture or with the Murphy button? This is a matter of individual preference— the results are about the same. Another open question is whether the anastomosis should be made to the anterior wall of the stomach (Wolfler) or to the posterior (von Hacker). We have performed the anterior 57 times, with good re- sults. In the last year we have made a few posterior — 10 in all; they have done as well in every respect. The point of great import- ance is that in either operation the juncture should be at, or near, the greater curvature and at the bottom of the gastric pouch. By actual observation of our own cases we found that the opening, properly placed, either anteriorly or posteriorly, came within one-half inch of the same point in the gastric cavity. The attached wall of the stom- 138 WILLIAM J. MAYO ach promptly assumes a funnel shape, and regurgitant vomiting is seldom seen. Lastly, shall the two limbs of the bowel be separately anastomosed (Braun) as a primary operation? If the above method of securing a low union with the stomach be carried out, it is usually unnecessary. It sometimes happens that either by adhesions or angulation distress will be occasioned by small amounts of bile passing into the stomach, which may or may not be vomited. I wish to call especial attention to this condition following gastro- enterostomy. In going over the literature it would seem to be met with only immediately after operation, frequently causing death from the so-called "vicious circle." Carle showed by ex- perimentation that a small quantity of bile was found in the stom- ach for a long time after gastro-enterostomy, but that it caused no trouble. We have met with this condition months after the operation, but it did cause distress, although not usually attended with vomiting of food. At times a mouthful of bile-stained fluid would be regurgitated or could be washed out of the stomach by gastric lavage. I am uncertain whether this is due to regurgitation through the open pylorus or through the fistula. In the three cases referred to the operation was performed for the rehef of ulcer with unobstructed pylorus. At the second operation the fistula was found much reduced in size, with some angulation. Entero- anastomosis cured these cases. It would seem that primary re- gurgitation of bile and pancreatic juice could be prevented by a low location of the fistula, and that secondary regurgitation would not take place in pyloric obstruction, but with an unobstructed pylorus, as when the operation is done for open ulcer, entero- anastomosis as a primary operation would be indicated to prevent these delayed but unpleasant symptoms. In the future we will make an entero-anastomosis in every case in which gastro-enterostomy is performed if there be no obstruction of the pylorus. That this is clearly indicated is shown by the three secondary operations in 13 cases. In obstructed pylorus we did not have a single instance in 54 operations, and therefore it could not be considered in this, the larger, group. PROBLEMS RELATING TO SURGERY OF THE STOMACH* WILLIAM J. MAYO In the preparation of the subject to be discussed I have thought it wise to confine myself to the practical aspects of gastric surgery, using for this purpose the material obtained from a single hospital. This method of treating the subject is not due to a lack of appre- ciation of the work of the pioneers in this branch of our art. but rather with the hope that the limited experience of an observer in a somewhat distant field might be of greater interest. Gastric surgery, to a large extent, is still in the developmental stage, and this is due to the lack of definite knowledge upon which to base a surgical diagnosis. Volumes have been written upon diseases of the stomach from a medical standpoint, but as the state- ments made are based upon the symptoms of the patient or the results of postmortem examinations, we gain but little in that great middle ground in which the surgery of expediency will find its field of usefulness. The debatable territory is now being explored, and we shall shortly have more exact knowledge concerning it. Our own experience would seem to indicate that in the medical diag- nosis there are four important lines of inquiry to be pursued: (1) The history of the patient; (2) the size and position of the stomach; (3) or tumor localizing point of tenderness; (4) interference with the progress of the food. The examination of the stomach-contents has corroboratory value, especially with reference to the stagna- tion or retention of ingesta. The chemical and microscopic find- * Read before the Surgical Section of the Suffolk District Medical Society, February iG, \{)0i. Reprinted from "Boston Medical and Surgical Journal," May 1. hoi, vol. cxlvi. \o. 18, pp. 451-456. 139 140 "U'lLLIAM J. ^L^YO ings are unreliable in the early phases of disease, but possess some significance later in its course. Examination of the blood, the urine, the feces, etc., is of interest and helpful. The use of the gastroscope, gastrodiaphanoscope, ar-ray, etc., is still experimental. The mechanics of the stomach is the most interesting feature to the surgeon ; from this point of view the function of the stomach is largely mechanical. It absorbs fluids, equalizes the temperature of the ingesta, and the weak solution of hydrochloric acid and pep- sin which is secreted breaks up the food-masses, forming a homo- geneous material which is fed down into the small bowel, where the real work of digestion and absorption takes place. Any interference with the outlet promptly produces symptoms corresponding with the degree of obstruction, while ulceration or other disease involving the wall of the stomach, preventing it from acting as a reservoir, is also quickly resented. The distress in each case causes the patient to unconsciously try to adjust the quality and quantity of food to the loss of this peculiar function of the stomach. The result of obstruction at the pylorus is to in- crease the capacity of the stomach, and this is often the only ob- jective sign to which our attention is called before operation. Di- latation is to be expected in the first group, of which pyloric stenosis is the t^'pe, but unless the disease of the wall is sufficiently near the pylorus to add mechanical features, it is not present in the second group, of which ulcer is the chief example. Dilatation, due to benign obstruction at the pylorus, is followed by increase in the muscular wall of the stomach, the hypertrophy enabhng the damaged organ to carry on its function. This degree of compensation is often aided by the patient through a selected diet. In these cases compensation, alternating with dilatation and its discomforts, gives a cHnical picture which may be aptly com- pared to cardiac insufficiency. ^Miy is it that these patients, with far greater symptoms than would be tolerated in either the appen- diceal or gall-bladder regions, are allowed to go unrelieved.' It is not only that we are unable to know before operation the exact nature of the trouble, but that we also distrust our ability to make a diagnosis at the operating table. I'UOHLKMS UELATINf; TO SUIK.'KUY OF STOMAfU I 1. 1 In the beginning, every operation upon the .stomach partakes of an exploratory incision, and too often the proposed operation stops upon exposure of an extent of disease beyond intervention. This is particularly true of cancer. The surgical exploration of the stomach may not prove easy. The pylorus and anterior wall are open to inspection, and gross lesions of all parts can be ascer- tained, but not so the more minute forms of disease, such as the round ulcer. Our plan has been to explore by sight and touch the more accessible portions of the stomach-wall. Then, by opening into the lesser cavity of the peritoneum through the gastrocolic omentum, to pass the hand behind the stomach and search its posterior wall (Tiffany). To explore the interior of the gastric cavity a transverse incision is made three inches in length through the anterior wall half way between the pylorus and cardiac orifice. Into this a short rectal speculum, two inches in length and one and one-half inches in diameter, is inserted, and the fluids removed by suction. With the hand behind the stomach, nearly the whole of its mucous surface can be passed in review before the end of the speculum under direct light. (This is a modification of the method first brought out by Maylard at the International Congress, 1900.) With a considerable lesion one may often doubt whether the trouble is cancer or ulcer; or, not infrequently, the seat of an ulcer has undergone carcinomatous degeneration, leading to uncertainty. This is especially true of the pyloric region. We have had two cases in which the thickening about a pyloric ulcer was so great that even after incision, with the parts open to inspection, we were unable to tell the difference macroscopically. Enlarged lymphatic glands, unless distinctly cancerous, do not help us. It has been our experience that in the majority of diseases of the stomach marked by retention and fermentation of the food enlarged lym- phatic glands are to be found in the omenta. The most common forms of dilatation of the stomach are due to the healing of a gastric ulcer causing stenosis, or to malignant disease involving the py- lorus. The only cases of cancer of the stomach we have been able to diagnosticate sufficiently early to extirpate were the cases in 142 WILLIAM J. MAYO which obstruction and dilatation were present. Upon opening the abdomen, this factor is easily seen and needs no comment. There is a large group, however, of chronic cases of dilatation of the stomach giving rise to symptoms which, upon careful surgical exploration, show no adequate cause for the condition. In other cases hypersecretion or hyperchlorhydria is the cause of chronic gas- tric distress in which operative relief is indicated. We may say that pyloric spasm exists, due to a microscopic ulcer. It is a convenient term, not capable of either proof or refutation. In this connection I have examined over 100 stomachs in the course of other operations, with especial reference to the pylorus. Under anesthesia, if the normal pyloric opening be compressed between the thumb and fin- ger, invaginated into the stomach and duodenal walls on either side, the lumen will permit easy meeting of the opposing digits, and gives the feeling that an opening exists about the size of an old-fashioned silver three-cent piece. Comparing this with the cases which I had previously diagnosticated as having pyloric spasm, little differ- ence could be detected. In only four cases could a definite thick- ening be demonstrated in the pyloric ring. In these patients the slight abnormality was situated posteriorly, but in one only did incision reveal an ulcer. It was also noted that in these four cases there was very little if any dilatation of the stomach. In a number of other cases angulation was present, that is, a high-lying pylorus somewhat firmly fixed, with a sharp bend of the stomach downward immediately proximal to it. In 1896 I de- scribed five cases of this condition as a cause of dilatation of the stomach, under the title of "Valve Formation." I have seen a number of cases since. Adhesions outside the lumen of the stom- ach, the result of a perigastritis from gastric ulcer or a cholecystitis, may be the cause. (Robson, Cabot, and others have described a number of such cases.) We have met with this condition most frequently in connection with work on the biliary tract. In some cases, however, when no apparent cause for the dilatation exists, the stomach-wall is hypertrophied, and for this reason we must con- clude that in some manner obstruction does exist. The most perplexing cases are those of neurotic origin, in which Fig. 8 — Traction weight of small bowel, producing funnel shape of stomach at site of anastomosis. Fig. 9. — Showing proper and improper locations of opening: a. Proper position, leaving no pouch; b, usual position, forming intragastric pouch. I'HOJiLKMS UKLATING TO SLUCiKUY OF STOMACH \ V.i class I would i)l;ice the various grades of gaslroj)tosis. Some of these invalidji have also an accompanying dilatation, but usually of the atonic variety, that is, without increased muscular thicken- ing of the wall. On surgical exploration such a stomach is usually found to be empty and contracted, although j)revious examination has siiown it to be dilated. In tlie purely neurotic variety, while there may be little or no change in the size or position of the stom- ach, symptoms of ulcer may be so perfectly simulated as to lead to an exploration which proves negative. On manipulating such a stom- ach it nuiy contract in small areas, and for a moment look as though an ulcer existed, to as suddenly disappear, or the whole stomach may undergo vermicular contraction until it is no larger than the colon. In two such instances I have seen the pylorus suddenly dilate until two or more fingers could be invaginated through it. These cases must be classified surgically with movable kidney, movable retro- version of the uterus, varicocele, etc., usually occurring in neu- rasthenic individuals, and occasionally demanding an operation, which may be followed by benefit. I have examined a number of such stomachs, and we have operated upon a few, in wdiich dilata- tion coexisted or a mistake in diagnosis w'as made. One has a feeling that we should reject, surgically, this whole group, yet even neurasthenics are not exempt from actual disease, although we naturally subject them to a most careful and painstaking prelim- inary examination in which subjective symptoms are accorded but little weight. Speaking from an operative standpoint, dilatation with retardation of the passage of the food out of the stomach is the most important surgical indication. When does this condition demand an operation? It is largely a personal equation between the experience of the surgeon and the disability of the patient. The value of gastric drainage in these cases is apparent and needs no argument. The desirability of drainage of the non-dilated stomach is based largely upon clinical observation. Theoretically, it would not strike one that a well-drained or contracted stomach, even if ulcer be present, would be benefited by such a procedure. It is claimed that gastric drainage, especially gastro-enterostomy, rests the stomach, permits escape of secretions, and increases the 144 WILLIAM J. ?HAYO nutrition, thereby aiding recovery. Our own experience, while Hmited, in a general way seems to bear out this conclusion, but not wholly so. If we divide our cases of ulcer into two groups, in the first place all the cases in which ulcer existed in the pyloric end of the stomach, and in which the capacity of the stomach was in- creased from any cause, and in the second group place all the cases in which the contrary existed, we find the results much less favor- able in the latter. In cases of ulcer in the small stomach excision offers a more satisfactory means of cure, although I regret to say we have had but two cases in which we were able to do this. Gastric drainage can be best established in two ways : (1) Pyloro- plasty, and (2) gastro-enterostomy. Pyloroplasty, after the method of Heineke-Mikulicz, has been made fifteen times in St. Mary's Hos- pital; in four of these cases failure permanently to relieve the symp- toms necessitated a secondary gastro-enterostomy: the stomach was greatly dilated, and, by a more rapid stretching of its greater curva- ture, also pouched. The plastic operation on the pylorus in each in- stance was found to have been successful so far as enlarging its cali- ber was concerned, but the degenerated muscle-fiber of the stomach- wall had been unable to elevate the food from the gastric pouch to the high-lying pylorus, and the symptoms were largely unabated. As at the secondary operation the pylorus was always found adherent, it occurred to me that perhaps after the plastic opera- tion, if the pylorus could be anchored down in the vicinity of the umbilicus and allowed to become adherent at that point, it would drain the stomach better. We have practised this in five cases, but as we have also been more careful in the selection of non- pouched stomachs for the operation, I am uncertain whether the better results have been due to the method or to the care in selec- tion : There were no deaths in this group of cases and no later com- plications. Pyloroplasty will have a limited field of usefulness in cases in which dilatation of the stomach is not great. If it can be shown that the pyloric spasm is a large factor in the clinical course of gastric ulcer, pyloroplasty, which destroys the sphincter action of the muscle, might be the operation of choice in the group of cases without dilatation, or in which the stomach is contracted, as it is PROBLEMS RELATING TO SURGERY OF STOMACH 145 in these cases that gastro-enterostomy has been the least beneficial and late complications have occasionally arisen. All in all, we have found gastro-enterostomy to be the most satisfactory operation on the stomach. This operation was per- formed 80 times, with 8 deaths. For cancer, 21 gastro-enteros- tomies with 4 deaths, the greatest length of life was nineteen months, with ability to carry on manual labor for more than sixteen months. With few exceptions, however, the palliation has been of such short duration as liardly to justify the operation. The hope of the future for cancer of the stomach is early exploration and extirpa- tion. For benign conditions, gastro-enterostomy has the great ad- vantage in that it drains the stomach from the lowest point, re- lieving the retention of obstruction equally with the painful con- tact which the food causes in gastric ulcer. Twelve cases of chronic intractable ulceration in the vicinity of the pylorus, in which some narrowing of the orifice was produced and dilatation was present, were benefited, the cure in the 11 that recovered from the operation being fairly good. In 6 cases of ulcer in which the pylorus was of normal size and the stomach contracted, relief was less certain and slow to come about. In 3 cases of ulcer subjected to gastro-enter- ostomy a secondary operation became necessary for angulation at the site of the anastomosis. This took place after some weeks or months, and was found to be due to a contraction which fol- lowed at the anastomotic opening. This subsequent narrowing is of no consequence so far as the stomach is concerned, but as one- third the lumen of the small bowel is involved, the reduction may be a serious matter, causing an angulation later. The attempt at obliteration comes on after the stomach has resumed its function, and takes place only in the cases in which the pylorus was unob- structed, nature making the usual endeavor to close an unnecessary fistula. The symptoms of this complication are attacks of burn- ing pain in the stomach, with nausea, and perhaps a little bile- stained fluid may be regurgitated, or at times be washed out of the stomach, but there is usually no stagnation or vomiting of food. Entero-anastomosis relieves the condition. It would seem wise VOL. I — 10 146 WILLIAM J. MAYO to make an entero-anastomosis at the primary operation whenever gastro-enterostomy is performed in a case in which the pylorus is unobstructed. In no case of permanently obstructed pylorus has contraction of the anastomotic opening followed in our cases, so far as we know. We have never seen the opening completely closed. The splendid drainage established in the dilated stomach, which presupposes some interference with the passage of the food, leads us to use gastro-enterostomy for temporary purposes in a normally drained stomach. Even in these cases it cures many and relieves the majority. As to the method of performing gastro-enterostomy, there are still a few questions to be settled: (1) Shall we use the suture or the Murphy button? So far as I can judge, the results are about the same. (2) Shall it be on the anterior or posterior wall of the stom- ach .f^ Here again there is little choice. We have made 69 an- terior and 11 posterior, with equally good results. Theoretically, the posterior operation would seem the better, as one can secure the jejunum at a higher point. We have made the posterior within 6 inches of the origin of the jejunum, and it takes 14 inches to form a loop for the anterior method. Making the posterior opera- tion so close is not a safe procedure, since, should it become neces- sary to do an entero-anastomosis later, there is not sufficient room on the proximal side of the anastomotic opening for this purpose. We lost one case from this cause. In either operation, from 14 to 16 inches of intestine should be left on the proximal side. The main thing in gastro-enterostomy is that the opening should be low down, near the greater curvature, in either operation. We have had little trouble with primary pernicious vomiting (vicious circle) for more than four years since we began this practice. The anterior operation is usually made about half-way between the lesser and greater curvatures, and where there are but few blood- vessels. This is a bad practice, since it leaves a pouch into which the bile and pancreatic secretions can easily enter, and it encourages vicious circle. In doing the posterior operation the inferior border is more accessible, and one naturally places the opening lower down. The anastomosis should be effected in such manner that its inferior Fig. to. — lorminj; an apron of the omentum, attachinf: to thf ~toniacn arM)\c inc anailomosis. PROBLEMS RELATING TO SURGERY OF STOMACH 147 edge shall be at the bottom of the stomach pouch, on a line with the greater curvature in either the anterior or posterior operation (Fig. 9). In two of our cases of gastro-enterostomy the bowel detached spontaneously from the stomach, — once on the seventh and once on the tenth day, — with resultant leakage and death, contrary to Chlumsky's experiments, in which it was shown that union was firm after the fifth day. It was noted at the autopsy that it was the superior edge of the union only that detached; the lower edge, being just at the origin of the gastrocolic omentum, was so pro- tected as to be of unusual strength. After making the anterior inferior anastomosis we grasp the omentum upon either side and pull it upward in such a manner as will not tract upon the transverse colon. The two upper free ends are fastened together and then to the stomach-wall, not less than one inch above the anastomosis, with fine catgut. The edges of the omentum are then united to each other for two and one-half inches, forming an apron which completely covers the site of union, protecting the weak point, yet having no connection with it (Fig. 10). Should the omentum drag in the future, the strain would come above the opening upon the stomach and increase that funnel shape (Fig. 8) which the stomach should assume after the operation is properly completed. In the posterior inferior operation a few sutures attaching the margins of the torn mesentery of the transverse colon to the stomach will furnish the same protection to the union. Gastro-enterostomy for late cancer of the pylorus will be followed by bad results without regard to method, and if ascites be present, union will probably not take place. COMPLICATIONS FOLLOWING GASTRO- ENTEROSTOMY* WILLIAM J. MAYO During the past ten years 98 gastro-enterostomies have been performed in St. Mary's Hospital, with 9 deaths. The mortaUty in the malignant cases was 20 per cent., and in the benign cases, 6 per cent. During this time 14 pylorectomies and partial gas- trectomies have been made, with 2 deaths — 14 per cent. Of these, 9 were excisions with complete closure of both the stomach and duodenal ends, communication being established by means of an independent gastrojejunostomy of the usual type. One death. These 9 cases, added to the 98 cases above mentioned, give 107 gastro-enterostomies with 10 deaths — an average mortality of 9 per cent. The causes of death were as follows: Exhaustion, 3 cases ; exhaustion in which pernicious vomiting was a prominent feature, 2 cases; progressive pneumonia, 3 cases; detachment of the anastomosed intestine from the stomach- wall, 2 cases. The deaths from exhaustion were the result of starvation at the time of the operation. The patients appeared fairly well until the fourth to seventh day, when a gradual failure of the vital forces appeared and death ensued in from twelve to twenty-four hours. The postmortem showed the abdominal condition to be good. Cachectic subjects bear rectal feeding badly, and giving nourish- ment by the stomach should be practised early, when possible. The two cases in which regurgitant vomiting hastened death were among the early operations, in which the intestine was joined to the anterior wall of the stomach halfway between the * Reprinted from "Annals of Surgery," August, 1902. 148 Fig. II. — Showing proper and improper locations of opening: a. Proper position, leaving no pouch b, usual position, forming intragastric pouch. COMPLICATIONS FOLLOWING GASTRO-ENTEROSTOMY 149 greater and lesser curvatures, causing an intragastric pouch to form, which • contrihulcd to tlie unfortunate conii)lication. In neither case could it be said that the vomiting itself caused death, but in the feeble condition of the patients it certainly was a factor. It will be noted that nearly one-third of the total death-rate was due to bronchopneumonia. There have been many explanations as to the frequency with which lung complications occur following operations upon the stomach. These complications were believed to be the result of general anesthesia, but experience has shown that they are rela- tively as frequent after the use of a local anesthetic. The situation of the incision in the epigastrium, preventing coughing and ex- pectoration, is thought to be an element in causation, yet similar incisions in the gall-bladder region have no such effect. The latest hypothesis is that some of the venous blood returning from the stomach does not pass through the portal vein, and in this way infected emboli are carried directly into the circulation and pass at once to the lungs. In two of the three cases a chronic bronchial cough was present at the time of operation, and the patients were in bad general condition. In one case material was aspirated through the trachea from the esophagus, causing pneumonia. It is difficult, by means of the stomach-tube, thoroughly to cleanse and empty the greatly dilated stomach in debilitated subjects. In this case, on elevating the stomach out of the abdominal incision, some of the jQuid contents gravi- tated into the esophagus. This should be avoided by elevation of the head and thorax at this time. The recumbent posture fol- lowing operation is apparently injurious in some cases, and we now encourage the old and feeble to sit up early. It is evident that there is as yet no entirely adequate explanation for the pro- duction of the pulmonary complication. There are probably several contributing causes in most cases. In the two cases in which the anastomosed intestine was detached, causing death from leakage, one took place on the seventh day after gastro-enterostomy for malignant pyloric 150 WILLLIM J. MAYO obstruction. There was a small amount of free fluid present in the abdomen at the time of operation, which would usually contraindicate a plastic procedure, such as gastro-enterostomy. In the second case, detachment on the ninth day followed an epileptic seizure. This was in a patient with benign obstruction, who had up to that time done unusually well. He had suffered from epilepsy for years, and the aura began in the epigastric region. In a violent contraction of the stomach such a detach- ment might easily take place. Chlumsky's experiments on pre- sumably healthy animals went to show that after five days the union was perfect. That this is not true in diseased states in the human subject is shown by these two cases. Of the 97 cases which recovered from the operation, 5 benign cases came to secondary operation on account of changes at the anastomotic orifice. The most important feature in the mechanics of the anasto- mosis is that the union shall be at the inferior border of the stomach, close to the greater curvature, and at the bottom of the gastric pouch, giving a funnel shape. Properly placed, the anastomotic opening should have its inferior border at the bottom of the stomach, and as to whether the opening shall extend from this point upward anteriorly or posteriorly is really of little mo- ment. (See Fig. 11.) The anterior operation has usually been placed relatively higher than the posterior, to avoid the blood- vessels, causing an intragastric pouch to form, which has been one source of pernicious vomiting. The posterior operation, for technical reasons (easier exposure), is usually placed nearer the greater curvature. The union in the 107 cases under discussion was made to the anterior wall of the stomach 8.3 times, and 24 times to the posterior wall, with equally good results, so that location of the opening on the anterior or posterior wall cannot of itself be essential. In our experience one operation is as easy as the other. For thin subjects with a long mesocolon we prefer the posterior method. If the mesentery is short or contains much fat, or if the vascular loop, from the superior mesenteric artery, which supplies the transverse colon, is small, bringing the opening CO.Ml'LKATIONS !• OIJ-* )\VI Nf ; (;A.STKO-IAr I .UOSTO.M V 1.51 in the poslerior layer of tlic j^'aslroeolir ornciitMrn iti rlose prox- imity to it, the aiiU'rior operation is preferre*!. After i)0.stcrior gastro-enterostomy the torn edj^es of the mesentery are sutured to the posterior wall of the stomach, as advised l)y Willy Meyer, to prevent downward displacement and interference with the loop, as hapi)enc(l to Meyer, Czerny, Korte, and others. These sutures arc introduced in such a manner as to provide a short flap of the mesenteric marf:^in, which drops over the anastomotic opening, furnishing further i)rotcction. After the anterior operation, the edges of the omentum are caught each side of the anastomosis and sutured to each other and to the stomach-wall one inch ahove the opening. The edges are united to each other downward for three inches, forming an apron over the anastomosis, yet having no connection with it; and as this is done with a fine catgut suture, the adhesion is not of itself permanent. This makes the omentum available if leakage occurs, and in time the omentum returns to its normal situation if no accident happens. This may seem an unnecessary precaution, but when it is considered that 20 per cent, of the deaths were due to se])aration of the bowel from the stomach at a time (ninth and tenth day) when neither suture nor button would furnish adequate support, it is not unreasonable. Both of our fatal cases followed anterior operations, and it was the superior edge of the union which gave way, as shown by post- mortem. The inferior margin, being protected by the origin of the omentum, was exceedingly firm. We have used the Murphy button in all, except one of our cases, in which the suture and the Robson bone bobbin were employed to meet a special indication. Case I. — Gastro-enterostomy; Reoperation Four Years Later for Secondary Ulceration; Recovery. — Mrs. H. H. O., aged thirty- eight years, Scandinavian, mother of three children, housewife, was admitted to St. Mary's Hospital May, 1899, with the follow- ing history: Has had symptoms of ulceration of the stomach for several years ; for the j)ast two years the trouble has been constant. The vomiting, which at first occurred immediately after taking food, is now delayed a number of hours, but the larger part of the nourishment is eventually rejected. She eats as small an amount 152 WILLIAM J. MAYO as possible and is limited entirely to liquid food. Has lost 35 pounds or more in weight. Personal and family history good. Physical Examination. — Emaciation marked; skin dry; pulse and temperature normal. Upper abdominal region distended. On inspection, peristaltic waves can be seen passing from left to right. Splashing phenomenon easily developed. On air disten- tion the greater curvature of the stomach found to lie on a line with the crest of the ilium. Test-meal shows free acid. Diagno- sis, benign pyloric obstruction due to the cicatrization of an ulcer. Operation. — Irregular cicatrix involving pylorus, three-fourths of an inch in diameter and one and one-fourth inches in length. Anterior gastro-enterostomy. Recovery uneventful. For three years she remained in splendid health, gaining over 40 pounds in weight. April 1, 1902, was readmitted to hospital on account of return of previous symptoms of obstruction, which had begun suddenly three months before, and were supposed to be due to an attack of appendicitis. Patient had lost much flesh and was on a liquid diet. The trouble was evidently due to some inter- ference with the outlet of the stomach. Ojperation kpxW 2, 1902: Amass of adhesions was encountered to the right of the median line, due to an ulcer of the stomach just above the anastomotic orifice, and involving the opening above and upon the right side. Perforation had occurred and the adhesion to the abdominal wall had prevented leakage. The transverse colon was closely adherent and much reduced in caliber where it passed under the anastomosis. The entire ulcerated area was excised, leaving a large opening with only one-fourth of the gastro-intestinal union on the left side intact. This defect was sutured, and the gastro- enterostomy completed by suture over a^Robson bone bobbin, the large plastic being protected by the omentum. The Murphy button was found in the stomach somewhat corroded, but in fairly workable condition. Pylorus completely obstructed. The stom- ach was dra'VNTi down into a funnel at the site of the anastomosis. I am under the impression that at the time of the sudden symp- toms the button became impacted and caused the ulceration. This is surmise, as it was found in the fundus of the stomach. If the stomach-wall is thick, the muscular and peritoneal coats should be incised before the suture is placed, and the suture should grasp only a small portion of these structures, otherwise the button may be held in position too long. In many cases in COMPLK'ATIOX.S FOI-LOWIXG GASTRO-EXTEROSTOMY 153 which the button passes, vomiting, with symptoms of obstruction, may appear in the second or third week while it is in transit. Gastric lavage and rectal feeding for a day or two will cause these symptoms to subside. The suture oi)eration for gastro-enterostomy is undoubtedly just as good as the button, and, so far as can be judged, the results are about the same. Among men of great experience Kocher uses the suture and the posterior method; Czerny, the button and the posterior; Mikulicz prefers the suture in benign cases and the button in malignant cases, and uses the anterior operation alto- gether. He finds that an entero-anastomosis is necessary in the suture operation to prevent pernicious vomiting, but does not find it necessary with the button, which tends to prevent angula- tion while in sitii, and this is during the dangerous period. Rob- son's bone bobbin acts in a similar manner. Kelling found that with the suture a ring of mucous membrane projected into the stomach, diminishing the caliber of the opening. The opening is less perfect with the suture, and entero-anastomosis is more often necessary to prevent pernicious vomiting. These advan- tages in favor of the button are counterbalanced by its tendency to drop into the stomach and remain there (Case I). This usually does no harm, and in malignant disease, at least, does not counter- balance the advantage. In our earlier experience with gastro-enterostomy, the ope- ration was performed entirely for pyloric obstruction, and in but two cases (IV and V) did any secondary complication develop with regard to the orifice, except its occasional occlusion by an advancing malignant growth. Two cases of malignant obstruc- tion, examined postmortem after the lapse of some months, showed no marked contraction of the opening. For non-malignant pyloric obstruction patients in the best of health, all the way from the present time up to eight years after the operation, demonstrate the permanence of the artificial opening. In two benign cases dying of other causes six months and three years respectively after the operation, and representing an anterior and a posterior location of opening, there was no contraction. In a case reported by 154 WILLIAM J. MAYO Cordier after six and one-half years death from other cause allowed a postmortem examination, and there was no contraction of the anastomotic opening found. Without going into detail, it may- be said that if permanent obstruction at the pylorus exists, no marked contraction of a properly formed gastro-enterostomy may be feared, unless by accident (Case V). About three years ago gastro-enterostomy for the relief of ulcer was first performed at St. Mary's Hospital, and since that time with increasing frequency — about 25 cases in all. In a ma- jority of these cases the pylorus was not mechanically obstructed, although the ulcer was usually in the pyloric region, and in some cases ultimate cicatrization might be expected materially to reduce the caliber of the normal opening. In three of these cases angulation and obstruction at the site of the anastomosis occurred at a later date (Cases II, III, and IV). In these cases secondary exploration revealed a marked contraction of the orifice, reducing its size to that of a lead-pencil or less, although in no case was obliteration complete. There was found an angulation of the jejunum at the attachment, causing a spur which accounted for the symptoms. The reduction, so far as the stomach was concerned, was of little moment, but a contraction involving one-third of the lumen of the small bowel was serious and caused valve formation. Case II. — Entero-anastomosis Thirteen Months After Gastro- enterostomy; Recovery. — Miss G. C, aged twenty-one years, American, seamstress, was admitted to St. Mary's Hospital May 9, 1900, with a typical history of ulcer, which had existed for more than a year and defied ordinary methods of treatment. Hemat- emesis had been a prominent feature, and on two occasions was so copious as to threaten life; she had lost 25 pounds in weight. Family and personal history otherwise good. Physical Examination. — Marked anemia from the hemor- rhages; organs other than stomach normal. A painful point the size of a silver dollar in the epigastrium. Stomach-contents not examined, it being feared that the necessary manipulation might cause a return of the hemorrhage. May 10th, anterior gastrojejunostomy; Murphy button. COMPLKATIONS lOIJ.OWIXG f;ASTUO-i:.\TKKr).ST(>.M V \~)') Stomach small, ijyloriis miobstructed, ulceration on lesser cur- vature of irrcf^ular outline, an inch in diameter, shown hy indura- tion, and covered hy i)erijj;aslric adhesions. JJulton passed duriiif^ third week. Dischar^'cd in the fourth week. Rapid gain in weight and complete disappearance of symptoms for four months. She then began to have attacks of burning pain in the stomach; tiiese became more fref[uent, and occasionally a little bile-stained fluid would be vomited. Xo great loss of weight or strength. In June, 1901, exploration revealed the fact that the gastro- intestinal fistula had contracted to the size of a lead-pencil or smaller; this produced a kink of the jejunum at the site of the anastomosis. Entero-anastomosis between the afferent and effer- ent limbs of the jejunum promptly relieved the symptoms. Patient now in good health. Case III. — Secondary Gastro-enterostomy and Entero-anasto- mosis Twenty Days after Primary Gastro-enterostomy ; Recovery. — P. D., male, aged thirty years, German, farmer, was admitted to St. Mary's Hospital March 21, 1901. History of chronic ulcer of the stomach extending over six years, which had obstinately resisted treatment. During most of this time he had been inca- pacitated for labor. To relieve the pain, semi-starvation had been practised. Personal and family history negative. Physical Examination. — An emaciated man of sallow com- plexion, dry and leathery skin. Heart, lungs, kidneys, etc., in normal condition. Tenderness just above umbilicus. Stomach moderately dilated; free acid and some retardation of food. Operation March 2'-2d: Anterior gastrojejunostomy; Murphy button. Ulcer on posterior wall and adherent to pancreas. The latter enlarged and thickened; no mechanical obstruction at the pylorus. For two weeks patient did very well, then began to vomit biliary and pancreatic secretions; button passed on six- teenth day. Vomiting at first intermittent, and no food returned unless given during the period of active regurgitation. Twenty days after the primary operation the abdomen was reopened. The anastomotic opening had contracted to the size of a lead- pencil, and spur formation of the small bowel was marked. As it seemed improbable that the ulcer should have permanently cica- trized in this short space of time, anterior gastrojejunostomy was again ]>erformed with the Murj)hy button, and an entero-anas- tomosis short circuiting the l)iliary and pancreatic secretions below both openings was made by means of a small button. 156 WILLIAM J. IVIAYO Discharged in three weeks. The patient rapidly gained in weight and strength. He is now in good health and able to perform manual labor. Entero-anastomosis promptly relieved the condition in these two cases. In the third, for reasons referred to later, death ensued. Contraction of the anastomotic opening is to be expected if the pylorus is unobstructed; but that it does not always pro- duce symptoms was shown in a fourth case, in which gastro- enterostomy for an active ulcer had promptly relieved a most serious condition. At a secondary operation for a pelvic tumor, some months later, a great contraction of the orifice was found, but without unpleasant symptoms arising therefrom. In Case V, after the first entero-anastomosis failed to relieve, the writer was under the impression that perhaps the kinking caused the bile to accumulate in the duodenum, and that the regurgitation was through the pylorus. For this reason the pylorus was excised, with complete closure of both the duodenum and stomach ends; yet this failed to check the biliary vomiting, showing conclusively that it was the spur at the opening alone which was responsible for the trouble. Von Eiselsberg reports cases in which he has closed the pylorus by a circular purse-string suture, evidently with the same idea which proved fallacious in this case. The question of the reduction of the opening taking place in the greatly dilated stomach pari passu with the contraction of the stomach itself has been pretty well settled by Robson, Korte, and others. The stagnation is promptly relieved, but the hyper- dilated stomach does not contract greatly, and the lesser degrees of dilatation which regain normal size do not materially afiFect the anastomotic opening. Carle and Fantino have shown con- clusively that small quantities of bile are to be found in the stomach after gastro-enterostomy, and that it does not lead to trouble. Ferrier and others have connected the gall-bladder directly with the stomach without interfering with digestion. The pancreatic juice cannot be the cause, as Stendel has experimentally divided the jejunum, fastening the open end to the stomach and closing COMPLICATIONS FOLLOWING GASTRO-KNTKUOSTO.MY 157 the duodenum completely at the severed point, causing all the biliary and pancreatic secretions to pass through the stomach, yet no harm resulted. This was also true of JMoynihan's case, in which this procedure was carried out on the human subject. McGraw believes that the views of Kelling are correct, and that it is the distention of the duodenum which is rcsponsil)le for the bad effects. The fact remains that entero-anastoraosis between the proximal and distal loops of the intestine short circuiting these secretions relieves the condition. The possibility of secondary spur formation following gastro-enterostomy for ulcer in which the pylorus is open must be borne in mind, and, if possible, excision of the ulcer is to be preferred. This the writer has been able to do three times for gastric ulcer and once for duodenal. It has been advocated, especially in this country by Robert Weir, to perform an entero-anastomosis in all cases of gastro-enterostomy at the ])rimary operation. This is certainly logical in cases under consideration in which the pylorus is open. We have preferred the simple operation of entero-anastomosis rather than the more elaborate methods of Roux and others, and in only one case, that a posterior operation, has relief failed to result. This was due to the fact that the jejunum was anasto- mosed so close to its origin as to prevent proper drainage from the proximal side through the interintestinal fistula. Case IV. — Gastro-enterostomy Followed by Entero-anastomosis, Pylorcctomy, Eniero-anasiomosis ; Roux's Operation; Death. — Mrs. J. M., aged forty-two, Scandinavian, housewife, one child, was admitted to St. Mary's Hospital on June 19, 1901. Typical history of chronic ulcer of the stomach. For three years symptoms nearly constant — vomiting, ])ain, loss of weight and strength; confined to bed for several weeks previous to admission to the hospital, and for some months has required opiates more or less constantly. Personal and family history good. Physical Examination. — Emaciation marked. Painful area in epigastrium. Stomach not increased in size. Free acid. June !20th, posterior gastrojejunostomy; ]Murphy button. Attachment to jejunum about six inches from its origin. Stomach not dilated, pylorus open; location of ulcer could not be accurately 158 WILLIAM J. MAYO determined on account of perigastric adhesions. Gall-bladder contained one stone, evidently "slumbering," as there were no evidences of disease about this viscus. Stone removed and gall- bladder drained through stab wound on the right side. Patient discharged in good condition on the twentieth day, July 24, 1901, readmitted; one week before had commenced to have attacks of burning pain in the stomach, and since had regurgitated a little bile- stained fluid at frequent intervals. Gastric lavage failed to relieve the symptoms. June 25th, entero-anastomosis. Operation difficult and un- satisfactory on account of the short length of the afferent intestine, and when completed, the interintestinal fistula was on a level with the gastric opening and only about two inches from it. Gastro- enterostomy contracted to less than the tip of the little finger, and angulation of the attached jejunum. The patient's condition improved rapidly, and for a time she was apparently relieved. October 12th she was readmitted, with all the old symptoms in an aggravated form. Under the impres- sion that the biliary and pancreatic secretions entered the stomach through the pylorus, on October 14th pylorectomy was performed, and the duodenum and stomach completely closed by a purse- string suture. No relief. October 18th a second button entero- anastomosis was made. This was a mistake, as between the previous entero-anastomosis and the origin of the jejunum there was less than three inches. Some relief was experienced for a few days. October 30th the previous symptoms had returned with increased severity, and, as the patient was becoming ex- hausted; as a final resort the operation of Roux was performed. The adhesions from the previous operations rendered this ex- tremely difficult. The jejunum was divided as closely as possible to the last entero-anastomosis, and the distal end turned in by a purse-string suture. Less than an inch of jejunum projected on the proximal side. A Murphy button was inserted and with some difficulty secured in position. A loop of bowel 16 inches below was attached laterally. Patient returned to bed in bad condition and died thirty-six hours later. This brings up the question as to how long a loop of jejunum should be made above the point of anastomosis. Robson says that for the anterior method 12 inches is about right, and for the posterior somewhat less. Mikulicz says that 15 c.c. is the neces- sary amount for the posterior operation and 50 c.c. for the anterior COMPLICATIONS FOLLOWINC; CASTRO-ENTEROSTOM V l.l!) operation. Wo have averaj^'ed ahout 14 inches for the anterior method, and since the unfortunate termination of the case referred to, not less than ten inches for the posterior. Meyer reports a ease in which fifteen centimeters proved to be too short for con- venience at a secondary operation. Peterson, from the Heidel- berg clinic, calls attention to the fact that the origin of the jejunum is at a higher level than the site of the anastomosis in the posterior operation. This would place the pro.ximal portion of the jejunum above the opening, and he believes that the absence of pernicious vomiting in the cases in Czerny's clinic is due to this cause, although it is evident that the location of the opening on the posterior wall of the stomach must in these cases have been at an inferior point, and it is probable that the advantage may lie in this feature of the operation. If the obstruction at the pylorus be permanent, there can be no objection to the short length of jejunum above, but if an open pylorus threatens contraction and spur formation, this may prove unfortunate. The last com])lication to be brief]y referred to is the possibility of the small bowel passing through the loop of intestine above the anastomosis. This danger is much greater with the anterior than with the posterior method. Case V so well illustrates this condition as to need no further comment. Case V. — Anterior Gastro-enterosfomy; Secondary Operation for the Relief of a Twist at the Anastomotic Opening Caused by Small Intestine Passing Through the Loop. — R. X. S., male, aged forty- one years, American, barber, was admitted to St. Mary's Hospital January 1, 1901, with the following history: For several years has suffered from attacks of burning pain in the epigastric region lasting a few minutes at a time, but recurring at intervals of several hours. Much worse when at work at his trade. These "cramps" lasted several weeks at a time, after which there would be an interval of weeks or months of good health. For several months he has had more or less stomach trouble, and occasionally vomited. The distress caused him to eat sparingly, and he has lost '■2o pounds in weight. He had an attack of appendicitis with an abscess some years ago; the abscess was incised, but the appendix was not removed. He has had a right inguinal hernia for many years. Examination. — A spare man, six feet one inch in height, emaciation noticeable. With the exception of the stomach, no 160 WILLIAM J. MAYO feature of interest. Painful point in epigastrium. Free acid; greater curvature of stomach three inches below the umbilicus. Diagnosis. — Pyloric obstruction from ulcer. January 2d, an- terior gastrojejunostomy; Murphy button; appendectomy and Bassini operation on hernia. An ulcer existed at the pylorus, extending to the lesser curvature, irregular contour, size of last phalanx of forefinger. Evidently partially cicatrized and obstruct- ing pylorus. On the fourteenth day symptoms of intestinal obstruction occurred, lasting forty-eight hours. Condition re- lieved by gastric lavage and rectal feeding. Button passed on the sixteenth day; evidently cause of symptoms. Discharged January 18th; gained rapidly in weight and strength. For a year he remained in good health, although complaining that if he stood erect he had a "drawing feeling" in his stomach. From this time to May 14, 1902, when he was readmitted to the hospital, he had slowly developed all the former symptoms of obstruction at the outlet of the stomach, and had a constant pain in the abdomen, centering below the umbilicus. May 15th abdomen opened. Gastrojejunal orifice nearly obliterated and stretched to an inch in length. Jejunum twisted at the site of anastomosis, one-half turn from the left to the right. Somewhat more than one-haK of the small intestine had passed through the loop of jejunum between the origin of the jejunum and the attachment to the stomach. The point of entrance was on the right side, beneath the transverse colon. The traction weight of the intes- tines upon the mesentery at the inferior margin of the loop had caused the volvulus. The mesentery at this point was much thickened. The intestines were replaced, the gastrojejunal fistula divided, and the opening into the stomach closed. The opening into the jejunum was inclosed by a purse-string suture, half of a Murphy button introduced, and a posterior gastrojejunostomy made. The pyloric stricture was nearly complete; the ulcer evi- dently cicatrized. It is probable that the part of jejunum imme- diately below the anastomosis passed through the loop first, pro- ducing the twist which was so prominent a feature on opening the abdomen. When this happened is hard to tell — probably not for some months after the operation. When the process once began, it might be expected to continue until such an amount of intestine traveled over the loop as to pull the mesentery taut, the symptoms increasing as the condition gradually developed. It is possible that at the time the juncture was effected a slight twist may have occurred. THE RADICAL CURE OE CANCER OE THE STOMACH* WILLIAM J. MAYO ('uiK-er of the sloiiiiuli is a hopeless malady unless il can be cured hy operation. The only excuse for the apathy of the medical profession as regards this condition is a belief that it is hopeless, whether operated upon or not. I do not believe that such a con- clusion is justified by the facts. In cancer of practically all the other organs operation is urged even enthusiastically by men who look with disfavor upon the surgical relief of malignant disease of the stomach. W. G. Macdonald found records of 43 cases of un- doubted cure after extirpation of cancer of the stomach. Murphy collected 189 cases operated upon radically by Kronlcin, Maydl, Rydygier, Czerny, Morison, Bevan, and Mayo, with 26 deaths. Of these, 17 survived three years— about 8 per cent. This per- centage was reduced to 5 per cent, by recurrences after three years, but as many of these cases were alive and apparently well less than three years, the law of averages might be expected to at least main- tain the 8 per cent, or better it in due time. Are the results of radical operations for cancer of the cervix uteri, for example, nuich better than this? If 8 per cent, of cases of cancer of the stomach can be cured by radical operation, this dis- couragement is not justified, neither is it creditable to us as a profession. One-third of all cases of carcinomatous disease is located in the stomach. The condition has not received a fair .share of attention from a surgical standpoint. Take another point of view: What has been the result in the *Read by invitation before the New York Medical Association October ii, imH. Reprinlod from "The St. Paul Medical Journal," December. 1904. VOL. I— 11 Kil 162 WILLIAM J. MAYO cases in which recurrence has manifested itself? Kronlein's statistics are the most available for consideration. The unope- rated usually died within the year, gastro-enterostomy prolonged life on an average three months, and radical operation gave an increase of fourteen months over the unoperated. Mikulicz in 100 cases had an average duration of life following operation of one and one-third years. The relief usually lasts until shortly before death, and there is not that prolonged and hopeless illness which characterizes the unoperated cases. The patient has not only the hope of cure, but a possibility of it. Why are the results of extirpation of cancer of the stomach not better? Because the diagnosis is not made sufficiently early. The stomach is a favorable organ for operation; it has a large blood-supply from several sources. It can be rendered relatively sterile, and at least 60 per cent, of carcinomata are situated in the pylorus, the most movable and accessible part of the organ. The tendency to lymphatic infection is probably less in cancer of the stomach than in similar conditions in the breast, uterus, or rectum (Macdonald) . In 802 collected cases McArdle found 40 per cent, relatively free from important lymphatic involvement, and Shaw demonstrated that 18 per cent, of cases of cancer of the stomach which came to autopsy had absolutely no secondary involvement. The physician has been taught to believe that the presence of a tumor contraindicates an operation, and unless one can find a tumor, the diagnosis is uncertain. He is put in the position of be- ing asked to make a diagnosis of cancer and take the responsibility of a perhaps unnecessary exploration. We should not ask for a diagnosis of cancer, but we should ask that a suspicion of malig- nant disease demands a surgical consultation quite as often as appendicitis or ectopic pregnancy. In cases of doubt the patient should be permitted to exercise a choice as to an exploration. Some years ago an eminent authority expressed the opinion that the presence of a tumor of itself demonstrated the incurability of the disease. This dictum must at least be qualified. A small movable tumor in the pyloric region with obstructive symptoms RADKAL rURE OF CANCER OF STOMACH 1(53 is a favorable consideration. Many patients die from starvation without secondary involvornont, and the niovahility argues for a freedom from adliesions wiiich complicate operations and increase the primary mortality. From our own experience we would judge that, with symptoms warranting an exploration, the lack of defi- nite signs usually means an extensive and hopeless involvement of the body of the stomach. Had the pylorus been involved, obstructive phenomena would have given far earlier warning. A large tumor or one indefinite in outline or of fixed character would be a different question and lead to a doubtful opinion. Another source of delay is the prolonged and usually unavail- ing attempt to make a diagnosis through various tests of the stom- ach-contents. After long and painstaking experience with lab- oratory methods for the early diagnosis of cancer of the stomach we have been forced to the conclusion that they do not amount to much, although late in the disease they are fairly reliable guides. I do not wish to discourage such examination — it is an effort in the right direction, and I hope will lead to increased precision in the future, but in any event it should not lead us to procrastinate. Given a previously healthy person of middle age or beyond, who is beginning to lose weight and strength with loss of appe- tite and slow and more or less painful digestion, we should sus- pect cancer. If, after thorough examination by all the methods at our command, adequate reason for the condition is not forth- coming in a few weeks at most, we should advise exploration. Should this become the practice, cancer of the stomach would give as favorable a showing as cancer of the breast or uterus. If the profession awakes to its responsibility, the general public will become educated. I would emphasize the necessity of obtaining a careful history. A very large proportion of the cases which we have had gave an early history of ulcer of the stomach. Graham gave this his attention, and found it to be the case in over one- half of all the patients coming under our observation. In several instances operation revealed the cicatrix of former ulceration to be undergoing malignant degeneration. This is more forcefully brought to mind by the frequency with which competent men take 164 WILLIAM J. MAYO just the opposite view — that a history of previous ulcer argues for a return of the ulcerative process and against malignancy. In our experience, cancer of the pyloric region has been the only form of the disease which we have been able to diagnosticate sufficiently early to attempt radical relief. Many times the proc- ess is distinctly ring like, with comparatively little lateral involve- ment. These cases give early evidence of retardation of the prog- ress of food from the stomach. The repeated finding of material left from the evening meal in the morning washings from the stomach is an easy way to determine that obstruction exists or the test-meal method may be used. By distending the stomach with air, not only can its outline be palpated, but often a tumor may be brought up within reach of the palpating hand. With an ordinary stomach-tube and a David- son syringe this can be safely done, air being permitted to escape or more pumped in at will. This procedure is more efl:"ective be- cause always under perfect control. The bicarbonate of soda and tartaric acid test is painful and does not always last sufficiently long, neither has it the element of safety. Behrend reports three deaths following its use, the sudden and uncontrollable distention doing fatal mischief to the diseased gastric wall. These simple means for practical examination are at the com- mand of every practitioner, and while he may not arrive at as ac- curate results as the trained specialist, he may at least have his suspicions aroused in time to give the patient a chance for his life. The surgical exploration presents no difficulty as to the pres- ence of gross lesions, and usually the malignant nature of the trouble is manifest. Occasionally, however, an ulcer will have so great an amount of new tissue about it as closely to resemble can- cer. We have been misled twice — once removing an ulcerous pylorus with the idea that it was cancer, and once removing a py- lorus with an epitheliomatous ulcer, supposing it to be a simple ulcer. This leads us to the conclusion that, other things being equal, every ulcer which can be located should be removed either as a complete operation or as a part of the operative procedure indi- cated. It also has the advantage of removing the original source RADICAL CTRE OF CANCER OF STOMAL II l()0 of disease and preventing secondary malignant degeneration of tlie soar tissue involved. In deciding as to the operable possibilities, the two most ini- f)()rlaiil considerations arc the extension of the disease to the neigh- boring tissues and the lymphatic involvement. As a rule, exten- sion to neighboring viscera contraindicates operation. Extensive adhesions greatly complicate the operation and increase the pri- mary mortality. Haberkant had a mortality of 72.7 per cent, in cases with adhesions, as contrasted with 27.3 per cent, without ad- hesions. Lymphatic involvement, unless localized to the vicinity of the growth, is a hopeless complication. In a general way the surgically removable glands lie in four groups — first, along the lesser curvature; second, along the greater curvature; third, about the head of the pancreas; and fourth, in the greater omentum. In all these positions in favorable cases extirpation may be justly at- tempted. It should not be forgotten that a certain amount of glandular hyperplasia may be found which is not due to malig- nancy. Such enlargement is often, if not usually, present in any form of disease which permits fermentation of the gastric contents. Fenger described simple adenopathy in the submaxillary glands in cancer of the lip, and Halsted has recorded cases in which microscopic examination of enlarged lymphatic glands did not show carcinomatous change in cancer of the breast. jMikulicz believes that the relief afiforded by pylorectomy and gastrectomy justifies the operation in otherwise favorable cases, even if all the glands cannot be removed. Extension along the lesser curvature is more unfavorable than along the greater curva- ture. Robson advises that in every case the resection shall extend along the lesser curvature to the gastric artery, to enable removal of glands which are very prone to infection and otherwise liable to be overlooked. It is needless to say that secondary involvement of the viscera, shown by ascites, etc., precludes operation. There is but little tendency for the growth to invade the duodenum. In a fair j)roportion of hopeless cases, present or imminent obstruction indicates gastro-enterostomy. After purely exploratory operations for inoperable disease I 166 WILLIAM J. MAYO wish to draw attention to the value of the permanent buried su- tures of silk or silver wire in the strong fascia in order that the patient may be allowed to get about at once and leave the hospital in a few days; such patients do not bear confinement to bed well, while waiting sound abdominal union, and often die without being able to return to their homes and friends. In St. Mary's Hospital during the past ten years 213 operations have been made upon the stomach: 126 for non-malignant disease, with a mortality of slightly less than 5 per cent.; 92 for malignant disease, with 9 deaths — 10 per cent. Over one-fourth of the latter group were explorations, but in no case did death follow such examination. Radical operations upon the stomach owe their inception to the master mind of Billroth, although Pean and Rydygier preceded him in the actual performance of pylorectomy. Billroth placed the operation upon a sound footing, and brought it prominently before the profession. He was not enthusiastic about it, and in speaking of his results said: "All the patients left the operating- room in shock, from which some of them recovered." His mor- tality was 55 per cent. The operation which takes the name of Billroth consists in exsecting the diseased part and narrowing the cut end of the stom- ach to the size of the duodenum, and then suturing the intestine into the gap. The point at which the three angles came together so often gave way, with resultant leakage and death, that it re- ceived the name of the "fatal suture angle." Kocher was one of the first to see the defect in the Billroth operation, and made a radical departure from it, suturing the cut end of the stomach completely and inserting the duodenum in an independent posi- tion upon the posterior gastric wall. This is a most excellent method, but it occasionally happens that the traction necessary to adjust the duodenum to the greatly shortened stomach has a ten- dency to obstruct the opening. In seven operations with the Kocher method we had one such case, and were compelled later to do a gastrojejunostomy. In 13 pylorectomies since that time we have closed both the duodenum and the stomach ends, making a , i£Ai)i( Ai. rrui: OF cANc i;h of sroMAfH 167 gastrojejunostomy of the usual type at once. All in all, this has proved the most satisfactory method of ojjeraticjn. Pyloreetomy and partial gastrectomy are rK)t more difficult than similar opera- tions upon the })reast, uterus, or rectum. We have founrl the fol- lowing technic simple and satisfactory: Through a median incision, having its inferior angle at or slightly to the left of tiie umbilicus, the stomach is exposed, and the gastrohepatic omentum tied as far to the left as the gastric artery. This mobilizes the diseased part, and by passing the fingers behind the pylorus in the lesser cavity of the peritoneum, the gastrocolic omentum can be tied off without danger to the superior mesenteric artery. A gauze pad is drawn under the freed part, and a heavy clamp placed on the duodenum with sufficient force to form a groove. A catgut ligature is tightly tied in this groove, and a clamj) placed on the stomach side and the duodenum divided just beyond the ligature. A silk purse-string suture about the duodenum, three-fourth inch below, enables the tied end to be invaginated in a similar manner to the stump of the appendix. The stomach is turned up and to the left, and heavy clamps placed an inch from the growth. In the sulcus formed by the pressure a continuous catgut shoemaker's stitch is placed, using one thread with a needle at each end. The diseased part is then severed, and a continuous silk Gushing suture rolls in the first suture line. In- dependent gastro-enterostomy with the Murphy button is then performed, either upon the anterior or the posterior wall of the stomach, as preferred. The operation is done quickly without loss of blood or opening the stomach. Our experience of 20 pylorectomies with 3 deaths is too small to draw conclusions from, yet in a general way we can say that the results of these radical operations upon the stomach have proved as satisfactory as upon the breast, uterus, or rectum. The mor- tality of IG per cent, cannot be considered excessive when it is noted that this includes those early cases in which misfortune so often attends inexperience. One of our cases of pyloreetomy and. partial gastrectomy lived three years and five months and then died after a short illness from secondary involvement of the liver. 168 WILLI Ail J. iL\YO While we have no other case which hved beyond the three-year limit, we have several favorable cases which maj^ be expected to afford good results. As to complete gastrectomy^, there is not much to be said. The cases are too few and the time too short. We have twice removed practically all the stomach, but in each instance enough of the gastric wall remained to make the jejunal attachment. In neither of these cases was there that sagging of the stomach and dragging down of the esophagus which made the operation of Schlatter, Richardson, and others so successful. In conclusion, let us put this question to the practitioner of medicine: Can you cure a case of cancer of the stomach? If not, why withhold the only known means of effecting such a cure — a surgical operation? Again let us ask, can you diagnosticate can- cer of the stomach early enough for surgical relief? if not, why withhold an exploration, the only certain means of diagnosis? THE PRESENT STATUS OF SURGERY OF THE STOMACH* WILLIAM J. MAYO We have recently gone over the records of somewhat over 900 operations upon the organs contained within the upper abdominal zone. In reviewing the histories of these cases a number of points of interest were noted. One of the most notable features was the exceedingly close relationship which existed between the gall- bladder and bile-passages, the stomach and duodenum, and the pancreas. The association of function and the continuity of mucous surface make the disease of any one organ a menace to the integrity of the others. The gall-bladder and stomach were the organs primarily affected in tlie large majority of cases. The duodenum, from its peculiar situation, seemed to act as a buffer, and was secondarily involved from gall-stone disease and gastric ulcer in about an equal proportion of cases. In the latter instance the ulcer was in the first portion of the duodenum, seemingly due to the acid gastric secretions eroding the mucous membrane at a point above the opening of the common duct, with its alkaline fluids. The pancreas was also usually affected secondary to gall- stone disease or a duodenitis. It is not to be wondered that it has been difficult to arrive at the exact pathologic diagnosis in many of these cases, particularly since this field of work is comparatively new. Abdominal surgery owes much to pelvic surgery, which was first in the field; the ease of diagnosis and the remarkable results of operative procedures *The .\nmial Address in Surgery, read at the annual meeting of the Stiite Medical Society of Wisconsin, Milwaukee, June 4, 1903. Reprinted from "Xorlh- western I>jincet,"' July 1,5. 1903. Hi!) 170 WILLIAM J. MAYO in the pelvis, as contrasted with the fatahty of the surgery of the abdomen, gave an impetus to abdominal work which at once placed it to the front and rendered it the pride of professional achievement. Experience, gained ofttimes as the result of mis- taken zeal in the performance of unnecessary and mutilating operations upon the ovaries and tubes, and later the uterus, has since been turned to good account in the territory of gastro-intes- tinal disease. Pelvic surgery has reached its place — much more conservative and less often resorted to than it was five to ten years ago. The appendix, after much discussion, has also reached its proper position in surgery, and as much can be said of gall-stone disease, the extent and importance of which have only of late come to be appreciated. I believe that the stomach is destined to play a great part in the surgery of the future. Just as the appendix followed pelvic surgery — to be accepted and, in turn, replaced by the diseases of the gall-bladder, so now the stomach, and vrith it the correlated duodenum and pancreas, is to be the center of surgical observation. It is one of the curious phases of modern medicine that surgery leads the way. With the accumulated experience of thousands of postmortem revelations, how little did we know of diseases of the ovaries and tubes, extra-uterine pregnancy, and appendicitis.' Gall-stones were looked upon as innocent autopsy findings, and today the correct interpretation of gastric phenomena is in an equally unsatisfactory condition. Surgery of the stomach has been slow to advance — it had to live down a bad name. The early operations of the Billroth school for cancer, before modern abdominal surgery was properly understood, gave a discouraging mortality, the statistics of which still hamper and embarrass our work. Billroth, with the courage of conviction, attacked malignant disease of the gastric cavity, because then, as now, it was the victim's only chance. Nor was the condition of benign disease much better. Operations for the deadly complications of ulcer, such as perforation and hemorrhage, gave a frightful mortality; no wonder that this work has been looked upon with dread by the physician, and that he has con- PRESENT STATUS OF Sl'IUiEKV OF TIIK STOMATH 171 liiiiicil "tlic sins of medical omission," lia\'iii^ the results of "siir- gieul c-oinmissioM" before his eyes. Gastric surgery has been judged by its results iu late operations for cancer and fatal coniplieations occurring in the course of benign disease. We miglit as well condemn operations for apix-n- dieitis by the results of surgical operation for the general suppu- rative peritonitis which it sometimes caused. As a matter of fact, the stomach is one of the most favoral)le organs for operative attack. Easily accessible to a large extent of its surface, with a splendid blood-supply from four sources, it permits a freedom of work and a certainty of healing not exceeded by any organ in the abdomen, and by no means least in importance is the fact tluit its cavity can be rendered reasonably sterile before operation. The great difficulty to be surmounted is in obtaining a correct diagnosis. We have seen "gastralgia" disappear after the removal of gall-stones, and "dyspepsia" relieved by the excision of the chronically inflamed appendix. The field of gastric surgery is be- ing gradually relieved of a number of these parasites of diagnostic obscurity. It will no longer do to give a blanket diagnosis of "stomach trouble," and the symptoms of the patient, instead of being considered the disease, must be referred to their proper pathologic source. Compare the present knowledge of appendicitis with the comj)lex symptomatic diagnosis of former days — obstruction of the bowels, inflammation of the bowels, peritonitis, etc.; it would be tedious merely to name them. They have gone the way of "pelvic cellulitis." It can be laid down as an axiom that ignor- ance breeds complexity, and one need not be a prophet to fore- tell the disappearance of mo7"e than half of the so-called "diseases of the stomach" of our medical text-books. Upon what shall we base a diagnosis of a surgical lesion of the stomach? Our diagnostic means can be classified into three groups: First, the history of the ])atient; second, the physical outlines of the stomach, including the use of the stomach-tube to develop stagnation or retention of food; and, third, laboratory methods, including test-meals, chemistry and bacteriology, and 172 WILLIAM J. MAYO microscopy. Graham and Millet have the records of nearly 1100 examinations of the gastric contents in conjunction with the history of the patient and the size and position of the stomach and pylorus. Of these patients, over 300 came to operation. The diagnosis, based on the history and physical outlines of the stomach and the rational signs and symptoms, gave a correct diagnosis in over 80 per cent, of the cases. The laboratory examinations, including test-meals and so forth, were of value in only a small percentage of cases, and then only as corroboratory evidence. I do not wish to discourage such tests, but simply to protest against the time spent in waiting until certain chemical phenomena appear before recommending operation. This is especially true in the early diagnosis of cancer. Time and again have we had patients held for weeks waiting for these supposedly valuable diagnostic appearances, until the developing tumor and cachexia discouraged the enthusiastic diagnostician and sent a hopeless patient to the surgeon. Exploratory incision is the only hope for the patient with cancer of the stomach, and a suspicion of this disease should compel the physician to explain, and let the victim and his friends choose between exploration and procrastination. In developing the outlines of the stomach, an ordinary Davidson syringe and a stomach-tube are sufficient. In this way air can be pumped into the stomach and allowed to escape again and again, until a careful mapping out is accomplished. The carbonic acid gas test is often painful, and does not give time enough for thorough work; there are also a number of deaths recorded from the sudden distention rupturing a diseased gastric wall. The use of the stomach-tube is also very valuable in showing the actual results of gastric function — not in test-meals necessarily, but as to whether the ordinary meal is retained too long in the stomach. For example, remnants of the evening meal found in the stomach in the morning on several occasions indicate mechanical obstruction at the pylorus. Inspection for gastric waves, so commonly seen in obstruction at the pylorus, splashing sounds, and so forth are also of value. PUKSKNT ST.\T( S OF St U(;ERY OF TIIK .STO.MA< II 17.'} It is not my (Irsirc to j^o into the (|iiosti()n of differential dia^'- nosis, l)iil I do wish to call the at tcntion ol" the ^'<'neral practitioner to the fact tliat the n^fiiieineiils of teelinieal diai^iiosis are often useless and occasionally harmful in causing delay, and that the sensible practitioner, with the few siiuple means at his command, is perfectly capable of arriving at a reasonable diagnosis, and will at least be able to direct the majority of his patients needing surgical treatment to the surgeon in time to be benefited. There are two important groups of surgical lesions of the stomach: (1) Those of benign origin and usually inflammatory in character, with ulcer as the type. ('-2) Cancer of the stomach. Ulcer of the stomach has been heretofore studied from its complications, such as perforation and hemorrhage, and from autopsy findings. This has been unfortunate, as it has exaggerated fatal secondary phenomena and has thrown but little light upon the condition during the period of chronic disability, in which surgery would be a well-planned attempt to cure, rather than a last resort. Acute ulcer is usually diagnosticated. The fre- quency of sudden perforation and hemorrhage in this variety of lesions has enabled more accurate understanding of the subject; but we must not take these symptoms to be the standard for chronic gastric ulcer. In this lies the frequent failure to recognize chronic ulceration. Pain is tlie most characteristic symptom of the old ulcer, and the nearer the lesion to the pylorus, the more cramp-like the pain, while in the duodenum it may almost exactly simulate gall-stone colic. Chronic ulcer causes slow and painful digestion; the pa- tient is careful of his diet, and has usually one of his own, the value of which experience has taught him. The symptoms are not steady, and days or weeks of distress are followed by intervals, more or less j)rolonged, of comparative comfort. The pain symptom is often variable, even during the period of one day, one or more meals giving no trouble, and the next one, perhaps, causing much suffering. Unlike acute ulcer, hemorrhage and vom- iting are not common symptoms; the latter, a late phenomenon, is usuallv due to secondarv contraction and obstruction at the 174 WILLIAM J. MAYO pylorus. Acute ulcer, and occasionally the chronic ulcer, is usu- ally cured by rest; but the latter gets its name because it has not been cured, and, like chronic appendicitis and gall-stones, the same case may be medically cured a score or more of times. Leube puts the death-rate of ulcer at 25 per cent., and says that, if curable, four or five weeks should be the limit of required time. In a very recent study of 500 cases, treated at the London Hospital between 1897 and 1902, 211 had had previous similar attacks; in other words, were known to have had ulcer, wnth intervals of apparent cure. In the remainder the symptoms had been more or less continuous — 18 per cent, died and 42 per cent, wxre not cured at the time of discharge. A total of 60 per cent, died or were not cured, and of the 40 per cent, of patients supposed to be cured, who can tell their future course? .^he large majority of cases of chronic ulcer of the stomach are surgical, and, after a reasonable trial of medical treatment, should be so considered. The function of the stomach is interfered with mechanically, first, by obstruction at the outlet, which prevents proper egress of the food; second, by disease of the pyloric portion, which interferes with its muscular action; and, third, disease of the fundus of the stomach, which prevents proper reservoir function. Pyloric obstruction gives unmistakable evidence of its presence, by gastric dilatation, stagnation, and retention of food causing fermentation, late vomiting, and emaciation. These cases are seen in all stages, from the slight temporary interference with digestion to the most marked degree of disability. The condition can be aptly compared to a valvular heart lesion; spells of dila- tation alternating with compensation through hypertrophy give periods of comparative health after a more or less prolonged train of gastric insufficiency. Benign obstructions are usually the late results of the healing of a gastric ulcer, and many times the symptoms are such as to lead to the diagnosis of cancer. I have no doubt many victims are allowed to die under the impression of a necessarily fatal ending. It must not be forgotten that, like cancer, chronic ulcer is a PRESENT STATIS f)I' SlIUMOKV <^)F THE STO.MA(H 1 I.) cliseusc ol' adiilL life, in lliis rcsijccL (liU'eriiig from llic aculo form, which is moMt common in adolescent females. Drainage oper- ations give m:ir\('loiis relief in hcnign obstructions at the pylorus; a patient at liie xcrge of starvation in a few weeks becomes a picture of lieallli. (iastro-enteroslomy has been the operation of choice, but in certain cases the newer pyloroplastic method of Mikulicz or the gastroduodenostomy of Finney is also indicated, and gives in some respects even more desirable results. The second group of cases is that in which ulcer exists in the muscular pyloric region, preventing the pro|)er mixing of the food material. In over ^OO o])erations for benign disease of the stomach we have found ulcers capable of giving symptoms requir- ing operation, nearly always in that segment of the stomach lying to the right of the cardiac orifice, and having the lesser curvature as its sui)erior border. Its inferior border on the greater curvature does not extend so great a distance to the left of the pylorus. The fundus of the stomach has but feeble muscular action; it slowly compresses the food into the pyloric antrum, and the latter, by a piston-like action, forces the food backward into the fundus, as well as forward into the duodenum. Ulcer in the pyloric region gives rise to great pain and distress from the so-called "pyloric spasm." The latter is not confined to the pyloric sj)hinc- ter, but to all or any part of this muscular region. The symp- toms are pain and indigestion, gas, hypersecretion, and so forth — the "pyloric syndrome" of Hartmann. Many of these cases heal in time, and the resulting cicatrix produces the obstructions which were discussed in group one. The proper surgical treat- ment of such cases is not settled, and no one method will api)ly to all cases. The dilatation is not extreme, as the obstruction is due to spasm and is essentially temporary in character. Gastro- enterostomy on the cardiac side, and usually at a point directly opposite the esophageal opening, gives splendid temporary results; but with the healing of the ulcer, which quickly takes ])lace as soon as the food is prevented from passing through the ulcerated area, the spasm relaxes and the pyloric region begins to func- tionate normally. The gastro-intestinal fistula often contracts, or 176 WILLIAM J. MAYO the double stomach drainage of itself gives future trouble. The plastic operations about the pylorus, so useful in benign strictures, are probably not efficient, since, no matter how large the outlet, the food must still pass through the ulcerated area before reaching the pylorus, and that obstruction has no influence in the produc- tion of ulcer is shown by the typical examples in the duodenum. Excision of the ulcer itself is sometimes feasible, but, unfor- tunately, the lesion is often multiple, and may be hard to detect. Rodman suggests that, in certain cases, excision of the pyloric end of the stomach should be the operation of choice, as it would not only permanently cure the condition, but also prevent the possibility of secondary malignant degeneration, which has oc- curred not infrequently. The third group, in which the reservoir function of the stomach is interfered with, is usually the result of extensive ulceration and cicatricial stenosis in the body of the stomach, causing hour-glass contraction. It is interesting to note that hour-glass stenosis may be multiple, and that a contraction of the pylorus also very commonly exists, so that a combination of gastrogastrostomy and gastro-enterostomy is usually indicated. A few words in regard to some dilatations of the stomach not of organic origin, such as the so-called atonic dilatations often found in neurasthenic individuals, and without the pyloric syn- drome of Hartmann. In these cases there is no retention and little stagnation of food. As a rule, these patients are not bene- fited by operation. This is also true of gastroptosis, which we have found to be present in over half the cases of movable kidney. Relaxed conditions in the neurasthenic state are not often per- manently benefited by surgical operation. It is one of the mis- fortunes of surgical progress that neurasthenic symptoms are too often mistaken for organic disease. We have but to look back on the discredit thrown upon surgery by the mutilating operations upon the pelvic organs of women to impel us to go slow in that numerous class of neurasthenic stomachs, and before we operate for ulcer let us be sure the lesion exists elsewhere than in the mind of the patient. PRESENT STATLS OK SURGERY OF TIIJ-: STOMAfll 177 In conclusion, I wish to speak hriefly in regard to cancer of the stoniacli. Karly operation is a prerecjuisile, and diaf^nostic exploratory incision is necessary. We lia\e operated upon 113 cancers of the stoniacli, of wliicli '■27 were radical extirpations; 5 died within a month, and one later from another cause — too early to know the ultimate result. Of the 21 who recovered, the average length of life was over a year. It is suri)rising how few failed to live twelve months or more. One lived three years and seven months, several are alive now after more than two years. I am convinced that cancer of the stomach will in five years give as good ultimate results after excision as operations on the breast. Sixty per cent, of cancers are located in the pyloric portion — that is, in the movable part of the stomach. The lymphatic arrange- ments are the same as the vascular, and the dome of the stomach is isolated from this portion, having a different vascular and lym- phatic connection. If all of the lesser curvature with the corre- sponding lesser omentum, and all of the body and greater curva- ture to the left gastro-epiploic artery, be removed the results in cancer of the pylorus should nearly equal what might be expected after complete gastrectomy. The remaining portion of the stom- ach enables intestinal anastomosis to be made with considerable ease, and the gastric pouch rapidly enlarges to assume the function of the stomach. VOL. I — \i A REVIEW OF 303 OPERATIONS UPON THE STOMACH AND FIRST PORTION OF THE DUODENUM* WITH TABULATED REPORT OF 313 OPERATED CASES WILLIAM J. MAYO Functionally, the small bowel begins at the entrance of the common duct of the liver and pancreas, a point which marks the primitive division between the foregut and the midgut (Hunt- ington) . The first portion of the duodenum may be said to be the vestibule of the intestinal tract, and its diseases resemble in char- acter those of the stomach, rather than the intestine. In the large majority of instances lesions at this point cannot be diagnosticated accurately from similar diseases in the stomach which may be due to the same causes. For this reason I have associated all the cases of this description into a single group for the purpose of study. Total number of cases, 303. Of these, 286 are taken from the records of St. Mary's Hospital, and the remainder are from the records of the Minnesota State Hospitals for the Insane at Roches- ter and St. Peter. The average age was forty-two; males, 42 per cent.; females, 58 per cent. , Duodenum: 26 cases, 2 deaths — 7.6 per cent. Lesions of the first portion of the duodenum can be divided into two groups: first, those due to ulcer, and, second, those associated with gall- bladder disease. Ulcer limited to the duodenum was found 11 times — 1 acute perforating, 2 chronic perforating protected by adhesions, 5 active, * Read before the Philadelphia Academj' of Surgery, May 11, 1903. Reprinted from "Annals of Surgery," July, 1903. * 178 OPERATIONS UPON STOMACH AM) DLODENUM 1 7f) and 3 cicatricial contraction with ohstructivc symptoms. Two died after operation — one from pneumonia following excision of the ulcer, one from exhaustion after gastro-enterostomy. In 3 cases the signs and symptoms were not to be distinguished from gall-stone disease, and the operation was undertaken with the supposition that the trouble was in the gall-l)ladder. Five times ulcers existed upon both the duodenum and the stomach. Of the 16 patients in this group, 14 were males. The duodenum was fre- quently associated with gall-stone disease, and usually secondary' to it; but in 11 cases the duodenum was the prominent feature. Five were due to gall-stone perforation, requiring intestinal suture. In 3 of these the gall-bladder was completely separated function- ally from the bile-tract, and had become an appendage to the duodenum. Four times crippling adhesions to the gall-bladder, but without stones or evidence of cholecystitis, were encountered, requiring dissection to loosen — a periduodenitis of unknown origin. In one case an inflammation of an accessory lobe of the pancreas was the cause of dense adhesions. All but one of the cases in which the gall-bladder was involved occurred in females. There were no deaths in this group. In no instance was the duodenum the seat of primary malignant disease, and in but two was there any evidence of extension from pjdoric cancer, and then it was not marked. In two patients the diagnosis of lesions originating in the duodenum was made previous to operation. The differen- tiating features of these cases were good appetite, delayed pain, general absence of vomiting, and in only one was there hemat- emesis. In two cases there was evidence of blood in the stool. Otherwise the signs and symptoms were similar to lesions of the stomach or gall-bladder, and, even in the light of operative invest- igation, points of differentiation did not become evident. Our experience leads us to believe that surgical diseases of the duode- num are much more frequent than has been supposed. The subject of perforating and bleeding ulcers of the stomach has been so thoroughly dealt wnth by Keen and Foote, Weir, Rob- son, Rodman, and Andrews, and lesions of a similar character in the duodenum by \Yeir, Murphy, and others, that it seems un- 180 WILLIAM J. MAYO necessary to dwell upon the few cases whicli have occurred in this series. The present discussion will be confined to the results ob- tained, and some practical deductions based upon two large classes of cases: First, gastric ulcer and associated causes of serious dis- turbance; second, cancer of the stomach. Stomach: 277 cases, 28 deaths — 10.1 per cent. In the benign group there are 168 operated cases with 11 deaths (6.5 per cent.), and nearly all these operations were performed for chronic ulcer and its late cicatricial results. Included in this class are all the non-malignant obstructions. The conditions calling for operation were gastric pain, with or without acute exacerbations, repeated hemorrhages, emaciation from inability to retain sufficient nourish- ment. In a few cases dilatation due to known or unknown cause gave mechanical reasons for interference. ^ There is no doubt that perverted stomach secretion is the most important manifestation in the majority of cases. This is shown by the almost constant association of excessive secretion in ulcer, and the fact that similar ulcers in the duodenum are in that part of the intestine not protected by the alkaline juices poured in through the common duct. In this connection most interesting informa- tion is furnished by those reported cases in which a typical peptic ulcer has developed in the jejunum, immediately below a gastro- jejunostomy made for the purpose of drainage, the lesion in the jejunum in every particular resembling the original ulcer for which the gastro-enterostomy was performed. In operating upon cases of this description the excessive amount of gastric secretion is con- stantly in evidence, and the results of drainage operations in re- lieving the distress and healing the ulcer bear out the importance of this view of the case. Attempts to classify ulcers of the stomach have been based largely upon postmortem experience and accidental complica- tions, such as perforation and hemorrhage. Such classifica- tions tend to exaggerate the importance of fatal complications, which render surgery a desperate resource rather than a well- planned effort at cure. Further surgical observations are necessary to clarify the con- OPERATIONS UPON STOMACH AM) IHODKNTM ISl fusion which siiitoiiikIs ^aslric ulcer. In ;iMcui|)l iii<^ to ^Toup our ojx'ratcd cases, \vc found tlial llicrc were such wide \ariations in the coiiditions present Ihat no orderly chissilicalion couhl bo made on a purely clinical basis. In a general way, the following answered the purpose most satisfactorily: (1) Round and fissure ulcers: (a) Acute; (6) chronic. They have the distinguishing feature that there is but little thickening about the base of the ulcer. Many amount to little more than a fissure, and are closely associated with group (2). (2) Mucous erosions; a condition which nuist be accepted with caution. (.'3) Chronic ulcer with a thickened base, usually irregular in form, and probably an extensive variety of the chronic round ulcer (4) Benign obstructions without regard to cause, although usually of infiamuiatory origin. In our experience at the operating table the last two varieties are most frequently met with. The acute round ulcer of Cruveil- hier occurs by preference in the chlorotic type of adolescent fe- males, and usually responds to medical treatment. Operation is most often called for in the acute cases by that peculiar perforation so graphically portraj^ed by Rokitansky — "cut out by a punch"; or by severe hemorrhage from the stomach. Chronic round ulcer and fissure ulcer do not often lead to harmful cicatricial contraction, on account of their small size. Near the pylorus they may be the starting-point for a band-like stenosis incircling the pyloric ring. Chronic round ulcer is usually found in adults, and in our experience has been more frequent in females. It would seem that there is little difference between the chronic round ulcer and the chronic cicatricial ulcer, except that as the outer coats are involved the extent of ulceration increases and loses its characteristic round or oval form, while usually a healing process is apparent in some part of its extent. A subvariety of this group is the "pore-like" ulcer described by Murchison, which is met with more often in adults and gives rise to grave hemorrhages, and yet is so minute that it is difficult to locate, even at postmortem. The mucous "erosion," limited to a small area or several such patches, was seen in a few 182 WILLIAM J. MAYO instances. The large "mucous erosion" described by Dieulafoy as giving rise to alarming hemorrhages was not met with. I am unable to say just how much importance is to be attached to the surface erosion of limited extent. In the first place, the detection is difficult. The whole question of the surgical exploration for round ulcers and erosions is one surrounded with difficulty and uncertainty. There are usually no external manifestations which lead to location of the lesion, and the only way a diagnosis can be established is to open the stomach and, with a short, wide speculum explore the interior. The margin of the instrument may and fre- quently does produce a traumatism to the superficial mucous layers, and the result is very like the pathologic, erosion. We have seen undoubted and typical examples covered with a membranous film of mucous character which, when brushed off, allows the nature of the trouble to become apparent. The chief obstacle to accurate diagnosis lies in the surgical indications. Round ulcers and ero- sions are often multiple, and, as a rule, do not cause cicatricial contraction at the pylorus. Clinicar experience has demonstrated that drainage is the best method of surgical treatment with which we are acquainted, therefore an exploration, however attractive to the surgeon, is often not completed; but the surgical indications are fulfiled by some form of gastro-intestinal operation, and the diagnosis remains unproved. The surgeon hesitates to expose the patient to even a slight risk, for purely diagnostic purpose. The old adage, "a good prognosis is better than a good diagnosis," leads to operations based upon symptoms. If round ulcer is found, ex- cision is the proper course; but there is always the chance that the ulcer excised is not the only one, and that others may exist unde- tected or in an inaccessible situation. We may well ask ourselves in such cases. Does an ulcer exist? Usually, we can answer yes, and base the diagnosis upon such symp- toms as would establish a medical diagnosis. Clinically, these cases come to us after medical treatment has failed utterly, and either the diagnosis is unquestioned or there is secondary inter- ference with motility, resulting in retardation or retention and gastric dilatation, giving mechanical reasons for interference. Fig. 12. — Showing line for incision in case of ulcer of the stomach. OPKIfATIOXS UPON' STOMACH AND DlOlHCNf.M 183 The theory of jnloric si)asin is most interesting, ulthongh it slioiild be called a hypothesis rather than a definite condition. I have examined the pylorus in over 300 eases at the pperating table with a view to estal)lishing a normal pylorus under anesthesia. Usually, tlie normal pylorus in the anesthetized i)atient will allow the thumb and the forefinger to meet nicely about the caliber of a silver dime, and under some conditions of deep anesthesia it may be found di- lated to the diameter of a silver twenty-five-cent piece. I am satisfied, however, that spasm of the whole or some part of the pyloric portion of the stomach may and often does take place, and that it is one of the causes of the retention of the excessive secre- tions and distress; but I am by no means sure that it is confined to the pyloric sphincter. The so-called "chronic ulcer" of Robson has a thickened base, and is frequently of large size and irregular outline, in this respect differing from the chronic round and fissure ulcer, in which there is but little new tissue deposit about the ulcer. Does the round ulcer lead to the chronic cicatricial ulcer? It is probable that the differ- ence is merely one of degree, although the fact that the latter is much more common in males is rather against this hypothesis. The majority of operations were for thick-based chronic ulcer of the stomach or its late results, and these cases were very satis- factory, the irregular thickened patch of stomach or duodenal wall often locating the process with exactitude. As a rule, the ulcer was located near the lesser curvature, and not infrequently at the pylorus. The posterior wall was affected more often than the anterior, if only one surface was involved. On the duodenum the anterior wall was more often the seat of ulceration. The youngest patient was a girl of seventeen and the oldest a man of sixty-four. In 60 per cent, of our malignant cases a previous history of ulcer was obtained. In 2 cases malignant degeneration of the margin of a chronic gastric ulcer was demonstrated — certainly a strong ar- gument for the excision of such ulcers, when possible. We found conditions favorable for excision of ulcer in only 3 cases. On 6 occasions we either excised or turned an ulcer in bv suture, in com- 184 WILLIAM J. MAYO bination with pyloroplasty or gastroduodenostomy. In 2 of these cases three-fourths of the pylorus was excised and closed by suture. Lund has pointed out that "sentinel" enlarged lymph-nodes in either the lesser or greater omenta may aid the surgeon in locat- ing the ulcer. We haA'e found this a valuable observation. In all the ulcers of every description which we examined the upper two inches of the duodenum, pylorus, pyloric antrum, and that part of the stomach lying to the right of a line dra^m downward from the esophagus was the seat of disease, and in only a few in- stances of extensive hour-glass contraction did the ulcer extend to the left of this line. In handling the stomach during operation, limited contraction of the wall could often be noticed in the pyloric third, but not toward the cardiac end. Cannon's experiments are very interesting in this respect. He demonstrated with bismuth and the a:-ray that the fundus of the stomach did not contract strongly, but that the pyloric portion, by a backward action, kept up a current in the fundus. Ulcers occur in all parts of the stom- ach, but in the cardiac end it is a question if they are often the cause of chronic symptoms calling for operation. Twelve chronic dilatations without ulcer or obstruction were operated upon. In all the cases the wall of the stomach was of normal or increased thickness, indicating that an obstruction existed, either from a high -lying but non-stenosed pylorus, or beyond the py- lorus. In 1895 I reported several cases of interference with free gastric drainage by valve formation, due to a short gastrohepatic omentum holding the pylorus high, the body of the stomach sagging sharply downward. More than half of the cases were of this descrip- tion. In a few instances the medical diagnosis was extreme atonic dilatation; but even in these cases there was no great thinning of the gastric wall. We have not considered simple gastroptosis suffi- cient cause for operation, but in a few cases exploration revealed this condition, and in all the cases the wall of the stomach was either of normal thickness or thinner than normal. In three of these cases shortening of the gastrohepatic ligament was done after the method of Beyea. Cancer of the stomach: 109 cases, 17 deaths — 15.6 per cent. Robson. Moynihan. Mayo. Hartmann Mikulicz Fig. 13. — Lines of incision practised by different surgeons in the removal of cancer of the stomach. OPERATIONS ri'OV STOMACH AM) DrOOKNI \I 1 S.> Late diagnosis and cachexia make the aspect of this group dis- couraging. Palliative operations predominate, with considerable immediate mortality and no great prolongation of life. The hope of the future lies in early exi)l()ralory incision, and the necessity for this depends upon clinical ohservation rather than laboratory methods, which too often only become valuable when the extent of the disease is beyond cure. Given a patient of middle or later life, who begins to lose flesh and appetite and suffer from indiges- tion without ai)parent cause, the possibility of cancer should be considered; and if the source of the symptoms cannot be shown within a few weeks, the situation should be exy)lained to the pa- tient, and the choice between exploration and procrastination allowed him. AVhen we consider that early operation is the only hope, we may not wait on our own responsibility. The public in this way will soon become educated and cures will be more fre- quent. Gastrojejunostomy for malignant disease, in our hands, has had an increasing mortality, due to the fact that the better cases are selected for gastrectomy, and the late hopeless obstruc- tions are given the meager benefits of gastro-jejunostomy — 34 cases, 10 deaths — 30 per cent. Is there a hopeful outlook for cancer of the stomach? We know of the prime necessity for early operation: it now remains to demonstrate how the procedure can be made more effective. In a general way, the lymphatics of the stomach lie in three groups: First, the lesser curvature and lesser omentum; second, along the greater curvature and the gastrocolic omentum; third, in thegastro- splenic omentum. The main lymphatic channels follow the direc- tion of the blood-vessels to the deep glands about the celiac axis. The dome of the stomach, as pointed out by Robson, has no main lymphatic channels and few lymphatic glands. If all the stomach excepting this portion be excised, the remaining part will be ade- quately nourished on the right side by cardiac branches derived from the gastric artery, which joins the stomach at a point from one to one and one-half inches below the esophagus. On the left, the vasa brevia, given off from the splenic artery distal to the origin of the left gastro-epiploic vessel, — a distance of four and one-half to 186 WILLIAM J. MAYO eight inches from the esophagus, — give an adequate blood-supply. These vessels anastomose with the inferior phrenic vessels. There- fore, excision of all the stomach lying below and to the right of a line drawn between the gastric artery and the left gastro-epiploic vessel is the logical operation. The advantage of this line of sec- tion is obvious. All the main lymphatic connections are removed at the primary operation. We know clinically that the remaining portion of the stomach is seldom involved unless the primary lesion is at the cardiac orifice, and the retention of the dome of the stomach enables comparatively easy intestinal anastomosis. One reason that only from 5 to 8 per cent, of gastric cancers have been cured by extirpation lies in the fact that a part of the organ has been retained in which the vascular and lymphatic connections with the diseased area have not only been close, but direct. In the dome of the stomach, the lymph-current is feeble through small vessels, and, most important of all, is in the other direction. IVIiku- licz has already called attention to the necessity of removing the whole of the lesser curvature, with its gastrohepatic omentum, and has done much to elucidate the question of lymphatic infection by showing that in 20 cases of gastric cancer only 1 was completely free from lymphatic involvement, although, in a total of 189 glands examined, 110 were found to be without contamination. In making this radical operation we have proceeded as follows: First, ligate the gastrohepatic omentum from the pylorus to the gastric artery, which is tied. The section is made as close to the liver as possible, and includes nearly the whole of the lesser omentum. This mobilizes the pyloric end of the stomach, which is drawn down and out. Second, with the fingers in the lesser cavity of the peritoneum, the gastrocolic omentum is ligated at a safe distance. The duodenum, on the one side, and the pylorus, on the other, are doubly clamped and divided between with the cautery knife. A purse-string suture of silk is placed around the duodenum, three-fourths of an inch below the divided end, and, after suturing with catgut through the cauterized area, the stump is inverted and the purse-string suture drawn tight. This disposes of the duodenum permanently. Third, ligation of the gastrocolic Fig. 14. — The completed operation for cancer of the stomach. (M'I:K AlIONS ll'ON" STOMACH \\l) IHODKNfM 1 S7 oiueiiLuin lo u |)<>iiiL near Llii^ orij^ln of Llic IcfL gu.slro-epiploic artery, which is lied. Fourth, a groove is made by heavy pressure forceps, separating the doinc fiom the hahiricc of the sloiiiach, and with catgut on two needles, a shoemaker stiteh in the i)ressure furrow renders section with the actual cautery bloodless and avoids opening the portion of the stomach to be retained. This line of suture is turned in by a continuous silk Cushing suture, sup- ported occasionally by an independent Ilalsted stitch of the same material. In this step of the operation we sometimes use the Kocher clamp and suture each layer separately. P'ifth, gastro- jejunostomy between the gastric pouch, which is just about large enough for the purpose, and the jejunum. Sixth, entcro-anas- tomosis between the two limbs of jejunum, short circuiting the biliary and pancreatic secretions as nearly as possible at the same level as the origin of the jejunum. It took two deaths to teach us the value of this mana?uver. The deaths were not from regurgi- tant vomiting, but when the anastomosis was effected in some cases, the intestine was sharply bent at the site of union, being drawn upward and to the left in such a manner as to leave from 14 to 16 inches of jejunum hanging upon the anastomosed area, a situation in which peristalsis does not materially aid in onward flow of the biliary and pancreatic secretions. The proximal loop becomes distended with these juices to the level of the anastomosis, giving a traction weight of a column of fluid the diameter of the distended intestine. In one patient on the fifth day and one on the ninth day union suddenly gave way entirely, or in part, when the patients were apparently doing well. This did not happen in every case — two out of eight only; but in at least half of the cases the bad mechanics of the situation was evident on inspection. Seventh, the remains of the gastrocolic omentum are attached to the poste- rior wall and the abdomen closed. This operation should give all the benefits of complete gastrectomy in pyloric cancer. (I find that Moynihan has recommended and practised a similar pro- cedure.) In view of the splendid work of Hartmann and Cuneo, it is a question whether the operation outlined should be the routine 188 WILLIAM J. MAYO one, or for exceptional cases only. That the whole of the lesser curvature, with the glands in the corresponding portion of the lesser omentum, should be removed is the copelusion of all surgeons of large experience; but the advantage of removing the major part of the greater curvature is open to debate. Cuneo demon- strated that the lymph-current along the greater curvature was from the left to the right, and that in pyloric cancer not only is there comparatively little tendency to lymphatic involvement, but that it is confined to the glands in the immediate vicinity of the growth, and does not extend to the left of the pyloric portion. Hartmann, therefore, bases his line of section upon this fact, and removes all the lesser curvature and saves as much as possible of the greater curvature. We have several times made an operation very similar to that described by Hartmann, as it is certainly much easier than the one which we have outlined, and, as the mechanics of the anastomosis are better, entero-anastomosis is unnecessary. Occasionally, however, growths or glands are found to the left along the greater curvature. It may be said that such cases are inoperable, yet we have had two such patients live beyond a year. In the 8 cases operated upon by the radical method given above there were 3 deaths, while there were but 2 deaths in the 18 remain- ing cases operated by various methods from simple pylorectomy to the operation of Hartmann. The former group comprises only a small number of the worst cases, and some of the deaths might have been avoided by better technic. Be this as it may, some form of radical extirpation has been the only reasonably satisfactory op- eration which we have performed for cancer of the stomach : 27 cases, 5 deaths — 18.5 per cent.* One patient lived three years and seven months before recurrence. Several are alive and well over two years, and the general average has been over a year. It is sur- prising how few of those recovering from the operation have failed to live a year or more. It may not be out of place to discuss briefly the merits of the * Since completing this paper, one case died after five weeks from abscess of the lung, making 6 deaths — 22.5 per cent. OI'KUATIONS UPON STOMACH AM) l)\ OUKMM 189 llirce chief inclliods of improviiif^ gaslric drainage, namely, py- loroplasty, gastro-cnterostoniy, and gastroduodenostomy. Nineteen cases were subjected to tlie pyloroplasty of Ilcineke- Miknlicz; of these came to secondary gastrojejunostomy through failure of the operation adequately to drain the stomach. The remaining cases are well. There were no deaths. The opening can be made of sufficient size, but the increase in caliber is not in the line of gravity drainage, or, at least, the enlargement of the opening is as much above the pylorus as below it, and the greatly dilated stomach, with its overstretched and degenerated muscu- lature, is unable to elevate the food, and the stagnation is not en- tirely relieved. Again, in the 6 reoperated cases, the pylorus was found adherent at a high level, due to the abstinence of food and other causes of downward traction during the healing process. In 3 cases we fastened the pylorus, after plastic operation, to the neighborhood of the umbilicus by suture, to secure a low point. These patients have remained well; but as we were also careful to choose only moderate dilatations, the value of the manoeuver is uncertain. Gastro-jejunostomy was done 168 times, divided as follows: Gastrojejunostomy, ISl; gastroduodenostomy after Finney, 26; independent gastrojejunostomies in connection with pylorectomy and gastrectomy, 22. Of the 121 cases of gastrojejunostomy made purely for drainage purposes, there were 17 deaths. The per- centage of mortality in the benign cases was 8 per cent.; in the malignant, 30 per cent., the great mortality of the latter being due to the choice of favorable cases for radical operation, the hope- lessly advanced and cachectic coming to gastro-jejunostomy, and, could the condition have been known beforehand, an operation would not have been undertaken in some of these cases. Gastrojejunostomy for benign obstruction at the j^ylorus is one of the most satisfactory operations with which we are ac- quainted. It rapidly drains from the lowest point, and if the obstruction at the pylorus is permanent, the new opening does not contract materially. Again, if the opening be made at the bottom of the stomach pouch, at or near the greater curvature, 190 WILLIAM J. M^O regurgitant vomiting will not take place, and entero-anastomosis is unnecessary, providing either the Murphy button or Robson bone bobbin be used mechanically to maintain an opening during the early critical period. In some instances a feeling of distention or vomiting after operation may take place, and, under such cir- cumstances, we promptly direct gentle stomach lavage. We now use the posterior suture operation over the bone bobbin for benign obstructions, and the Murphy button for malignant disease, and in the latter instance the anterior method. However, as between the suture and the Murphy button and the anterior and posterior operation, we have been unable to see any marked difference in results beyond the occasional retention of the button in the stom- ach, which seems to be of no practical importance. During the recent visit of Professor Mikulicz to this country (May, 1903), he was kind enough to do a posterior gastro-jejunos- tomy in our clinic by a method which I believe is greatly superior to the one we had been in the habit of doing. It avoids the possi- bility of angulation, since it does not form a loop, with its attendant dangers. The operation, as performed, depends on two simple principles: First, the origin of the jejunum lies above the greater curvature of the stomach. After opening the transverse mesocolon and fastening it to the posterior wall of the stomach, the upper three or four inches of the jejunum lie directly in contact with the gastric wall, hanging perpendicularly with its free border (opposite the mesentery), facing the wall of the stomach. Second, by making a transverse incision in the jejunum three or four inches from its origin and an incision close to the greater curvature of the stomach, a suture anastomosis is made in which the stomach is drained at the lowest point without the possibility of kinking the intestine. The whole trouble has been that in making a longitudinal incision in the intestine it was necessary to form the misfortune-breeding loop. The scheme of the operation is much the same as that used by Czerny. The good mechanics of the procedure has been es- pecially dwelt upon by Peterson, of the Heidelberg clinic. Gastrojejunostomy, if the pylorus be unobstructed, is far from satisfactory. In a paper read before the American Surgical Asso- OPERATIONS UPON STOMACH AND DUODENUM 191 ciulion, June, lOO'^, I rcporlcd 1 cases in which contraction at the site of the anastomosis took phace, and we have reoperated upon 4 simihir cases since that time. In (5 of these cases we did a secondary entero-anastomosis between the hmbs of the loop. Four times the entero-anastomosis was effected with the Murphy })utton, and two of these patients died from sudden separation of the anastomosed area at the end of the first week. Death did not take place in two suture operations. In all these cases the proximal limb of jejunum from the point of anastomosis to its origin looked enlarged and thickened, a condition that might be called water-logged, and in marked contrast to the bowel immediately distal to the anas- tomosis. In this condition of the afferent loop lay the reason for the failure of the plastic union after the button, and merely illus- trates the well-known danger of setting up pressure necrosis in damaged tissues. Primary entero-anastomosis with the button is safe, but not so secondary operations. If the obstruction at the pylorus is com})lete, this condition of the jejunum above the gastro- intestinal anastomosis is not found. A large number of cases of benign affections of the stomach without pyloric stenosis require operation. This is particularly true in ulcer, and relapse after this operation has been frequent. Our observations would seem to show the following course of events. After the operation there is at least temporary healing of the ulcer. The pylorus begins to functionate normally, and the unnecessary gastro-intestinal fistula contracts. There is renewed irritation from retained se- cretions, followed by reopening of the ulcer, return of pyloric spasm, and failure of the operation to effect a permanent cure. In some cases the double stomach drainage seems to give rise to unpleasant symptoms without contraction of the fistula. In 28 cases of gastrojejunostomy with open pylorus, 8 came to secondary operation from contraction of the gastro-intestinal opening, while in all cases with permanent ol>struction at the pylorus there were no cases of secondary operation from this cause. This has also been the experience of Ochsner, who points out the fact that if re- lapse takes place, symptoms will arise within four months. To obviate this sequela, in one case, at the primary o|)eration, we 192 WILLIAM J. MAYO divided the pylorus and closed both the gastric and duodenal ends by suture, thus creating the favorable condition of complete ob- struction. Once we sutured the pylorus high up under the liver, causing valve formation, as first suggested by Cordier. Once we placed a circular purse-string suture about the pylorus, closing sufficiently tight to obstruct the opening. This idea was adopted from Dawbarn. I may say that all of the methods proved satis- factory; but there was the grave objection of too much operating for a benign condition, and it introduced unnecessary elements of danger. In June, 1902, Finney introduced his method of so-called pyloroplasty, but which is in reality a gastroduodenostomy. The opening is downward in the line of gravity, and in most of the suit- able cases for this operation the gastric dilatation is not extreme. In two cases of rather extensive dilatation and pouching we com- bined with it shortening of the gastrohepatic ligament, as described by Beyea. The operation of Finney is especially adapted to those cases in which there is little disease about the pylorus. It enables careful examination of the pyloric end of the stomach, and excision of a neighboring ulcer can be easily combined with it. We had two such cases. It is less suitable if there be extensive involve- ment of the pylorus; but it is in just this class of cases that gastro- jejunostomy is at its best. The question to be settled by further experience is, whether the operation of Finney will as rapidly cure active ulcer of the stomach as gastrojejunostomy. In the latter operation the drainage is from the cardiac end to the left of the muscular pyloric portion; while even if the pylorus be made of ample size by the Finney procedure, the food and secretions must pass the ulcer site before it leaves the stomach, and we know that obstructions are not at all necessary to the formation of ulcer, as they exist beyond the pylorus in the duodenum. In 26 cases oper- ated upon by the method of Finney we had one death, and that from avoidable cause. Were it not for the mortality, resection of the muscular pyloric portion of the stomach would be indicated in gastric ulcer, as in this way the ulcer-bearing area would be per- manently disposed of and an absolute cure insured. This pro- (Jl'KUATIONS t |•(^^ STOMACH AM) DCODENUM D.'J <'('(lun' was firsl .siiH7 of tlio mucous nuMubraiie only; socoikI, roinid ami (issun- ulcers, which ponclrafc thn)U DUODENUM 190 trie toluiiy is a .somcwliat rare condition, wliidi, however, exists in a lar^aT proportion of cases than has been supposed. Heretofore, only tlie more severe grades have been diagnosed, and the larger percentage of these cases have died. Frankl-IIochwart found 10 deaths out of 11 cases, and Albu found 31 out of 40. There are many mild cases in connection Avitii the more .severe grades of gastric dihitation siiown by muscular cramps, prickling in the extremities, but without typical contractions, and these .symptoms should call attention to the necessity of immediate operation for drainage purposes. Of laboratory methods of diagnosis, excess of hydrochloric acid is one which has proved to be of value, and this only as corroboratory evidence. So much for chronic ulcer without complications. The large majority of cases develop changes in the size or position of the stomach as the result of complicating cicatrices, adhesions, or obstructions. In fact, many cases do not have symptoms upon which a diagnosis can be based before the development of these changes. Of all these conditions, narrowing or fixation of the pylorus is the most important. An ulcer in the vicinity of the pylorus, by contraction or spasm, may mechanically obstruct the progress of the food, with resulting dilatation of the stomach. This produces the so-called pyloric syndrome of Hartmann, pain, indigestion, gas, and hypersecretion, and in many cases per- istaltic waves can be seen passing from left to right over the gas- tric area. With a stomach-tube and an ordinary Davidson syr- inge the stomach can be dilated. If the lesser curvature is in its proper position and the great curvature lies below the umbilicus, some degree of dilatation is present. If the entire stomach is pro- lapsed, the question of dilatation is easily ascertained by noting the relative position of the curvatures on air distention. Pyloric obstruction gives unmistakable evidence of its presence, dilatation, stagnation, and retention of food causing fermentation, late vomit- ing, and emaciation. These cases are seen in all stages from the slight temporary interference with digestion to the most marked degree of disability. The condition can be aptly compared to a valvular heart lesion. Spells of dilatation alternating with com- 200 WILLIAM J. :\L\YO pensation through hypertrophy give periods of comparative health after a more or less prolonged term of gastric insufficiency. The best practical test as to the loss of motility is the finding of remnants of food in the stomach upon using the tube before breakfast. The various test-meals have some corroboratory value. Of over 1200 cases in which careful examination of the stomach-contents, including test-breakfast and so forth, was made, nearly 400 came to operation. The clinical diagnosis, based on the history, physical examination of the stomach, with the use of the stomach-tube to develop the outlines and remove reten- tion products, gave a correct diagnosis in the large majority of cases. The chemical and microscopic examination of the gas- tric contents proved of little value. The only one upon which any reliance is to be placed is that high values for hydrochloric acid argue for ulcer and low values for cancer; but only as cor- roboratory evidence, since the exact opposite may be true. With- out going into the question of differential diagnosis, I wish to call the attention of the general practitioner to the fact that refinements of technical diagnosis are often useless, and occasion- ally harmful, in causing delay, and that the sensible practitioner with the few simple means at his command is perfectly capable of arriving at a reasonable diagnosis, and will at least be able to direct the majority of his patients needing surgical treatment to the surgeon in time to be benefited. The prognosis of ulcer of the stomach has given rise to much discussion. Tricomi believes that 20 to 25 per cent, will die under medical treatment. Brinton estimates that about 50 per cent, are cured by medical means. Debove and Remond state that 25 per cent, die directly from the lesion itself (perforation and hemorrhage), and 25 per cent, additional from different compli- cations, such as pulmonary tuberculosis due to the chronic anemia. Leube, who has given the subject of gastric ulcer careful study, says that 25 per cent, die from the direct effect of the lesion, and that cases curable medically should be cured in four to five weeks' time. A recent study of 500 patients treated at the London Hospital between 1897 and 1902 is most interesting: 211 had had attacks CIIHONIC !!.( KR or ST().MA( II AM) DUODENUM 201 prcN ioiisly, that is, were known to have had ulcer with intervals ol" ai)i)areiil enre, 18 per cent, died, and 4£ per cent, were not cured at the time of discharge. A total of GO per cent, died or were not cured. EHniinate chronic gastric ulcer, and the cases ol" chronic dyspepsia, gastralgia, and cardialgia, not due to gall- stones or the appendix, will be reduced to small proportions. The development of cancer upon chronic ulcer is also a risk, the full significance of which has only of late become apparent. Du- plant has been (juoted as against the theory of cancer formation after ulcer; but Audisten, in going over practically the same material recently, has shown that pyloric cancer often begins at the margin of an ulcer. Lebert says that 9 per cent, only become cancerous; but this refers to those cases which pass directly from the one condition into the other. How about the early ulcers followed in later years by malignant change? Statistics upon this point are not available, but most observers place cancer- grafting upon ulcer base much higher. In 135 cases Graham found a good ulcer history preceding cancer in 60 per cent, of cases; but in many cases years had elapsed between the ulcer and the cancer. From the standpoint of a pathologist, FUtterer comes out very strongly for the cancer upon ulcer theory, and especially as secondary to the so-called "fish-hook" ulcer. Dunn says that, reasoning from analogy alone, one must conclude that ulcer is a frequent precancerous condition, and he believes that this can be demonstrated from his own experience. We have specimens showing the two conditions existing in the same case- Cases with early history' of ulcer, which develop decided gastric symptoms later, should be looked upon as suspicious of malignant disease. Before taking up the discussion of the surgical treatment, let me add a few words in regard to some dilatations of the stomach not of organic origin, such as the so-called atonic dilatations often found in neurasthenic individuals, and without the pyloric syndrome of Ilartmann. In these cases there is no retention and little stagnation of food. As a rule, these patients are not benefited by operation. This is also true of gastroptosis, which we have founti 202 WILLIAM J. MAYO to be present in over half of the cases of movable kidney. Re- laxed conditions in the neurasthenic state are not often perma- nently benefited by surgical operation. It is one of the misfor- tunes of surgical progress that neurasthenic symptoms are too often mistaken for organic disease. Uncertainty of diagnosis will render the majority of opera- tions primarily an exploration, either as to the actual condition present or as to its extent and surgical indications. The incision is placed in the median line, between the ensi- form cartilage and the umbilicus. This enables the movable por- tion of the stomach to be drawn out of the abdomen. The exami- nation should be thorough, and should include a digital explor- ation of the fundus and the cardiac orifice in every case, to avoid overlooking high hour-glass contractions. The duodenum should be inspected and palpated in its upper four inches, and the gall- bladder should also be examined, as complicating stones are not infrequent. If one ulcer is found, search should be instituted for others. Usually even- small ulcers can be located by a slight thickening of the gastric wall, perhaps a little place where the mucous coat is glued to the muscular tunic, preventing the normal sliding of one upon the other. Ulcers involving the muscular and peritoneal coats are easily recognizable by the milky or opaque appearance of the peritoneum, usually smooth and having the thick, stiff feel of scar tissue, with lessened vascularity, unlike the nodular "feel" of gastric cancer. Lund has pointed out that enlarged glands in the omenta may aid localization, and we have found this sign of great value. Strange to say, such lymph- nodes are most usually located in the gastrocolic omentum, rather than in the lesser omentum, a situation so peculiar to cancer. To locate ulcers on the posterior wall it may be necessary to open the lesser cavity of the peritoneum suiBficiently to introduce a finger for exploration. This can be done either through the lesser or the gastrocolic omentum. Exploration of the gastric cavity can be best accomphshed by a longitudinal incision into the stomach, half way between the curvatures. Through this opening a short wide speculum can CIIRO.VK' VLCFM OK STfXMACII AM) DTODKNUM 203 lie inlroducvd, and, with u fiiif^ci- heliiiid. Llic greaLcr part of the niucous membrane can he gone over. Small ulcers may be diffi- cult to locate, and in many cases prolonged search for an ulcer, the medical diagnosis of which is established beyond (jucstion, is not advisable. On one occasion a thorough exploration did not reveal an ulcer which had bled repeatedly for weeks, and it wa.s only by accident that sponging the mucous surface started up the hemorrhage from a little fissure previously undetected. The indication for the surgical treatment of gastric ulcer is of a two-fold nature: first, the question of the ulcer itself, and, second, the relief of the comphcating dilatation, distortion, ad- hesions, etc. The keynote in the treatment is drainage, because it is largely a question of mechanics. In some cases the ulcer- ation has already terminated, and the problem to be solved is purely mechanical. Gastro-enterostomy is the operation of the widest range of application, but excision of the ulcer, pylorectomy, and pyloroplasty have each a limited field of usefulness; w^hile in hour-glass contractions gastrogastrostomy combined with gas- tro-enterostomy is necessary to -establish a cure. Excision of the ulcer would appear to be the indicated pro- cedure; yet this has some disadvantages. In the first place, in a considerable percentage of cases there is more than one ulcer present, and one or more may be undetected or exist in an inac- cessible situation. Again, the tendency to ulceration exists, and new ulcers may manifest themselves later. There is also the liability of future contraction, the bands having their origin at the site of a former ulcer. It is probable that the radical pro- cedure of Rodman will be indicated in an increasing number of cases. He advises complete excision of the ulcer-bearing area, that is, the muscular pyloric region, to be followed by independent gastrojejunostomy. This procedure would meet all the indica- tions, and also prevent a possible malignant degeneration of the ulcer base. Pyloroplasty has a small field of usefulness in narrow stric- tures of the pylorus, provided the ulcer has healed. The objec- tions to it are that there is great liability to fixation of the pylorus 204 WILLIAM J. MAYO after operation, and also a tendency to recontraction. However, it is a safe operation; in 19 cases we had no deaths, but we had to reoperate upon 6 cases. The gastroduodenostomy of Finney is a far better procedure. It gives a very large opening, and as the enlargement is down- ward in the line of gravity drainage, the results are much better than in the unmodified pyloroplasty. We have made this oper- ation 34 times, with 1 death. It is probable that the operation will often fail to cure in open ulcer, since obstruction has no part in its causation, as is shown by the existence of ulcers in the duodenum beyond the pylorus. It can be readily seen, therefore, that no matter how large the opening is made, the food and irritating gastric secretions must pass over the ulcer area to reach the intestine. The conditions calling for the Finney operation are late cicatricial stenosis and contractures. All in all, gastrojejunostomy is the operation of choice. It drains the stomach rapidly from its lowest point, and from the cardiac side to the left of the ulcer-bearing pyloric region. How- ever, the procedure has some drawbacks. If the pylorus is not permanently obstructed, the anastomotic opening may contract. This can be obviated by excising the pyloric ring or running a silver wire purse-string suture about the duodenum immediately below the pylorus, thus producing the favorable condition of permanent obstruction. If the latter condition exists, there is little danger of recontraction. Anterior gastro-jejunostomy is liable to secondary peptic ulcer of the jejunum from the irritating gastric secretions. The posterior operation, while not free from this disaster, is much less liable to it. In 15 cases which Mikulicz found recorded, all of them followed the anterior operation. All the secondary ulcers were found either on the intestinal side of the anastomosis or immediately below the opening. Watts found but 2 out of 13 after the posterior operation. The anterior operation also has a tendency to drag on the attachment, and contraction is more frequent than after the posterior operation, especially if the Murphy button be used. The suture methods provide a larger permanent cicatricial area of adhesion, which ( IIHOMC ri.CKU OK S'ln.MAril AM) IJIODKM \t '"-iO.'J |)ri'vciit.s I he (livcrliciihua t'oriiiat ion al the site of atliiNloiMo.si.s in the jejunum so often seen in the former c-ases. In (jur experience the i)Osterior suture gastro-jejunostomy is the hesl method known at this time. As to the manner of per- forming llie ojjeration, the Murphy button makes the most per- fect opening, but is Hable to be retained in the stomach, although this is seldom a cause of serious after-trouble. The suture meth- ods are favoral)le, inasmuch as the opening can be made of large size to guard against contraction. However, there is some ten- dency to the formation of a diaphragm. To a certain extent this can be avoided by excising the nuicous membrane rather freely at the opening, as advised by Moynihan. The McCiraw ligature is a safe method, and a good opening can be secured. There is, however, a considerable amount of scar-tissue formation about the ring, and occasionally a bridge of mucous adhesions forms across the opening. In our experience, it takes from six to eight days for the opening to become established by the cutting through of the rubber ligature. We have made the McGraw ligature ope- ration 15 times, with 2 deaths, but we have used it in the worst class of cases, in which other methods would have rendered many of the operations inadvisable. This procedure is particularly indicated in cancer, and by the anterior method. It gives time for plastic union, and requires little disturbance of tissues, and, as the hydrochloric acid is decreased in cancer, there is little or no danger of secondary jejunal ulcer. Any of these methods will give good results. In 238 gastro- jejunostomies for all causes which have been made (C. H. and W. J. Mayo), there was a mortality in the malignant cases of 25 per cent.; in the non-malignant, of about 7 per cent. The button gave 7.5 per cent., the Mikulicz suture, 5 per cent., the McGraw ligature, 13 per cent., and the Finney operation, 3 per cent. The percentage of secondary operations is also interesting. In the total number of cases (238) there were 14 secondary opera- tions (G per cent.). In a general way it can be said that the anterior method of gastrojejunostomy gave a slightly smaller mortality, but there was relatively a higher percentage of reoperated cases. 206 WILLIAM J. MAYO In other words, other things being equal, the posterior method gave the best permanent results, with but a slight increase of mortality. In conclusion, I wish to speak briefly in regard to cancer of the stomach. Early operation is a prerequisite, and diagnostic exploratory incision is necessary. We have operated upon 135 cancers of the stomach, of which 34 were radical extirpations; 5 patients died within a month, and 1 later from another cause, too early to know the ultimate result of the operation. Of the 28 who recovered, the average length of life was over a year. One lived three years and seven months, and several are alive now, after more than two years. Sixty per cent, of cancers are located in the pyloric portion, that is, in the movable part of the stomach. The lymphatic arrangements are the same as the vascular, and the dome of the stomach is isolated from this portion, having a different vascular and lymphatic connection. If all the lesser curvature with the corresponding lesser omentum, and all the body and greater curvature to the left gastro-epiploic artery, be removed, the results in cancer of the pylorus should nearly equal what might be expected after complete gastrectomy. The remaining portion of the stomach enables intestinal an- astomosis to be made with considerable ease, and the gastric pouch rapidly enlarges, to assume the function of the stomach. The operation can be made nearly bloodless by tying at proper points the four blood-vessels which nourish the stomach, much as is the case in hysterectomy. By making the division of the stomach and duodenum with the actual cautery, wound inoc- ulation is prevented, and by suturing all the coats with catgut through the cauterized area, distal to the holding clamps, neither the stomach nor the intestinal canal is opened, with the attendant risk of infection. Lastly, by completely closing both duodenal and stomach ends permanently, gastrojejunal anastomosis can be effected in the usual way at a healthy situation on the gastric wall. RADICAL OPERATIONS FOR THE CURE OF CANCER OF THE PYLORIC END OF THE STOMACH* WILLIAM J. MAYO Seventy per cent, of all pistric carcinomata involve the pyloric portion of the stomach, and 60 per cent, have their origin at the pylorus or within three inches of it. Considering the fact that the radical operation was successfully performed in the time of Billroth (1881), before the inception of modern abdominal surgery, and that during the succeeding twenty-two years more or less work has been done in this field, it is curious to note that pylorectomy and partial gastrectomy have not as yet achieved an accepted .surgical position. There have been a number of reasons for this anomaly: First, a belief that the diagnosis could not be made be- fore the case had advanced beyond the possibility of cure, and. .second, that the operation was difficult, prolonged, and bloody, with an almost prohibitive mortality. To a certain extent the first consideration is true, but not entirely so, as we have in ex- ploratory incision the one diagnostic resource which is reliable, and which must be resorted to in the large majority of cases before a surgical diagnosis can be made. Without exploration the truth is but slowly established, and at the expense of progressive hopeless involvement for the patient. Exploration can be safely accom- plished through a small incision and with a short period of dis- ability. It may be said that the patient will not submit to an ab- dominal incision upon suspicion, but herein we do the intelligence of the public an injustice. We have seldom been refused the op- * Reprinted from "Annals of Surgery." March, 1904. 207 208 WILLIAM J. MAYO portunity, when the matter has been fairly and candidly laid before the patient and his friends. The plea for delay has more often come from the attending physician. The writer is of the opinion that an early diagnosis must be based upon clinical phenomena, the result of observation and ex- perience. In attempting to solve some of these problems, we have en- countered a number of misleading statements which appear to have been generally accepted by the medical profession. Three of these are of sufficient importance to deserve brief discussion : (1) The value of laboratory methods of diagnosis. (2) The sig- nificance of palpable tumor. (3) The history of previous ulcer. 1. Laboratory methods of diagnosis are based chiefly upon the chemistry of the gastric secretions (test-meals, etc.) and the mi- croscopic examination and chemical reactions of gastric findings, as well as the urine, feces, and blood. In the surgical stage these examinations have little value, but they gain in diagnostic im- portance with the progress of the disease and become of the greatest value when the patient is in hopeless condition. In the examina- tion of somewhat over 1500 stomach and duodenal cases, of which 430 came to operative demonstration, Graham and Millet showed this beyond question. These examinations should be made, but exploration should not be delayed by reason of the inconclusive nature of the results. 2. Tumor. Many years ago it was believed that the presence of a tumor demonstrated inoperability. This is by no means true; a small movable tumor in the pyloric region may be a favorable indication. The early diagnosis of cancer depends in a great measure upon the introduction of mechanical phenomena from obstruction at the pylorus, with or without palpable tumor; and it is the interference with gastric motility which early calls the attention of the patient to his trouble, and not the presence of the cancer itself. Without these symptoms a surgical diagnosis would seldom be made. In our experience the patient with marked symptoms of cancer of the stomach, but without any evidence of pyloric obstruction, proves, on exploration, to be the victim of ad- OPERATIONS FOR CANCER OF PYLORIC EM) OF STOMACH 209 vaiiccd and lio]K'le.s.s disease of llie bod\' of the stoniacli, in wliich there were no symptoms during the operable period. 3. A history of previous ulcer with complete recovery during a prolonijcd period is aj)t to be taken as an indication that the present gastric trouble is due to a recurrence of the ulcer, and leads the patient and attending physician to postpone interference. Usually this is true, but too often the renewal of symptoms is due to cancer development uj)on an ulcer base. We have had this occur a number of times, and the writer has become a convert to the belief that cancer frequently develops upon an old ulcer scar. In 145 cases of cancer of the stomach which came to operation at our hands Graham found a previous history of ulcer in 60 per cent, of the cases, although years may have elapsed after the healing of the ulcer before the cancer began. Lebert says that 9 per cent, of ulcers develop cancer, that is, pass directly from the one condition to the other. Ochsner, Futterer, Dunn, and others believe that the irritation of healed ulcer defects in the mucosa furnishes the starting-point for the majority of cancers. Murphy states that precancerous lesions can usually be demonstrated in the history of the case. It is to be noted that the geography of cancer and ulcer is nearly identical. The second proposition concerns the ulcer itself. There are two local manifestations of the malignant process upon which the advisability of operation depends — (1) Local extent of disease; (2) lymphatic involvement. 1. Movability of the growth is a very important factor in judging the extent of disease. Limitation to the pyloric end of the stomach is also of importance. Extension to neighboring organs usually contraindicates operation, with the occasional exception of the transverse mesocolon. The duodenum is rarely involved to any considerable extent. Adhesions are a serious complication, not only because they are the advance guard of the cancerous proc- ess, but in that they add to the difficulties and dangers of the operation. Haberkant found a death-rate of 73 per cent, operated upon in the face of extensive adhesions, and 27 per cent, without such com])lication. Mikulicz had a mortality of 70 per cent, when VOL. 1—14 ■210 WILLLIM J. iL\YO there was close adhesion to the pancreas. A moderate amount of adhesions which permit free motihty of the growth has not ma- terially influenced the prognosis, in our experience. 2. Lymphatic infection. This is the most important element in the attempt at cure of cancer of the stomach, because of the dif- ficulty in estimating its extent. The mere presence of enlarged lymph-nodes does not necessarily imply cancer. Glandular hy- perplasia occurs with great frequency in ulcer as the result of in- fection, and the location of such IjTnph-nodes may lead to the site of ulceration, as pointed out by Lund. Llcerating gastric carcinomata may give rise to infected glands without epithelial invasion, but in practically all cases of gastric cancer the lymphatic structures are involved. In the Breslau clinic, 20 out of 21 cases showed glandular involvement. In a general way the lymph- channels follow the blood-vessels. On the lesser curvature the blood- and lymph- vessels lie in the wall of the stomach itseK, and, as pointed out by Mikulicz, it is necessary in every case of pyloric cancer to remove all the lesser curvature to the gastric artery. For convenience, this situation on the lesser curvature for the be- ginning of the line of excision may be called the "Mikulicz point of election."' We owe a debt of gratitude to Cuneo for his masterly exposition of the lymph drainage of the stomach. He showed that there are but few lymph-glands along the greater curvature, and these are confined to the pyloric region (Fig. 15). These glands, with the blood-vessels, are set at some distance from the greater curvature, thus enabhng rapid expansion and contraction of the stomach, without interference with the circulation. The lymph-stream in this situation flows from left to right, and does not drain more than one-third of the adjacent stomach, two-thirds going into the lymph- channels of the lesser curvature. In the immediate vicinity of the pylorus, however, it drains its fair share. The lymphatics of the greater and lesser curvatures enter the deep receiving glands about the celiac axis on the anterior surface of the aorta. Cuneo prac- tically demonstrated that the fundus and two-thirds of the greater curvature are free from lymphatic involvement in cancer of the Willi/iM J.t-Uro. --^^ f Pneumo^siric Left- - I Ganglion Pneumojasinc Rijht Hepatic artery -^^ Ganglio Gastro-cpiploic vein Fig. 15.— Showiii}; an;iti)my of tlic slDniach, with cspociul rcfcrtnce to clistribiitinn of the lymphatics. OPERATIONS FOR (AXf KR OK PYLORIC KM) OF STOMACH '■^' 1 1 pylorus. IIurLinanu at oik-c seized upon lliis hasic princijile and fixed the point of election for the line of section upon the greater curvature at a healthy i)lace on the gastric walK to the left of these glands. The distance to the left is regulated by the extent of disease. In a previous communication the writer called attention to the lymphatic isolation of the dome of the stomach. This has also been noted by Robson and Moynihan. It is evident that the extent of this free zone along the greater curvature is much wider in i)yloric cancer than was at that time considered possible. The retention of this portion of the stomach relieves the operation of many serious difficulties without loss of completeness. The patient's stomach should be cleaned the day before rather than immediately previous to operation, as it may prove to be rather trying to one unaccustomed to the process. A small amount of liquid nourishment may be given after the lavage, but nothing on the morning of the operation. The teeth and mouth should have been previously cleansed as well as possible. A preliminary hypodermatic injection of morphin, to enable the anesthetic to be reduced to a minimum, may be of value. The operation itself can be divided into: (a) Incision and exposure; (6) control of hemorrhage; (c) closing of the stomach and duodenal stumps; (d) reestablishment of the gastro-intestinal canal; (e) avoidance of infection; (/) measures for preventing shock. (a) A small incision is made in the median line, half-way be- tween the ensiform cartilage and the umbilicus; through this two fingers are introduced for exploration. If the condition is inoperable, the incision is closed, and a sufficient number of buried non-absorbable mattress sutures of silk, linen, or wire introduced into the aponeurotic structure of the linea alba to enable the patient to get about at once and to return to his friends within a few days. Non-absorbable sutures, buried in fixed structures, such as fascia and bone, seldom give trouble, and furnish immediate strength. In muscle and movable tissues, atrophy necrosis may occur. We limit their use, however, to the hopeless cases of ex- ploration for malignant disease. If operation is decided uj^on. the 212 WILLIAM J. ilAYO small exploring incision is rapidly enlarged to four or five inches, and a sufficiency of the gastrohepatic omentum is tied off at once close to the liver. This opens the lesser cavity of the peritoneum and mobilizes the pyloric end of the stomach ^dth tumor. The entire area is now packed off with gauze pads. (b) Control of hemorrhage. The pyloric end of the stomach is suppHed by four blood-vessels — the gastric and superior pyloric above, and the right and left gastro-epiploics below. By ligating these four vessels early, the operation is rendered practically blood- less. The gastric is doubly tied about one inch below the cardiac orifice, at a point where it joins the lesser curvature, and divided between the ligatures. The superior pyloric is doubly tied and divided. The fingers are passed beneath the pylorus, raising the gastrocolic omentum from the transverse mesocolon, and in this way safe hgation behind the pylorus of the right gastro-epiploic artery, or in most cases its parent vessel, the gastroduodenal, is secured (Fig. 16). The left gastro-epiploic is now tied at an appro- priate point, and the necessary aniount of gastrocolic omentum doubly tied and cut. Sometimes the right margin of the omen- tum becomes very much congested from the venous obstruction produced in this way. In a few cases it has seemed wise to excise the devitahzed omentum, especially if drainage is to be used, with its attendant possibilities of secondary infection. In one such case a considerable amount of omental tissue sloughed, although, fortunately, the patient recovered. If drainage is not used, it will act as an omental graft and give no trouble. It is important that, in ligating the gastroduodenal vessel and the gastrocolic omentum, the fingers should raise the structures away from the middle colic artery, which runs immediately beneath in the transverse meso- colon (Fig. 17). The lesser cavity of the peritoneum is a potential rather than an actual space, as the two layers of peritoneum are in contact, and the middle colic has been accidentally caught in tying the vessels from without inward. As this vessel usually supplies the entire transverse colon, ligation may result in gangrene of the transverse colon, as pointed out by Kronlein. This has hap- Mikulicz-Hart- mann line. Fig. i6. — Showing ligation of pastrohcpatic omentum and superior vessels in such manner as to leave all the lymph-nodes attached to the part of the stomach to be e.xcised; also lines of division of duodenum and stomach. WilliAm J Mayo. Fig. 17. — Showing methods of excision. Note that all the glands on the greater curvature are removed in every case OI'KUATIONS I ()U CANCKIt OF PVI.OHIC DM) Ol' STO.NJACH '2]'i jx'iu'd u muiil)<.'r of tiiifs. Tlic (•(jiiLnjl ol' Ju-inorrliaj^c is \cry similar to the ligalioii ol" tlic loiir vessels concerned in abdominal hystcreclomy and t'lilly as easy. (c) The duodcmmi is doubly damped and divided between willi the actual eaulery, to pn^vent inoculation of the cut surfaces with cancer (Fig. !(>)• ' l'*' 'I'lodenal stump should be left one-fourth inch long, and, before removing liie clamj), a running suture of cat- gut is introduced through the scared stumj) and tied as the clamp is removed. A purse-string suture of silk or linen, three-quarters of an inch below the stump, enables inversion in a similar manner to the stunij) of the aj)pendix (Figs. IG and 17). A long Kocher holding clamp is now placed from the tied gastric artery at Miku- licz's point of election in an oblique direction, so as to save as much as possible of the greater curvature to Hartmann's point of election on the greater curvature (Fig. 17). The blades of this clamp should be covered witli rubber tubing, and the compression sliould be just sufficient to retain the tissues in its grasp. A second clamp is applied on the tumor side to prevent leakage. The tissues between are severed with the Paquclin cautery, one-quarter of an inch from the holding clamp, and, as the tissues are divided, several catch forceps are caught on the projecting stump, to prevent retraction of some part of the gastric wall from the grasp of the Kocher clami). The pyloric end of the stomach, with the tumor guarded against leakage by the clamp at each end, is removed. The cauterized stump projecting beyond the Kocher clamp is rapidly sutured with a catgut buttonhole suture, from the greater to the lesser curva- ture, through all the coats of the stomach, and in the same manner directly back, and tied at the starting-point; this prevents hemor- rhage as well as leakage (Fig. 18). The doubling of this form of suture holds the approximated edges evenly in line. The Kocher clamp is now removed, and any bleeding point caught and tied. The final suture of silk or linen is now introduced, made aiter the right-angled plan of Cushing. It is taken sufficiently far from the catgut suture line to enable easy approximation of the sero- muscular layers without tension (Fig. 19). Steps b and c can be varied sometimes to advantage. We have 214 ^VILLIAM J. MAYO frequently tied off the gastrohepatic ligament and the superior vessels, and at once double clamped and divided the duodenum. By pulling upward on the stomach side the gastroduodenal artery is easily caught, tied, and divided, and the operation proceeded with as before. In a few cases we have begun on the stomach side, ligating and dividing the gastric and left gastro-epiploic vessels first, then clamping, di^nding, and suturing the stomach as before. Complete the duodenal end with its vessels last. This is favored bj" Hartmann. If there are adhesions, however, the first plan mobilizes the stomach much better, and enables more accurate work and greater exposure of that part of the stomach, which at the line of section lies naturally deep under the costal arch. (d) Restoration of the gastro-intestinal canal was first ac- complished by Billroth, by joining directly the cut surface of the duodenum to the shortened stomach, the opening of the latter viscus being partly sutured to reduce it to the size of the duodenal end. The angle where the three suture lines came together leaked so often, especially if there was the least tension, that it was called the "fatal suture angle." Kocher saw the defect in this method, and began implanting the cut end of the duodenum to the posterior gastric wall at a sound point, and completely closed the stomach. This method gives excellent results, if there be no tension, in bringing the parts into easy, apposition. Unfortunately, this tension occurs often. Billroth's second operation is the operation of choice: complete closure of the duodenal and stomach ends with an independent gastrojejunostomy of the usual type. It has the two chief requi- sites of gastro-intestinal anastomosis; there is no tension, and the parts to be united have not been injured. Either the anterior or posterior method can be used, and the Murphy button or suture operation be performed. If the patient is in good condition and the operation has been completed promptly, we prefer the posterior suture method; if the patient's condition is poor, the anterior button operation is chosen (Fig. 19). (e) The question of cancer infection grafted upon a raw surface is an important one. We have seen carcinomatous nodes i;,llM-nJ M>F STCj.MArH 21.3 develop in llic alxloiiiinal incision and in I lie alxioniinal need!*' [jiUK'lnres made in snlurin^ llic alxioniinal wall al'lcr partial gaslrccloniy. Disseininalion of carcinonia hy rougli liandlirif^ or allowing infected cells to escape into the wound is not uncommon. It is for this reason that all .sections of the disea.sed part are made witli the actual cautery, which prevents inoculation of raw surfaces and checks capillary hemorrhage, and leaves the ajjfjroximated ends in an asej)tic condition until they are digested hack to the outer suture line. Pyogenic infection is prevented by the clamps placed upon each side of the excised stomach, sealing against escape of contents, while the exposed edges beyond the clamp are sterilized by the use of the cautery in making the section. In addition to this the gauze pads are arranged in two rows — an outer deep layer, which is not changed until final removal, and an inner superficial layer, which is being constantly renewed. Upon removal of the final gauze pack the entire field is carefully gone over and any little bleeding point checked by ligature. After sponging the surfaces with a moist saline gauze pad, the abdominal incision is closed. In some cases drainage seems wise on account of accidental soiling. This is seldom necessary, but if in doubt, drain, with a cigarette drain placed at the lower angle of the external wound, entirely away from the visceral suture lines. The internal end of the drain should reach a situation just above the transverse colon, which acts as a dam when the patient is placed in the proper position in bed, head and shoulders elevated. In this half-sitting posture the little pouch formed by the transverse colon is not unlike an artificial pelvis into which any fluids gravitate. If but a limited area is to be quarantined, the gauze should be brought out in the most direct manner possible. (/) If the patient is in good condition, there is practically no shock, because there is no l)lood loss and but little exposure of abdominal contents. The operation proceeds systematically, and can be done in a suitable case by the average operator, from the beginning of the abdominal incision until it is closed, in from fifty minutes to one hour and fifteen minutes. If the patient is in bad condition, owing to early obstruction, the chief danger 216 WILLIAM J. MAYO comes from the lack of fluids in the body. As suggested by Dudley Allen, this should be made up by subcutaneous injections of saline solution, 40 to 60 ounces a day, usually 20 to 30 ounces every twelve hours, for two days previous to the operation. This is continued for several days following operation, if necessary. For subcu- taneous injections we prefer the ordinary Davidson syringe, to which is attached an aspirating needle. The hand-bulb regulates the inflow. The injection can be given by a nurse as easily as an enema. In debilitated patients very little anesthetic is used — just enough to enable the surgeon to open and close the abdomen. All the visceral work can be done without pain. The previous administration of morphin keeps the patient from becoming nervous. An enema of 6 ounces of coffee is given as soon as the patient is put to bed. If necessary, morphin, strychnin, or like remedies are given. The after-treatment is simple — the head and shoulders of the patient are raised by four or five pillows, rectal alimentation is instituted, hot water by mouth after twelve hours in tablespoonful doses, increased to an ounce every hour. After thirty-six hours the usual experimentation with liquid foods is begun. To recapitulate, there are six important stages to the operation as outlined: 1. Open the abdomen. 2. Doubleligateanddivide the gastric artery; ligate and divide the necessary amount of gastrohepatic omentum close to the liver, leaving most of its structure attached to the stomach. Double ligate and divide the superior pyloric artery and free the upper inch or more of the duodenum (Fig. 16). 3. With the fingers as a guide underneath the pylorus, in the lesser cavity of the peritoneum, ligate the right gastro-epiploic or gastro-duodenal artery, and progressively tie and cut away the gastrocolic omentum distal to the glands and vessels up to the ap- propriate point on the greater curvature, and here ligate the left gastro-epiploic vessels (Fig. 17). 4. Double clamp the duodenum, divide between with the OPERATIONS FOR fAXfER OF PYLORIC END OF STOMAf H -217 cautery, lcii\in^ oiio-l'ourlli iiuli projection. \N itli a rumiiiifj; su- ture of catgut through the scared stump the end of the duodeuuin is closed as the clamp is removed. A purse-string suture about the duodenum enables the stump to be inverted (Figs. 10 and 17). The ])roxiinal end of the stomach is double clamped along the Mikulicz-Hartniaiin line (F'ig. 17), and divided with the cauterj", leaving one-fourth inch projection. Suture through the seared stump with a catgut buttonhole suture. This is again turned in after removal of the clamp by a continuous silk or Gushing suture (Figs. 18 and 19). 5. Independent gastrojejunostomy (Fig. 19). 6. Closure of the wound. Forty-one radical operations on the pyloric end of the stom- ach, 37 for cancer, and 4 for inveterate ulcer, have been performed by C. H. Mayo and the writer. Of these, 13 have been done es- sentially by the plan outlined above, with 1 death. There were 6 deaths in the remaining 28 cases, performed by various methods. In the last 11 cases this technic was used practically as given, and there were no deaths. Making all due allowance for increased ex- perience and possibly a better selection of cases, the difference is too marked to be entirely accidental. It is hardly necessary to say that this is a composite operation, and in no sense to be con- sidered original. In a previous contribution on this subject, published in the "Annals of Surgery" July, 1903, a somewhat similar operation was recommended by the writer, only that it was far more ex- tensive, removing all the stomach except the dome. With in- creased observation and experience I believe that the former opera- tion, with a mortality of 3 deaths in 8 cases, is unnecessarily severe for the average case of pyloric cancer. The operation described at that time has a place in surgery, and should be used in the cases of extensive disease involving the body of the stomach. In these cases it has practically all the advantages of complete removal of the stomach, and should be used as a substitute for total gastrec- tomy, where possible. ULCER AND CANCER OF THE STOMACH FROM A SURGICAL STANDPOINT* WILLIAM T. MAYO Disease of the stomach is one of the most important subjects before the medical profession today. Heretofore these conditions have been considered almost entirely from a medical point of view, and it is only of late that their surgical possibilities are beginning to be recognized. The treatment of gastric disorders has been based upon chem- istry, and for two decads hydrochloric acid and pepsin have formed the chief ammunition which has been fired at diseases of the stom- ach. During this time a feeling has gradually developed that these means were inefficient, and while many investigators have turned toward new pharmacopeial combinations, there has also been an intelligent effort toward a better understanding of the functions of the stomach, and with this, a different version as to its pathology. The mechanics of the stomach are usually at fault, and not its chemics, and it is for this reason that surgery is rapidly invading the field. I would not be understood as saying that even the greater proportion of diseases of the stomach are surgical; but that the majority of serious lesions may become so cannot be doubted, and it is also true that a very large proportion of chronic gastric disorders are due to faulty mechanics and must have sur- gical interference. While the stomach absorbs fluids and equalizes the temperature of the ingesta, it does not digest the food; but with a weak solution of pepsin and hydrochloric acid it macerates the ingested material, and the pyloric portion breaks up the food-masses by a grinding * Reprinted from "The Medical News," April 16, 1904. 218 I L( I;K .wo ( ANTKU ok SJOMAril 219 inolioii. As llic pnjccss is coinplcled tlie pHxiuct is turned into tlie small intestine, where digestion and assimilation are accom- plished. The mnscuhir action of the stomach is of a twofold nature: the fundus contracts rather slowly upon the food material, forcing it toward the pyloric end, which grinds and pulverizes it into a harmonious whole, the fundus being the hoj)per, and the j)yloric [)ortion the grinding-stones. It can readily he seen that an ob- struction at the outlet produces dilatation with retention or stag- nation of the food, and this often results in fermentation and ab- sorj)tion of deleterious products. The extent of the disorder depends upon the degree of obstruction and the compensatory hy- pertrophy of the musculature of the stomach. Many of these cases develop a degree of hypermotility which equalizes the ob- struction. In other cases a period of compensation alternates with muscle failure, exactly like valvular heart lesions and their effect ui)on the cardiac muscle. We see the same phenomena in obstructions at the outlet of the gall-bladder and at the pelvis of the kidney, and again in the urinary bladder — all organs with a temporary storage function and a limited outlet. Even with moderate obstructive dilatation the victim begins to diet, and in the main "diet" means restricted food-supply — an attempt to reduce the amount of work to be done by the stomach to the minimum. In the extreme cases the solids are first eliminated from the dietary because, as has been pointed out, the stomach does not assimilate such articles of food, and the bulk increases the difficulty of passing the obstruction. Fluids, on the contrary, are absorbed to a certain degree, and also more easily pass the con- tracted outlet. Therefore, foods with much fluids and little resi- due are chosen. The stomach may become enormously distended, and even the absorption of fluids be reduced to small proportions. So true is this that Cramer was able to demonstrate, by the quantity of urine passed, the degree of dilatation and the amount of gastric function. He divides these cases into three grades, as shown by the urine collected for twentv-four hours: First degree, 1500 to 220 WILLIAM J. MAYO 1000 c.c; second, from 900 to 500 c.c; and third, all cases falling below 500 c.c, and further says that lavage of the stomach which contains much material, with urine below 500 c.c, may dilute the toxic products in the stomach and render them more readily ab- sorbable. Tetany has its onset immediately following the use of the tube in many reported cases. The stomach should be emptied some hours before lavage is practised in such cases. That these products of retention and fermentation are poisonous cannot be questioned. We had one case in which gastro-enterostomy turned a great amount of such material into the small bowel and the patient promptly died from toxemia, which followed the ab- sorption. The symptoms resembled the effect of cadaverin upon animals. Gastric tetany probably depends for its origin upon the absorption of retention products. The severe grades are the only ones which we have heretofore recognized. They have been ex- ceedingly serious. In 40 cases Albu found 31 deaths. Mild grades, however, are much more common, and are apt to be over- looked. Muscle-twitching, prickling sensations, and so forth in connection with serious dilatation are to be looked upon as warnings and surgery should not be long delayed. Cramer has recently written exhaustively upon this subject. The popularity of the stomach-tube depends upon the mechan- ical removal of food remnants, which would otherwise stagnate in the stomach ; but it does not cure any more than the catheter which is used to relieve urinary retention or residual urine cures the prostatic hypertrophy upon which the condition depends. The stomach-tube relieves the temporary obstruction, and may aid restoration of compensation by giving time for muscular de- velopment; but our faith in the ability of gastric lavage, either plain or medicated, to cure mechanical obstructions has been shattered beyond repair. It is not my purpose to discuss in detail the nature of the various obstructions, but rather to indicate the effect of such interference upon gastric motility, and to call attention to the fact that serious or progressive obstructions are surgical conditions, and that dila- tation of the stomach and its degree is the sign-post which tells us LTXKU AND (.ANCKU OK ST(jMA( || !2'-21 the projjor course of action. If the lesser ciir\'atiire is in normal position and the «^reater curvature of the stomach lies l)elo\v the umbilicus, some degree of dilatation is i)resent. If the stomach is prolapsed, the distance between the curvatures, demonstrated by air inflation, will readily show the amount of dilatation. The tartaric acid and bicarbonate of soda test is sudden, and in certain cases there is danger of rupturing a weakened gastric wall. W itli an ordinary Davidson syrin()() casc.N uliicli liaxc Ix-cii ex- amined by Graham and Milld, ITS came to operation. The only test which had even corroboratory worth was that higJi values for hydrochloric acid ar^uc for ulcer and low values for cancer; but even this is not to l)e relied upon. Otherwise, the microscopic and chemical findings, with the excei)tion of the occasional presence of blood, were i)ractically worthless. It should not be forgotten that ulcer of the j)yloric portion of the stomach may form a well-defined tumor, and the anemia from chronic blood loss develop a cachexia which, with the tumor, may give the appearance of gastric carcinoma in a hojieless stage. Xo doubt many of these patients have died from a condition which was essentially benign. We have but to look over the literature to see the frequency of this mistake, a gastro-enterostomy permanently curing a patient in whom a supposed malignant tumor with stenosis j)roved to be benign. Prognosis. — The mortality of chronic ulcer is given by Leube at about "io per cent., and he states that about half of the remainder will be cured. Debove and Remond estimate that 25 per cent, die directly from the lesion (hemorrhage, perforation, and so forth), and 25 per cent, additional from complications, such as tuberculosis induced by the anemia. In 500 cases reported recently at the London Hospital 211 were known to have had previous attacks, 18 per cent, died, and 42 per cent, were not cured at the time of dis- charge, leaving 40 per cent, supposed to have been cured. There is also the further danger of cancer grafting upon ulcer base which is undoubtedly of frequent occurrence, although years may have elapsed between the ulcer and the malignant change. In 157 cases of cancer of the stomach which came to operation at our hands Graham found a good ulcer history in over GO per cent., although perhaps the malignant degeneration developed years after the ulcer had healed. FUtterer, Dunn, and many others also insist upon the etiologic importance of ulcer in cancer of the stomach. Surgical Indications. — When should gastric ulcer be considered surgically.^ Leube says that from four to five weeks should be VOL. I — 15 226 WILLIAM J. ]MATO time enough to cure gastric ulcers which can be cured medically. For those cases which have not developed obstructions, it would be best to try the Leube treatment (rectal feeding, etc.)- The bulk of these patients have employed various forms of treatment, with temporary benefit, but they have finally reached the chronic state. Such cases should be considered surgically, especially if secondary complications, such as obstruction, dilatation, adhesions, and deformities, have developed. Indications for the surgical treatment of gastric ulcer can be divided into two distinct classes : first, the ulcer itself ; second, the complicating adhesions, deformities,| obstructions, etc. There are clinically three varieties of gastric ulcer: (a) The mucous erosion, limited to the superficial epithelium of the mucous membrane; (6) the round, fissured ulcer, which is probably the most frequent and is limited to the mucous coat, excepting in perforative cases. This ulcer has the distinctive feature that it cannot often be located from the exterior of the stomach; sometimes a little thickening can be felt or a point where the mucous coat does not glide on the peritoneal and muscular tunics in the normal manner. This variety of ulcer is often multiple, and may lead to embarrassment at the operating table, where an ulcer accurately diagnosticated cannot be demonstrated; (c) the large, irregular ulcer, invading all the coats of the stomach and easily recognizable by its thickness and the milky or opaque appearance of the peritoneum. This variety is often mistaken for cancer, especially as enlarged lymph- nodes are to be found in the omenta, particularly in the gastrocolic omentum, as pointed out by Lund. The complications of gastric ulcer are many. For example, adhesions, the separation of which may give relief, but is open to the objection that it leaves the ulcer which caused the trouble un- cured, so that the adhesions usually reform. Separation of ad- hesions must be done most carefully, as they may be protecting a perforation which it would be unfortunate to open. Obstructions usually exist at the pylorus, but may occur in the body of the stom- ach, forming an hour-glass contraction. Gastric Drainage. — The keynote to the surgery of gastric ulcer ULCER AM) ('AN( Kit OK STOMACH 227 is drainage, and llie best place for this is to the left of the muscular pyloric portion. There are four methods of inducing gastric drainage: (1) The Heineke-Mikulicz pyloroplasty, which gives satisfactory ultimate results in about 70 per cent, of the cases. This procedure has the misfortune of leaving the enlarged opening at a high level, so that if the stomach be greatly dilated, the de- generated muscle nmst elevate the food to the high-lying outlet and also there may be firm adhesions binding it to the surrounding tissues. (2) Finney's gastroduodenostomy. This is a most ex- cellent operation, giving a large opening in the line of drainage. It is most useful in narrow strictures, and least successful in open ulcer, as the food must pass the ulcerated area to reach the outlet. (3j Rodman's operation. Excision of the ulcer would seem to be indi- cated in a large number of cases; but if so, it must be combined with gastrojejunostomy — (a) because there may be more than one ulcer; (h) because a stricture often follows at the site of incision, no matter how made; (c) because it leaves the tendency to ulcer unchecked and new ulcers maj^ form. Rodman suggests that the most sensible thing to do is to excise the entire pyloric end of the stomach, — the so-called ulcer-bearing area, — closing the duo- denum and stomach end permanently, and then do a gastrojejunos- tomy in the usual manner. This would meet all the objections to the excision of the ulcer and avoid the possibility of secondary cancerous change. (4) Gastrojejunostomy. All in all, this method meets the indications most perfectly, although it is prob- able that Rodman's operation will be more often considered in the near future. Three hundred and twenty-eight operations for the improve- ment of gastric drainage have been performed by C. H. ]Mayo and the writer up to April 1, 1904. This number is divided as follows: 20 pyloroplasties, 7 secondary operations, no deaths. Finney gastroduodenostomy, -10 operations, no secondary operations, 1 death. Gastrojejunostomy, 271, mortality in the benign series 6 per cent., with 6 per cent, of secondary operations; of these, 5 were subjected to Rodman's operation. In the malignant cases there was 23 per cent, mortality after gastrojejunostomy. The 228 WILLIAM J. MAYO posterior operation, made by means of the suture is preferred by us for ulcer, reserving the anterior operation either with the Murphy button or the McGraw Hgature for the cases of cancer and such cases of benign disease as present unusual difficulties to the posterior operation. Cancer of the Stomach. — Gastric carcinoma is the most fre- quent form of cancer in the human body. Statistics ranging from those given by Welch as 21.4 per cent, of the total, to Virchow, who places it first at 31 per cent., and Haberlin, at 40 per cent., of the total number of cases. Why have there been so few attempts to cure these patients by the only known means, i. e., a surgical operation.^ There are three reasons for this conservatism: first, a belief that cure cannot be accomplished; second, that the mortality of radical operations is almost prohibitory; third, that the diagnosis cannot be made until the case is hopeless. Let us discuss these reasons more in detail. First. — Can cure be accomplished by operation.^ Macdonald found 43 cases cured by operation which could not be disputed; Murphy collected 189 cases operated upon radically by Kronlein, Maydl, Rydygier, Czerny, Morison, Bevan, and Mayo, ^^-ith 26 deaths. Of these, 17 survived three years — about 8 per cent. This was reduced to 5 per cent, by recurrence after three years; but as many of the cases were alive and well more than two years, enough of these cases would, by the law of averages, survive to bring the percentage up to 8 per cent, or more. Not only that, but Kronlein's statistics demonstrated an average prolongation of fife of fourteen months over the unoperated cases. ^Mikulicz, in 100 resections, had an average duration of life of one and one-third years. In our own experience of 43 radical operations for malig- nant disease of the stomach, all but two who sur\dved the operation lived beyond the year; one lived three years and seven months, and several are alive and well more than two years after. Second. — Is the mortality prohibitory? The mortality of the operation depends largely upon the case. In an early operation upon a patient not materially reduced, the mortaHty wiU not ex- ceed 10 per cent., and the average mortality at this time, taking ULCER AM) CANCKIt OK STOMMll 2'2J) llic ^r()()(l and had cases as llicy come, is iiol ahovc '■-!() lo '^i') per <('iil. Much of llie (lau^'cr (IcikmkIs upon I lie adhesions to ucij^dihoriti},' viscera. Ilaherkaiil had a iiioilaHlN' of l^i.l jx-r (ciil. with ad- hesions, and ^27.') per cent, without adhesions. Mikulicz had a mortality of 70 per cent, with adhesions to the pancreas and '^.'i..3 per cent, without such adhesions. In the 189 cases operated on by seven surgeons and quoted by Murphy, the mortality was only about 15 per cent. Tliird.^Ciin the diagnosis be made in time by medical means? Vi) this I say, most emphatically, no; the condition can be sur- mised, but not often dia<;iiosticated with certainty. Exi)loratory incision is practically harmless, and a suspicion of gastric cancer .should cause us to lay the facts before the patient and his friends and let them decide whether the only sure means of making a diagnosis shall be undertaken or not. There are two points in the diagnosis to which I wish to call particular attention, because I believe that their significance has been exaggerated: First, as to the presence of a tumor. It has been said that the presence of a tumor marks inoperability. This, however, is not true; a small movable tiimor in the region of the pylorus is a favorable indication, because, while cancer does not give evidence of its presence in an early stage, the tumor does give such evidence by obstructive symptoms which call attention to the mechanical conditions present. As a matter of fact, it is only those diseases located in the pyloric region which are amenable to surgical intervention. This part of the stomach is accessible to palpation and also to operative procedures. Pyloric cancer, by its location, introduces mechanical features which enable an early diagnosis. The pylorus normally is near the middle line, rather than to the right, as we have been led to believe, and the stomach lies in a much more vertical position than has usually been pictured, so that the pylorus is not much above the lowest point of the normal stomach. In our experience, the larger number of cases with symptoms of gastric cancer in which no trace of a tumor was present at the pylorus and without .symptoms of obstruction have proved to be hopeless cases of cancer of the body of the stomach which gave no early symptoms. 230 WILLIAM J. MAYO The second point upon which I wish to speak is that a labora- tory diagnosis of cancer of the stomach is of Httle importance dur- ing the operable period, because the disease is too slight in extent to interfere with the secretions. When it does so, it usually means a hopeless involvement. Radical extirpations of gastric cancer follow modern lines. The necessity of removing the regional lymph-structures is recognized just as it is in cancer of the breast. The first great improve- ment was due to the researches of Mikulicz, who showed that the blood-vessels and lymphatics of the lesser curvature lay in the wall of the stomach, and that it was necessary in every case to remove all the lesser curvature, even if the growth was small and confined to the pylorus. Cuneo demonstrated that the glands of the greater curvature were near the pylorus, and that the lymph-current of this region was from left to right, showing that it is not necessary to be so radical in operating on the greater curvature, especially as the blood-vessels are set at some distance from the greater curva- ture to allow rapid expansion and contraction of the stomach. The lymph-nodes in this situation are set free from the gastric wall with the blood-vessels. As to the operation: by tying the four blood-vessels supplying the stomach at proper points, the extirpation can be made practically bloodless, much as in abdom- inal hysterectomy. The duodenum beyond the growth and the stomach at a healthy point proximal to it should be doubly clamped, and the diseased part cut out with the actual cautery one-fourth inch from the holding clamps, preventing inoculation of the cut surfaces with cancer. The seared stumps should be sutured with catgut at once before removing the clamps, thus checking the hemorrhage and preventing soiling of the field from the otherwise open viscera. Lastly, both the duodenal and stomach ends are buried by silk or linen sutures and an independent gastrojejunos- tomy made. Such an operation in a case suitable for extirpation can be done in an hour without loss of blood or shock. We have made pylorectomy and partial gastrectomy 43 times, with 7 deaths . Cancer of the stomach can be cured without excessive mortality only when exploratory operation is undertaken. THE ASSOCIATION OF SURGICAL LESIOXS IN THE UPPER ABDOMEN* WILLIAM f. MAYO The liistory of medical advance from the empirical treatment of symj)toms to scientific expectancy has been based upon post- mortem study and modern research work. The spirit has been fine, although somewhat pessimistic. The great advances have been along the line of preventive medicine, and in the elucidation of problems connected with the etiology of disease. To a certain extent, surgical progress has been made through similar channels; the spirit, however, has been one of magnificent optimism. The great advances in surgery have come from the clinical side, modern technic having enabled examination of pri- mary conditions during hfe. As a groundwork for tlie study of medicine the pathologic laboratory and autopsy room offer unrivaled advantages, but they reach end-results only. Who of us has not admired the precision with which a trained observer makes a diagnosis.' A few days later one visits the autopsy room and here observes that this al- most clairvoyant analysis was correct. At the same time, we must also acknowledge that the patient has not profited; it means a hundred pages of pathology and two meager lines of treatment. ITow many times do we observe m the autopsy revelations that the original lesion has been but slight, and at one stage easily curable, and the pathology present largely the result of secondary complications and terminal infections? * Oration on Surgery at the Fifty-fifth Annual Session of The American Med- ical Association at Atlantic City, .June 7 to 10, 1904. Reprinted from "Jour. Amer. Med. Assoc," June 11, 1904. 231 232 WILLIAM J. MAYO The postmortem has been an mstitution for years, yet what did we know about appendicitis until surgery led the way, or about gall-stone disease, supposed to be an "innocent postmortem find- ing," until operation opened up the field? Did knowledge of extra- uterine gestation come from the autopsy? By no means. It was obtained through means of the surgeon's knife, and why? Be- cause secondary conditions so often mask the original lesion at the postmortem. On the contrary, our understanding of acute conditions causing sudden death has been brought about by post- mortem investigations; for example, fatal hemorrhage and per- foration placed acute gastric ulcer on a sound foundation. Not so chronic gastric and duodenal ulcers with late death from complica- tions, i. e., cholangitis from gall-stone disease, and chronic pan- creatitis from the same cause. These conditions could not be correctly studied until death occurred from accidental cause, and real progress lay dormant until surgery invaded the field. From research work we have profited greatly; we are gaining rapidly by this means day by day, but we cannot compare the animal to the man, nor can we always artificially create similar conditions. Like the postmortem, it has its limitations, yet these two fundamental methods have enabled us to advance. Without them we would still be in the middle ages. In the light of these investigations the time has conie to view conditions from a new standpoint, — at a time when the lesion is in its infancy, while the patient may be benefited, — and the research work applied to the elucidation of living problems. The medical man must haunt the operating theater as he has haunted the autopsy room and the laboratory. The times have changed: we must have more treatment and less pathology. It is here that surgery wins its triumphs. We should not forget, how- ever, that to physicians with pathologic training we owe the knowl- edge which enables us to approach the subject. Courvoisier was the real father of gall-stone surgery. It was Balzer, Fitz, Opie, and others who cleared up the pathology of pancreatitis, and sur- gery followed their lead. It is in the union of the internist and surgeon that progress is most rapidly made, and in the readjust- SURGICAL LESIONS IX UPPER ABDOMEN 28.'$ inciil (ji' .science the i'(>niier will he the architect and the latter the master builder. Once more the j)hy.sician and surgeon will come together, and the mistakes of endeavor in new fields will receive timely correction at the hands of the pathologist and exjjerimental worker. Today the surgical borderland lies in the upper region of the abdomen, a locality until recently considered almost purely medical. What are the reasons for this invasion, and have the results jus- tified the attempt? Surgery must be judged on three grounds: First, the mortality of the operation and the question whether this is greater than the expectant plan; second, the permanence of cure contrasted with medical treatment; third, the question of disability, either intro- duced by the operation or the natural length of time which the healing process involves. To each one of these considerations we must answer j-es, and we can go still farther and say that earlier operation would reduce the mortality, increase the permanence of cure, and lessen the disability. Certain parts of the body are so closely related in their anatomy, function, and pathology as to be almost necessarily considered as part of the same system. The generative organs of women form so distinct a field of work as to have built up a specialty. Can we separate diseases of the kidney from those of the ureter, bladder, and urethra? By no means. Each may stand in an etiologic relationship to the others which cannot be ignored. In the upper abdomen we have attempted to study the stomach independent of the associated organs, i. e., the liver and bile- passages, the duodenum and pancreas. The result has been a confusion in diagnosis and treatment. The palm of a hand may cover a serious lesion of any one of these organs, and that, too. at the point of greatest liability; moreover, any one of this group may start a pathologic process which may extend to any one of the others, and with a frequency fulh' as great as occurs under similar conditions in either the generative or the urinary system. Note the disturbance of the stomach which occurs with gall- stone disease; the adhesions to the duodenum and the pancreatitis. 234 WILLIAM J. MAYO an association direct as it is vital. Again, let me call attention to chronic ulc^r of the stomach with adhesions to the pancreas, secondary ulcer of the duodenum adherent to the bile-passages or gall-bladder. These are not fanciful pictures, but drawn from every-day work. I have no hesitation in saying that with an operative experience of over 1400 cases of this description, mistakes in exact diagnosis are still common, and in many instances un- avoidable. The history may be the only valuable diagnostic resource when the patient comes to us, and we all know how un- reliable that may be. Given a history of painful attacks which have been very severe, but which have completely ceased, with tenderness on deep palpation in the epigastrium, and we may have disease of any one of these four organs, and not infrequently an association, either direct or indirect, of the pathologic process. If we clearly understand the possibilities of error, we are better pre- pared to meet complications or execute a change of front and ope- rate on one organ when another procedure was planned. In the majority of cases a pathologic diagnosis is possible, and one can say with certainty, "this is gall-stone disease," or "this is ulcer of the stomach"; but in a considerable minority a surgical diagnosis is the best that can be made. That is, we can say: "In this locality is a diseased process which requires operative treatment, the exact nature of which must be determined by incision." The patient does not come to us for the purpose of having a certain operation performed, but seeks relief from suffering and disability. Let me call your attention to the anatomic diagram, showing the nearly vertical position of the stomach, with the pylorus in the middle line of the body, and but little elevated above the lowest point of the gastric cavity. It is turned upward and to the right just enough to prevent the weight of the gastric contents bearing directly on the sphincter apparatus. The only portion of the duodenum in which we are interested is the four inches lying between the pylorus and the papilla of the common duct of the liver and pancreas. This may be called the vestibule of the small intestines. Its position subjects it internally to the perils of ulcer from the acid gastric juices which its thin tunics but SURGICAL LESIONS IN IPPEU ABDOMEN iSo inadequately resist. Externally, its function is often interfered with hy adhesions to the j,'all-l)latlder and bile-tract, secondarj- to gall-stone disease. This unoffending bit of intestine is so often offended against as to cause it to become the most frecjuently dis- eased portion of bowel of the same length. The remaining eight inches of the duodenum is prot^ted by the alkaline secretions of the pancreas and liver. Its fixed position and peculiar horseshoe shape, with its delivery point nearly as high as its origin, enable it mechanically to slow the ingested material during the mixing proc- ess for which its large caliber affords accommodation. The anatomy of the bile-tract is equally interesting, and es- pecially the relations of the common duct of the liver to the pan- creatic duct and the duodenum. There is continuity of mucous surface, each protecting itself from the secretions of the others by the mechanical washing effects of its own secretion, the joint dis- charge, and a feeble sphincter apparatus preventing entrance of the duodenal contents. Since the gall-bladder furnished the initial lesion in more than one-half of the diseases of this series, it is of great interest. Like the appendix, an obsolete organ of storage function and limited outlet, it gives rise to a variety of troubles, which we are only of late beginning to appreciate. The gall-bladder has a capacity of about an ounce, and as we find almost universally, in organs of storage function, the neck is raised shghtly above the lowest point, to prevent the weight of contents resting directly against the out- let. The little pouch thus formed may be called the pelvis of the gall-bladder (^Brewer). It is here that the obstructing stone is so frequently lodged in cystic impactions. Murphy observes that the fundus of all organs has but few lymphatics, while the region of the neck has an abundant supply; hence, even with septic contents, there is but a mild reaction when the pelvis is ob- structed, as compared with the startling temperature-curves of duct stones. The pancreas, composed of two originally separate parts, in nearly half of the specimens which have been examined has two patent ducts — that of ^Yirsung, which is the important one, uniting 236 WILLIAM J. MAYO with the common duct of the hver. The minor duct of Santorini, however, has a possibility of useful function in certain diseased processes, as pointed out by Opie. The pancreas was originally an intraperitoneal organ, becoming retroperitoneal by a later evo- lution (Huntington), and in this anatomic peculiarity Brewer be- lieves lies one of the reasons for the diffusion of fat necrosis result- ing from acute pancreatitis. Mikulicz has called attention to the fact that adhesions to the pancreas in gastric cancer gave a mor- tality of 73 per cent, in his resection cases. Robson also notes the pancreatic mischief occasioned by perforating gastric ulcer on the posterior wall. The blood-supply of this group of organs is almost entirely from a single source in the celiac axis. It has been developed experimentally that the severance of all connections of the pancreas excepting its blood-supply does not check secretion if food is placed in the gastro-intestinal tract. In the nerve-supply from the pneu- mogastrics and sympathetic ganglion we find the same direct rela- tionship involved. If we study the function, we see the same asso- ciation. The stomach can be compared to a mill, the fundus the hopper, in which the food is macerated in a weak solution of pepsin and hydrochloric acid, and the muscular pyloric portion the grind- stones in which the masses are broken up into a homogeneous whole. The entrance of food into the duodenum causes the outflow of biliary and pancreatic secretions, the absorption returning to the liver by way of the portal vein. This is so elementary that you wonder that I should refer to it, yet the causation of the common surgical lesions lies in perversion of these fundamental functions and is just as simple. Mechanical injury of the pyloric portion and excessive acidity of the gastric secretions, under anemic condi- tions, give rise to ulcer and lie behind the precancerous lesions which Ochsner notes are found in the history of cancer of the stomach in the majority of cases. The acidity of the gastric secretions renders the contents of the stomach, when turned into the intestine, rela- tively sterile, but increases the liability of ulcer of the duodenum. The sterility of the upper intestinal tract is still further increased by intestinal absorption, as shown by Adami, the bacteria being SI ii<;n \i, hKsioNs i\ i rrKi; ahdo.mia' 237 picked u|) and dc'slroycd in |)aii l)y llic glands. Many ^('rni> arc, however, carried to the li\-cr, and here eitlier annihilated or screened out ol" the hl.ood in the portal \('in and discliar^'ed with the Itile. We iniist h)ok on the hile as ai\\a\s contaiiiinfi a few bacteria, and it is pr()l)ahl\- this alttMiiiatcd infection of l)ile retained in the la.sly must be discarded. W liilc \\c had no deaths in 20 operations, we liad 7 reUipses. In calling attention to the inali^'nanl di.sea.ses of this grouj) of or^'ans the writer would enij)hasi/,c the possil)ilities of cure hy means of operation. In the gall-bladder we found that 4 per cent, of cases at the operating-table had malignant disease, and all the.se cases had gall-stones present or evidence that they had been present at one time. Since it has become the practice to remove all thick-walled gall-bladders as useless and a possible source of future trouble, many cases of malignant disease in an early stage have in this way, as one might say, been accidentally cured. We have met several such instances. Cancer of the stomach is the most common type of malignant disease in the human body, constituting one-fourth to one-third of the total number. The radical treatment of gastric carcinoma is now on assured ground, with a mortality of 10 per cent., or less in favorable cases, to 20 per cent, in late but still operable cases. The only necessary factor to insure success is an early diagnosis, and this must be on clinical grounds, supplemented by early ex- ploratory incision. In 46 gastric resections for pyloric cancer we had 7 deaths. The profession may well .look upon the surgical achievements in this new field of w^ork with pardonable pride. That there are many shortcomings must be admitted, but in the history of surgery there has never been a territory opened up with equal rapidity, nor one in which the physician and surgeon have worked together in such harmony for the common good. DUODENAL ULCER* A CLINICAL REVIEW OF 58 OPERATED CASES, WITH SOME REMARKS ON GASTROJEJUNOSTOMY WILLIAM J. MAYO Duodenal ulcer has been considered a rare malady, and sur- gically it has not received the attention its importance merits. In a paper read before the American Surgical Association in May, 1900, Weir analyzed the cases reported in literature, and with observations drawn from his own experience placed the subject on a sound foundation. Interest has been still further quickened by a number of other investigators, notably Murphy and Brunner. It has been stated that nearly all duodenal ulcers are secondary to gastric ulcers, and that the two are usually combined in the one case. This has not been entirely borne out by our experience — at least the gastric ulcer, if present, has not been of the same grade and character as the duodenal. It is, of course, possible that a round or fissured ulcer of the stomach might have existed without recognition from the exterior of the stomach-wall. Based on the same examination, in 10 out of our 58 cases of duodenal ulcers there was a separate distinct ulcer found upon the gastric wall. In 18 cases the pylorus was involved by a lateral extension of the duodenal ulcer, making 28 out of 58, or about 50 per cent. In 100 cases of chronic gastric and duodenal ulcers recently reported by Moynihan,t 22 involved the duodenum, in 9 the lesion was confined to the duodenum, and in 13, separate and distinct ulcers existed upon both gastric and duodenal walls. Up * Reprinted from "Annals of Surgery," December, 1904. t" Annals of Surgery," May, 1904. 244 DIOOKNAL I l.CKIt 245 to two years a^o 11 per cent, of tlio <::asliic and diKKlciial ulcers wliicli came under our care involved tlie duodenum, SO per cent, t lie sloniacli. I )urinii I lie past year, wit li more careful ohserxat ion, we find "il per cent, duodenal alone, or comhined with pistric This includes a numerous j^roup of duodenal ulcers which extend up to and involve the pyloric rintr. We have seen a numher of pyloric iiIchms due to lateral extension and involvement of a i^astric ulcer, and in se\eral of these the duo- denum was attacked on one margin. In others the gastric wall was involved in a duodenal ulcer, the classification of gastric or duodenal being based upon the extent of the involvement. In all the cases of duodenal ulcer with 5 excej)tions the ulcerated area was easily identified as a thick, opatpie spot, puckered in ap])ear- ance, and usually covered by peritoneal adhesions, closely resemb- ling the large, irregular gastric ulcer of Robson, and furthermore it has been more frequent in adult males. In this variety of gastric ulcer Seymour Taylor found 72 males to 28 females. In the series of 58 duodenal ulcers herein reported, 43 were in males and 15 were in females. In 2 of the acute jierforating cases the ulcer was clean, clear cut, and set in normal tissues; in 4 the perforation was through a thickened area. In the chronic cases a few completely surrounded the duodenum; in others the outlines were irregular and of various sizes and shapes, and in the smallest at least 1 cm. in diameter. ^lost of them involved a considerable extent of in- testinal wall. The 5 cases without any appreciable thickening are of great interest; of the 2 acute perforations just referred to, 1 gave a history of four years' chronic trouble; the second, of but two weeks; no other ulceration of either stomach or duodenum could be detected. In the third there was chronic hemorrhage with acute exacerbation; the fourth case, in which gastro-cnterostomy was done, died six months later from another cause, and post- mortem did not disclose macroscopic evidence, at site of previous slight thickening, of any defect in the mucous membrane; the fifth case was buried in adhesions; evidently there had been a minute perforation, which, however, could not be identified. This would seem to indicate that typical round ulcer of many years' 246 WILLLVM J. ilAYO standing may exist without involvement of the outer coats, and therefore give little or no external evidence of disease, just as hap- pens in the stomach. It is probable, however, that most duodenal ulcers are of the cicatricial type, and in our series there has been a relatively greater tendency to perforate than in gastric ulcer. This is shown by the peritoneal adhesions which are so often found, and by the frequency of what may be called chronic perforation, protected by a mass of adhesions to the liver, gall-bladder, or gas- trohepatic omentum. These structures often form a plaster over the perforation and protect against extravasation of bowel con- tents. Such chronic perforations were found in 10 of the 58 cases. In but 2 patients did there appear to be more than one ulcer of the duodenum, and in one of these there was some question as to whether there was not some connection between. In 2 cases of supposed gall-bladder disease which we had opened and drained on account of adhesions thought to be due to cholecystitis without stones, no relief followed, and reoperation became necessary. At this time more careful investigation revealed duodenal ulcer. In a previous paper * the 'WTiter referred to 4 cases of periduo- denitis of unkno^Ti origin, operated upon for supposed gall-bladder disease, and in which the condition of the gall-bladder did not bear out the presumption. One of these cases has since been reoperated and duodenal ulcer found. Considering the known errors and the possibilities springing from a predetermined gastro-enterostomy and imperfect examina- tion of an ulcer situated in the pyloric region, it is evident that duodenal ulcer is a far more common condition than has been supposed. The situation of the duodenum and the thinness of its tissues render it especially liable to erosion from irritating gas- tric secretions. Its sheltered situation fortunately enables pro- tective adhesion in many cases, while its limited capacity and freedom from obstruction beyond prevent tension. The normal condition of relatively sterile contents, especially in the class of cases under discussion, is also a favorable circumstance. All the cases of duodenal ulcer occurred in the first two and *" Annals of Surgery," July, 1903. DIOOEXAL ULCER 247 one-half iiulics of the bowel, entirely above the entrance of the eomnion duct, and therefore in an acce.ssil)le situation. Errors in examination should he largely eliminated, and in all doubtful eases of gall-stone disease and gastric ulcer in which duodenal ulcer is possible the first portion of the duodenum should be in- spected. For this reason we now employ a longitudinal incision, one inch to the right of the median line, through the rectus muscle. This enables careful examination of the duodenum, gall-bladder, stomach, and pancreas. If more space is needed, Bevan's lateral curved prolongation of the incision at either the upper or lower end, or both, gives additional access to these organs. Longitudinal incisions through the body of the muscle close well, and are more reliable against hernia than when located in the median line. The operative indications are few. The causation and con- tinuation of duodenal ulcer depend on the irritating gastric secre- tions. These should be diverted by a gastro-enterostomy. If acute perforation exists, suture of the opening and cleansing of the infected area in the peritoneal cavity, combined with gastro- enterostomy, if the patient's condition warrants it, best fulfil the indications. Should there be extensive peritonitis, pelvic drain- age and the nearly sitting posture (exaggerated Fowler position) should be instituted for a few days following operation. For convenience, the 58 operated cases are divided into 5 groups: (1) Acute perforation; (2) hemorrhage; (3) chronic ulcer with gastric complications; (4) chronic perforating ulcer w4th gall-bladder and liver complications; (5) chronic ulcer requiring operation for relief of pain and distress. Group 1: Acute perforation, 6 cases, ^ deaths. Acute per- forating ulcer was found 6 times, in all but one a complication of chronic ulcer with a history of four to twenty-one years' standing. In 4 there was some attempt at adhesions, the acute perforation evidently occurring at a site of a partially protected area. In 4 of these patients there was a considerable sized opening found. In 2 already referred to there was a clean-cut perforation through what was otherwise normal bowel-wall. In 4 suturing was easj% in 2 difficult and unsatisfactory, requiring gauze packing. One 248 AVILLLOI J. :VL\YO of these patients died from inanition from prolonged leakage, although a gastro-enterostomy was done at the same time. In the other the gauze pack was left undisturbed eleven days, and rectal feeding employed for five days. The leakage was but slight and the fistula soon closed. In 4 the suturing held perfectly, but, unfortunately, 1 died from pneumonia on the tenth day. In this patient a gastro-enterostomy had also been done. The condi- tions were unusually favorable for an operation, which was per- formed within two hours of the perforation and the gastro-enteros- tomy was considered ad\'isable, as it appeared certain that stenosis would follow, since it had existed previously. Superficially, this would seem to argue against gastro-enterostomy in acute perfora- tion, but in each case the postmortem showed the gastro-enteros- tomy to be perfect. Group 2: Hemorrhage, 1 case, 1 death. There was a single case of prolonged and repeated hemorrhage in a chronic ulcer of three years" duration; during a two weeks' period of observation the stools showed constant evidence of blood. There was a single hematemesis. The ulcer was easily recognized as a little, thick- ened patch of otherwise normal bowel-wall, and was excised with pyloroplastic enlargement. The man was markedly anemic and a poor subject for anything but a forced operation. Valuable time had been lost in attempting to build him up. Death from pneumonia occurred on the fifth day. Group 3: Duodenal ulcers with gastric complication, 28 cases, 1 death. Chronic ulcer with compHcations from interference with gastric drainage was the most frequent form met, and gastro-en- terostomy was performed, with recovery in each instance except one, in which acute obstruction of the transverse colon followed anterior gastro-enterostomy. At postmortem a long prolapsed transverse colon was found hanging over the jejunal loop as it would over a clothes-line. The symptoms were not acute until a few hours after death. Operation should have been done. Group 4-' Duodenal ulcer with gall-bladder and liver complica- tions, 11 cases, 1 reoperation, no deaths. In this series gastro- enterostomy was performed in 7 cases, with successful outcome. DUODENAL VlAKli 24!) In 4 cases plastic ojjcralioii was rcs(jrtc(l to, with or \vitli(jut ex- cision of the nicer. In ii ji mocHHed pyloroplasty, 1 after the plan of Heineke-Mikulicz with a poor resnlt, a secondary gastro-enter- ostoniy heinff re(jiiire(l within three months, the other 2 after tin- plan of Kiiirie>-, with snccess. In the tliird the (Inodeinini, at a point 2 inches from (he pylorus, was acutely flexed upon itself hy adhesions to the liver, due to a closed adherent perforation. This made it possible to do a plastic operation upon the first portion of the duodenum without disturbing the adhesion. Group 5: Thirteen cases, no deaths. In this small series there were sym{)toms calling for operation which did not involve the stomach or gall-bladder, and further emphasized the fact that it was usually the complication which hastened operation. In practically all these cases adhesions marked prev^ious attacks of regional peritonitis. Gastro-enterostomy was performed in each, with good results. To recapitulate, there were 59 operations in 58 cases Of these, 7 were for acute conditions developing, with one exception, upon chronic ulcer, with 3 deaths. Fifty-one operations for chronic conditions, with 1 death. At the present time posterior gastro-enterostomy would appear to be the operation of choice in the chronic cases, but the last word has not yet been said. The time elapsed since operation in the majority of the cases herein reported suggests the possibility of further sequelte, particularly in those cases in which there is no obstruction, and in which experience has taught us that at least partial closure of the gastro-enterostomy opening may take ])lace. It is almost certain that even with a large gastro-enterostomy the food will pass out by preference through an unobstructed pylorus by muscular action, the apparent gravity advantage of a low-jioint gastro-enterostomy being equalized by intra-abdominal tension. Gastro-enterostomy performed for gastric ulcer is open to the same objection if there be no stenosis. For this reason, when the ulcer does not cause at least partial obstruction, it may be necessary artificially to block the pyloric outlet. An occasional complication following gastro-enterostomy is 250 WILLIAM J. MAYO bile regurgitation into the stomach. Acute vicious circle will sel- dom be seen if the opening be made at the bottom of the gastric cavity. We had but one case in 316 gastrojejunostomies (exclud- ing our first 14 cases), and that one in a patient eighty years old. Chronic bile regurgitation is a more frequent condition, beginning, as Ochsner points out, within three months, if at all, although it may be a year or more before it becomes troublesome. Carle and Fantino have shown that a little bile is to be found in the stomach at times in nearly all cases after gastrojejunostomy, and state that it does no harm. We have had a number of pa- tients complain bitterly of the distress occasioned. As a rule, the food passes out quickly, but there will be attacks of biliary regur- gitation at intervals of days or weeks. We have observed this phenomenon after the various methods of operation, and have been compelled to reoperate a number of times to check the disturbance. If the patient is in good general condition, we now perform a posterior suture gastrojejunostomy with a nine- or ten-inch loop, after the clamp method introduced into this country by Moyni- han.* Four inches below the completed gastrojejunostomy an entero-anastomosis with suture between the two limbs of the bowel is made, using the holding clamps. This adds ten minutes to the operation. A medium Murphy button is preferred by many surgeons for making the entero-anastomosis. To do this, the intestinal clamp is removed after the two posterior rows of sutures are introduced in the gastrojejunostomy, and half the button passed down inside each limb through the incised jejunum to a point previously marked with the knife. The intestine is nicked, and the proper part of the button forced through and junc- ture made without a puckering suture (Weir). This adds about three minutes to the time. A few mattress sutures should be placed as a protection about the button if it is employed. The only patient lost in the last 61 gastrojejunostomies at St. Mary's Hospital was one in which the button union gave way suddenly on the sixth day, the patient having no bad symptoms up to that time. Secondary laparotomy *" Transactions of American Surgical Association," 1903. WJ.MAyo. Fig. 21. — Ulcer of duodenum. Pylorus blocked by infolding method A. Sutures in place, but not tied. !\Iethods B and C at points marked X X- \''iJU\\yo Fig. 22. — Posterior gastro-enterostomy with entero-anastomosis and infolding; sutures placed for obliterating intestinal interspace. X marks site of silver-wire constriction or complete division. Note that the drawing shows stomach and colon drawn outward and upward as in actual operation. Replacement of viscera reverses position and brings intestinal opening at the bottom of gastric cavity. 1)1 OIJKNAI. ru'Kit 2.51 was pcrforiiic'd ten hours hiler, huL Ihc piilicut dit-d. Alter coiii- pletinj:; the l)uttori cntero-anastomosis the gastrojejunostomy is finished in the. usual manner, and the opened mesocolon all ached to the jjosterior wall of the stoniaeh in several places. To prevent hile arising to the level of the stoniaeh, and also to cause the food always to pass out the efferent bowel, the afferent intestine between the entero-anastomosis and the gastrojejunos- tomy should be closed in one of three ways: Method A: Infolding may be practised after the plan of Scott- Matolli, a continuous linen or silk suture an inch and a half in length turns the periphery of the intestine into the lumen (Fig. 22). Method B: Fowler accomplishes the same result by passing a No. 20 silver wire twice about the afferent loop at point X (Fig. 22), and twisting tight enough to obstruct without injury to the circulation, the twisted ends of the wire being turned closely into the wire loop.* Methods A and B prolong the operation about three minutes. Method C: We have in some secondary operations completely divided the afferent intestine at point X (Fig. 22), closing both in- testinal ends by a circular suture (Doyen), making the separation absolute; operation prolonged five to eight minutes. In all cases the open space between the two limbs of the intestinal loop should be partially closed by a few sutures at its lower part to prevent a coil of bowel herniating into the opening (Fig. 22). A comparison shows the infolding method to be the easiest, but Crile had a case in which, after a time, the infolded intestine straightened out and required another operation. The Fowler method is evidently more certain. The Doyen operation is, of course, sure, but takes a little more time and adds somewhat to the gravity of the procedure. When finished, however, it has all the advantages of the "Y" operation of Roux. Closure of the pylorus to divert all the food to the gastro- enterostomy is under consideration, and probably should be done in the large majority of cases if there is no cicatricial obstruction. The three methods already described for closure of the afferent *" Transactions of American Surgical Association," 1902. 252 WILLIAM J. MAYO intestine apply equally to the pylorus. The infolding method is shown in Fig. 21. The point for the application of the Fowler loop or complete division is shown by X (Fig. 21). We have either infolded or divided for the purpose of obstructing the py- lorus. Ochsner has used the wire loop a number of times success- fully for this purpose. The entire time of the combined operation should not exceed thirty-five to fifty minutes, according to the method chosen, in- cluding opening and closing the abdomen. In view of the fact that gastrojejunostomy is no longer a last resort, but an operation of choice to promote comfort and relieve disability, we must not only give a low mortality-rate, but also a high percentage of per- manent cures. A REVIEW OF 500 CASES OF GASTRO-ENTER- OSTOMY, INCLUDING PYLOROPLASTY, GASTRODUODENOSTOMY, AND GASTRO- JEJUNOSTOMY* WILLIAM J. MAYO The writer includes in this series all the cases in which an inci- sion was made into both the intestine and the stomach, and plastic union established between the two organs, with the intent to in- crease gastric drainage. The term "gastro-enterostomy" is used in its literal sense — the formation of an artificial passage between the stomach and intestine; the terms pyloroplasty, gastroduodenos- toniy, and gastrojejunostomy being used as expressing more ac- curately the exact method employed. The histories of the cases in this series have been worked out with a view to showing the actual results of operation, both as to mortality and as to the percentage of secondary operations. The method of computing the mortality is to charge as a death from operation every case dying in the hospital without regard to cause of death or length of time after operation. The series includes pa- tients dying as long as three months after operation, from coinci- dent chronic nephritis, etc., which might be called the combined mortality of operation and disease. This classification is hard on the statistics, but it eliminates the personal equation. The statis- tics include all the cases which have been operated upon in our clinic (C. H. and W. J. Mayo) up to June 20, 1905; they comprise the early cases showing a particularly high mortality. The secon- dary operations were repeated in some instances two to five times * Read before the American Surgical Association, July, 190j. Reprinted from "Annals of Surgery," November, 1905. 253 254 WILLIAM J. MAYO , before good results were obtained, so that the number of operations was nearly double the number of cases reported. Pyloroplasty, 21 cases, no deaths. Seven secondary operations (33^ per cent.). Gastroduodenostomy, Finney, 58 cases, 4 deaths (6.9 per cent.). Two secondary operations (3.4 per cent.). Gastrojejunostomy, 421 cases. Benign, 307 cases, 19 deaths (63^2 per cent.). In the last 140 cases there w^ere 4 deaths, a mor- taUty of 2y per cent. The last 80 cases gave but 1 death. Malig- nant, 114, with 21 deaths (18 per cent.). Of these 114 cases, 63 were in connection with pylorectomy and partial gastrectomy, with 8 deaths (13 per cent.). The very unfavorable cases of cancer obstruction were subjected to gastro-enterostomy, so that this operation gives a higher mortality than radical excision. In the last 40 gastrojejunostomies for malignant disease the mortality was 8 per cent. In the 421 gastrojejunostomies there were 21 reoperated cases (5 per cent.). Pyloroplasty The pyloroplasty of Heineke-Mikulicz, in our experience, has but little risk in suitable cases, but it is open to objection. The procedure enlarges the caliber of the opening as much in an upward direction as downward in the line of drainage, and the extent to which this enlargement can be carried out is limited. Following this operation, the pylorus is exceedingly prone to become ad- herent, so that the opening remains at a high level. The stomach, if greatly dilated, must elevate the food to the high-lying outlet, and it frequently happens that the degenerated muscle-fibers are incapable of the muscular effort, and, as a result, the patient is not materially benefited. In 3 cases we fastened the pylorus, after operation, to the region of the umbilicus by suture, to secure a low drainage point, taking advantage of the fact that adhesion after operation was the rule, to secure fixation at a more favorable situa- tion. These 3 cases have continued in good health, but there are valid objections to the procedure. In the 7 cases which came to secondary operation, the adhesions were most marked. Gastro- REVIEW or 500 CASES OF GASTRO-ENTEROSTOMY 55.5 jejunostomy in each case resulted in cures. In the remaining 14 cases cure resulted. In !• ca.ses an ulcer was exci.sed at the same time, with favorable result. Pyloroplasty has a small field of use- fulness, l)ut in performing it the later method of Mikulicz should be adopted. The incision should be curved downward up'on botli the stomach and duodenum, much like the Finney method, the result being an increased caliber over pyloroplasty as ordinarily performed, and establishing better drainage lines. The principle in plastic union, established by pyloroplasty, is one of the first importance, and widely used in surgery. It is especially valuable in choosing the line of closure after excising gastric ulcers, etc. Gastroduodenostomy Strictly speaking, this operation implies a separate opening between the stomach and duodenum, such as the Kocher operation; but, as the method of Finney more easily answers the same purpose, we have followed this plan in the entire group of 58 operations, with 4 deaths and with 2 secondary operations (mortality, 6.9 per cent.; secondary operations, 3.4 per cent.). In the first 46 cases there was only 1 death; in the next 12 there were 3 deaths. It does not seem fair to count "2 of the deaths, as one was from pneumonia after complete recovery and one from embolus due to an old endo- carditis. We had an opportunity to reexamine the operated field in 3 patients after a number of months — there were extensive ad- hesions present in '2. Patients should be carefully selected for operation. Extensive disease, adhesions, a short gastrohepatic omentum, and especially the presence of scar tissue should be considered contraindications, since it is in these varieties that gastrojejunostomy gives the most satisfactory results. Two of the 4 deaths were due to suture leakage on account of tension in scar tissue. In open ulcer the food must still pass the ulcer area to reach the pylorus, and ulcer does not depend on obstruction, as shown by the frequency of duodenal ulcer beyond the possibility of obstruction. Reasoning on this ground, we would not expect the curative results 256 WILLIAM J. MAYO from gastroduodenostomy in active ulcer which we would get from gastrojejunostomy made to the left of the ulcer-bearing area, and our experience bears this out; in this class of cases it has not given the same measure of relief. The importance of this objection is somewhat minimized by the fact that the line of enlargement is not only downward in the line of drainage, but also along the greater curvature, which is seldom involved in ulcer. The opening can be made of ample size, and it avoids the risks habitual to gastro- jejunostomy, as it leaves the outlet at its proper situation. In 4 cases we were able to combine with the Finney operation an exci- sion of the ulcer. In selected cases the Finney method is the one of choice. There were two secondary operations in this group; in both individuals bile came into the stomach, causing distress. We had made the opening too large, as shown at reoperation. One case had severe hemorrhage from insecure suturing, and required re- operation in twenty hours, with recovery. Gastrojejunostomy Gastrojejunostomy, 421 cases. Benign, 307 cases, 19 deaths (6 per cent.). Secondary operations, 20 (6^/2 per cent.). Malig- nant, 114 cases, 21 deaths (19 per cent.). One secondary operation (0.9 per cent.). Of these operations, 63 were made in connection with pylorectomy and partial gastrectomy. The writer has been greatly interested in gastrojejunostomy. No operation has conferred greater benefits in suitable cases than this one. Unfortunate experience, however, sharpened the inves- tigation as to the causes of deaths, and the complications which we found to be more or less inherent in every method with which we became acquainted. In all but 3 of the fatal cases a postmortem examination as to the cause of death was obtained. Not until recently have we secured a method which could be depended upon to give good results steadily with a sufficiently low mortality to justify the employment of gastrojejunostomy in cases in which disability, rather than impending death, was the spur to operative relief. KKVIKW OK 500 CASES OF GASTKO-P^NTEROSTOMY 257 The first chiiin to invc.sti/:iation comes with the question. Shall the o[)eration be made anterior, after Wiilfler, or posterior, after von Hacker? Qf the total number of cases, 126 were anterior and 205 j)osterior. The mortality in the anterior c direct anastomcses hy the Kocher method, between tiie duodenum and stomach after partial gastrec- tomy, a second oi)eration was re(iuired in eight weeks to relieve angulation and obstruction. McGraiv Ligature. — Total, 36 operations, lienign, 17, ^2 deaths (11.7 per cent.). Malignant, 19, 3 deaths (15.7 i)er cent.). The McGraw ligature method anterior has been very free from bile regurgitation and is exceedingly safe. It can be placed in bad tissues and can be used in ])oor subjects. We have used this operation four times with a hemoglobin of less than 25 per cent.; once 20 and once 24 per cent., in bleeding ulcer, with re- covery; once with hemoglobin of 24 per cent, in cancer with acute obstruction, w-ith recovery; once with hemoglobin of 10 per cent, in cancer with hemorrhages simulating ulcer. The latter patient was scarcely conscious at the time of operation. He lived three days, and although a stout, heavy rubber cord was used and tightly tied, there was no sign of an opening at the postmortem. It requires some vitality to cause the tissue to cut through. This man did not have sufficient resistance to set up atrophy necrosis, and the result was the same as in a cadaver. Two cases of can- cerous obstruction with a considerable quantity of free fluid in the abdomen recovered after a ligature operation. Tissues which are of poor vitality, but which have some power of repair, will do so after a McGraw ligature. The button might set up an uncir- cumscribed slough, or the suture become easily infected, if these methods were chosen. The McGraw operation, including open- ing and closing the abdomen, can be done in twelve minutes, with- out hurry. The disadvantages are that it should be, or has been, used with the loop, and, like any loop operation, the opening may contract. Again, it does not allow immediate feeding. This fact and the uncertainty of the time of the ligature cutting through render the method one for the occasional rather than the average case. We had one case of acute regurgitant vomiting after the 260 WILLL\M J. MAYO McGraw ligature, which was reoperated on the fourth day. The ligature had been badly placed, and the opening lay at one side of the center of the bowel. Posterior Suture. — Total, SSS operations. Mahgnant, 10, with 2 deaths (20 per cent.). Benign, 218, ^\dth 11 deaths (5 per cent.). We do not do an anterior suture operation. The increased risk of contraction and jejunal ulcer which unavoidably attends the anterior method would, mth the suture, also increase the chances of bile regurgitation. In May, 1901, Mr. Robson demonstrated in this country the bone bobbin operation, with the suture on a posterior 10-inch loop. We did 15 operations by this method, with 1 death. In June, 1903, we began using the method of jSIikuhcz, making the opening within 3 or 4 inches of the origin of the jejunum, and using a transverse incision. We made 43 operations by this method, with 4 deaths, 2 of which could be fairly excluded. Four patients required a second operation at our hands, and to a large extent because we departed from the originator's technic. It came about in this way: The transverse intestinal incision Hmits the size of the opening to one-half the diameter of the intestine, less about one-fourth inch suture line, and the opening could seldom be made larger than would admit the invaginated thumb. We tried to enlarge this by encroaching on the bowel, and caused a valve to form which turned the bile into the stomach. These patients gave us a lot of trouble; the short upper limb of the loop made an ordinary entero-anastomosis of the two arms of the bowel impossible. We finally united the intestine each side of the open- ing in exactly the same manner as the Finney operation at the py- lorus. The result was good. This was our first experience with the short proximal loop; the patients which recovered after this method have remained in splendid condition, despite the small opening. In October, 1903, Charles H. Mayo did two operations vnih. a longitudinal intestinal opening without a loop and as short as possible — ^practically the operation we are doing now. Both recovered and remain well. Kit,'. 2.5. — ShoNviriR posterior wall of stomach drawn lhrou);h a rent in the transverse mesocolon. Note slight separation of Rastrocolic omentum from its attachment to the stomach, permitting anterior wall of stomach to appear, and insuring drainage at lowermost level. Black lines mark site of proposed anastomosis; the jejunum shows at its origin. Fig. 24. — Forceps in place and anastomosis half completed by suture. REVIEW OF 500 CASKS OF CASTRO-ENTEROSTOMY SOI III the siiiiiincr of 1!)0.'{ Mr. Moyiiiliaii (Icmonstriitofl to us tlie iiietliod he Avas usiiif^' with the clamp, using the oblique; j)osterior incision of the stomach-wall and a 9-inch looj). The use of the (•lamps simplified the technic, and the opening <-ould he made very large. \Ve made 53 operations after this method, with 3 deaths. The i)rimary results with the Moynihan operation were good. Not a case of acute vicious circle, but in the course of a year 7 patients recpiired a second operation for the distress caused by bile regur- gitation, either occasionally in large quantities or frequently in smaller amount. In June, 1904, we began the operation on the Roux principle, doing a posterior gastro-enterostomy on a 9-inch loop with entero-anastomosis. The proximal loop was then ob- structed — (a) By Scott-Matolli suture; (6) Prowler's silver wire; (c) division of intestine, turning in both ends after Doyen. There were 48 of these, with 2 deaths. Two required a second operation; in one the silk suture in the longitudinal plaiting had passed into the intestine. The infection caused adhesions, angulation, and bile regurgitation. In the second, in which the proximal intestine had been divided, the cut end intussuscepted through the upper part of the lateral anastomosis, causing obstruction. This com- plicated operation was of too serious a character to apply to every case, requiring from forty-five to fifty minutes for its performance, and on January 1, 1905, we began the routine use of the posterior suture operation without a loop, in the same manner employed in the two eases operated upon by Charles H. Mayo in October, 1903. The operation of choice is without a loop. Fifty-six operations, 1 death (1.8 per cent.). This operation became popularized by the writings of Peterson, of the Heidelberg clinic. Czerny used the method for years, usually Avith the Murphy button, and with splendid success. At the time Peterson brought out the favorable features of the method we had practically abandoned the button for the suture in benign disease, and the operation could not be easily done with a longitudinal o})ening without the holding clamps, which Moj'nihan and Little- field later popularized. Mikulicz, as already pointed out, was doing the operation with the transverse incision. 262 WILLIAM J. MAYO Properly to appreciate the advantages of the "no-loop" method, some physiologic and anatomic facts must be understood (Fig. 23). The stomach is not a bag, but a muscular organ, and when empty, the pylorus is not far from the lowest point, and lies nearly in the median line of the body. As the stomach distends, the organ becomes more nearly horizontal by the elongation of the greater curvature. The pylorus passes to the right of the median line, and relatively passes above the greater curvature. The gastro- epiploic artery sets away from the greater curvature about three- fourths of an inch when the stomach is empty, and sends its gastric branches upward on the anterior and posterior gastric wall, which it meets above the actual line of the greater curvature. This ar- rangement of the blood-vessels enables rapid distention of the stomach, without interference with the blood-supply. The lesser gastric curvature is more fixed in its position and can be divided into two parts : the perpendicular portion, which drops nearly ver- tically from the right margin of the cardiac orifice (about one and one-half inches to the left of the midline), and the horizontal or slightly curved portion, which turns sharply to the right and ends at the pylorus. Ordinarily, the concavity of the lesser curvature is from two to three inches, the corresponding point on the greater curvature being three and one-half to four and one-half inches, making the convexity of the pyloric segment. This is the grinding muscular portion of the stomach, the part subjected to the greatest amount of traumatism, and over 80 per cent, of all the lesions for which we are called to operate are either in this part or in the first two and one-half inches of the duodenum. The gastrojejunal opening should be placed just to the left of this portion of the stomach (Fig. 24) . The inferior margin of the gastrojejunostomy should lie at the greater curvature, on a line opposite the juncture of the transverse and longitudinal parts of the lesser curvature. This is rather farther to the right than has usually been practised. The duodenum passes through the mesocolon, nearly on a perpen- dicular plane with the cardiac orifice, one and one-half inches to the left of the midline, and when the stomach is empty, its lower border lies nearly on a line with the origin of the jejunum. When IU;\IK\V OF 500 C'AHF:S ok fJASTKO-IlNTKHOSTOMY 203 the gastric ciix'ily is (iislcndcd or dilated, it dc.scoiids and covers this point. The j)roi)er situation of the gastric opening should be ol)Hciue on the jjostcrior wall, hcj^inniiig on the body of the stomach between the lesser and greater curvatures, and extending flownward to the very bottom of the stomach — Moynihan's line (Figs. 24 and 25). To insure that the opening shall be at the very lowest point at its right margin, we slightly separate the omentum from the greater curvature, and pull one-fourth inch of the anterior wall out posteriorly, pushing the gastro-epiploic vessel out of the way (Fig. 26). In previous contributions to this subject we have called atten- tion to the changed nature of the proximal arm of the jejunal loop after gastrojejunostomy. It becomes succulent and thick. With the loop operation, the food could and did pass into the proximal arm, while the {)eristalsis of this short end was inefficient, and herein lay most of the difficulties. Peterson pointed out that if the jejunum were attached short without a loop, it would require a reverse peristalsis to carry food into the duodenum. Peterson's point of jejunal election lies within from one to three inches of the origin of the jejunum, varying as necessary to enable easy attach- ment to the stomach (Figs. 23 and 24). It will be noted that the line of proposed union is a natural one. The jejunum attaches to the stomach without kink or bend in the line of gastric activity (distention and contraction). The intestine comes off the bottom of the stomach as though it were mortised on, the opening extend- ing upward and to the left. As Cannon and Blake have experi- mentally j)roved, and we have clinically demonstrated, the food will pass out of the unobstructed pjdorus after any method of gastro- jejunostomy; with this method of operation it makes little differ- ence. Spasm of the whole pyloric end of the stomach, which quickly follows ulcer or other irritation, no longer holds back the food and secretions. The gentle compressing action of the cardiac end is quite sufficient to turn the secretions and delayed ingesta out the gastrojejunostomy. Relief of the pyloric obstruction, no matter what the character, permits normal progress. Interruption of this calls the new opening into use. It is not at all necessary for 264 WILLIAM J. ilAYO the chyme to enter the duodenum to stimulate pancreatic and bil- iary discharge; this happens whenever the gastric product enters the small intestine at any point. In all the loop operations, more or less of the jejunum, at its most important situation, is thrown into a by-channel. The opening of the common duct Hes four inches below the normal pyloric entrance of food. This operation brings the common duct opening eight inches above the gastro- jejunal food entrance, and the constant presence of biliary and pan- creatic alkaline secretions will certainly render secondary jejunal ulcer less frequent than primary duodenal ulcer. To recapitulate: 1. The gastric opening should be placed on the posterior wall, obliquely from above downward, and left to right (Fig. 23) (Moy- nihan's line). 2. The lowest point of the gastrojejunostomy should be at the lowest point of the stomach, on a plane perpendicular with the cardiac orifice (Fig. 23). 3. To insure this effect, the gastric incision should extend one- fourth to one-half of an inch on the anterior wall (Nos. 1, 2, 3, and 4). 4. The incision in the intestine should be longitudinal, opposite the mesentery, and begin from one to three inches from the origin of the jejunum, measuring on, the anterior surface. (Peterson's point of election, Fig. 23.) The exact distance depends on the ease of attachment — as short as can be conveniently done ^dthout tension. A description of the operation is, briefly, as follows: (a) The abdominal incision is made four inches in length, three-fourths inch to the right of the middle line, the fibers of the rectus muscle being separated. The lower end of the external wound lies op- posite the umbilicus. This opening also enable inspection of the duodenum and gall-bladder and guards against hernia when closed. (b) The transverse colon is pulled out, and the mesocolon made taut by traction upward and to the right, in this manner bringing the jejunum into view at its origin, (c) About three to four inches of the jejunum opposite the mesentery are drawn into Fig. 25 — Completed operation ln)ni behind marfjin of torn mesocolon attached by several interrupted sutures to line of union. Fig. 26. — Completed operation from in front. Anastomotic opening sliows througii as darkened area on posterior wall. Note that it goes to the bottom of the gastric cavity and slightly anterior, as indicated by suture line in the omental attachment. REVIKW OF oOO CASES OF (JASTRO-ENTEROHTOMY '2(!.3 ji sli^'lilly ciirwd clainp. Tlie liaudk-s of the chiin|).s should he to the riglit, to enable a short grasp on the intestine. Three- fourths of the circumference of the bowel is pulled through; tlie posterior border is not included, which prevents entanj-de- ment of the suture with the redundant posterior mucous mem- brane. The holding clamps are ai)plied sufficiently tight to check hemorrha<;c and prevent extravasation of intestinal contents. (d) The ligament of Trcitz is a short muscular mesentery cov- ered by a variable [)eritoneal fold (too variable for a reliable lantl- mark), extending upward from the origin of the jejunum on to the mesocolon. This j)eritoneal fold lies at the base of the arterial looj) of the middle colic artery which supplies the transverse colon. The mesocolon is opened within the vascular loop, and the posterior inferior border of the stomach pushed through. A small separation of the greater omental attachment to the stomach enables the anterior gastric wall to be drawn out posteriorly. The posterior gastric wall is drawn into a clamp, with the handles to the right, in such a manner as to expose the anterior wall at the base. (e) The two clamps are laid side by side and the field carefully protected by moist gauze pads. ^Yith fine, celluloidal linen thread, on a straight needle, the intestine is sutured to the stomach from left to right by a Gushing suture at least two and one-half inches. (/) The stomach and intestine are incised one-sixth inch in front of the suture line, and the redundant mucous membrane excised flush with the retracted peritoneal and muscular coats. With a No. 1 chromic catgut on a straight needle the posterior cut margins of the entire thickness of the gastric and jejunal wall are united by a buttonhole suture from right to left; at the extreme left the suture changes to one which i)asses through all the coats of each side alternately, from the peritoneal to the mucus, then directly back on the same side from the mucus to the peritoneum. This acts as a hemostatic suture, and also turns the peritoneal coats into apposition. It i)asses around the anterior surface and is tied to the original end, which has been left long for the jiurpose. If silk or linen is used for this suture, it may hang in situ, suppurating for months, (g) The clamps are now removed, and the linen thread 266 WILLIAM J. MAYO continued around until it is tied to the original end, firmly catching the blood-vessels in sight along the suture line. The parts are carefully cleansed and inspected. If necessary, a suture or two is applied, accurately to coaptate or to check the oozing, (h) The margins of the incised mesocolon are now united to the suture line by three or four interrupted sutures, and the parts returned into the abdomen. When the patient is placed in bed, a glass female douche point is passed just above the internal sphincter ani, attached to a gravity bag filled with one-half strength normal salt solution. The elevation should not be greater than six inches. The small stream passing into the rectum is readily absorbed without irritation. One or two quarts are taken up in an hour (Murphy). The pa- tient is then placed in the semisitting posture. Beginning at from sixteen to twenty hours, an ounce of hot water is given every hour; this is rapidly increased, and in thirty-six hours the usual experi- mentation with liquid feeding is instituted. Rectal feeding is un- necessary. The operation is, in all of its essential parts, that of Mr. Moynihan. CHRONIC ULCER OF THE STOMACH AND FIRST PORTION OF THE DUODENUM, WITH ESPECIAL REFERENCE TO THE SURGICAL TREATMENT* WILLIAM J. MAYO Chronic ulcer of the stomach is certainly a more frequent condition than clinicians would lead us to believe. Compare the results of autopsy findings with the clinical diagnosis upon ad- mission to the hospital, for example. Take three hospitals in Philadelphia — Blockley Hospital, giving 1.42 per cent, as the result of autopsy finding; University Hospital, clinical findings 0.48 per cent.; Pennsylvania Hospital, clinical findings 0.13 per cent. (Francine). In other words, in two hospitals of exactly the same character in the same city ulcer is found clinically nearly four times as often as in the other, while both fall short of the post- mortem from 3 to 11 times. Bettman finds that a diagnosis of gastric ulcer was made but 24 times in 27,567 Cincinnati hospital admissions (0.08 per cent.). Howard, in comparative tables, shows that New York city autopsy records give 1.42 per cent, of gastric ulcers, while the records of clinical admissions show only 0.44 per cent. Boston does better: autopsy, 1.84 per cent.; cHnical, 1.28 per cent. Francine says: "I entirely agree with Dr. Howard's statement that we cannot base accurate or conservative conclusions on data obtained from clinical observation." In 10,841 autopsies in 7 large American cities (Howard) the percentage of gastric ulcers was 1.32, while in Ix)ndon it was 4.6 per cent., and in continental Europe 8.54 per cent. Welch gives 5 * Read in the Section on Surgery and Anatomy of the American Medical Association at the Fifty-sixth Annual Session, July, 1905. Reprinted from "Jour. Amer. Med. .\ssoc.," 1905, vol. xlv. 268 WILLLOI J. :^IAYO per cent, in 32,052 autopsies in the Prague, Berlin, Breslau, Dres- den, Erlangen, and Kiel Hospitals. There are two explanations of the greater apparent frequency of gastric ulcer in Europe than in America. One is that it is not apparent, but real, and another that the pathologic departments of European cities have been permanently estabhshed for a long time and postmortems are obligatory, while until of late years the work in American cities has been under changeable management and subjected to a var- iable personnel; also that consent of legal representatives must be obtained, and therefore autopsies are relatively much less frequent in this country. As showing the eflFect of searching with a definite purpose, Griinfeld, of Copenhagen, found 11 per cent, of gastric ulcers in 1150 autopsies, and in the next 450, examined more care- fully, found 20 per cent. In regard to the relative frequency of gastric and duodenal ulcers, we have but few statistics. Francine found 38 cases of gastric ulcer in 2830 autopsies, 2 duodenal, also 2 duodenal and gastric, practically only 10 per cent, of duodenal. As compared with surgical findings, this is too small, and does not bear out the relative frequency sho'v\Ti by statistics of acute perforations. In 22 cases of acute perforating ulcers Moynihan found 15 gastric and 7 duodenal. Our experience with acute perforating ulcer is relatively smaller than with chronic ulcer. In 13 acute perforat- ing ulcers 6 were gastric and 7 duodenal. Brunner collected 600 cases of acute perforation, of which one-fourth were duodenal. He also showed that 90 per cent, of acute perforating ulcers occur through the site of chronic ulcers, and that diagnostic symptoms usually exist previous to perforation. The duodenum above the opening of the common duct of the liver and pancreas is exposed to the same ulcer-producing causes that exist in the stomach, ^\dth the possible exception of trauma- tism; and as its tunics are thinner, it is even more readily affected by irritating secretions and ingesta. Statistics appear to show that ulcer of the duodenum is a rare malady, but the data on which the supposition is based have been furnished either by postmortems or are the results of notoriouslv defective clinical examinations. CIIHOMC \lAK\i OK STOMACH AM) l)l()l)i:Nr.M — SL'HC;KHY 269 Postinortein study has c-ortaiii disa(lvanta{j;('s due to influences wliicli lia\(' i)erluii)s l)C(.-onie active either shortly before or at the time of death, and which often mask lli<> primary lesion and cloud the condition as it existed in life. Secondary clianj^es and ter- minal infections may prevent a correct interpretation of the signs jind symptoms which were manifested during the early stages of the disease — the curable period. This is shown by the revelations of surjfery in ap])ondicitis, extra-uterine pregnancy, and chole- lithiasis. The same potent force is now at work in the field of ulcer of the stomach and duodenum, and the first fact which has been demon- strated is that those forms of ulceration which affect all the coats of the viscera and which, by reason of their large size and thick, scar-like appearance, can be easily demonstrated, are nearly as common in the duodenum as in the whole of the stomach. The postmortem statistics which have been gathered are certainly erroneous in their summing up as to the relative frequency of gastric and duodenal ulcer. The only conclusion which one can arrive at is that the examination was general and did not cover the duodenum with the same care that it did the stomach. The duodenum has received but little attention and therefore it has not been subjected to close scrutiny during general autopsy. Admitting that clinical observation falls short of the autopsy findings, in what way do results of surgical work compare with post- mortem records ? Within fourteen years nearly 800 cases of gastric and duodenal disease have been operated in St. Mary's Hospital. Eliminating gastric cancer and all cases in which the necessity for operation did not immediately arise from ulcer, we have 384 opera- tions for gastric ulcer and its results, and 84 operations for duo- denal ulcer, or about 78 per cent, gastric and 22 per cent, duodenal; but this is not fair to the duodenal disease, because it has only been within a short time that we have recognized duodenal ulcers, and many of our earlier cases marked pyloric may have been an extension from the duodenum. We have thought it wise to narrow the limits of this to a consideration of the cases in the last two and one-half vears, from Januarv 1, 1903, to Julv 1, 190o, and also to 270 WILLIAM J. MAYO consider only the cases subjected to gastrojejunostomy, exclud- ing all of the ulcers excised or subjected to Finney's operation, pyloroplasty, etc. This gives us 231 cases, 119 males and 112 females, of which 158 were gastric, 60 duodenal, and 14 duodenal and gastric ; 20 of the duodenal ulcers extended up to and involved the pylorus. In others words, out of 231 gastric and duodenal ulcers, the duodenum was involved 74 times: 55 times in males and only 19 in females. Classification of Gastric and Duodenal Ulcers For clinical purposes I will classify all the ulcers operated on into two groups, the indurated and the non-indurated: First, the indurated ulcer, which involves all the coats of the organ and which usually shows evidences of cicatrization in some part of its extent. The diseased area is a thick, milky-white patch, easily identified from without the gastric or duodenal wall. In the stomach it involves the pyloric portion in the great majority of cases; it is frequently saddle-shaped, riding the lesser curvature, and extend- ing flap-like down the anterior and posterior walls. In such cases the pyloric portion beyond is usually thickened and gives rise to more or less obstruction, even if not actually involved in the ulcerative process. In about 20 per cent, more than one ulcer was found. In the duodenum the first 23^ inches is always in- volved well above the entrance of the common duct, with its alka- line discharges, and the ulcer extends up to the pylorus or within ^ inch of it. In only 3 instances was more than one duodenal ulcer shown. Sixty-eight of the 74 duodenal ulcers were of the indurated variety, and 151 of the total 231 cases of duodenal and gastric ulcers were so classified. Of these, 95 were males and 56 females. In 100 cases Seymour Taylor found 72 males and 28 females. Associated with this group of indurated ulcers are the benign pyloric obstructions of inflammatory origin, hour-glass stomachs, adhesions, and deformities arising from protected chronic perfora- tions. The non-indurated ulcer has also been termed medical or clinical, because although they give undoubted evidence of the W.J.Ma/o. \ \ V ^'tf*^* Fig. 27. — Showing forceps passed through from behind and grasping anterior gastric wall near the greater curvature at the lowest point. Saddle ulcer of lesser curvature near pylorus. Fig. 28. — Posterior wall of the stomach drawn through opening torn in transverse mesocolon. Forceps still marking low point. Dotted lines on stomach and Jejunum show situation of proposed anastomosis. CHRONIC ITLCKIl OK STOMACH AM) DLODK.VUM SURGERY 271 disease, there is nolhinj,' to show the ulcer site from the exterior of the stomach upon exploration. Tlie reason for this is that the lesion involves only the mucous coat. In some cases a little thick- ening can be discovered (Mikulicz), or a glueinf^ of the mucous to the nuiscuiar coats, preventing the normal slipping of one on the other (Moynihan). But in the typical case prolonged search of the interior of the stomach may be necessary to find the diseased process. Many individuals have bled to death from an ulcer so minute that it could be found only with the microscope. Bram- well says that many of these cases heal so minutely that no evi- dence can be found at autopsy. We have had a number of patients come to operation after years of trouble, — hemorrhages, stagnation and retention of food, etc., — who were cured by operation, but in whom no sign of ulcer could be shown on the exterior of the stomach. In some instances we have oi)ened and searched the interior of the gastric cavity to find an ulcer from which the patient had bled repeatedly, but we have not always found it. It is possible, or indeed probable, that in some of these patients an indurated ulcer may have existed in a situation not accessible to palpation or inspection. Non-indurated ulcers are of two varieties: (a) The mucous erosion of Dieulafoy, in which only the superficial epithelial layers of the mucous membrane are involved, and (b) the typical, round, peptic and fissure ulcer. 'In our experience the fissure-like ulcer has been of frequent occurrence. In one patient bleeding at the time of operation was found to proceed from a small fissure which could be detected only by bending the mucous membrane sharply, the little defect showing as the weave would show on folding a piece of velvet. Eighty of the 231 cases belong to this group of non-indurated ulcers, 56 being females and 24 males. One in- teresting diagnostic feature was first pointed out by Lund, He noted that an ulcer could sometimes be located by an enlarged "sentinel" gland in the omentum, tributary to the lesion. We have noted for a long time that in nearly all oi)en ulcers the tributary lymphatic glands were definitely enlarged to the size of a lima-bean, from 1 cm. to 1.5 cm. in diameter, usually in the gastro- 272 WILLLVM J. MAYO colic omentum, in this respect being unlike cancer, which affects the glands of the lesser curvature by preference. The enlarge- ment is soft and shows simple adenitis. It is probable that this may be a valuable diagnostic sign, and that we should find en- larged glands in all cases of ulcer, mucous or otherwise. The value of the sign is somewhat lessened because we have seen adeni- tis in the same situation in cases of cholecystitis; but in these cases enlarged glands were also found along the common duct. The question of non-indurated ulcer needs further elucidation. The very fact that the condition may not be cleared up at the operating table prevents us from gaining knowledge by the ex- perience. We know that the majority of well-selected subjects recover; but a minority give less favorable results, and operating for purely medical indications leads to unscientific and at times indiscriminate resort to operation. Especially is this true of that vast army of neurasthenics with gastric symptoms depend- ing on a neurosis — a complaint simulating ulcer: a prolapsed, splashy stomach, a too ready diagnosis followed by an ill-advised gastro-enterostomy. The fact that the stomach appears to be normal is explained by the inability always to locate a mu- cous ulcer. This is not a fanciful nor overdrawn picture. There is seldom a week but several such cases present themselves at our clinic and are refused operation. Many of them have already had their movable organs fixed (kidney and uterus), and the removable ones removed (ovaries, appendix, etc.), and now are anxious to secure relief by a further resort to the knife. That such cases are frequently operated on cannot be questioned, and that they will tend to bring surgery of the stomach into disrepute is equally certain. The so-called atonic dilatations are not often greatly benefited by operation unless there is marked and persis- tent delay of food in the stomach. This group is closely allied to the neurasthenic class, and the individual case must be carefully considered, with a prejudice against operation unless it can be clearly shown to be indicated. Fortunately, the field of gastric surgery is too large for a few of these unfortunate instances to affect the general results, but I would urge on the profession the Fig. 2Q. — Stomach and jejunum drawn into clamps (or suturing. Small forceps still marking low point of stomach. Fig. 30. — Shows relation of the duodenum to the stomach. CHKONK' 1:L( Kit OF STOMACH AND UUODEXUM — SURGERY 273 necessity of eliminating the neunithenic from the field unless the signs and symptoms of ulcer are distinct. Of course, the nervous condition is no. valid reason for refusing to relieve actual disease. There are some problems closely related with the non-indurated ulcer which are but little understood, and one of the most impor- tant of these is pyloric spasm. Numerous observations at the op- erating table have convinced us that pyloric spasm is not due to a contraction of the pyloric sphincter alone, but of any part or all of this end of the stomach. There is undoubtedly, also, such an entity as chronic contrac- tion of the pyloric muscle without actual demonstrable lesion, and to such a degree as to be the onh' evident cause of gastric dilata- tion, stagnation of food, and chronic distress. A serious form of interference with gastric motility has been noted by Ochsner, Finney, ]Munro, and others, in which there is chronic dilatation of the stomach and duodenum as far as the common duct of the liver, giving rise to many of the symptoms of obstruction. Ochsner believes this condition due to a pathogenic contraction of a normal excess of muscle in the second portion of the duodenum, which he has demonstrated anatomically. Another variety of interference with gastric motility is the not rare condition of " valve formation," in which a high-lying pylorus is held taut by a short gastrocolic omentum. I have met with a small number of such cases, three of which were reported before the surgical section of the American Medical Association at Atlanta, May, 1896. No final conclusions can be drawn from our personal experi- ence, but at present we do not advise operation in any case of acute ulcer, although certain complications, such as perforation, hemorrhage, and grave obstruction, may compel its speedy perform- ance. We do not advise operation in chronic ulcer or its associated diseases until careful and prolonged medical treatment has failed to cure permanently, and we strongly advise against operation in neurotic individuals with prolapse of the stomach. We advise operation in all cases of stagnation and retention of food depend- ing on mechanical causes, such as pyloric obstruction, and in cases of exhausting hemorrhages. We ad\ise and practise operation in VOL. I — 18 274 willia:m j. ^l^yo that considerable group of chronic cases "«'ith acute exacerbations, in whom frequent relapses with their attendant disabilities prevent the patient from the enjoyment of good health. It is this latter group which reminds us forcibly of the early days of appendicitis in which great divergence of opinion was made manifest, from the practitioner who rarely saw a case, to the equally honest man who saw them frequently but always cured them without trouble. We have gone through the same controversy as to the surgical treat- ment of gall-stones and other diseases. There are a number of careful observers who predict that the ultimate field of gastric surgery will be small, and that the diagnosis of surgical conditions cannot often be made; but this was equally true of the early days of appendicitis, of gall-stone disease, and of pyosalpinx. I think no unprejudiced person can doubt the con- clusion that gastric and duodenal ulcers and associated disorders are more frequent maladies than we have been led to believe, and also that ulcer or some of its numerous compHcations may and often does produce a train of symptoms which medicine is power- less to cure permanently. What percentage of gastric and duodenal ulcers may be ex- pected to be cured bj" medical means? Five hundred cases treated medically in the London Hospital in the five years from 1897 to 1902 gave a percentage of 18 for the death-rate, and 42 per cent, were not cured at the time of discharge. As 211 of the 500 had been cured one or more times of previous attacks, who can predict the future history of the 40 per cent, discharged as cured? Of the medically treated ulcers in the Massachusetts General Hospital, Greenough and Joslin showed that only 55 per cent, were discharged as cured, and 56 per cent, supposed to be cured were dead or still suffering at the time the report was made, five years later. Mumford reports only 4 per cent, of gastric dilatations resulting from ulcer as cured medically out of 122 cases in the Massachusetts General Hospital. Russell's statistics, derived from a large number of out-patients, show that 42.6 per cent, of patients with gastric ulcers recovered; but as it was the first attack in 27.7 per cent, and might, therefore, be called acute, this gives a recovery CHRONIC ULCER OF STO.MA( 11 AND DUODENUM SUROKHY "i t O of only 14.9 per cent, of the elironie cases. The hahince either died or continued to suffer (Blake). Munro remarks, "Is it not surprising that an increasing number of ulcer subjects are sponta- neously seeking surgical relief?" It is wise to be conservative and to compel each new departure to bring its own proof. What results can surgery show in this field? First, it has denionslrated the clinical frequency of ulcer of the stomach and duodenum, not a new thing, because it has been shown for years in the autopsy records; second, it has de- veloped a symptomatology which enables the diagnosis to l)e made, and has demonstrated the operative curability of ulcer and certain associated disorders. Surgery has brought back to a safe ground a large number of ulcer victims who, after repeated medical cures, had taken to fakirs, patent-medicine venders. Christian (?) Science (?), or were making the best of their condi- tion and using patent foodless foods and a restricted diet. In doing even this much, surgery has been open to sound criticism, first on the occasional selection of an unfortunate case for opera- tion, and, second, on the occasional unsatisfactory results of opera- tive interference both as to mortality and to permanence of cure. It is the surgeon's duty to overcome this prejudice by furnishing better results. The history of successful gastric surgery is not more than five years old, and the best of it not of over two years' duration. The medical man must discard the older statistics as to technic and mortality which have become merely venerable relics, and do not at all represent advanced surgical thought on the subject. It is certainly discouraging to turn to the newer works on medicine and find not the slightest attemjit made to show the advance in sur- gery, and the question of surgical relief being arbitrarily deter- mined by the achievements of tAvo decads ago. Surgery is essentially mechanical, and must benefit the patient in a mechanical way to a large extent (Fig. 26). Most surgical questions connected with chronic ulcer arise from interference with good gastric drainage, either by actual obstruction or by muscular spasm, so that the food and secretions are subjected to delay in 276 WILLIAM J. MAYO that part of the stomach lying to the left of the pyloric muscular portion, and the method of rehef "^'hich has the largest field of usefulness consists of gastrojejunostomy made on a line perpendic- ular with the cardiac orifice of the stomach (Fig, 27). This "will usually be found to be the most dependent portion. The opening must be placed on the posterior wall, at the very bottom of the gastric cavity, and should extend anteriorly 34 inch, so that the jejunum is mortised on to the stomach (Tig. 27). The line of the gastric opening should be that of ^Moynihan, obHque from above doT\Ti and left to right (Fig. 28). The jejunum should be anasto- mosed within three inches of its origin (Fig. 28), so that there shall be no loop (Fig. 29). After an experience of somewhat over 500 gastro-enterostomies, including gastroduodenostomies and pyloro- plasties, we have come to Peterson's conclusion that the loop has been responsible for the greater part of the evils arising after gastrojejunostomy, such as bihary regurgitant vomiting (vicious circle) . The intestine should be secured so high that there can be little loss of nutrient absorbents. The straight drop of the bowel gives protection against secondary jejunal ulcer by the constant presence of the alkahne bihary and pancreatic secretions, and also adds to the security against future comphcations. There is no doubt that contraction of the opening is less liable to take place if there is no loop to make traction. As to mortahty, we have had less than 3 per cent, mortahty in our last 150 suture operations, and in the last 81 cases of benign disease there has been but 1 death. These results are no better than those of Ochsner, Murphy, Munro, Deaver, Robson, Moynihan, ]\likuhcz, Kocher, Hartmann, and others. The operation here advocated has given us better results than any other which we have tried, but we have had two cases of chronic bile regurgitation occur, e^adently due to faulty technic. In both patients there has been great reUef of the original symptoms, and in neither has the complication as yet been of sufiicient mo- ment to require a second operation. Next to gastrojejunostomy, the operation of gastroduodenos- tomy, devised by Finney, is of the greatest value. It is espe- CHRONIC ULCEK OF ST( J.MAC II AM) IJUODEN L M^SLUGKUY 277 ciuUy suited to narrow strictures. In open ulcer it does not drain the stomach to the i)roxinial side of the nuiscuhir pyloric region, and the food must still pass into the area of the ulcer to reach the outlet. The pyloroplasty of Heineke-Mikulicz is now but little practised, and the method of closure is of the utmost importance after an ulcer is excised. As a matter of fact, the i)ylor<)plaslic principle has been one of the great factors in modern plastic sur- gery. The operation of Rodman, consisting of a complete ex- cision of the entire ulcer-bearing muscular pyloric end of the stom- ach, with independent gastrojejunostomy, will gain ground in the future. Graham has found a good precancerous history of ulcer or associated disorders in 36 per cent, of our operated cases of cancer of the stomach, and clear evidence of cancer development on ulcer in 30 per cent, of the last 40 pylorectomies and partial gas- trectomies — certainly an argument for the radical operation. Excision of the ulcer may be of value in a small group of cases if there be no obstruction and one is sure that only one ulcer exists. Admitting that the technic and mortality of gastric surgery are satisfactory, have the patients been relieved? Excluding some cases of bad selection, I can conscientiously say that we are doing no surgery today which gives more pleasing results in properly selected subjects than in the field of chronic gastric and duodenal ulcer and associated disorders. The disappoint- ments have been due to inability to secure and maintain good gastric drainage through imperfect technic, rather than failure of a properly executed operation to relieve. While gastrojejunostomy has the largest field of usefulness, we must not look upon it as a "cure-all." It is purely a drainage operation. If the stomach is not dilated and the pylorus be unobstructed, the food will continue to pass out the normal outlet and the patient will not be benefited. For this reason indurated ulcers with definite mechanical lesions give far better results than non-indurated ulcers in which obstruc- tions are not found, and it is this latter group which gives a con- siderable percentage of secondary operations and complications. We cannot agree with the opinion which has recently been ad- vanced that gastrojejunostomy should be done almost regardless 278 WILLIAM J. IVIAYO of the condition present. There is nothing mysterious about this valuable operation. It permits retained secretions and ingesta to escape readily. If motility is normal, it has little function. In conclusion, let me call your attention to the vast importance of this subject, as it enables us to differentiate the benign from the mahgnant diseases of the stomach. Nearly one-third of all can- cers in the human body are in the stomach. In 70 patients in whom we excised a large part of the stomach 4 "v\dth cancer lived more than three years, 3 are still alive and without return. A number of patients are alive over two years, and the majority live a year. The average mortality was 12 per cent., and in the last 40 cases there were but 2 deaths— a mortality of 5 per cent.* * "Annals of Surgery," March, 1904. THE SURGICAL TREATMENT OF CANCER OF THE STOMACH. REPORT OF 100 GASTRIC RESECTIONS* WILLIAM J. MAYO In the history of medicine we have no recorded example of a cancer of the stomach cured by medical means. Yet for some reason or reasons such cases are sent to the medical men, are entered in the medical wards of hospitals, and subjected to treat- ment which must result in 100 per cent, mortality. This is so true that, while cases of suspected cancer of the breast, the uterus, or the rectum are sent at once to the surgical side from the out- patient department, the possible victim of cancer of the stomach, even with a suspicion amounting almost to a certainty, is still sent to the medical ward. Yet of all the diseases of the stomach cancer is the one which should be treated surgically. External carcinomata may be treated badly, it is true, by plasters, but occasionally with success. Super- ficial epithelial growths sometimes disappear, at least for a time, under the x-ray. But gastric carcinoma has not even the small chance of relief afforded by these uncertain agents. Appendicitis is now universally conceded to be a surgical disease, yet some cases of appendicitis are known to recover spon- taneously and to remain cured. Extrauterine pregnancy may re- sult in a pelvic hematocele, with spontaneous absorption; but cancer of the stomach has no such possibilities. These examples are adduced merely to show the inconsistency of looking for a medical side to this question. It is worse than a blunder: it is a crime. *Reprintcd from "Jour. Amer. Med. Assoc.," .\pril 7, 1906. 279 280 willia:m j. aliyo The practitioner of medicine is not to blame for this state of affairs. He has retained these cases because the surgeon has shown Httle or no disposition to reheve him of the responsibility. There is no controversy, and no one is more anxious to turn these unfortunate victims of a medically incurable disease to the surgeon than the internist. Cancer of the stomach is the most frequent form of cancer found in the human body, and can be conservatively estimated at 30 per cent, of the total. WTiy has the medical profession been so slo"^ to apply surgical methods to the cure of this common malady? There are two important reasons : First, the frightful mortahty of the earher operations, which discouraged the patient, the physician, and the surgeon; second, the difficulties and uncertainties of estabhshing an early diagnosis. Review of Surgical Treatment The radical removal of cancer of the stomach was first per- formed by Pean in 1879, by Rydygier in the year follo\s*ing, and by BiUroth in 1881, but his was the first patient who recovered. Pean and Rydygier did not reahze the importance of the condition, and it remained for the master mind of Billroth not only to see its possibilities, but to establish the principles of operative relief. These remain today much as he left them, the changes being in technic rather than in new discoveries. Almost equaling Billroth in the importance of his early contributions to the operative treat- ment of gastric carcinoma stands the name of Kocher, and in selected cases the Kocher operation is the method of choice not only in the hands of its distinguished originator, but of practical surgeons the world over. The death-rate foUo^sang these early operations was appalling. Billroth's average mortahty at the time of his death was over 60 per cent. In 1896 Haberkant collected statistics of 257 pylorec- tomies, with a mortality of 64.4 per cent, before 1887, and 42.8 per cent, after that time. Goffe showed that the operative mor- tahty among the English and American surgeons was 76 per cent, before 1890 and 28.5 per cent, after that time. Guinard collated Fig. 31. — Showing cancer of pyloric end of stomach, with enlarged glands and the four blood- vessels tied, and line of gastric section: A. Left gastro-epiploic ligated; B, gastric artery ligated; C, suixrior pyloric ligated; D. gastroduodenalis ligated. '//J Mayo Fig. 32. — Showing duodenum clamped and pyloric end of stomach separated, ready for resection. SUItCIICAL TUKAT.MKNT (JF CANCER OF HTO.MAfU \iHl statistics of 201 cases between 18f)l and 18f)H, with a death-rate of .'3,5. .'i j)er cent. The lack of enthusiasm of all the parties concerned is not to he wonderecl at; nevertheless, there has been steady progress, and since 1900 the improvement in operative technic has been so marked that the mortality has become reduced to a remarkable degree. There have been comparatively few workers in this field, and the work has been so quietly carried on that the bulk of the i)rofession do not realize the enormous progress which has been made. To- day the mortality is probably not above 10 per cent, in the operable cases handled by men of experience, and in suitable cases nearer .3 per cent. Operations undertaken with the patient in extreme exhaustion from starvation and hemorrhage will continue to show a large death-rate. But these disasters should no more militate against the opera- tion in suitable cases than general suppurative peritonitis should stand in the way of early operation for appendicitis; they should lead us rather to an increased effort to secure the patients for opera- tion during the curable period. The stomach is a most favorable organ from an operative point of view. It has an abundant blood-supply from four sources, and the certainty of early wound healing makes plastic surgery safe. The immediate ligation of these four vessels makes radical opera- tion bloodless and devoid of shock, exactly as in cases of abdominal hysterectomy. By the use of clamps the entire area can be cleanly excised practically without opening the gastric cavity (Figs. 31 and 32). The gastric envelops are thick, with but a loose attachment between the combined peritoneal and muscular coats and the mucous membrane, so that a firm hold of the outer tunics can be secured which insures reliable union, while the mucous coat can be sutured separately. Running sutures are particularly effective and save much time (Fig. 33). The entire operation of pylorectomy and partial gastrectomy can be performed in from forty minutes to an hour and ten minutes, including opening and closing the abdomen; at this time only 282 WILLIAM J. SL^TO will an anesthetic be urgently demanded. In poor subjects the entire visceral part of the operation can be done without pain and without anesthesia. The preliminary administration of morphin hypodermically in the latter class of cases is a valuable adjunct to the anesthesia. We have done the Billroth No. 2, that is, the complete closure (Fig. 35) of both duodenal and gastric stumps and independent gastrojejunostomy, 76 times, and the Kocher operation 15 times (Fig. 34), and the Billroth No. 1, 9 times. Each method has its own field of usefulness in selected cases. For the average case the Bill- roth No. 2 is the operation of choice. Granting that the statistics of operative procedure are now ■ndthin reasonable hmitations, is the rehef afforded sufficiently great to make it worth while? General statistics are unsatisfac- tory, and we have, therefore, taken only those of Kocher, Kronlein, Mikulicz, and our own. In 1903 Kocher reported 75 cases, with an average mortality of 29.3 per cent. Of the 53 patients who recovered, 21 were alive at the time of the report and 6 had already lived more than three years, 1 alive and well after thirteen years and 1 after eight years. In the last 24 cases there were but 4 deaths (16 per cent.), and 2 of these, Kocher beheves, could be fairly excluded, giving a mortality of 8 per cent. Matti, of Kocher's clinic, brings the Berne statis- tics up to 1904, giving a total of 100 gastric resections. There was considerable improvement in the percentage of those remaining cured over three years, and it was further shown that those who died of recurrence averaged eighteen months of comfortable exis- tence, since gastric drainage was maintained to the end. In 1902 Kronlein reported 50 radical operations, with 14 deaths — 28 per cent. At the time of the report 22 patients were living — 4 three years and upward, 7 more than two years, and 13 more than one year. In 1901 Mikuhcz reported 100 gastric resections, with 37 deaths; 58 of the patients who recovered were traced. Seventeen were alive more than one year, 10 more than two years, and 4 more than two and one-half years. VUtuMO. l^'K- 33- — Showing closure of stomach, and the throuRh-and-through catgut suture two-thinls com- pleted, and outer linen continuous Cushing suture just started. ViJ I A/.(o Fig. 34. — Showing restoration of gastro-intestinal canal after Kocher method. Seromuscular linen suture in behind, and through-and-through catgut just beginning. SlKliii.il caii.il ri>l\ ii'liiniulciU |)n>iirior Kaslrojujunostiimy. Billroth No. :. Duixloniim anil iijiimim dotted in as they lie behind. SUHCK At. TKKAT.MKNT OF C'AN'f'ER OF STOMACH 28.5 ciinislaiKc.s. 'J'lio clinical liislory, with tlic clicinical and hiolofjicai exaininalion of tiie stomach-contents, can only lead to a suspicion; and we nmsl, act upon this il" we arc truly conservatixc. (liven a patient in tiic niiildle jx-riod of life who, without ap- parent cause, begins to lose flesh and strength, is unable to eat as before, and whose digestion is delayed, we have a right to suspect gastric cancer. If, in conjunction with this, we find loss of motility and a delay of food in the stomach, with evidences of blood and reduction of hydrochloric acid, a tentative diagnosis of carcinoma is justified. We should be esi)ecially suspicious if symi)toms of old or recent ulceration are obtained in the history. In our last 39 cases 56.4 per cent, showed direct evidence of carcinoma developing on ulcer. Graham shows a clinical history of ulcer in over 50 per cent, of the cases of gastric carcinoma which have come under his investiga- tion, although years may have elapsed between the two diseased processes. It is possible that a larger number of patients in Avhom cancer develops secondary to ulcer consult the surgeon than those without this history. The growing frequency of operation for ulcer brings patients to the operating-table for the relief of obstructions, deformities, and adhesions, and in a considerable number of these patients the gastric ulcer has undergone cancerous degeneration. Murphy says: "The history of the majority of patients with cancer of the stomach will show precancerous symptoms." This question cannot be settled by postmortem evidence. For example, suppose we were told that the postmortem examina- tion of 1000 women who died of cancer of the cervix uteri did not show a single one who had had cervical laceration. Would not the query at once arise : If the cancer was so extensive that the patients died of the disease, how would it be possible for any one to know by such postmortem examination whether they had ever had lacera- tion or not? Is this not equally true of ulcer? Before the ]iatient dies all trace of the ulcer would be lost in the gross extent of the disease. The presence of a tumor is not necessarily a contraindication to operation. A small movable growth in the pyloric end of the 286 WILLIAM J. MAYO stomach is rather a favorable indication, since the early obstruction attracts the attention of the patient by producing distressing symp- toms which might not have come on at all if the tumor were in the body of the stomach. Fortunately, 80 per cent, of all gastric carcinomata are in the pyloric end and along the lesser curvature. Seventy per cent, are so situated as to interfere mechanically with motility, and are, therefore, operable, while 10 per cent, are situated around the cardia, giving evidences of esophageal obstruction, and 10 per cent, are in other parts of the stomach. The earlier the mechanical symptoms appear, the better the prospect of early diagnosis and cure. There are some contraindications without exploration. One of the chief of these is finding typical carcinomatous glands in the supraclavicular fossa, particularly on the left side. This occasion- ally happens where the diagnosis may be plain, but the question of operation is less plain. Fixity of the growth and the presence of ascitic accumulations are also contraindications. Much has been written about the value of blood examination in cancer, especially as to the hemoglobin. We have had pylorectomy cases recover with the hemoglobin as low as 30 per cent. Again, some of the worst cases with obstruction may give a high percentage of hemo- globin due to concentration of blood from their inability to absorb fluids. One cannot help believing that more persistent attempts to improve the early diagnosis of cancer would have followed if better operative results had been obtained. This excuse no longer exists with an improvement in the mortality of about 10 per cent, and 25 per cent, of the operative recoveries living more than three years. The time has come for energetic action. All other means have failed, and exploratory incision of the suspected case is the only known means of early diagnosis. This should not discourage us, but should rather encourage better directed efforts toward securing: less formidable means of ascertaining the truth. SURGICAL TREATMENT OF CAXCEIi OK STOMACH 287 Steps of tiik Oi'KiiATio.v 1. Anesthesia. — We prefer ether anesthesia, giving a hypo- dermic injection of 36 grain of morphin tliirty minutes previous to its a(hninist ration. During the major i)art of the operation no anesthetic is required, since there is no j)ain experienced during the progress of the visceral work. 2. Exploration. — A short incision is made in the midline, half- way between the umbilicus and the ensiform cartilage. Two fingers are introduced, and the growth is explored with reference to other structures. Next the extent of glandular involvement is ascertained. If the case seems fairly reasonable for operation, the incision is rapidly enlarged and the growth drawn out of the abdo- men. This manoeuver permits careful examination of the lesser curvature, especially as to whether the infiltration in this vicinity extends beyond the possibility of removal. The transverse meso- colon is then inspected, as it is often infiltrated from behind. The posterior surface of the stomach and its relation to the pancreas are palpated with fingers passed through a rent in the gastro- hepatic omentum. We have dissected into the superficial surface of the pancreas a number of times without that fatality to which Haberkant (76 per cent.) and Mikulicz (74 per cent.) have called attention. 3. Mobilization of the Lesser Curvature (Figs. 31 and 32). — The stomach is drawn firmly downward and to the right, the left lobe of the liver raised by the fingers of an assistant, and the gastric artery tied with catgut on a needle at the highest possible point well beyond the lymphatic nodes. A pair of clamps are caught on the opposite side, and the artery and that portion of the gastro- hepatic ligament which has been ligated with it are cut. With a few nicks of the knife the pedicle is partly detached from the stomach and allowed to retract. This permits mobilization of the gastric wall and obtains a clear space near the esophagus for the division of the stomach. The superior pyloric artery and the re- mainder of the gastrohepatic ligament are now doubly tied and cut between, leaving the glands attached to the duodemmi. This 288 WILLIAM J. MAYO mobilizes the entire lesser curvature and makes the remainder of the work outside of the body. 4. Separation of the Pyloric End of the Stomach {Figs. 31 and 32) . — The hand is passed into the lesser cavity of the peritoneum be- hind the stomach, adhesions are carefully divided, and bleeding points ligated. Hot moist gauze pads are now placed in this space. Two pairs of narrow crushing clamps (Ferguson) are now placed on the duodenum well below the disease (as a rule, an inch below the pylorus), and the duodenum is divided between. The glands lying in the omentum immediately below the pylorus are carefully dissected upward, so as to remain attached to the pyloric end of the stomach, and a few bleeding points caught and ligated. The forceps on the stomach side with these glands is now lifted sharply upward, exposing the gastroduodenal artery in the groove between the head of the pancreas and the duodenum; this vessel is doubly tied and divided between ligatures. The glands in this region are dissected upward with the fat and hot gauze compresses placed in the space. 5. Freeing the Greater Curvature {Figs. 31 and 32) . — The gastro- colic omentum is tied and divided in sections below the inferior coronary vessels, care being taken to avoid the middle colic artery; accidental inclusion of this vessel has caused gangrene of the trans- verse colon, of which it is the sole blood-supply in 15 per cent, of the cases (Kronlein). Injury to the middle colic has necessitated re- section of the transverse colon in a number of instances (Kocher). The lymph-nodes lie close to the blood-vessels, and at a point well beyond these structures the left gastro-epiploic vessel is caught and tied. Care should be taken not to destroy its branches to the stomach beyond the point of ligation, as it will be the sole blood- supply for the contiguous stomach- wall. 6. Removal of the Diseased Structures {Figs. 31 and 32). — ^Light elastic holding clamps are now placed on the stomach an inch or more back of the proposed line of resection, a second pair grasping the tumor side and the growth with the glands and fat removed en masse. As it is cut loose, several catch forceps should be appHed to the margins of the cut gastric surface, projecting beyond the srUCKAL TUKAT.MKNT OI-' CANCKIt OV STOMAf'II 28!) clamp to prevent relraelion. This elaiiii) is straight, cpiite elastic, and covered with rubber so that it will not crush or injure the gas- tric wall. We have found those of Seudder very satisfactory'. The cut gastric wall is now lightly gone over with the actual cautery, I)articularly at the upper {)art, at which point we are most liable to fail to get well beyond the disease. 7. Suture of the Gastric Stump (Fig. 33). — After rearranging the hot moist packs to furnish ample protection, with No. 2 chromic catgut on a straight needle, beginning at the greater curva- ture, a running suture is placed through all the coats after the method of Charles H. Mayo. The needle enters on the peritoneum at one margin, passes through to the mucous coat and directly back on the same side from mucous coat to peritoneum. By doing this alternately, first on one side and then on the other, by a single suture the peritoneal surfaces are rolled into contact, the parts being firmly brought into apposition and the hemorrhage checked. On api)roaching the lesser curvature it will usually be found that the clamps are too close to the edges of the wound to permit of this manoeuver, and it may be necessary to unclasp them in suturing the last inch. As this situation is also under considerable tension, it is well to place one or two mattress sutures of linen at the upper end, completely and permanently to secure it, rolling the first catgut suture in by a wide grasp of the gastric wall far enough back to permit union without tension. Any point not well turned or showing a tendency to ooze is secured by an independent mattress suture of linen. Beginning now at the greater curvature, a fine linen continuous Gushing suture turns in the gastric wall without tension over the first row. 8. Restoration of the Gastro-intestinal Canal. — (a) After the Method of Kocher (Fig. o.J). — After careful cleansing, the stonuich is drawn toward the duodenum. If it is sufficiently mobile, the Kocher operation is performed, the duodenum being loosened up for the purpose. The posterior wall of the stonuich near the greater curvature, at a distance of 1} 2 to 2 inches from the gastric suture line and j)aralIol with it, is suturetl to the posterior duodenal wall just l)el()w the original clamp on it by a running suture of linen. VOL. I— 19 290 WILLIAM J. MAYO One-sixth inch in front of this, and just opposite the duodenal clamp, an incision is made through the peritoneal and muscular coats of the stomach to, but not through, the mucous coat. The clamp on the duodenum is now removed, its cavity opened up and sponged out. The posterior cut wall is firmly sutured with chromic catgut on a curved needle in front of the posterior linen suture through all the coats of the duodenum and stomach, using a Connell or button- hole stitch until the posterior inner row is completed half-way around. The mucous membrane of the stomach, which has been sutured behind without opening, is now cut through and its sutured lower margin inspected for hemorrhage or lack of apposition, and one or two interrupted sutures of catgut applied if necessary. The through-and-through catgut suture is now continued around the anterior surface, uniting the end of the duodenum to the stom- ach in a similar manner to that previously described in closing the stomach, the suture passing from peritoneum to mucous coat and back from mucous coat to peritoneum on the same side alternately and tied to the original end. The linen suture is now continued around to the starting-point, completing the second row. The entire suture line is inspected front and back, and several extra mat- tress sutures of linen used to reinforce at points of tension. If the stomach has a tendency to drag on the duodenum, the gastrocolic omentum close to the stomach is caught and anchored to the peritoneum on the left margin of the wound. The stumps of the gastrocoHc omentum are brought together with a couple of catgut sutures and the entire field inspected and sponged. The deep gauze compresses are now removed. If these have been carefully placed and renewed at intervals, there will have been no contami- nation or exposure, (b) Closure of the Duodenal Stump and Inde- pendent Gastrojejunostomy, Billroth No. 2 {Fig. 35). — If the stomach cannot be approximated to the duodenum, the duodenal stump is turned in by a circular suture after ligation in the groove made by the forceps and a posterior gastrojejunostomy is performed without a loop, that is, within three inches of the origin of the jejunum. The opening in the stomach, however, should run from above down, right to left, so that the proximal end of the jejunum shall lie close SURGICAL TREATMENT OF CANCER OF STOMACH 291 to the siituro lino, the distal oiul at the lowest point and f)assing to the left. After eoniplelion of the gastrojejunostomy in the usual manner the jejunum at once drops down into the left iliac fossa in its normal position. A few sutures close the rent in the transverse mesocolon in such fashion as to protect the suture line. If the j)ationt is in a poor condition, an anterior or i)osterior Murjjhy button operation can be made to save time. The button must be protected, however, by at least four mattress sutures of linen at intervals to prevent separation. .9. After-care. — After resection the patient should be placed in bed, the head and shoulders elevated to the semisitting posture, and a glass female douche point introduced above the internal sphincter, through which from one to four quarts of one-half strength normal saline solution is allowed slowly to enter the rectum for absorption from a gravity bag, thirty minutes to three hours being used in this process (Murphy). This is repeated in twelve hours with a lesser amount. From one-half to one ounce of hot water is allowed by the stomach every hour after sixteen hours, and the usual experi- mentation of liquid foods begun after twenty-four to forty-eight hours, the rectum being used as an auxiliary for four or five days. Palliative Operations The results of palliative operations for cancer of the stomach are relatively unsatisfactory. The statistics of gastro-enterostomy for the relief of obstructions due to inoperable malignant disease show as high or a higher mortality than gastric resection; the comparison, hoAvever, cannot be directly made, as gastro-enteros- tomy can be applied in cases in which radical incision cannot be performed. The average prolongation of life after gastro-enterostomies is not over four to six months, and the fact that patients live beyond this time gives rise to the query: Might not radical operation have given a cure? In 143 cases of gastro-enterostomy for malignant disease reported by Mikulicz the mortality was 33 per cent.; the average prolongation of life, (5.4 months. In 74 cases reported by Kronlein the death-rate was 24.3 per cent., and the average jiro- S92 WILLLIM J. iLVYO longation of life was but three months. In 140 of our cases the death-rate was 15 per cent, and the average prolongation of hfe, so far as known, was less than five months. It is true that the mortahty in recent cases is very much less, perhaps not over 10 per cent., but leaving the ulcerating, bleeding mass in the stomach to its own devices is unsatisfactory. The operation merely prolongs a chronic invalidism by a few weary months which are without hope. The judge who says to the pris- oner "I sentence you to death after five months" has not given the prisoner a desirable intervening existence. For cancerous obstruction of the cardiac orifice gastrostomy offers the only means at our command, a palHation which is not frequently demanded by the patient when the facts are placed plainly before him. In our series there were 18 gastrostomies for cancerous cardiac obstruction, with 3 deaths — 16.6 per cent.; average duration of life is about the same as after gastro-enterostomy. Of explorations with the discovery of hopeless gastric carcinoma there were 72, with 1 death in the hospital. The average stay of patients explored for incurable disease is less than five days, the deep wounds being closed with catgut, and the strong aponeurotic structures are braced with buried mattress sutures of linen, silk, or silver. It will be seen that of the total 313 cancers of the stomach operated up to February 1, 1906, only 26 per cent, were early enough to permit of radical extirpation. In conclusion, let me urge upon the profession the merits of radical operation on suitable cases of gastric cancer. THE TECHNIC OF (^,ASTROJEJUNOSTOMY WILLIAM J. MAYO The results following gastrojejunostomy arc as good as can l)e reasonably exjjected, and depend more upon the condition of the patient at the time of operation than uj)on the technical difficulties of the operation itself. The mortality is no longer the question considered. This is particularly true of the posterior suture operation, a procedure we used 130 times in sixteen months with l)ut 1 death. These results are not exceptional, and have been and are being duplicated by many other surgeons. Neither need we consider those serious cases of regurgitant vomiting of biliary and pancreatic secretions during the first week (vicious circle), since this complication has practically disappeared with the evolvement of better methods. While we can congratulate ourselves upon the immediate safety of the operation, we are not yet free from certain embar- rassing complications which may arise some days or weeks later. The most common subsec^uent condition is the chronic regurgita- tion of bile which comes on at intervals in a small percentage of patients. The symptoms vary from a temi)orary burning in the stomach, due to the entrance through the fistula of biliary and pancreatic secretion, to the most distressing vomiting of great f[uantities of such fluids. Ochsner has pointed out that this com- plication develops usually within ten weeks following the opera- tion. On reoperation the condition is found to be due to a partial kinking or obstruction from twisting, adhesions, or other cause, just as the acute "vicious circle" was due to early and more com- plete obstructions. * Reprinted from "Annals of Surgery," April. l!)0(l. 294 WILLIAM J. MAYO Since January 1, 1905, Charles H. Mayo and I have discarded all "loop" operations, T\-ith or without entero-anastomosis or closure of the pylorus, the anastomosis being made as close to the origin of the jejunum as possible. The results as compared with all previous methods in our hands have been infinitely better in every respect. The few "loop" operations that have been per- formed during this time have been made to meet special indica- tions. From January 1 to July 1, 1905, there were 56 of these "no- loop" operations, with but 1 death, which occurred in a patient practically moribund at the time of operation. Two patients, however, developed chronic bile regurgitation of a serious char- acter. These two patients are the ones referred to in a paper on "Chronic Ulcer of the Stomach and First Portion of the Duode- num," read before the American Medical i^ssociation, July, 1905, and published in the "Journal of the American Medical Associa- tion," October 19, 1905. Each patient had gained in flesh and weight, being reHeved of former symptoms, but in each occasional regurgitation of quantities of biliary and pancreatic secretions was a source of great discomfort and considerable disability. Reopera- tion in both cases during the past summer showed that the cause of the trouble was an angulation of the jejunum at its gastric attachment. In all the 56 cases referred to the anastomosis 'of the jejumma to the stomach was made in the line of peristalsis, that is, the proximal portion of the jejunum was attached to the posterior gastric wall to the left and above, and the distal end of the jejunum to the right and lower part of the stomach. In this partial twist- ing lay the secret of the comphcation (Fig. 36) . The question at once arises: Is the idea of continuity of peri- stalsis between the stomach and jejunum a matter of conjecture, or has it some practical significance? The writer has gone over in a large number of hving subjects the anatomy of this region, and the anatomic facts can be briefly stated as follows: For convenience we will take the origin of the jejunum as being at the point in which the duodenum passes through the •_uij^-i^ y Fig. ,^6. — Showing kink in jejunum resulting from changing normal direction ol its uppermost portion, in ''no-loop" gastrojejunostomy after posterior method. X and X mark commeDCcment of jejunum. W. -J. ''>^'-i^ Fig. 37. — Showing no kink in jejunum resulting from preserving normal direction of its upper- most portion, in "no-loop" gastrojejunostomy after posterior method. X and X mark commence- ment of jejunum. THE TECHMC i)V GASTKOJEJL'XOSTOMY 29.5 transverse mesocolon. Tlie distal end of llie duodenum lies liehind the stomach when the latter is moderately distended, and ahcjut 1^2 inches to the left of the midline, and 13^9 to 2 incites ai>o\e the umbilicus. Its horseshoe shai)e has the concavity directed to the left and upward, and the exit is within about two inches as high as the pylorus. The transverse portion of the duodenum passes forward over the prominent vertebral column and backward to the left side of the spine to the opening' in the transverse mesocolon. The terminal inch which marks the duodenojejunal juncture is directed upward and to the left, the mesentery of the proximal jejunum lying behind, and the free surface of the intestine directed forward. The jejunum from its origin drops at once into the left abdominal fossa. Not only does it pass to the left, but it gravi- tates backward into the left kidney pouch underneath the splenic flexure of the colon, so that at a point four inches from its origin it lies on a plane to the left and posterior. This can be shown in a \'ery practical way by drawing the transverse colon out through the abdominal incision, pulling it upward and to the right until the mesocolon is taut. This brings the beginning of the jejunum into view. It will readily be seen therefore that if the attachment is made to the stomach, so that the proximal portion of the gastro- jejunostomy is to the left and above, and the distal portion is directed to the right and below, we have introduced two serious displacements. The jejunum no longer falls in the normal manner to the left and backward, but is artificially caused to pass not only to the right, but forward, as it must ride the vertebral column or the structures immediately contiguous.* The active propulsion of the stomach lies in the pyloric end, in that part bounded above by the horizontal portion of the lesser curvature. The five-sixths lying to the left has mainly storage function, and its muscular action is less forceful. The proper site for the gastric incision is to the left of this point, on a line with the longitudinal part of the lesser curvature, with its lower end at the bottom of the stomach, under the cardiac orifice. * For variations in the origin of the duodenojejunal angle see Mumford, Testut, and Cunningham. 296 WILLIAM J. MAYO The writer has been unable to see that it made any difference in the results of a "no-loop" gastrojejunostomy whether the peri- stalsis of the stomach is the same as that of the intestine or not, as, with the exception of a tendency to contraction, there have been no complications introduced that have been other than intestinal in origin. Since the first of July, 1905, we have abandoned reversing the jejunum, and in a larger number of cases (65) we have had no trouble and no deaths. We apply the jejunum to the posterior wall of the stomach from right to left, exactly as the intestine lies under normal conditions. The distal portion of the jejunum passes from the bottom of the stomach directly back into the left fossa, as occurs normally (Fig. 37). Two drawings, made from sketches of the actual operations in our clinic, explain the mechanical conditions very perfectly. It is hardly necessary to say that the idea of the reversal of the peristalsis is not original with us, but will be found in the litera- ture on this subject to have been advocated at various times. As a matter of fact, in this operation it is of no importance. With any "loop" operation, four inches or more in length, the objections which we here make to the mechanics are not so appar- ent, but mechanical difficulties of some kind are so frequently in- troduced as to render gastrojejunostomy with entero-anastomosis the method of choice with the larger number of operators. The "no-loop" operation directed to the left, as outlined above, has given us vastly better results than any other method with which we have become acquainted. Steps of the Operation. — For benign disease the abdomen is opened from three-fourths to one inch to the right of the median line, splitting the fibers of the rectus muscle. The transverse colon is drawn out of the abdominal incision, and by a steady traction to the right and upward the mesocolon is brought out until the jejunum comes into view, and the intestine is grasped at a point three or four inches from its origin. On drawing the jejunum tight the fold of peritoneum which covers the ligament of Treitz (a small band containing muscle-fibers) is developed. This peri- THE TPXIINH' OF CiASTUOJh:JUNOSTOMY '2!)7 toiioal hand lias its ori^Mi; on llic transverse mesocolon, and extendi down on to the l)e<,'innin^' of the jejnnum, acting as a suspensory ligament; it will he i'oiind to lead to the base of the vascular anh of the middle colic artery, and accurately marks the place when- the transverse mesocolon is torn through to secure the posterior wall of the stomach. The stomach is drawn through this opening and the anastomosis performed, beginning at a point one inch above the greater curvature, on a line with the longitudinal por- tion of the lesser curvature, and ending at the bottom of the stom- ach, i2' 2 inches to the left. To secure a proper low point a small opening is made in the gastrocolic omentum, and one-half inch of the anterior wall j)ulled through behind. Having these features in view, a considerable portion of the posterior wall is drawn into a pair of light elastic curved holding clamps. (We prefer the Doyen.) The handles lie to the right and about transverse with the axis of the body. Beginning 13-2 to 3^2 inches from its origin, the jejunum is drawn into a similar pair of clamps with handles to the right. It will thus be seen that the left low point on the stom- ach lies in the tip of the clamps, and the distal point of the jejunum lies also to the left. By placing the two clamps side by side, the operation is completed in the usual manner by two-row suturing, chromic catgut suture being used for the inner through-and- through mucous stitch, as silk or linen may hang ulcerating for months before passing away. In applying this suture on the pos- terior row behind we use the Connell or buttonhole suture. On the anterior we use the method advised by Charles H. Mayo, which consists in entering the needle on the peritoneal side through to the mucous, and directly backward from mucous to peritoneum on the same side. By doing this alternately, first on one side and then on the other, with this first chromic catgut suture, the peri- toneal surfaces are rolled into contact, the parts to be united are held firmly in apposition, and the hemorrhage checked. The outer row consists of No. 1 celluloid linen (Pagenstecher), which we have used with great satisfaction since it was introduced for this purpose by Robson. It is very strong, smooth, and has no capillarity. Flattening the intestine (Cannon and Blake) shouM 298 WILLIAM J. MAYO be avoided by grasping the intestinal wall close to the margin of the incision -^dth the suture, so as to turn in a narrow seam from the intestinal side. On the gastric side, on the contrary, one need not hesitate to take a free grasp of the tissues. The rent in the meso- colon is fastened to the suture line mth three or four mattress sutures of hnen. This should grasp the peritoneal coat close to the margins of the rent in such manner that, when tied, all the raw surfaces shall be turned in behind the stomach, and the per- itoneum folded smoothly against the gastrojejunostomy opening, so there shall be nothing to cause adhesions between the meso- colon and the jejunum beyond the anastomosis. This short communication is to supplement the paper on "Gastro-enterostomy," read before the American Surgical Asso- ciation, July, 1905, and pubhshed in the "Annals of Surgery," Xovember, 1905, in which article credit has been given to origi- nators and promoters of useful suggestions in perfecting this operation. THE SURGICAL TREATMENT OF GASTRIC AND DUODENAL ULCER AND ITS RESULTS* WILLIAM J. MAYO The contributions to the literature of ulcer of the stomach and duodenum liave become so numerous that it will be impossible in a paper of this character to give credit for the many valuable ad- vances which have been brought out by individual workers. I wish, however, to express my appreciation of the work they have done. For more than twenty years the complications of gastric and duodenal ulcers have received surgical attention. Operations for benign stenosis of the pylorus and some acute secondary mani- festations have occasionally been performed, but not until within the last five years has there been a surgical invasion into the field of unhealed ulcers with a view to giving relief from pain, under- feeding, and chronic disability, which so frequently accompany the disease. Acute ulcers of the stomach and duodenum properly belong to the domain of internal medicine, and surgery has to do only with the complications, such as perforation, hemorrhage, and obstruc- tion. The surgical treatment of acute perforations of the stomach and duodenum is now on a sound footing, the results depending on speedy diagnosis and prompt operative relief. Patients operated on within the first five hours after perforation usually recover; * Read in the joint session of the Sections on Practice of Medicine and Surgery and Anatomy of the American Medical Association, at the Fifty-seventh Annual Session, June, ll)0(i. Reprinted from "Jour. Amer. Med. Assoc," September ii, 1900. 299 300 WILLIAM J. MAYO after ten hours the majority of them die. We have had 7 gastric with 2 deaths and 9 duodenal with 3 deaths. Suture of the perforation and pelvic drainage, with or without irrigation, the patient maintained in the semisitting posture (exag- gerated Fowler's) for several days, gives the best results. Whether or not a gastrojejunostomy shall be done at the same time is a moot question, since a considerable percentage of patients will develop stricture, adhesion, or other deformity in the healing proc- ess following perforation. If the perforation lies in the duodenum or close to the pylorus, it should be sutured transversely to avoid interference with the lumen of the part involved. Generally speaking, if it seems probable that stricture or other secondary con- dition will result and the patient is in good condition, gastrojejun- ostomy should be performed, providing it can be done without spreading the infection which is already present. If in doubt, a living patient is better than a completed operation at a greatly enhanced risk. Hemorrhage from the stomach occurs in two forms, acute and chronic. Recurring acute hemorrhages are best treated by open- ing up the stomach or duodenum, locating the bleeding point, and suturing the part firmly with catgut on a curved needle from the inner (mucous) side. The outer surface is then exposed and pro- tected over this area by a few mattress musculoperitoneal sutures of linen (modified Andrews method). We have not found gastro- jejunostomy for acute hemorrhage a reliable procedure. Out of 6 patients, we have had 1 bleed to death within two weeks after gastrojejunostomy and a second was saved by opening up and se- curing the bleeding vessel, while 5 who have been treated by pri- mary operation on the bleeding point, with or without excision of the ulcer, recovered. Chronic hemorrhages, on the contrary, can be cared for by gastrojejunostomy, especially if the ulcer is situated in the duo- denum or near the pylorus. In a number of instances we have, however, excised the ulcer and closed the defect by plastic opera- tion, planned in such way as not to interfere with drainage by StI<(;i(AL TREATMFA'T OF OASTHK \\l) Ul DlH.NAI. I I,< Kit .'501 .su}).sequent contraction. In several of these cases the procedure ainoiintod to gastric resections with end-to-end suture. Chronic Ulcer Chronic ulcer of the stomach and duodenum is more common in men, and is essentially a disease of adult life. It entails upon the victim years of invalidism, and in fully 25 per cent, is the direct cause of death, while indirectly, through anemia, it causes general infections, tuberculous or otherwise, and thus doubles the mor- tality. Surgical interference should be considered when the failure of medical treatment is made manifest by a continuance of the symptoms or frequent relapses. Our experience (W. J. and C. H. Mayo) covers 600 operated cases up to May 1, 1906 — 136 stomach, 13o duodenal, and 28 stomach and duodenum. In 46 cases classified as duodenal the duodenum was primarily involved, the stomach being involved only at the pyloric ring. It was rare that a gastric ulcer did not stop abruptly at the pylorus. Gastric ulcers were multiple in less than 15 per cent, of the indurated, and estimated at about 20 per cent, of the mucous, lesions. The most conmion form is the saddle ulcer of the lesser curvature above the pylorus, extending flap like down the anterior and posterior wall (prepyloric). Duodenal ulcers involve the two inches immediately below the pylorus, and extend up to within at least three-fourth inch of it. In all but o cases the bow'el lesion was single. Of the 163 duodenal ulcers, 77 per cent, were males and 23 per cent, females, and in all but 7 cases the site of the ulceration was indurated, the thin intestinal wall enabling ready identifica- tion. In the stomach, on the contrary, the thick tissues sometimes prevented accurate localization of a mucous lesion, the proportion being 70 per cent, indurated and 30 per cent, non-indurated. Three of the acute hemorrhagic cases were of the mucous variety, while all but one of the acute perforations occurred through the scar of a partially healed ulceration. The predominance of the indurated over the non-indurated 302 WILLIAM J. MAYO ulcer is even greater than would appear from these percentages. During the past year more experienced surgical examination has shown that over 85 per cent, of the gastric ulcers coming to opera- tion have been indurated, and less than 15 per cent, non-indurated. There is no question but that the frequency of non-indurated mucous lesions in the stomach and duodenum has been and still is greatly exaggerated. The very large majority involve all the gastric coats and are capable of demonstration at the operating- table. Gastric ulcers were found with nearly the same frequency in males and females. In this connection I would again call attention to the fact that ulcer of the duodenum is nearly as com- mon as ulcer in the whole of the stomach, and as more than three- fourths are in males, the increased frequency of occurrence of gastric and duodenal ulcer in the male sex is accounted for. In our earlier work duodenal ulcers in the vicinity of the pylorus were classed as pyloric and added to the gastric group. In the past two years we have learned to differentiate more accurately, and in the last 100 gastric and duodenal ulcers operated on 47 were duodenal, 44 gastric, and 9 had an independent ulcer of each organ; 62 were males and 38 females; 87 out of the 100 were of the indurated variety. That a very large proportion of chronic ulcers are medically incurable is recognized by the majority of unprejudiced investiga- tors. It now remains to be shown whether operative treatment is justified by its results. Gastrojejunostomy Gastrojejunostomy is the operation which has justly earned the most prominence in the treatment of gastric and duodenal ulcers. It is based on the common-sense principle of giving rest to the diseased part by diverting the food and gastric secretions to a new outlet, which should be on the storage side of tjie stomach, at its lowest point under the cardiac orifice. The method of choice is the one without a loop, made on the posterior surface through an opening in the transverse mesocolon. The distal end of the je- junum is attached to the lowest point and to the left, the proximal V/.I/vlAYO. Fig. 38. — Dotted lines show posterior "no-loop" gastrojejunostomy WJJvlAYO. /r y^^ ^-A- _«^c Fig. 39. — Resection of hour-glass stomach, with fields in gastrohepatic and gastrocolic omenta ligated. Dotted lines show site of proposed resection. SURGICAL TREATMENT OF GASTRIC AND DUODENAL ULCER 303 part to the rij^ht, three-fourth inch above the greater cur\'ature, givin<,' jui oi)eniii<,' somewhat oblique from above down, ri^'lit to left, not less than two inches in length (Fig. 38). The torn edges of the transverse mesocolon are attached to the suture line by three mattress sutures in such manner as to leave a perfectly smooth j)eritoneal border, the ragged and fatty margins being tucked up underneath so as not to cause adhesions. It is imjjortant that the distal end of the jejunum be attached as it lies naturally, that it may at once drop to the left and posteriorly in its normal anatomic position. The operation thus briefly outlined has a nominal mortality. We had but 1 death in 135 "no-loop" operations. The anterior method, preferably with a Murphy button or McGraw ligature, is occasionally demanded by reason of posterior adhesions or ab- normalities in the mesocolon or duodenojejunal juncture; but as it sacrifices 18 to 20 inches of the most important part of the upper jejunum, it cannot be considered a close rival to the posterior operation. For the past year and a half we have abandoned all "loop" operations unless forced to them, and in no case do we practise entero-anastomosis as a primary procedure. Of the total 600 gastric and duodenal ulcers, 383 (64 per cent.) were subjected to gastrojejunostomy. The question arises: What results can be expected from gastro- jejunostomy? In the early stages of any subject certain proce- dures gain a reputation which later experience does not always bear out. While gastrojejunostomy is by all odds the most useful means of relief, it is not a cure-all. It is purely a drainage opera- tion, and as in a large majority of cases the ulcers are situated in the grinding pyloric end of the stomach or upper duodenum, the alleviation aflForded is certain and speedy. But if the lesion does not involve permanent interference with natural drainage, the food will eventually pass out the normal opening rather than the gastrojejunostomy, and a certain amount of shrinking of the anastomotic stoma or angulation may follow. The dangers from this source are reduced to a minimum with a "no-loop" operation. The more serious the interference with 301 WILLIAM J. ^L\YO the normal motility, the greater the relief afforded by gastrojejun- ostomy, so that for indurated ulcer it is the method "par excel- lence." On the contrary, in mucous lesions, "^here the interfer- ence with gastric drainage is intermittent and due to muscular spasm, the results are less certain. Gastroduodexostomy The only operation on the stomach of any consequence which involves a new principle in surgery since the time of Billroth is that of Finney. It leaves the opening at the natural situation, and the enlargement is doTNTiward along the greater curvature away from the ulcer area. Gastroduodenostomy is especially applicable to those mucous lesions which interfere with drainage through spasm. It supplements gastrojejunostomy admirably, as its scope includes those cases in which the latter method has been less certain of cure, and it has superseded pyloroplasty. We have had 72 cases with 4 deaths, and the very large majority of those who recovered have remained well. Excision of the Ulcer Theoretically, excision of the ulcer is the logical procedure, but it must be shown by careful examination that the ulcer excised is the only one present. The saddle ulcer of the lesser curvature, if it does not interfere with drainage, should be excised if practicable because, as has been pointed out, if the stomach is not dilated and the pylorus is open, muscular action -^dll continue to force the food along the normal channel rather than out of the artificial stoma. Therefore, a direct attack on the diseased area under such circumstances, es- pecially when the margins of the ulceration are hard and calloused, would seem to be sound practice. We have excised ulcers 1-t times — 6 times of the lesser curvature and 8 times in connection with pyloroplasty and gastroduodenostomy, with no deaths. W./Mayo. Fit;. 40. — Rescct'on of hour-slass stomach, inner row of sutures nearly completed (catgut), outer row completed on posterior surface and beginning on upper anterior surface (linen). surgical treatment of gastric and duodenal ulcer 305 Rodman's Operation Some four years ago it was suggested by William Rodman that since, in the large majority of instances, ulcer of the stomach was in the pyloric portion, it would be wise to resect as one would for cancer, closing both duodenal and stomach ends completely and reestablishing the gastro-intestinal canal by an independent gas- trojejunostomy. We have performed this operation nine times with great satisfaction. All the cases recovered and have re- mained well. This procedure is chiefly indicated in indurated lesions in the vicinity of the pylorus. Resection of the STOiL\CH The treatment of hour-glass stomach by some form of plastic gastrogastrostomy has been a popular operation, but in the ma- jority of instances the ultimate results are unsatisfactory', as it leaves a large amount of scar tissue, and if it happens that the py- lorus is involved to any extent, adequate drainage is difEciJt to secure. Gastrojejunostomy on the proximal pouch does not prevent a certain amount of food passing into the distal loculus, where it stagnates on account of loss of motility. Multiple gastrojejunos- tomies, one for each loculus, with entero-anastomosis beyond, is an unnecessarily complicated procedure, especially if more than two lociili are present. In l'-2 cases without a death we have resected the affected por- tion of the stomach in the following manner (Figs. 39 and 40) : The gastrohepatic and gastrocolic omenta are divided and sep- arated from the diseased area; a straight elastic holding clamp is placed on the proximal side, across from the greater to the lesser curvature, about one inch back from the proposed line of resection. On the distal side the clamp is applied obliquely from above down, right to left, to increase the diameter of the cut surface, saving from the greater curvature. In this way we have been able to secure on the distal side an opening two-thirds the size of the proxi- mal one for suturing. By ha\'ing one inch or more of the tissue VOL. I — ^0 303 WILLIAM J. MAYO projecting beyond the clamps, the slack on the large or proximal side is taken up with each suture, the bite of the thread taking only two-thirds of the amount of tissue on the distal portion, a difference of diameter of one-third being disposed of in this manner without seam or noticeable pucker. If the ulcer is situated close to the pylorus, an end-to-end union is quite easy to obtain. In two cases we have been able to excise an ulcer of the duodenum close to the pylorus, with direct union of the amputated end of the duodenum to the stomach, the pylorus being removed in both instances. The last word on ulcer of the stomach has not been written, and it is evident that no one operation will be applicable to all varieties. At the present time it would seem that gastrojejunostomy has the largest field of usefulness, especially in those cases in which there is permanent interference with gastric mobility by reason of ob-. structive lesions in the pyloric end. For those cases in which obstruction is due to spasm or other non-mechanical cause the gastroduodenostomy of Finney is the operation of choice. In connection with this latter procedure the excision of an ulcer of either the stomach or duodenum in close proximity to the pylorus may be done with great satisfaction. Gastric ulcers which do not interfere with drainage and in which there is no loss of motility should be directly excised if possible. Calloused ulcers of large size and thick hard margins, whether hour-glass or not, are best treated by some form of gastric resection, as we have frequently found carcinomatous degeneration taking place in these cases. The large majority of duodenal ulcers give indications for gastrojejunostomy. There are few exceptions in which excision or resection will give better results. It can be shown that more than 90 per cent, of patients suffer- ing from gastric and duodenal ulcer who have been subjected to operation have been cured. Failures are more often due to tech- nical errors resulting in bad mechanics than to the inability of a properly, executed operation to cure the disease. SURGICAL TREATMENT OF (lASTHlC AM) 1)1 ODK.NAI. I I-( KR f5()7 '1\) one wlio cxpecls to practise sur^eiy of the stomacli we would stroiif^ly recoiiiiiKMid a close study of tlie ^^astric manifestations of tlie neurasthenic stale. Atonic dilatations, prolapse of the sloni- ach, and tlie various <^astric neuroses jnay sinuilate ulcer very closely, and u careful differentiation is absolutely essential, hccausc all of these latter conditions are unimproved and often ac- centuated hy operation, bringing' discredit on surgery as a whole. THE PRINCIPLES UNDERLYING THE SUR- GERY OF THE STOMACH AND ASSOCIATED VISCERA* WILLIAM J. MAYO Clinical surgery in the past as applied to the stomach has lacked an accurate physiologic and pathologic basis. Certain diseased conditions have been relieved successfully, but often the reason for the results have not been satisfactorily explained. This is especially true in regard to function. Our knowledge has been founded largely upon the scientific study of the dead, but this kind of information is grievously open to error, inasmuch as secondary complications and terminal infections frequently obscure the ini- tial lesion, and from the postmortem evidence we are not always able to get a clear mental picture of the disease as it existed during the curable period. The immediate cause of death may, appar- ently, be dissociated from the primary lesion, which, although completely healed, had set in motion the fatal pathologic process. Medicine is greatly indebted to postmortem pathology as a founda- tion, but as a superstructure it has a less exalted position. Modern surgery has dealt with the practical side and estab- lished a "living pathology." It has cleared up those three great avenues of peritoneal infection, the Fallopian tube, the appendix, and the gall-bladder, and is now engaged in a surgical investigation of the diseases of the stomach. The problems to be elucidated, however, are very different from those just mentioned, as infec- tions, primarily, play but a small part in gastric pathology. The * One of the Mutter lectures delivered before the College of Physicians, Phila- delphia. Reprinted from the "Edinburgh Med. Jour.," January, 1907. ■SOS SURGERY OF STOMACH AM) ASSOriATi:i) VISfEIlA 300 ^'.ip Ix'lwccii I he kii<)\vl('(l;^e obtiiiiied from i>().sttnortem sliuiy jind tlic clinical fin(liii<^s inusl ho bridged by animal experimentation. We have in the past depended too much upon form without inquiring sufficiently into function. At present there is an extra- ordinary interest in physiologic experimentation, and its effect upon surgery has l)een most beneficial. The work of Carrel, Crile, Gush- ing, and especially Cannon, whose investigations in this field have added so largely to our knowledge, are conspicuous examples of the trend of recent progress. A vast amount of work in the practical and comparative anat- omy of the gastro-intestinal tract has been done by Cunningham, Huntington, and others, and in the experimental physiology of the digestive system by Pawlow, Starling, and a host of co-workers. It is with the view of bringing together and correlating some of the undigested facts in embryology, anatomy, physiology, and pathology, which have a bearing on sound surgical practice, that this pai)er is written. Embryology. — The study of the embryology of the gastro-intes- tinal canal is most interesting and instructive, and it shows how much more permanent is function than form. The primitive in- testinal tube is composed of three fundamental parts, the foregut, midgut, and hindgut. In form the small intestine begins at the pylorus and the large intestine at the ileocecal valve, but func- tionally they maintain the rudimentary type, the small intestine beginning in the duodenum, just below the common duct, probably at the muscle described by Ochsner, which he has shown to be found near the juncture of the second and third portions of the duodenum and the large intestine beginning near the splenic flexure of the colon (Keith). The ancients, in calling the cecum a "second stomach," were much nearer the truth than we thought. From the emliryologic foregut we get the tongue, the back wall of the pharynx, the esophagus, the stomach, and duodenum to its third portion, the liver and pancreas being developed from the duodenal end. The gall-bladder and common duct are direct invaginations of duodenal structures (Cunningham). It will be noted that all these organs having their origin in the foregut have 310 WILLIAM J. MAYO to do with the preparation of food for digestion, but are not them- selves capable of absorption. The stomach and duodenum have some selective action in absorbing certain things, such as stimu- lants, but so far as food and drink are concerned, they are prac- tically unable to absorb. In operating upon the upper gastro- intestinal tract soon after a meal I have frequently noted the milky lines made by the full lacteals in the jejunum, while there was no such appearance above this point. The duodenum, with its wide caliber and fixed position and with its delivery point nearly as high as its pyloric origin, enables the chyme from the stomach to be thoroughly mixed with the pancreatic and biliary juices, absorption taking place below Ochsner's muscle. The jejunum and ileum take up the solid por- tions of the food, absorbing at least 90 per cent, of the proteids, but by no means do they make way with all the fluids. The con- tents of the ileum at the ileocecal valve are still in the liquid state, though the ingesta have lost the bulk of their nutritive elements. The cecum absorbs the fluids which are held for this purpose between the ileocecal apparatus and a physiologic muscular con- tracture (the cecocolic sphincter) near the hepatic flexure of the colon. In this intestinal segment the contents are churned back and forth until the fluids have been reabsorbed and the waste solids driven further along into the transverse colon. It is easily seen that the midgut retains its primitive characteristics: all the absorption takes place in the jejunum, ileum, cecum, and adjacent colon. Man prepares his food with the organs which have their origin in the foregut, and absorbs his nutrition from the derivatives of the midgut, i. e., he eats with the jejunum and ileum and drinks with the cecum. The posterior "no-loop" gastrojejunostomy does not, therefore, deprive the patient of any appreciable amount of absorbing surface, and the nutrition is in every respect normal. The cecal function explains why cholecystenterostomy into the hepatic flexure of the colon has proved so successful an alleviation in complete obstruction of the common bile-duct, when the duo- denum for any reason could not be used for the purpose. When a permanent external cholecystostomy is made, the patient loses SURGERY OF STOMACH AM) ASSOCIATED VISCERA 311 with the bile discharge 20 to 30 ounces of fluid each day, thereby being compelled to drink a larger f|uanlify of liquid, and at best shows signs of dehydration, whereas when the bile fluids are turned into the cecum, they are quickly reabsorbed. The effect of the bile in digestion, especially in aiding the pancreatic ferments, is, of course, lost, but its usefulness upon the function of the large in- testine is j)reser\'ed. The value of water in the human economy is testified to by the fact that it is redistilled in the cecum and used in the system over and over again. The mechanical effects of fluids in carrying solids as far as the cecum cannot be overestimated, but beyond this situation liquids would interfere with the storage function. The primitive hindgut begins near the splenic flexure of the colon, and although largely a matter of convenience, this portion of the intestine has a very considerable nutritive function. The normal action of the colon is an antiperistalsis, except during defecation. In his most admirable address on Surgery before the British Medical Association in 190,5 ^Ir. Bond shows that particles of indigo-carmin, placed inside the anus, are carried upward by what he calls "reverse mucous currents." This process occurs close to the intestinal mucous membrane, and takes place in spite of bowel passages. Cannon, in his experiments, also finds that antiperistalsis is the normal movement in the large intestine. By selective action valuable food elements, and especially fluids, are carried back into the cecum. Great advantage has been taken of this physio- logic fact in giving saline infusions by the rectum. Murphy has shown that salines introduced very slowly into the rectum will be absorbed with great rapidity, largely by reverse peristalsis to the cecum. From two to four quarts, which otherwise could not be introduced into the body except by hyperdermoclysis or venous transfusion, can be taken up in this manner. This is especially valuable in connection with gastric surgery, when the stomach is not available and fluids are essential. Comparative ])hysiology is interesting as showing that in carnivorous animals digestion is practically completed in the small 312 WILLIAM J. MAYO intestine, while in the herbivorous the colon is equally important, on account of the liquid nature of the plant and grass juices upon which they depend for nourishment. The human species occupies an intermediate position. Anatomy. — The dome of the stomach follows the curve of the diaphragm, and rises above the esophageal opening. In searching with the hand in the stomach for the cardiac orifice one uncon- sciously tends to pass above the actual situation, especially as on irritation the cardia contracts, leaving the mucous membrane nearly smooth. A little dimple, however, can be found, and steady pressure at this point for a short space of time allows the finger to pass through the opening. The lesser curvature has a somewhat fixed position; in the anesthetized patient about two-thirds of its extent hangs nearly longitudinal with the long axis of the body, while the lower hori- zontal third turns sharply to the right and somewhat upward, varying from 2 to 3}^ inches in length. Distention of the stomach comes about almost entirely through alterations in the relations of the greater curvature. The stomach is a contractile organ, exercising its functions through muscular action, gravity playing but a small part in the onward progress of food. When distended, it extends downward and to the right, the pylorus being carried across the median line and upward to prevent the weight of the food resting against the pyloric sphincter. In all distended organs having storage function this elevation of the outlet will be found to prevent continuous exertion of muscular force for retention purposes, as shown in the urinary and gall-bladders. The stomach is emptied of its contents, not only by general contraction, but also through muscle-bands which extend down- ward from the fundus, grasping the greater curvature (Cannon). As these bands contract the greater curvature is shortened, and the pyloric outlet brought more nearly to the bottom of the cavity. In a general way it can be said that all the stomach lying to the left of the longitudinal part of the lesser curvature has storage function (fundus), about four-fifths of the whole (Starling); on SURGERY OF STOMACH AND A55SOCIATED \XSCERA 313 the right, having the horizontal part of the lesser curvature as its superior wall, is the antrum, the grinding fH)rtion of the stomach. There is said to be a slight thickening of the circular muscular fibers at the entrance to this cavity. Starling states that this is apparent rather than real, and that the antrum is a physiologic and not an anatomic compartment. By contraction of these circular muscle- fibers the food-masses are held powerfully in the pyloric end during the kneading process, some prepared chyme escaping through the pylorus, while a much larger amount of food is ejected backward into the fundus, which steadily compresses the whole. In pyloric obstruction the antrum loses its identity and becomes a part of the distended fundus, as in the cadaver, accounting for the perfect drainage of a gastrojejunostomy; whereas if the same operation is done and the pylorus be unobstructed, the artificial stoma fails to drain, and is often productive of harm, as the muscular action of the stomach propels the food out through the normal outlet without regard to the gastro-enterostomy. The terminal three-fourth inch of the stomach next to the pylorus can be classed with the pyloric apparatus, ha\-iiig but comparatively little to do with the food-grinding, and acting rather as a funnel with the apex at the pyloric ring (the pyloric canal of Jonnesco). The pylorus, under the muscular contractions of the stomach, projects into the duodenum, and, from the duodenal side, greatly resembles the vaginal portion of the cer\'ix. The importance of these muscular features is sho^ii by the fact that nearly 80 per cent, of all ulcers of the stomach are situated in this grinding pyloric end, the most common variety being the saddle ulcer of the lesser curvature, extending flap-like down the anterior and posterior surfaces (prepyloric). The ulcer crater, if one be present, will usually be found upon the posterior wall, or there will be two ulcers facing each other, one anterior and the other posterior, connected superiorly by a bridge of induration. The pylorus is seldom primarily involved in idceration, as the pyloric canal is normally contracted and is less exposed to mechan- ical injury or the presence of an excess of acid secretions. The large majority of so-called pyloric ulcers are in reality duodenal. 314 WILLIAM J. MAYO The latter ulcer extends up to the pylorus, or within three-fourth inch of it, in 96 per cent, of all the cases; error in accurate localiza- tion has led to mistaken identification. As a matter of fact, we find at the operating-table that no less than 40 per cent, of all gastric and duodenal ulcers are situated in the duodenum. On the peritoneal surface of the gastric side of the pyloric ring will be found a peculiar arrangement of the blood-vessels which is nearly constant. From the lower side a thick vein passes upward somewhat more than half-way upon the anterior surface. From the upper border a second vein reaches downward in the same line, nearly, if not quite, meeting the first. Not only is ulceration most frequent in the antrum, but the topography of gastric cancer will be found to be the same as ulcer. In 54 per cent, of 134 resections of the stomach we found malignant disease originating in the submucous tissue at the margin of an ulcer. Primary cancer of the duodenum, however, is rare. We have seen only two instances, although three times we have found carcinoma developing on the gastric side of a duodenal ulcer which had involved the pylorus. The pyloric antrum occupies a sheltered position under the left lobe of the liver, and this is also true of the duodenum above the common duct. As contrasted with the remainder of the stomach, acute ulcer perforations into the free peritoneal cavity are rela- tively uncommon in the antrum on account of the ease with which adhesions to the liver, gastrohepatic omentum, and suspensory ligament and gall-bladder are created, and these tend to prevent the gastric contents from escaping. The cardiac end of the stomach has no such protection, and although this region has but 10 per cent, of the total number of ulcers, acute unprotected perforations are more frequent here than in the entire antrum. Perforations of duodenal ulcers are more common than gastric, but the sheltered situation of the duodenum enables ready adhesive protection, while its contents are nearly sterile and relatively small in amount. The blood-vessels supplying the stomach are all from the celiac axis, and are four in number: the gastric, which reaches the stom- ach on the lesser curvature, just below the esophageal opening; the SURGKKY OF STOMArH AND AHSOCIATKIJ VISCEKA 31;5 superior pyloric branch of the licjialic artery at the p\loriis; the pistroduodciial, which gives rise to the right gastro-cpiploic; aiul the left gastro-epiploic, from the splenic artery. The blood-vessels of the lesser curvature lie in the wall of the stomach; those of the greater curvature are to be found at a considerable distance from the gastric wall, arterial branches passing upward from the gastro-epiploics upon the stomach, anterior and posterior, in a sawback manner. When the stomach is distended, this permits the greater curvature to sag downward toward the blood-vessels without compressing them, enabling rapid changes in size and position. Tortuosity of the uterine arteries makes possible the great size of the uterus during pregnancy, because it comes on slowly; but tortuosity would not allow the rapid changes to which the vessels of the greater curvature of the stomach are subjected. By tying the four blood-vessels, gastrectomy can be made bloodless, just as hysterectomy is made bloodless in the modern operation. The lymphatic arrangement is nearly the same as the vascular, but varies in one important particular. On the greater curvature no lymph-nodes are to be found to the left of its middle, and the lymphatic circulation of this region is from left to right (Cuneo). Therefore, in radical operations for cancer it is possible to save a great deal of the greater curvature. On the contrary, in the lesser curvature the lymphatic structures lie in the submucous tissues, rendering it necessary, in every case of carcinoma of the pyloric end, to remove all the lesser curvature as high as the gastric artery (Mikulicz). The glands about the pylorus are most numerous on the inferior surface, although several are found just above it in the line of the pyloric artery. The relation of these glands to the field of gastric cancer was long ago dwelt upon by Kocher. The dome, arising above and to the left of the cardiac orifice, is disconnected from the remainder of the stomach in its blood and lymphatic supply (Robson and Moynihan). In a general way the lymphatics of the stomach greatly re- semble the other hollow viscera of the bodv; the neck of organs 316 WILLIAM J. MAYO is the region of the lymphatics, the fundus being less abundantly supplied. Physiology. — Thanks to the investigations of a large number of physiologists, we know many of the secrets connected with the functions of the stomach and duodenum. One important fact developed is that the stomach not only has many ferments, but that these ferments are called into action by the character of the ingesta, and they can be increased or diminished by appropriate diet. Gastric function is carried on to a considerable extent inde- pendent of the nerve-supply; the chyme, when ready to leave the stomach, has attained a proper acidity, and this causes con- traction of the gastric muscles, the pylorus automatically opening to permit its escape. The duodenum also has control over the pylorus, and acidity is here again a feature influencing the pyloric closure. According to Kelling, the chemistry of the duodenal contents is the most important agent in the control of the pylorus. The mucous membrane of the antrum produces specific substances called hormones (Starling), which, acting with nerve impulses such as sight, taste, and smell, control the amount of food ingested and the necessary gastric secretion. Not only is much of this process the result of chemistry, but the same force, through the vascular system, stimulates the gland- ular activity of the liver and pancreas. It has been shown that the introduction of chyme into the duodenum causes pancreatic secretion by means of a product of the intestinal mucous membrane called "secretin," when the pancreas has no connection with the body except its blood-supply. The great protective agent in pre- venting self-digestion in the stomach is mucus, and this is ap- parently true of the entire gastro-intestinal canal. The stomach equalizes the temperature of the ingesta, macer- ates the food-masses in a weak solution of hydrochloric acid and pepsin, and converts the contents into a harmonious whole, the muscular action of the pyloric antrum being the active agent in the latter process. The cardiac end is a temporary storehouse which enables its possessor rapidly to place a quantity of material SURGERY OF STOMACH AND ASSOCIATED MSCERA 317 where it can be drawn upon as di^'cstion proceeds. To a large ex- tent, then, the stomach is a convenience, and obviates the necessity of continuous feeding, just as the urinary bladder and the large intestine beyond the splenic flexure are conveniences to prevent continuous eliininalion. Excess or changed secretions, especially acidity, seem to lie behind much of the pathology of the stomach, accounting largely for ulcer in the pyloric antrum and duodenum above the common duct, with its alkaline secretions. Trauma plays an important part in the jiroduction of ulcer. We have but to remember the frequency of this malady in the grinding pyloric end of the stomach, and to note that duodenal ulcer usually originates at the point which received the impact of the chyme forcibly ejected from the pylorus to appreciate the influence of local injury in gastric disease. It has been thought that gastrojejunostomy would pass the food too quickly from the stomach into the intestine. This belief, however, has been proved groundless, as the stomach does not contract in such a manner as to empty its contents until the proper chemistry has been reached and the food is ground in the antrum before propulsion is begun. It is altogether probable that modern methods of food prepara- tion have greatly changed gastric digestion, and that a considerable share of the now very prevalent diseases of the stomach are due to modern dietary changes from primitive conditions, just as the loss of necessity for the grinding action of the teeth has resulted in their premature decay. In aboriginal races many diseases of civilization, such as appendicitis, gall-stones, ulcer, and cancer, seem to be rare (Senn). The nerve-supply of the stomach is of two kinds — the vagus and the sympathetic. During fetal rotation the stomach turns upon its right side, which thereby becomes the posterior wall, and the left vagus lies anteriorly. Terminal filaments of the vagi join with the sympathetic fibers from the abdominal ganglion, and form the plexuses of Auerbach and Meissner, which lie in the gastro-intestinal wall. The control of the cerebrospinal nerves over the stomach is 318 WILLIAM J. MAYO limited, and has to do with food requirements, but, so far as actual digestion is concerned, the sympathetic is the controlling factor. Beyond the stomach the cerebrospinal nervous system is even less influential until the sigmoid and rectum are reached, where again conscious control is essential. The sympathetic nervous system, developed from mesoblastic tissue, is undoubtedly the primitive one. It is closely allied with the control that is inherent in the gastro-intestinal muscles, and which is myogenic in origin, and of the same nature as His' heart- muscle band, which regulates the heart-beat. The action of the gastro-intestinal canal is regulated largely by this mysterious myogenic force. The sympathetic nervous system has retained to a large extent its primitive control over those organs which have to do with the maintenance of the body, but it is altogether probable that it is losing its prominence on account of the overshadowing develop- ment of the cerebrospinal system, and just as other vestiginal organs, like the gall-bladder, appendix, and wisdom teeth, seem to develop a tendency to disease, so possibly does this mysterious nerve body. It is within the realms of possibility that abdominal ptosis and many forms of neurasthenia are characteristic expres- sions of evolutionary instability in the sympathetic nervous system. The one essential difference between the sympathetic and cerebrospinal nervous system is that the sympathetic is not seg- mented, and is, therefore, unable to prevent general disturbances upon irritation of any of its component parts. This is especially interesting in connection with certain diseases of the stomach, which are exceedingly misleading, and which apparently have a common origin, such as pyloric spasm, atonic dilatation, prolapse of the stomach, and gastric neuroses. Now that surgery of the stomach is occupying so important a place, it is vital that we should eliminate these conditions from the operating-room. We have had an opportunity, in a considerable number of cases, to examine the stomach during that contraction of the pyloric canal and antrum which constitutes "pyloric spasm." This muscular contraction, when present, causes the patient to suk(;kky of stomach and asso( iatkd visckua 319 oxporionco a sensation wliicli ho speaks of as "gas pain." It can he ohserved in operative examination under local anesthesia. The peculiar appearance of the stomach during pyloric spasm may give rise to the belief that physical disease exists, because f)f a rougliened, puckered appearance, which lasts a fraction of a minute, and then changes position or develops a rhythmic contrac- tion, confined to the antrum. In tlie most extreme degrees we have found it in connection with stones in the appendix, impacted gall-stones, tuberculosis of the intestine, and chronic intestinal obstructions of various kinds. Clinically, it would appear that irritation of any part of the gastro- intestinal canal and allied organs, the liver and pancreas, may produce j)yloric spasm, and that this condition may overshadow the local disease. The greatest care is necessary in differentiating these cases from gastric ulcer and other inflammatory infections. While it is possible that pyloric spasm may exist as the result of a mucous ulcer, as believed by von Eiselsberg and Doyen, such is not our conviction, although it will mimic gastric ulcer clinically, and is often diagnosed as such. "We are inclined to look with a great deal of suspicion upon any ulcer of the stomach which cannot be absolutely demonstrated not only to the operator, but to onlookers as well. Bacteriology. — The contents of the fasting stomach are rela- tively sterile, and this is equally true at the height of digestion, the chyme, when discharged into the duodenum, being nearly free from pathogenic organisms. This is due largely to the acid gastric secretion; while not actively germicidal, the general influence of all the secretions of the stomach are against germ Ufe. Cooking renders a large amount of the food consumed sterile, but many organisms during mastication are picked up from the mouth, which teems with bacteria of all kinds (Harrington). Cushing found that with sterilized food the discharge from a jejunal fistula was free of microorganirsms (providing the mouth was kept in good condition), and recommended, in addition to sterilizing the food, antiseptic oral washes and careful cleansing of the teeth previous to operations upon the gastro-intestinal tract. The 320 WILLIAM J. RIAYO vigorous use of the tooth-brush to those who are not accustomed to it, at the end of a few days may start up a gingivitis, and tem- porarily increase the virulence of the organisms. Such oral anti- sepsis in this class of patients should begin at least two weeks prior to the operation. As the average patient cannot be kept in prep- aration such a length of time, the practical importance of this step in the individuals just referred to is not great. A moderate number of bacteria, particularly bacilli, pass with the food into the duo- denum, and members of the colon group work upward from below. Adami and Ford have shown that leukocytes pass out upon the free mucous surface of the duodenum and upper jejunum, picking up particles of fat and microorganisms of various kinds, particularly bacilli, and by phagocytosis destroy them in the neighboring lymphatics. Some bacteria, however, are continu- ously carried to the liver and there annihilated, the pigments of the slaughtered organisms giving rise to the little pigmented areas sometimes found in the liver. A varying number of bacteria, however, are passed through the liver and excreted with the biliary secretion. The bile, therefore, must be looked upon as always infected, and it is probably this attenuated infection which gives rise to gall-stone disease (Lartigau). The relative sterility of the stomach and duodenum and the upper jejunum shows why gunshot wounds of this locality have been so much more often followed by recovery when operated upon than those further down in the intestinal tract. One interesting feature in this connection concerns the reason why gall-stone disease is so much more frequent in women than in men. In 1700 of these patients upon whom we have operated three-fourths have been women, while just the opposite has been true of duodenal ulcer. In 200 operated cases of the latter disease 73 per cent, were males, and only 27 per cent, females. Granting that biliary infection is equal in men and women, is this sex dis- proportion due to different mechanical conditions? We have been investigating, during life, a considerable number of patients, with a view to determining whether there was a difference in the arrange- ment of the duodenum and common duct which favored ascending SUHGKIiV Ol' ST().MA(H AM) ASSOCIATEO \I.sri:KA 3-21 ^alI-l)Ia(l(lor infection in women, and also as to whetlier the nic- fliaiiics in men were uni'avoralile for [jermittinj; the alkahne dis- charge from the common duct quickly to neutralize that excessive acidity of the chyme which seems to lie behind the etiology of duodenal ulcer. We have not been able to satisfy ourselves that tliere is sufficient anatomic difference to ex|)lain the phenomenon. There is no (luestion but that bacteria can travel uj) the common duct into the gall-bladder. Bond found that indigo-carmin placed within the anus could be detected in the gall-bladder in about forty- eight hours. It is probable that this .sex difference is physiologic rather than anatomic, and possibly due in some way to the poten- tial capacity of the female liver to care for mother and child. In the passage down the intestinal canal bacteria increase in numbers and in virulence, and in the large intestine bacterial growth adds considerably to the bulk of the stool. The acid change which takes place in the large intestine is probably caused by germ life rather than by intestinal secretions, as the large in- testine, when entirely free from feces, has, like the small intestine, an alkaline reaction (Bond). The problem of securing asepsis during operations upon the upper abdomen is most interesting. To one who reads about the gastric surgery of five and ten years ago the frequency of fatal results from pneumonia following operations upon the stomach is most noticeable, and had various explanations, the anesthetic and the aspiration of gastric contents regurgitated into the esophagus being the most generally accepted hypotheses. In Mikulicz's clinic it was brought out that part of the venous blood from the stomach, instead of passing through the portal vein so that the liver might sterilize it, returned directly through the vascular anastomoses about the cardiac orifice, and that this unsterilized blood was the cause of a pneumonia embolic in character. Muscatello demonstrated that the endothelial serous hning of the diaphragmatic area was exceedingly active in absorption, and, based ujion this, Clark elevated the foot of the bed after abdonu'nal operations to insure rapid absorption of septic products and prevent peritonitis. The unshed blood in the vessels has no VOL. I — il 322 WILLIAM J. MAYO germicidal properties; and this septic material, when absorbed before being acted upon by the proper tissues, was carried to the lungs, producing embolic pneumonia of the same variety that had been noted after operations on the stomach. Fowler pointed out that the inherited resistance to peritonitis in the pelvis, which had been brought about by tubal diseases in women and appendi- citis in both sexes, caused absorption in this locality to be slow, and therefore advised that, in peritoneal sepsis, the head of the patient's bed should be raised, and a drainage-tube introduced into the pelvis to drain away the discharges. In connection with general septic peritonitis we followed the Fowler method, except that we raised the patient's head and chest much higher. The betterment in the mortality in peritonitis, actual or impend- ing, was remarkable, although it was soon noticed that although the patient did very well, there was often no drainage from the tube, therefore drainage was not the sole factor. In work upon the upper abdomen on the left side any fluids at once gravitate toward the diaphragm; on the right, they are prevented from doing so by the liver. To prevent rapid absorp- tion of these unsterilized products the raising of the diaphragmatic area so as to drain this material into the pelvis is of the first im- portance, as here, by means of the omentum and pelvic colon, any septic particles can be rendered harmless by absorption through the lymphatics and radicals of the portal vein. So far as these organs which lie within the active radius of the portal vein are concerned, the liver must be looked upon as an adjuvant to the lymphatic glands in the destruction of microorganisms. The omentum, according to Dickinson, is the most important agent in developing phagocytosis and opsonins; its germinating endothelium is constantly producing lymphocytes, and is capable, under proper stimulation, of throwing both newly formed phago- cytes and those called from a distance into germicidal action. This process is aided by the vermicular and swaying movements of the intestines, which, in spite of gravity, brings all parts of the small intestinal wall in contact with the omentum, the epiploic tags having the same function for the more fixed large intestine. SI i{(;kuv of st().ma< II and assoc iatkd vise era 32.'5 Dudgeon and Sargent have shown that, from some source, the .stai)hyIoc-occu.s alhus first api)cars in jjeritoncal sepsis, and rapidly |)ro(hues a mild infection, and this furnishes tlie necessary stimulus to the protective endothelium, so that the phagocytes are attracted in time to destroy the more virulent, hut later developed l)acteria. Bond has demonstrated that the plastic exudate produced l>y the stai)hylococcus albus protects the weakened endothelium, and prevents the microorganisms passing directly into the blood- stream. When absorbed by sound endothelium, bacteria are carried into the lymphatics and destroyed. In closing, let me quote from Welch's recent address on the occasion of the dedication of the new buildings of the Harvard Medical School: "There is a highly significant and hopeful scientific movement in internal medicine and surgery today, characterized by the establishment of laboratories for clinical research, by the applica- tion of refined physical, chemical, and biologic methods to the problems of diagnosis and therapy, and by the scientific investiga- tions along broad lines of the special problems furnished by the living i)atient." r- LIVER AND GALL-BLADDER REPORT OF TWO OPERATIONS FOR THE RELIEF OF GALL-STONES AND ONE FOR STRICTURE OF THE COMMON DUCT OF THE LIVER* WILLIAM J. MAYO Case I. — Gall-stone. — E. B. B., age tliirty-six. male. Ma- chinist. History: For several years liad sufi'ercd occasionally from attacks of severe pain in the right side. About one year ago he had one of these attacks, but instead of recovery, it was fol- lowed by prolonged illness with great suffering, and eventually an abscess formed in the right luml)ar region, which was opened in several places by his attending physician. December 1, 1890, he was admitted to St. Mary's Hospital. He was emaciated, somewhat jaundiced, and required opiates to relieve his suffering. On the right side, extending from the free margin of the ribs to the iliac crest, was an ill-defined tumor, and in the right groin a small sinus. Operation, December 2, 1890. Abdominal section in the right linea semilunaris. Upon opening the abdominal cavity a mass of adherent tissue was encountered, surrounding a contracted and adherent gall-bladder containing a single stone. A rubber drain was placed, followed by free discharge of bile during the next two weeks. It was of interest to note that a laxative dose of sulphate of magnesia ])roduced a copious discharge from the biliary fistula, but of a watery character; calomel caused less discharge, but of thicker and darker bile. The fistula healed and the patient was discharged in three weeks. In a short time he gained 40 pounds in weight and is now working at his trade. The previous history of abscess must have been due to empyema of the gall-bladder with external perforation and discharge. Case II. — Gall-stone. — F. L., female, age thirty-three. Married. Four children. Admitted to St. Mary's Hospital * Reprinted from tlie " Xortluvestern Lanoel," April 1, 189^. an 328 WILLIAM J. ^L\YO June 19, 1891. Her attending physician gave us this history: "Nine days after the birth of last child, two years ago, she was seized with intense pain in the region of the gall-bladder which lasted some hours. Since that time she has never been entirely free from pain in the right side, and at frequent intervals has suf- fered from the most agonizing colics requiring large doses of mor- phin for relief. She has never been jaundiced. An indefinite tumor can be felt in the region of the gall-bladder." Operation, June 24, 1891. Abdominal section from the costal border of tenth rib downward three inches. The gall-bladder was found elongated and filled with clear mucus. Impacted in the cystic duct was a single large gall-stone which was extracted after great difficulty. The bladder was sutured to the abdominal wall and a rubber drain inserted. More or less bile was discharged until the fistula was healed, in fourteen days. Patient recovered completely and gained rapidly in flesh. Case III. — Stricture of Common Duct of Liver. — G. Z., female, age twenty-eight. Married. Five children. Referred to me with this history: "After the birth of the last child, one year ago, had some puerperal inflammatory trouble, with more or less jaundice and pain in the right side ever since." Upon examina- tion found a badly lacerated cervix, a cystocele, a ruptured per- ineum, and prolapse of enlarged ovaries and tubes into Douglas' pouch. November 9, 1891, admitted to St. Mary's Hospital. The uterus was cureted and irrigated, cervix repaired, an operation made for the cystocele, and a new perineum built up at one opera- tion with catgut sutures. This was followed by marked improve- ment in her general condition for about two months, although a certain amount of jaundice persisted. From this time on the jaundice gradually increased, with clay-colored stools, etc. The enlarged and adherent ovaries and tubes still prolapsed and painful. February 3, 1892, exploratory abdominal section was made from the costal end of the right tenth rib do^mward three inches. On account of the enormously enlarged liver this was extended downward; deep under it the moderately distended gall-bladder could be felt, but on account of the size of the liver could not be drawn into the wound. Following the cystic duct to the common duct a mass of inflammatory adhesions were encountered which were torn loose, and although the patient was thoroughly anes- GALL-STONES AND STRICTURE OF COMMON DCCT OF LIVER "i'Z*.) thclizcd, she at once Ix'^an to V(jiuit up hilc in large (jwantities. As it was evident that the duct was again patent, the incision was closed with sijkworni gut and a second section made in the Hnea alba for the removal of the diseased aj)pendages. The tubes were adherent, the ovaries enlarged, and small i)apillomata were found on the right broad ligament. During the succeeding twenty-four hours tlie patient vomited over two quarts of bile, and after this there was a ra{)id and uneventful recovery. The jaundice rapidly disappeared, and the patient was discharged in three weeks. COMPLETE OBSTRUCTION OF THE COMMON DUCT OF THE LIVER. ANASTOMOSIS BE- TWEEN THE GALL-BLADDER AND JEJU- NUM BY MEANS OF MURPHY'S BUTTON* WILLIAM J. AND CHARLES. H. MAYO Mr. H., age seventy-one. Referred to us with the following history: "A number of attacks of gall-stone colic during the past two years, increasing in frequency and severity. In February of this year he had an attack which was followed by a light jaundice; this cleared up in a short time. Three weeks later he had another attack of the colic, with complete obstruction of the common duct and almost constant pain since, with vomiting, emaciation, and jaundice of the most pronounced type." Admitted to St. Mary's Hospital May loth and operated upon May 17th. The usual incision was made at the right of the rectus muscle, four inches in length. The gall-bladder was found deeply placed under the liver; no stones could be detected. The ducts were traced downward to a point close to the duo- denum, where the obstruction seemed to exist. The incision was enlarged by a lateral cut across the rectus muscle, but the point of obstruction was too deeply placed for any accurate manipulation or inspection. The gall-bladder, which could not be drawn to the surface, was opened, and by means of the Murphy button was quickly fastened to a loop of the intestines as high as practicable. The incision was closed, and the patient made an uneventful re- covery. The jaundice quickly disappeared, and appetite and strength returned. Considering his advanced age and debilitated condition, no other method of operation could have been performed without great or insurmountable technical difficulties. * Reprinted from "Northwestern Lancet," June 1, 1893. 330 SURGERY OF THE GALL-BLADDER, CYSTIC AND COMMON DUCTS, WITH REPORT OF SEVEN CASES OPERATED UPON* WILLIAM J. MAYO Mr. President and Members of the Southern Minnesota Medical Societ7j: The rapid advance in abdominal surgery during the past ten years has brought the gall-bladder into the operative field, and in this time more definite knowledge in regard to its pathology and treatment has been gained than during the previous one hun- dred and sixty years. In 1733 Petit wrote the first of his classic essays, and during the succeeding ten years he placed the pathology of gall-bladder disease upon a sound basis, far in advance of his time, but his work was little appreciated until recent years. With the anatomy of the gall-bladder, its ducts, and their relative position to the liver and duodenum, you are all familiar. The physiologic function of the gall-bladder is a moot point — the commonly accepted belief is that it is a storehouse for bile to be discharged during digestion. So late an authority as Landois and Stirling authorizes this view. J. B. Murphy logically attacks this question, and, as a result of experimental and practical study, asserts that it has nearly the same function as the second bulb of a syringe in regulating the flow of bile, causing a steady stream, rather than an intermittent current, into the duodenum. Cholelithiasis is the mo^t common patho- logic condition of the gall-bladder for the relief of which surgery offers the only rational method. Gall-stones, as a rule, are formed in the gall-bladder as a result of precipitation and accretion. Under some circumstances stones * Reprinted from "Jour. Amer. Med. Assoc," August 26, 1893. 331 332 willia:\i j. mayo are also found in the hepatic ducts, especially in cancerous ob- struction. The diagnosis depends largely upon the history, char- acter, and location of the pain, and often physical examination will reveal a tumor in this region. Jaundice does not appear unless the common duct is obstructed, and is far more common in malig- nant disease, and for the same reason the color of the stool, upon which much stress has been laid, is usually of small importance. The diagnostic value of the finding of gall-stones passed with the stool is absolute, but I am inclined to think that such passage of gall-stones is less common than is generally believed, and certainly the onset, duration, and cessation of a colic is no indication that a stone has been passed, but meiely that the cystic duct has been obstructed, and that either the obstruction has been removed or the bladder has exhausted itself in the effort. When olive oil in large quantities was a popular remedy for hepatic colic, the soap- balls passed with the stool and resulting from the action of the intestinal alkalis upon the oil were erroneously supposed to be the oflFending bodies, and as such were exhibited to the sufferer and his friends. At the present time we hear much less about stones found in the feces. While stones may be passed through the ducts, or by ulceration, into the intestine, externally, or into any neighbor- ing viscus, or after causing years of suffering remain quiescent without producing further trouble, such fortunate outcome is very exceptional, and in the majority of instances operation is the only relief from a life of suffering or death from a complication. Septic infection of the gall-bladder, either as a result of stones or from extension upward through the ducts of a septic process, is a not uncommon occurrence, and may result either in chronic inflam- mation or empyema. Fenger has done much to elucidate this subject. Large accumulations in the gall-bladder are not infrequently confounded with right renal tumors or even with ovarian cysts, and many such mistakes are recorded, especially when dropsy of the gall-bladder exists, with great retention of catarrhal products as a result of duct obstruction. Injuries of the gall-bladder some- times occur. Some years ago, in my father's practice, I saw a case SUIKilOKV OF GALL-HLADDKU, ( VSTK' AM) COMMON DICTS 333 of undouhtcfl rupture of the f^'all-hhulder. A hoy twelve years of jif^e was tlirowu froui ji wa^'on, the wheel passing partly on to the right side of the ahdonieu. Ascites developed, and large (piaiiti- ties of thin bile were asi)irated at different times during a month. Comi)lete recovery took j)lace. Operations upon the gall-bladder may be divided into tiiree general classes : First, cholecystotomy, or the simple ofjcning and removal of stones. Lawson Tait usually performs this operation at one sit- ting, and the open gall-bladder is stitched into the incision, form- ing a temporary fistula. Since bile is not septic and does not cause peritonitis, other than the adhesive variety, slight biliary contami- nation of the i)eritoneum causes no harm, and this open method enables us to manipulate with the finger inside the abdomen, outside of the gall-bladder, which at times is a great aid in extract- ing stones. It also allows the subsequent escape of overlooked stones. If the contents of the gall-bladder be septic upon aspira- tion after the abdominal incision is made, it is far safer to stitch it into the incision and delay opening for several days until ad- hesive inflammation has shut it off from the general cavity. Blind aspiration of the gall-bladder has long been known to be of great danger; not, as supposed, on account of the escape of a few drops of bile, but because of the contents being often septic and the slight leakage setting up septic peritonitis. Suture of the gall-bladder after opening and removal of stones is seldom practised, and the method is condemned as unsafe. This, however, is not logical — the whole cjuestion of safety depends on the patency of the ducts. If there be no obstruction to the outflow, so that there will be no tension within the sac, suture and return is a safe procedure. Abbe tests the condition of the ducts by using a syringe and forcing water through them into the in- testine; if the fluid passes freely, he does not hesitate to trust to immediate suture and return. The second general class is where the gall-bladder is removed. Langenbeck first practised cholecystectomy and formulated in- dications for its performance; he gives much too wide a scope to 334 WILLIAM J. MAYO this procedure, and Greig Smith very properly hmits its applica- tion to single or double stone, where the fundus of the gall-bladder cannot be sutured to the abdominal wall, or to cases wherein the tissues are too thin or inflamed to bear a suture. It is mainly practised by a few Continental surgeons and is not a popular operation. The third class is a very important and often perplexing one — wherein obstruction exists either in the cystic or in the common duct. To suture such a gall-bladder to the abdominal walls is to invite a permanent fistula — in any case an annoyance, and if complete obstruction of the common duct exists, the escape of all the bile externally leads to debility and eventually to death. Fortunately, the recent work of J. Thornton, Mayo Robson, Robert Abbe, and Charles McBurney has given us meth- ods of opening these ducts and removing stones with either suture of the incised duct or drainage. McBurney has success- fully opened the duodenum and shelled an impacted stone out of the intestinal orifice of the common duct. There will remain cer- tain cases in which the obstruction in the ducts cannot be removed, and in these cases enterocholecystotomy is our only hope of success. Winiwarter first sutured the gall-bladder to the colon; while this was much better than an external fistula, yet most of the physiologic effect of the bile in digestion was lost. Gaston, of Georgia, by experiments upon dogs, developed a complicated method of suture to the duodenum; but it remained for a brilliant young western surgeon, J. B. Murphy, to invent his mechanical device by means of which an effectual back door for the escape of bile into the duodenum can be safely and quickly made. In con- clusion, I append a diagrammatic report of seven cases in which I operated upon the gall-bladder or its ducts. 2 -Sri ^n J3JS 5S lliiiH • e c ! « a 2p is = »; 5 ■iiasaH I 0$ o §o .S 60-a .2 a a g«3 c Z ii 6 - 2« i h K o ;:_: o V; o «: « = ill 8o8 b w C oj: o lis s u 9 S B A u s 2 • f ^ b2 E-5 w3 js^^H 5-S £ J- 3 a 3 i=-"5 e 3 "-3 ■=• cs _S5 2503 .rCo>.u«in ss :5?J^£' c. :2: y^ yl^22y=3 " = s s i 1 !^ = 2'>> 6 o 0-— .c6 o o ' oE 2 3 Sh ^1 bS §e ■ E i" IsseJI^e = "o S.£f £1^ = §•1 3 S, -•- gwE tig'g« M U Ul (S 2 H < s o ri -e j: = .2.5; cs J*" t:o- ■5::«s 8-2 i= 12 a O rt 412 o M 3 15^ '-CO. c I. » 3.£r2 S^'ti O-O w K fc O ri < V n O 4-1 ^ ".2 *-» (J S e J2 ■ - o - ■— - c -J 00 „jq c-'T'S CO £%rt rCE c " IS" ^. .S53 5 rtH C E o— ,S ^'b5 «> rt-ti CO ^rt 0." - "2 b rt 22 S- >oS 2 ^Sb!- 3 .ja |«= ^•7;« 3 < H O Oh I z 3 H Z o u Q OS E_ .(/3 S ;^'5 QU lie ^X Q IJ2 S5d CO ■^ j:.E Q^ c ^.Si c . ^ c ; > c 33r. 4i K.5 S ^ c -5 :S 3: S GALL-STONE DISEASE* WILLIAM J. MAYO With a view to shortening the time necessary to present the subject, I have made but brief reference to the history or Htera- ture of gall-stone disease. From recent investigations Webster estimates that 10 per cent, of males, 25 per cent, of females, and 36 per cent, of the insane of both sexes have gall-stones. From 1500 autopsies Schroder estimates that 12 per cent, of adults and 2 per cent, of children are afflicted with gall-stones. Xaunyn says that only about 1 per cent, of those who have gall- stones are made aware of their condition by symptoms. These statistics must be received with a certain degree of caution and probably are high. Autopsies performed in large hospitals are of necessity made upon many who were sick some time before their death, and their lowered vitality might permit bacterial infection of the ducts and gall-bladder. With the slug- gish bile-current which would be engendered by a protracted ill- ness this is a condition which would not exist in the average in- dividual free from disease, and tends to render these figures some- what unreliable. These factors are certainly important, as shown by the high average of gall-stones in the autopsies of the chronically insane. Kilbourne, of the Second ]\iinnesota Hospital for the Insane, informs me that they are present in at least a quarter of the postmortems made at that institution. Etiology. — Gall-stones are usually formed in the gall-bladder, but under exceptional circumstances may form in the hepatic ducts. They are always produced by the mucous membrane and *Reprinted from the "New York Med. Jour.," vol. Ixii, August 24, 1895. 336 GALL-STONE DISEASE 337 l)ile, and never from the blood. There are two essential factors in their formation: (1) Slow flow of bile. (2) A catarrhal condi- tion of the mucous membrane, frequently due to a bacterial in- fection, and usually caused by Bacillus coli communis. Welch found germs in the center of gall-stones, even when no trace of their presence could l)e detected in the mucous membrane. The Bacillus coli communis is a great secondary invader, waiting for an entrance to be effected by more active and shorter-lived germs, and it is altogether probable that the original invading germ may often have disappeared after sufficiently devitalizing the epi- thelium so as to permit the coli communis to effect a lodgment. As an indirect cause, the changing abdominal pressure in child- bearing women becomes an important factor in influencing the bile flow, and an accidental duodenal inflammation may light up an ascending catarrh of the bile-tract. The frequency of bile- stone in hospital inmates may easily be accounted for by their sluggish existence in effecting the bile flow and lowered vitality, allowing of germ infection. In spite of the fact that those physi- cally debilitated or of advanced years are most frequently affected with gall-stones, the majority of those who seek relief from the suffering engendered are otherwise in good health. I recently removed 132 stones from a girl under twenty years of age with a history of colic for three years previous. ISIany similar operations in the young have been reported. It is an interesting query as to whether a simple catarrhal jaundice might lay the foundation of future gall-stones. In several instances I have been able to ob- tain a history of simple catarrhal jaundice, particularly in child- hood, followed by gall-stone disease after a lapse of many years. The common faceted gall-stones are usually composed of bile- ]>igments and cement substances from the epithelium; occasion- ally of cholestcrin, when they are crystalHne in fracture, light- colored, and often but a single stone, or cholesterin may be the nucleus of an ordinary calculus; not rarely the stones are pure bilirubin and few in number. There are usually either a large number of small stones or a small number of large stones; all the way from 1 to 8000 have been found. The largest number I have VOL. 1 — 2i 338 WILLIAJI J. MAYO removed was 402, from a woman twenty-four years of age. Twice I have found but a single stone. The diagnosis depends largely upon the pathologic condition. Ordinarily, gall-stones lying free in the gall-bladder do not produce symptoms of a very definite nature. Neuralgic pains extending up into the region of the liver and toward the right shoulder, with indigestion and a certain degree of local soreness, may be present. Under some circumstances the irritation of the stones will aid an infective process in causing an ulceration which may end in a con- tracted gall-bladder, vnih adhesion to the surrounding viscera, or may even expel the offending stone into the intestine, possibly to cause intestinal obstruction when of sufficient size. It is altogether probable that many cases of dyspepsia of ob- scure origin depend upon gall-stones undiagnosticated. Colics are not infrequent, especially several hours after eating, as food passes down the duodenum. Obstruction of the cystic duct, either temporarily or perma- nently, at once brings more urgent symptoms. Colics are frequent, and if the obstruction is complete, the pain is severe and lasts from a few moments to several days, until either the accumulated fluid in the gall-bladder is forced past the obstruction, which ordinarily happens, or rarely a small stone may be pushed on, with relief. At times, however, the obstruction is not relieved, and the over- stretched bladder is unable to contract with such painful force and terminates either in a cystic accumulation or, if the infection is sufficiently active, eventuates in an empyema of the gall-bladder, which in turn may form a fistulous tract to the surface. I have examined two old cases of gall-bladder fistula self-cured after years of suffering from gall-stone obstruction, and have operated on two empyemas of the gall-bladder due to occlusion of the cystic duct by stones. Stones in the cystic duct mUst be distinguished from renal colic, from diaphragmatic pleurisy, from lead colic, and, when enlarged, from movable kidney. A little reflection will suggest the radical points of difference. Temporary jaundice may accompany stones in either the gall-bladder or the cystic duct. Under such cir- GALL-STONE DISEASE 38i> cumstances it is a valuable sign, due to swelling, but its absence does not militate against the diagnosis. Tenii)orary jaundice in obstruction of the cystic duct should be looked for with a certain degree of anxiety as a possible forerunner of obstruction of the common duct, which indicates a more serious state of affairs. Stones in the common duct are usually the result of the passage downward of stones from the gall-bladder, and not infrequently stones in the gall-bladder and cystic duct coexist. The larger size of the common duct enables the majority of calculi to pass readily into the duodenum, but at times the large size of the stone, or more frequently a small floating stone, more or less obstructs this large bile-duct and produces a well-marked set of signs and symptoms. As a rule, the obstruction is not complete at once, as the steady biliary flow prevents early complete obstruction. The so-called "floating stone," which Fenger has done so much to elucidate, is a very interesting phenomenon, depending on the fact that the duodenal o[)emng of the common duct is smaller than the duct itself, and that a stone too large to pass readily out may ob- struct the outflow of bile and yet change its position in the common duct, permitting an intermittent biliary outflow, thus causing temporary attacks of jaundice and coHc which are short in their duration. There may be a number of colicky attacks with a varying degree of jaundice, which nearly or quite clears up, only to return in a more marked degree. Eventually, the obstruction, through irritation and inflammatory swelling caused by the in- fective cholangitis which is prone to occur, brings on complete cholemia. The bile in the blood rapidly reduces the red corpuscles to below 3,000,000, or even to 1,500,000, with death from exhaus- tion, or even more rapidly from a suppurative cholangitis. These cases are frecjuently marked by fever, chills, and sweat- ing, at times of almost daily occurrence. The Germans have taught us that this is essentially septic in its nature, but the French have proved beyond question that it is due to the irritation of the stone and the absorption of bile without regard to an infective proc- ess. Under such circumstances the common duct is often of huge dimensions. It has even been opened under the impression that 340 WILLIAM J. MAYO it was the distended gall-bladder; its contents are, however, always bilious. The gall-bladder, as a rule, is small, and, while the liver is at first enlarged, it does not long remain so. The jaundice, with its accompanying itching, hypochondriasis, slow pulse, and white stools, with the light colics, the fever, chills, sweats, and lack of tumor, aid in the distinction of stone in the common duct. Ma- laria would not give rise to the persistent jaundice, while an en- larged spleen and the presence of the Plasmodium malarise would aid in the diagnosis. From cancerous obstruction due to tumor of the liver, pancreas, or duodenum the diagnosis may be difficult, particularly in the later stages. It should be borne in mind that, as a rule, the pres- ence of a tumor with jaundice is against simple stone in the com- mon duct; while, as pointed out by Mayo Robson, a history of slight colics and intermittent jaundice in the early stages, wdth fever, chills, sweats, and without tumor, favors simple stone ob- struction. The mistake of diagnosticating a tumor in case a tongue of liver projects downward in this region, as described by Riedel as a not uncommon phenomenon, should not be made. I have oper- ated twice on these indications, and have found them a reliable means of diagnosis. Fagge, in his great work on "Practice," long ago pointed out that cancer of the liver and biliary ducts was often complicated wdth gall-stones, and correctly interpreted their presence as the exciting cause of the cancer, rather than an acci- dental compKcation. Musser estimates that 90 per cent, of persons with cancer affecting the liver also have gall-stones, and the danger of the calculi acting as an irritative in the production of cancer should be borne in mind. Twice within a year I have operated in cases of cancer complicated \\dth gall-stones; in each, jaundice with an appreciable tumor gave indications of the fatal complication, while a marked history of gall-stone disease ante- dated the final illness by many years. The prognosis then depends much upon the situation of the stones. Stones lying free in the gall-bladder may never give a diagnostic symptom of their presence, although septic processes (;all-stone diskase .'J41 in the ^'all-hladdcr diu' to an inft'flion or retention may present, urgent syniptcjins for relief. Stones in the cystic duet, l)y prcKlue- ing changes, in the gall-bhidder, exhausting colics, or ulceration, require relief. Stones in the common duct are urgent and demand prompt removal. The chances of stones jiassing into the intestine have been greatly overestimated, because of the prevalent idea that a colic means the j)assage of a stone, which is seldom the case, and even if it did pass, it must be a small one, and the history of this class of patients is that in the small stone cases there are many stones, and there is no guarantee against future trouble. The idea of the frequency of the expulsion of gall-stones has also been due to washing the stools and finding small bodies supposed to be calculi, or the giving of some bland oil, such as olive, which forms soap- balls with the intestinal alkalis, and these are many times ex- hibited as the offending bodies. The possibility of secondary cancer should also be taken into account in the prognosis. Treatment of Gall-stone Disease. — It can be asserted without fear of contradiction that there is no medical treatment for gall- stones: it is wholly surgical. We cannot dissolve these calculi in the living body, although much can be done for the relief of the disease in a non-operative way when the stones are confined to the gall-bladder. The internal administration of remedies which increase the rapidity of the biliary flow has a deserved reputation in preventing the formation of stones, phosphate of sodium, or a course at the Carlsbad or other saline springs, being most popular. As the passage of food through the duodenum is particularly apt to bring on a colic, the use of remedies calculated to allay any existing duodenal catarrh may be of aid, and the avoidance of such articles of diet which the patient soon learns to know as most liable to start a colic. Glycerin has of late been lauded, as well as many preparations of enterprising drug firms. Such allegations are supported by uncertain evidence, derived from the fact that intervals of quiescence are frequent, or based wholly upon the imagination of the owner of the drug. Olive oil a I one time had a great reputation, the examination of the stools 342 WILLIAM J. MAYO after its administration showing quantities of soap-balls formed by the action of the intestinal alkalis and shaped by vermicular action. For the rehef of pain caused by a temporary blocking of the cystic-duct opening hypodermic injections of morphin, or even chloroform, may be required. Surgical Treatment. — While, as a matter of course, even the most enthusiastic surgeon would not advise operation in every case of gall-stone disease, it can be said that, other things being equal, every case of gall-stone disease causing marked symptoms should be relieved by removal of the calcuh, and if obstruction of the cystic duct has taken place, an operation is imperative, while if the stone is in the common duct, delay in operating would be criminal. The situation of the gall-bladder, as shown by Hamil- ton, is very definite in the male, but has a somewhat larger range of position in child-bearing women, for obvious reasons. The in- cision for gall-bladder operations is a matter of some importance. Lawson Tait, the leader in this as in other branches of abdominal surgery, prefers a vertical incision downward from the tip of the cartilage of the tenth rib. Musser and Keen make use of an in- cision skirting the margin of the costal cartilage. In a recent ar- ticle on Hernia in the "Annals of Surgery," Greig Smith calls attention to the advantages of an incision in the course of the ex- ternal oblique muscle, on a line running from the tip of the tenth cartilage toward the umbilicus, with separation of the fibers of the internal oblique, as giving a better closure of the wound, the essen- tial principle being the same as McBurney's incision for chronic appendicitis. Having observed a well-marked hernia following the vertical cut, I practised this incision in two instances to ad- vantage. For work on the ducts, an additional incision along the costal margin enables an ample flap to be turned downward and inward, giving a large amount of working space. Cholecystotomy is indicated for the removal of stones from the gall-bladder, and is an important part of the method of elevating stones from the cystic duct, not only as an aid in the removal of the impacted stones, but also as there may be stones behind the obstruction. Needling the gall-bladder for diagnosis, as originally done by Har- (JALL-STONE DISEASE .'US ley, is useless and (Itm^croiis, ciLher before or after tlie abdomen is opened. Even with the finger in the abdomen stones cannot al- ways be felt, through the gall-bladder wall, and it is aI\va3^s neces- sary to open this viscus before their absence is decided Uf)on. Some years ago, before modern methods of protecting the peri- toneal cavity temporarily with gauze were perfected, many sur- geons fastened the gall-bladder in the abdominal wound and waited several days for adhesions to shut the free peritoneal cavity off. This has the great disadvantage of not being able to use the finger in the abdominal cavity outside the gall-bladder and ducts as an aid in coaxing the calculus from its bed, particularly when im- pacted. It is probably the rule that if the contents of the gall- bladder are actively infectious, it is already adherent, and under such circumstances, if not so adherent, it would be wise to make a secondary opening. In two cases in which the gall-bladder was filled with the products of an infective inflammation as well as stones, I found it already adherent, which made the operation largely extraperitoneal, by reason of the adliesions to the omentum, colon, and parietal peritoneum. In his original operation Bobbs closed the wound in the gall-bladder with sutures and dropped it back, and this so-called "ideal" operation has had some adherents, but the objections to it are weighty. The following points are urged against suture of the gall- bladder in the wound: 1. The chances of a permanent fistula; but if the ducts are open, this should not take place. In one of my early operations prolonged biliary leakage gave some annoyance, and vigorous cauterization was necessary for cicatrization. This was due to a too accurate suture of the mucous membrane to the skin, leaving practically a mucocutaneous mouth. In all my later operations I have used the peritoneal and muscular coats of the gall-bladder, and carefully avoided the mucous membrane in the suture, thus getting a larger cicatrizing area between it and the skin, with the result that the fistulas have rapidly closed. 2. The danger of hernia by the interposition of the gall-bladder between the margins of the incision. In 'rZo of the l28 cases of ini- 344 WILLIAIVI J. aiAYO mediate suture reported by Elliot gauze drainage was employed, and this would have a parallel effect. In favor of drainage of the gall-bladder is the important fact that not infrequently stones will be found extruded into the dressings, although all were supposed to have been removed. In one of my cases, after removing about 200 small stones, during the next week over 50 more were discharged: in another case 2 stones were found in the dressings at the end of the week. Drainage also cures the catarrhal condition of the gall-bladder, as when sutured into the wound it usually shrinks up into a fibrous cord, or at least by forcible elevation of its normally dependent fundus furnishes free drainage and prevents any retention of fluids or new stone formation. At times it may be impossible to bring a contracted gall- bladder to the surface, but experience has shown that it can be safely drained through the incision, using the omentum, if possible, in forming a channel. Morison drains through the loin near the hepatic flexure of the colon, and maintains that fluids naturally gravitate to this point. I can imagine a case of gall-stone in which the certainty of complete removal and the healthy condition of the walls of the gall-bladder might render suture and dropping back good surgery, but, as yet, I have not met with such a case. The removal of impacted stones from the cystic duct is a matter of great difiiculty, and may require incision of the duct. In 2 out of 3 such cases I have been able to push the stone back into the bladder. In the third I incised the duct and removed the stone, but, being unable to suture the empty duct with any degree of accuracy, drained with a rubber drain and suture of iodoform gauze to the proper place by means of fine catgut which will hold long enough to cause adhesions at the desired point, and yet be absorbed sufficiently early to allow removal of the gauze. The value of fine catgut for the accurate placing of deep gauze tampons cannot be overestimated. The removal of stones from the common duct is often an operation of the greatest difficulty, and in the case of a floating stone may be impossible. GALL-STONE DISEASE 34.5 Fcngcr, in two oases, introduced his finder into tlie diluted (-(jni- inon duet, hnt was uiiahle to reach a floating stone which retreated upward into the hepatic ducts. If the stone can be found, the duet should be incised over it, and before removal the inij)ortant suggestion of Elliot should be carried out — that is, the placing of sutures after incision while the stone is yet in position. It is exceedingly difficult to i)lace the sutures after removal of the stone. In one of two eases of stone in the common duct I could not find the stone, and performed choleeystenterostomy with the Murphy button. The patient, an old man, completely jaundiced, is now alive and well, two years after the operation. The opera- tion of choleeystenterostomy in obstruction of the common duct by means of Murphy's mechanical device is of inestimable value, and in the feeble cholemic condition which this form of obstruc- tion fiuickly produces in the victim is a life-saving procedure. The Germans still do Winiwarter's suture method of chole- eystenterostomy. The use of this method for routine treatment of all forms of gall-stone is not logical, and with few exceptions the union of the gall-bladder and intestine should be confined to complete obstruction of the common duct. For simple stones without their removal, and for obstruction in the cystic duct, it cannot be considered good surgery. Cholecj^stectomy was ad- vocated by Langenbeck, but is now seldom done as a matter of choice. At times a small, friable gall-bladder which cannot be brought to the surface is tied off, and the region thoroughly drained to meet a pressing indication. Spasmodic attempts to popularize removal of the gall-bladder have been made at various times, but as a routine procedure it is uncertain. I have performed 16 operations for gall-stones — 8 for simple removal of unimpacted stones, 3 for stones in an empyemic or cystic gall-bladder, 3 for stones impacted in the cystic duct, and 2 for stones obstructing the common duct — with 1 death, in a case of suppurating gall-bladder, with perforation and septic periton- itis already present. The patient died within a few hours follow- ing operation. A CASE OF GUNSHOT WOUND OF THE LIVER. IMMEDIATE OPERATION. RECOVERY* WILLIAM J. MAYO H. C, a lad of fourteen years, while hunting October 3, 1896, was shot accidentally by a companion with a rifle. The rifle was only a foot away, and the bullet entered the body one inch below the ensiform cartilage and three-fourth inch to the right of the median line. There was no wound of exit. The accident occurred some distance from the city at about 12 M., and the patient was brought to the hospital at 3 p. m. in a condition of shock. Temperature was subnormal; pulse, 118, and very feeble. He complained of some pain in the abdomen, which was contracted and rigid. He was somewhat nauseated, and had vomited once after the accident, but without evidence of blood. Measures were taken to relieve the shock, and the patient was at once prepared for operation. Ether was administered at 4 p. m., just four hours after the injury. An incision was made from the tip of the ensiform cartilage to the umbilicus, opening the abdominal cavity freely. A lateral cut was then made to the right, so as to include the wound of en- trance. The bullet had cut half of its diameter into the edge of the right costal arch, and entered the liver just at the inner edge of the right suspensory ligament. On the surface of the liver was found a bit of his flannel shirt, which had evidently been caught against the costal arch and cut out like a pouch. From the hole in the liver dark blood was welling up, and the abdomen was filled with blood. The liver, as * Reprinted from "New York Med. Jour.," March 20, 1897. 346 (ASK OF (U'NSIIOT WULNU Oi' Ll\ KK li47 is usiKil in youiij; adults, was very large, extendiiifr downward nearly to the uinhilicus. A i)rol)e passed directly backward six inches was arrested by the posterior abdominal ^\all. With a finj^er introduced under the gastrohejjatic ligament through Winslow's foramen, search was made in the lesser cavity of the peritoneum for the probe which had been left in position, but it could not be felt. A i)iecc of iodoform gauze was i)assed along the finger under the lesser omentum and brought out at the lower angle of the incision to drain this space. With a finger the wound in the liver was rapidly searched for foreign bodies, and then packed deeply with iodoform gauze, bringing it to the sur- face in the track of the bullet. The bleeding was checked at once. The abdomen was not irrigated, the effused blood being left for absorption. The abdominal incision was closed, except at the points of gauze drainage. After vigorous stimulation the patient rallied. The drains were removed in six days, and he left the hospital fully recovered in twenty days. Eight weeks after the injury an .r-ray photograph was taken and the bullet located in the muscles of the back. It gave the boy no trouble and was not interfered with. SOME OBSERVATIONS ON THE SURGERY OF THE GALL-BLADDER AND THE BILE-DUCTS* WILLIAM J. MAYO During the past eight years 105 operations have been made on the gall-bladder and the bile-ducts in St. Mary's Hospital . For gall-stones in the gall-bladder or cystic duct, or both, cholecystotomy was performed 64 times, with 1 death; 4 of these cases required separate incision of the cystic duct for the removal of an impacted stone, and in 3 cases soft stones in the duct were crushed. Four of the 64 cases required secondary operations: 2 for the removal of gall-stones overlooked at the primary operation; one for the relief of biliary fistula, probably due to putting the gall- bladder on the stretch by too low an attachment to the external incision. The fourth had a cystic accumulation, caused by a stricture in the cystic duct. In addition, 3 cases had more or less trouble after the healing of the fistula, due to inefficient drainage through the cystic duct, and in two of these the wound reopened several times, with discharge of retained secretion. In 8 cases cholecystotomy was made for the relief of infection of the gall-bladder or the ducts, or both, wdth one death fourteen days after an operation for suppurative cholangitis. Two cases of infective cholangitis were operated upon; both recovered. In these 2 cases the gall-bladder was shrunken and contained no stones, although there probably had been a previous cholelithiasis. In the remaining 5 cases the suppurative process *Reprinted from "Annals of Surgery," October, 1899. 348 suitGEUY or c;all-ijlai)iji:u and uilk-ducts 3U) was confined to the ^all-hladdcr, and in 4 stones were present. In 1 a secondary operation has since been performed for retention cyst, and in another there is a mucous fistula present at this time. It will be noted that, in the total of 72 cases of cholecystotomy there were 2 deaths, and 7 cases in which the gall-bladder continued to give trouble from causes not connected with the stone-formation. In all these cases of failure to cure by cholecystotomy there has been an obstruction to drainage through the cystic duct. The ol)stacle is usually a stricture due to an ulceration in the duct from prolonged lodgment of a stone or, as pointed out by Fenger, to angulation of the duct. The mucous membrane continues to se- crete, and colicky pains attend the forced passage of the secretions. Harris has pointed out that the small gall-bladder firmly con- tracted on stones is especially liable to subsequent attack of re- current regional peritonitis, and this has been our experience. After the removal of the external drainage the thickened walls of the gall-bladder continue to contract, interfering with the drainage through the ducts from the islands of mucous membrane not pre- viously destroyed, and a condition results resembling a chronic appendicitis in many respects. In such cases, and in all cases in which a stricture is already present in the cystic duct, cholecys- tectomy should be performed. In the 7 cases in which further trouble was experienced chole- cystectomy, or removal of the mucous membrane of the gall- bladder, which amounts to the same thing, would have resulted in cure. Cholecystectomy was made 4 times — all the patients recovered: once as a primary operation in a case of acute gangrenous in- flammation of the gall-bladder, once for biliary fistula, and twice for recurrent cholecystitis. Choledochotomy was performed for stones in the common duct 11 times, with recovery in each instance. In 8 of these cases the gall-bladder was shrunken, and in only 5 did it also contain stones. One was a typical example of the ball-valve stone de- scribed by Fenger; all these cases have remained well, the gall- bladder giving no further trouble. The fact that the cystic duct 350 WILLIAM J. MAYO was able to pass the stone would indicate that drainage was free. As previously remarked, such a gall-bladder firmly contracted upon the stone would argue to the contrary, and clinical experience dem- onstrates the correctness of this view. Cholecystenterostomy was made once with the Murphy button for a stricture of the common duct which had caused jaundice for eight months in a man seventy-one years of age. The patient re- covered, and remained well for more than six years, eventually dying of other causes. In 3 cases a gall-bladder, fastened by adhesions to the adjacent viscera, gave rise to pain and gastric symptoms. In these cases the history of colics in previous years warranted the belief that stones had once been present. Liberation of the adhesions relieved the symptoms. In 7 cases an exploratory incision was negative, no gall-stones being present. In 2 of these a diseased appendix was found to be the source of trouble. Dietl's crisis from a movable kidney proved to be the real difficulty in another, and in 4 no cause for the symp- toms could be discovered. This gives 98 operations and explora- tions upon the gall-bladder and ducts for non-malignant disease, with a mortality of 2. For malignant disease involving the bile-tract 7 operations were performed, 4 cholecystotomies, 1 cholecystenterostomy, and 2 exploratory incisions, with 3 deaths. In 5 of these cases jaundice was present and the gall-bladder was distended with clear mucus. In 3 a small quantity of bile-stained serum escaped on opening the peritoneal cavity. The results merely substantiate the well-known fact that, in the great majority of cases of malignant disease involving the bil- iary apparatus, operation is contraindicated. The straight incision of Tait, either through the rectus muscle or at its outer margin, was employed until two years ago, since which time the Bevan incision has been made use of in the majority of cases. The latter gives a larger space for work and is easy of closure. The so-called "ideal" operation of cholecystotomy with immediate suture has not been employed. Theoretically, it does SURGERY OF GALL-KLADDKR AM) niLP>DUCT.S .'{.> 1 not appear to be a scientific procedure, as it does not furnish drain- age to the accompanying cholecystitis, and there is nothing to pre- vent the formation of stones in the future. The normal position of the gall-bladder makes the fundus de- pendent. Given an infection with a sluggish bile current and the deposit of sediment, and formation of stone is mechanically easy. It would seem that the cases of typical colics in which stones are not found migiit be due to the occasional emptying of the gall- bladder by muscular contraction, causing great pain as the sedi- ment or nmcus passes. When the fundus is sutured into the ex- ternal wound, the necessary elevation brings the cystic duct more nearly at the bottom, so that the mechanical conditions in pre- venting sedimentation and stone-formation are even better than in the normal individual. Our observation has been that the ca.ses of cholecystotomy in which all the stones were removed and the ducts free have had no further trouble, and this without any changes in the habits of the individual, which might contribute to the result. The fixation and elevation of the fundus permanently drain the gall-bladder through the cystic duct, which has become the most dependent part, and gravity alone is effectual. The gall-bladder was opened and the stones removed before suturing it into the incision, with the exception of two cases of suppurative cholecystitis. The after-treatment was prolonged in these two cases, and the writer believes, notwithstanding the suc- cess of Riedel with this method, that with proper protection even septic cases should be opened and carefullj'^ examined at the primary operation. The cystic obstruction frequently depends upon a stone, which can be removed much better at this time than later. If the gall-bladder is adherent to the omentum, colon, or other viscera, the adhesions are freed a sufficient space on the fundus for manipulation, and on the inner side to permit careful exploration of the ducts and to facilitate the removal of stones. No attempt has been made to free the adhesions at every point, as this would open up further avenues of infection and destroy temporary barriers of adhesions. If the cause of the cholecystitis is removed, the adhesions usually disappear in time, 352 WILLIAM J. AL\YO and if separated, reform temporarily. Should the history indicate that these adhesions have been causing trouble, it would, of course, be best to di^^de them and cover the raw surface by suturing to prevent reformation. It is a frequent experience to find a contracted gall-bladder deep under the hver, which it is impossible to suture into the wound. In such cases we have found a very satisfactory method of drainage to be as follows : Two or more long sutures of fine catgut are passed through the walls of the gall-bladder below the opening. A well of gauze strips 2 inches in width is now formed by passing one or more doubled thicknesses down just outside of the gall-bladder, and tying in place with the catgut sutures. The side of a rubber drain is caught at a Httle distance from its extremity by one of the threads of cat- gut on a needle, and the tube is then passed into the gall-bladder and tied in place. This firmly anchors the drains in position and prevents displacement; the external incision is partly closed up to the drainage in the usual manner. The catgut is absorbed before it is necessary to remove the drains. We have not hesitated to crush soft stones in the cystic duct with the fingers; but if the deposits were hard, the duct was opened by incision and the stones removed. The success of Mayo Robson in crushing stones impacted in the bile-ducts is encourag- ing, and probably could have been done with advantage in some of the cases in which cutting was resorted to. The time of drainage of the gall-bladder has varied with the case, if one is certain that all stones have been removed, and the walls are not greatly thickened nor adherent, and bile flows freely at once, three or four days is sufficient. If there have been many small stones or much fragmentation of soft ones, a quantity of bile-sand, or evidence of prolonged in- flammation, one or more weeks, is necessary. The suggestion of Knut Hoegh that suturing at a low point in the incision favors fistula formation is probably true, as it does not properly elevate the fundus and should be avoided. Of the 11 cases in which stones were removed from the common SURGKRY OF CALL-RLADDrOR AND RIIJC-DUCTS 3.53 (liict, in only 4 was an attcnij)t made to .suture the duct; in 2 of these there was no leakage. Seven cases were treated by drainage without suture, and the recovery was just as prompt as in the sutured cases. The stone is isolated, and the duct held between the finger and tluimh. By pressing the stone firmly against the duct-wall and tiicn relaxing it, veins can sometimes be located — as they fill — and avoided. A longitudinal incision is made, and the stone caused to present. Following the suggestion of Elliot, the stone is used like a stocking ball for the placing of a lateral suture of fine catgut each side of the opening; it is then removed and search made for others, which are removed, if found. A strip of gauze two inches wide is carried down each side of the opening and tied in jilace by the catgut. The edges of the incised duct are approximated with tissue forceps, and covered with several thicknesses of a third layer of gauze, and a rubber tube passed into the gauze well and fastened by one of the catgut threads. The gall-bladder is opened and drained in the usual manner, to pre- vent tension. Cholecystectomy should be more frequently employed. Hans Kehr is emphatic on this point. It requires a larger incision and more maniinilation, and, as performed, is a far more serious opera- tion than cholecystotomy. It is often necessary, as a secondary operation after the drainage of an infected gall-bladder, or after failure of cholecystotomy to cure from any cause. Considering that the mucous membrane is the only part of the gall-bladder which gives rise to after-trouble, in three cases during the past year we have opened the gall-bladder and removed the mucous membrane to the cystic duct. At this point the mucosa is cut across part at a time, and one or two bleeding points caught with forceps and tied, the duct being left open. The muscular and peritoneal coats are sutured into the incision and drainage established in the usual manner. It is surprising how easily this can be accomplished. The mucous membrane is tough and separates from the muscular coats readily. The adhesions to the peritoneal coat are separated only enough to explore the ducts. It does not require a long incision, nor does it necessitate VOL. I — iS 354 WILLIAM J. MAYO prolonged manipulation, and yet the essential part has been re- moved to the same extent as in the usual cholecystectomy. This modified operation adds little to the risks of an ordinary cholecystotomy. [Note. — Since writing this paper the author has, in two addi- tional cases, performed the modified cholecystectomy as detailed above. In one of these cases a stone impacted in the cystic duct was removed with the mucous membrane of the gall-bladder and duct to a point beyond the stone with perfect ease.] CHOLECYSTECTOMY, WITH ESPECIAL REF- ERENCE TO THE REMOVAL OF THE MUCOUS MEMBRANE OF THE GALL- BLADDER AS A SUBSTITUTE REPORT OF A CASE IN WHICH THE GALL-BLADDER WAS REMOVED FOR MALIGNANT DISEASE* WILLIAM J. MAYO Excision of the gall-bladder is clearly indicated in four groups of cases: (1) For traumatisms, such as gun-shot wounds or crushing injuries. (2) Phlegmonous cholecystitis and gangrene of the gall- bladder. (3) For malignant disease. In these three groups all the coats of the gall-bladder are involved in the diseased process, or injury and complete cholecystectomy is a logical sequence. (4) For the relief of permanent obstruction of the cystic duct, the common duct being patent. In this last group only the mucous membrane is at fault; if this were not present, there would be nothing to drain, and obliteration of the cystic duct would be harmless. It is this class of cases in which removal of the mucous membrane of the gall-bladder offers a quick and safe method of relief. In the complete operation of cholecystectomy the close relation which exists between the gall-bladder and the liver, and the deep situation of the pedicle at the cystic duct, introduce certain elements of danger which cannot be ignored, and which render the operation in every way more serious than simple re- moval of the mucous membrane. The latter procedure adds but little to the risks of an ordinary cholecystotomy. * Presented to the Section on Surgery and Anatomy, at the Fifty-first Annual Meeting of the American Medical Association, held at Atlantic City, X. J., June 5-8, 1900. Reprinted from "Jour. Amer. Med. Assoc," December 1, 1900. 855 356 WILLIAM J. MAYO Traumatisms to the gall-bladder requiring its ablation are but rarely met with, and are usually associated with grave injuries to the liver. A small number of cases have been reported, which not infrequently were the result of indirect violence. In the one case of this kind which I treated free drainage enabled repair of the ruptured gall-bladder to take place without excision. In acute phlegmonous cholecystitis and gangrene of the gall-bladder two courses are open: (1) Freely to drain the gall-bladder and pack the surrounding space with gauze, or (2) to remove the offending organ at once. The principles are essentially the same as in the treat- ment of appendicular abscesses by drainage, leaving the appendix, or to remove the appendix, if possible, at the primary operation. In three instances of this character the writer has excised the gall-bladder and drained freely, the bleeding surface of the attached liver and the infected stump being covered with gauze held firmly in position by sutures of fine catgut. The stitches hold the gauze in place until adhesions form, and the catgut is absorbed before re- moval of the drainage is indicated. In this manner sufficient pres- sure can be obtained to check the oozing and limit extravasation. The gall-bladder, with adjacent liver substance, has been re- moved on a number of occasions for malignant disease by cutting instruments, the Paquelin cautery, or by the elastic ligature. A case of like nature was operated upon recently by the author, and on account of its rarity is reported somewhat in detail as follows : Mrs. E. R., American, aged sixty-five, was admitted to St. Mary's Hospital April 18, 1900. History. — She has been in her usual health until within the past six months. During this time she has suffered from a boring pain in the right side, which of late has become almost constant. Gastric symptoms have been of moderate severity. There have been some loss of appetite and constipation, with a decrease of 15 pounds in weight. No jaundice nor history of colics. Examina- tion reveals a somewhat movable tumor in the right hypochondriac region, evidently connected with the liver. The mass has a nodu- lar feel. Exploratory incision April 21, 1900. A carcinomatous gall- CIIOLECYSTKf TO.MV 357 bladder involved the adjacent portion of the liver and the cystic- duct. There was some infiltration along the common duct, aiul extending to- the duodenum at one place was a considcrahic arc;i of adhesions. A few glands in the angle between the cystic and hepatic ducts were infected. The disease was so definitely cir- cumscribed with such slight glandular involvement that its re- moval was decided upon. The excision began at the common duc-t, two inches of which was removed with one inch of the hcjjatic duct. The vessels were caught and tied as divided. An area of adherent duodenum the size of a silver dollar was included in the excision. The opening in the intestine was closed by circular purse-string sutures, and the lower end being thus freed, the gall- bladder with the attached liver was removed with the Paquelin cautery knife. The larger vessels were grasped with forceps. The free venous oozing from the liver substance was not controlled by the cautery, although easily checked by slight pressure, the blood- current having but little force. A piece of sterile gauze the size of the wrist was placed in the cavity, and a continuous suture of fine catgut was run through the liver substance on each side of and around the gauze, effectively compressing the bleeding liver margins against it, and controlling the hemorrhage. The portal vein was exposed to a considerable extent in the bottom of the cavity. Adequate drainage was afforded, the bile being conducted to the surface. Recovery was uneventful. The gall-bladder contained a single stone, % inch in diameter. The fourth class of cases in which a permanent obstruction exists in the cystic duct are far more numerous. The obstruction may be the result of adhesive inflammation in the outer coats, causing angulation or the long lodgment of a stone in the cystic duct, with resulting ulceration and cicatrization, or stricture from any cause. Of 132 operations on the gall-bladder and bile-ducts which have been made in St. Mary's Hospital during the past nine years, 11 were cholecystectomies and 7 of these were for the relief of ob- struction in the cystic duct, which had caused mucous fistulas or recurrent attacks of colic, due to retention of the secretions in the gall-bladder. The indication in these cases is clear. It is the continuous secretion from the mucous membrane prevented by the olistruction from draining through the natural channel, which 358 WILLIAM J. MAYO causes the trouble; the peritoneal and muscular coats are harmless, and by removing the mucous membrane down to the obstruction, relief is afforded. In my own experience, obstruction of the cystic duct is met with either primarily — cystic gall-bladder — or occurs secondarily after operation for gall-stone disease in about 10 per cent, of cases. It seems unnecessary to say anything about technic. The mucous membrane of the gall-bladder is easily detached, and as aU the adhesions are to the peritoneal and muscular coats, the separation is readily effected. The gall-bladder partly inverts it- self as the cystic duct is approached, rendering easy removal of an impacted stone. If it is small and deeply placed, removal of the mucous membrane is more difficult, but can be accomplished more readily than complete extirpation. One or two small vessels re- quire ligation. The muscular and peritoneal coats are sutured to the upper angle of the wound in the abdominal wall, and drainage established as in ordinary cholecystotomy. As a secondary opera- tion, removal of the mucous membrane is most serviceable. Drain- age has failed to cure, and the adhesions formed by the previous union of gall-bladder to the external incision vastly increases the difficulty of complete extirpation. The operation in these cases is best accomphshed as follows: An incision is made into the abdominal cavity on the inner — median — side of the site of the former operation, but these external attachments of the gall-bladder are not severed. The adhesions are separated to a limited extent on the inner side, to enable careful exploration of the ducts. The adhesions in other directions are purposely left, and act as a protection to the outer and lower por- tion of the operative field. After proper gauze protection the gall- bladder is opened on the inner side in the explored area, about 13^ inches down, and this incision is carried outward toward the ex- ternal attachments. The separation of the mucous membrane is begun at the middle, and the enucleation carried down to the cystic duct, where it is di\aded at the point of obstruction. The separa- tion is then proceeded with from T\'ithin outward until completed. The scar tissue at the place where the gall-bladder is attached to the abdominal wall renders detachment difficult if commenced at f'HOLECYSTKC T(JM Y 359 that point, but by Ix'^inning well })(>l()\v, tlio muroiis mfinbrane can be readily sejjarated. The muscular and peritoneal coats are drained in the usual manner, a {)iece of sterile gauze being tacked about the inner divided wall by a few catgut sutures, which renders the drainage quite i)erfect. THE SURGICAL SIGNIFICANCE OF JAUNDICE* WILLIAM J. MAYO Jaundice as a result of diseased states of the blood has been eliminated as irrelevant to this discussion, and reference will be made only to those forms of icterus due to an impediment to the outflow of bUe through the liver-ducts. Obstruction to the in- testinal dehvery of the bile is caused, first, by swelKng of the mu- cous lining of the ducts, due to an infection; second, by gall-stones in the common duct; third, by disease or tumors involving the ducts; fourth, by disease or tumor of the head of the pancreas. Pressure from without the ducts, as in certain mahgnant growths of the pylorus, may also interfere directly with the biliary flow; and, again, some forms of cirrhosis, either of the atrophic or of the hypertrophic variety, may obstruct the smaller liver-ducts, cutting off from the main channel the secretion of the part, and result in jaundice. In one way or the other the common forms of jaundice are due to mechanical interference, and by careful at- tention to the history, with the physical examination, a differential diagnosis may be made; if not pathologic, at least a surgical, diagnosis may be made; that is to say, a condition which can be relieved surgically, although it may be any one of several forms of disease. Gall-stones in the gall-bladder do not produce jailndice except under rare circumstances. The rule is that jaundice with gall- stones means a stone in the common duct. The exceptions are, first, an infection of the gall-bladder which travels along the ducts, producing a cholangitis, usually accompanied by some inflamma- tory reaction, slight temperature, a quickened pulse, and follows an attack of colic, with or without the passage of a calculus; second, * Reprinted from "Northwestern Lancet," January 1, 1902. 360 TIIK srH<;i(AL SHiMl'ICANCK OF JAINDICIO .'}(> I an inflamed gall-bladder (cholecystitis), in which infection lias passed throu — soinelimes very little. 362 WILLIAM J, MAYO We have several times removed stones from the common duct, or a diverticulum of it, in which jaundice was not present. Jaundice from gall-stones should give a history of previous colics. The physical examination will usually be. negative. Con- trary to the prevalent opinion, the gall-bladder is shrunken in- stead of distended, in the majority of cases. Occasionally it happens that the gall-bladder will be found distended. In such cases it will probably also contain stones. The reason for the shrunken condition of the gall-bladder lies in the great inflamma- tory thickening of its coats and the contraction of this connective tissue after the stone is expelled. Jaundice from malignant disease may have a history of gall- stone attacks years before, and the cancer is often due to the irrita- tion of the calculi. The jaundice comes on after some indefinite history of abdominal disease, and is preceded by loss of flesh and progressive weakness, without much acute pain. The jaundice slowly but steadily increases, and does not change day by day in intensity, excepting to get deeper. On palpation the distended gall-bladder can often be felt. This enlargement is usually the rounded end of the gall-bladder, occasionally nodular, due to an extension of the malignant process. In the latter stages a little ascitic fluid can be detected in the abdomen. The age of the patient over thirty -flve is to be noted. Chronic inflammation of the pancreas is accompanied by jaun- dice in many cases, and is often confounded with malignant disease. The previous history may be that of gall-stones or obscure epi- gastric distress. ' The jaundice does not change in the early stages, as is to be expected in a gall-stone in the common duct, and there are fewer of the repeated colicky pains. It also differs from malig- nant disease in the fair degree of health maintained for a length of time beyond the expectancy of cancer. The gall-bladder can be felt distended, but without nodules, and the patient may be of any age. The hypertrophic cirrhosis of the liver of Hanot is a disease of young adults, and the jaundice is slight, but persistent. The great size of the liver makes the diagnosis. Atrophic cirrhosis and allied forms of hypertrophic cirrhosis are THE SLH(;KAL significance of JAtSDICE 363 usually found in adults. Tlio jaundice, if i>n'S(>nt, is a late symp- tom, and ascites is frequent. Here, again, physical examination of the liver and spleen, with the alcoholic history, comi)letes the tliagnosis. To recapitulate: First, jaundice is not to he expected in un- comi)licate(l gall-stone disease. Second, jaundice from a stone obstructing the common duct gives the history of previous attacks of gall-stone colic, varies greatly in intensity in the early stage, and these changes are ac- companied by colicky pains. An early history can usually be obtained of some little fever, and at times chilly .sensations or sweats. The gall-bladder cannot often be palpated. Third, in malignant di.sease the loss of flesh before jaundice, the age of the patient, and the unchanging (except for the worse) of the icteric hue, with a distended and, perhaps, nodular gall-bladder or adjacent tumor, completes the clinical picture. Fourth, jaundice from chronic pancreatitis will probably be confused with malignant disease. The po.ssibility of its existence leads to an examination as to the history, the age of the patient, and the duration of the symptoms. Fifth, the various forms of cirrhosis accompanied by jaundice are to be distinguished by the physical examination of the liver and, occasionally, the size of the spleen. Catarrhal jaundice, with which every practitioner is familiar, has not been mentioned. It is a disease most common in young adults, and is usually due to an extension of a mild infection from the gastro-intestinal tract. It is sometimes seen in an epidemic form. The age of the patient, the slow pulse, and the lack of general symptoms and short duration make the differentiation easy. Catarrhal jaundice may, in isolated cases, complicate appendi- citis, pneumonia, cancer of the stomach, and a host of common diseases; but as in these cases it is usually incomplete and tran- sient, it does not long mask the primary source of disease. How- ever, these secondary conditions are at times most serious, the catarrhal infection developing a purulent cholangitis, and death resulting from multiple abscess of the liver, with the usual symp- toms of sepsis, accompanied by a marked jaundice. CANCER OF THE COMMON BILE-DUCT. RE- PORT OF A CASE OF CARCINOMA OF THE DUODENAL END OF THE COMMON DUCT WITH SUC- CESSFUL EXCISION* WILLIAM J. MAYO Primarj^ carcinoma of the common duct is rare. In 4578 autopsies Kelynack found 8 cases of primary cancer of the gall- bladder, and but 2 having origin in the duct. Musser collected 100 cases of carcinoma of the gall-bladder and 18 of the bile-ducts. The site of the neoplasm in the common duct is usually either at the juncture of the hepatic and cystic ducts or near its duodenal termination. The 18 cases collected by Musser showed 3 in the hepatic duct, 1 at the juncture of the cystic and hepatic duct, and 14 in the common duct, and of these latter, 9 were at or near the papilla. In 17 cases of cancer of the ducts Rolleston found 15 in the common duct, and of these 10 at or near the papilla. As to the etiology of carcinoma of the common duct, there is some question. It is conceded that gall-stones are the most com- mon cause of cancer of the gall-bladder. In Musser 's 100 cases of gall-bladder carcinoma 69 contained gall-stones, and he found evi- dence that calculi had at one time been present in the majority of the remainder. Primary cancer of the gall-bladder and gall-stone disease are more common in females than in males, and in about the same proportion. In 22 collected cases of cancer at or near the papilla of the common duct Edes found gall-stones in but 4, and 3 of these in the gall-bladder. In 36 cases of cancer of the common duct Rolleston found gall-stones in less than half. Cancer of * Reprinted from "The St. Paul Medical Journal," June, 1901. 364 CANCKU or Tin; (om.mo.n iiiij;-i)i (T 3fi/> the common diicl is ('(jiially frccjiicnl in the male and female, which does not favor the belief that gall-stones are the cause of the ma- lignant process in the duct, although calculi an- und(>ul)l((lly an important ctiologic factor in cancer of the gall-bladder. The ex- tensive exi)erience of Mayo Robson, however, entitles his opinion to great weight, and he states his belief that gall-stones arc the most common cause of the malignant neoplasms in the biliary pas- sages, although the calculi have not remained in situ. The histologic variety of carcinoma of the ducts is always of the columnar cell type, although Robson states that secondary degeneration of papillomata occurs. Systemic infection is rare; the growth usually progresses by contiguity, and sooner or later the lymph-glands of the gastrohepatic omentum are involved. In some of the cases reported the growth was very small at autopsy, even after a year or more of marked symptoms, notably the cases reported by Edes, in which it was not larger than a bean. Death usually occurs from debility, the result of the jaundice and in- fection of the biliary ducts. The symptoms are not distinctive, and the diagnosis cannot often be made. The chronic jaundice and cachexia are not dissimilar to malignant growths of the head of the pancreas, and the occurrence of glycosuria and fatty stools is not sufficiently common in the latter disease to aid frequently in differentiation. In primary carcinoma of the common duct pain is not usually severe, in this respect differing from stones in the same situation, but as so many cases have or at one time had stones, the diagnostic importance of the pain symptom is not great. There is usually no tumor present, a sign commonly existing in cancer of the gall-blad- der. An exploration of the ducts in doubtful cases is the only way a positive diagnosis can be established. McBurncy first called attention to the case with which the duodenum can be opened for the purpose of removing stones im- pacted in the diverticulum of Vater, and first performed the oper- ation. We have several times successfully opened the duodenum for this purpose, and a large number of such operations are now on record. Carle, in an address before the Italian Surgical Congress, 366 WILLIAM J. MAYO strongly urges incision of the duodenum for removal of stones or growths from the duodenal end of the common duct, and cites cases in which stones were formed in the duct and might later give rise to carcinoma. Cancer higher up in the duct would necessitate union between the remaining fragment of the duct and the duo- denum, as Halsted succeeded in doing. As a palliation chole- cystenterostomy is the indicated procedure. The anastomosis may be made either between the gall-bladder and duodenum, or, if the latter is involved, with the transverse colon or jejunum. We have three times joined the gall-bladder to the transverse colon for inoperable obstruction of the common duct, and these cases did fully as well in every respect as three cases in which the duodenum was used as a receptacle for the biliary discharge. One case supposed to be malignant proved not to be so by living in good health six years after uniting the gall-bladder and colon. As a result of our own experience I see no reason why the transverse colon may not serve as well as the duodenum, provided the latter, more favorable, situation is not practicable. The proximity of the large bowel and the nature of its coats render anastomosis with the gall-bladder easy, and in palliation of malignant disease it is perhaps almost as good, for the purpose, as the duodenum. In the original work of Winiwarter it was the chosen method. A few cases of enormous distention of the common duct have been reported. Robson details an instance in which he had been able to suture such a cystic formation to the surface of the body. Summers, in a most interesting case, united the common duct to the duodenum with a successful result. In these cases the ob- struction was, however, non-malignant. The following case came under my observation: M. K., female, age fifty-nine, German. Admitted to St. Mary's Hospital November 1, 1900. History. — For many years patient has suffered from sudden attacks of pain arising in the epigastric and extending to the right hypochondriac region. The suffering has been very severe, lasting from two to six hours, and ending with an attack of vomiting accompanied by prostration. At times has been somewhat jaun- CANCEIl OF THK COMMON H1M;-I)I (T 367 diced alter lliese at lacks. Ahoiil one year a^o appelit*- liegaii to fail, distress in the stoiiiacli Ix-caine more constant, but less severe, and there has been a j)ro^ressi\e loss of weight — over 40 poninls in all. Family and personal history j^ood. J'Lvaniindlion. I'atient somewhat emaciated; there is a marked cachexia with a moderate jaundice. l*ulse, temperature, and respiration normal. Liver can be outlined just below the free margin of ribs; gall-bladder cannot be felt. There is a tenderness and some rigidity of the muscles in this region, otherwise examina- tion negative. Urine has a trace of albumin and nuicli bile. Test-meal developed free hydrochloric acid, and on distention with air the outlines of the stomach were normal. Stools con- tained traces of bile, })ut were light colored. The history was clear as to tiie presence of gall-stones, but the patient had a distinct malignant cachexia. Diagnosis. — Either gall-stones in the common duct or malig- nant disease. Xovcmber ■kl: Incision through the right rectus muscle. Liver somewhat larger than normal; gall-bladder enlarged, containing bile mixed with ropy mucus, and a single non-faceted, dark-colored gall-stone, the size and shape of a small pea. The cystic and com- mon ducts were moderately dilated, but no stone nor other ob- struction could be detected on most careful exploration. Gall- bladder drained to the surface after attaching to the parietal peri- toneum. The findings at the operation were unsatisfactory, and did not account for the condition of the patient. For forty-eight hours drainage of bile was free, but gradually increased in quantity up to two or more pints a day; the skin became greatly irritated from the discharge, and examination showed that a large part, if not all, of the pancreatic secretion was being discharged, with all the bile, to the surface. Stools contained no bile. A Jacol)'s self-retaining female catheter was inserted into the gall-blad- der through the fistulous opening, and in this way the drainage was directed into a receptacle without contact with the skin. It was evident that there was an obstruction which had been over- looked at the duodenal extremity of tiie duct. Patient was in a very feeble condition, and on November 30th was allowed to return home. Even with the continuous drainage she improved somewhat. The jaundice disappeared, and on January 29, 1901, she was readmitted to the hospital. On Jan- uary 31, 1901, an incision four inches in length was made to the inner side of the fistula. The adhesions were separated, and the common duct and duodenum thoroughly exposed. At the ex- 368 WILLIAM J. MAYO treme end of the common duct a hard body could be felt through the wall of the duodenum, the size of a filbert, which was supposed to be a stone lodged in the ampulla of Vater. An incision was made two inches in length in the anterior wall of the duodenum, exposing a grayish- white mass which was strictly localized to the site of the papilla of the common duct. Its size did not exceed the end pha- lanx of the forefinger. About one-third of its length projected into the free lumen of the duodenum, and two-thirds posterior to the intestinal wall. The tumor was excised, exposing the free end of the common duct. The removal was made partly with a knife and partly "^dth the Paquelin cautery, and finally the whole raw surface was seared with the cautery. The common duct was otherwise free of obstruction. The incision in the duodenum was sutured. No enlarged lymphatics were discovered, and no secon- dary nodules in the liver or pancreas could be detected. A small drainage wick was inserted and the wound closed. The attach- ment of the gall-bladder to the skin was left undisturbed. The discharge from the fistula rapidly diminished, and in three weeks had completely ceased. Stools became normal in color, and the gain in the weight and general appearance was most rapid. The specimen was sent to LeCount for examination, and his report is as follows: "Sections cut from all pieces sent after parafiin embedding were stained from varying levels. The tissue is from the duodenal wall, and some sections show portions of Brunner's glands. The Lieberklihn's glands may be traced to lower depths than normal through a very inflammatory mucosa that contains a few small lymph-nodes and small areas of hemorrhage. Certain of these glands are directly continuous with groups of epithelial cells that lie deeply ^dthin the mucosa and the muscular coats. The epi- thelium in these invasions are altered as follows : They lose their columnar shape, become possessed of larger and more deeply stained nuclei, possess karyokinetic nuclei in many instances, and do not retain their characteristic grouping, being, instead, arranged in disorderly clumps and bunches that vary in size; these deeply lying collections of epithelial cells always possess an irregular cavity that simulates a gland of the simple tubular type or a gland-duct." One may conclude from this that not only is the tissue from a cylindrocellular carcinoma, but that the structure is such that it strongly supports the contention of Lohmer ("Ziegler's Beitrage," 1900, xxviii, 372), who asserts, in opposition to the views of Ribbert, that, in glandular carcinoma, the new glands are produced by a direct proliferation of preexisting glands. CANCER OF THE COMMON' RILE-fJUCT 369 So far as the writer can ascertain, the only case in wliich a car- cinoma of the common duct has l)een excised previously to the one which forms the basis of this communication was that of Wm. S. Ilalsted. Tliis case being the first of its kind on record, and extremely interest iiig, I c|uote. with his consent, from the "Boston Medical and Surgical Journal," vol. cxli, December '28, 1899, No. 26, 645: "Primary Carcinoma of the Duodenal Papilla and Diverticulum of Voter, Succesfifully Removed by Operation; Cystico-enterostomy Three Months After the First Operation. — Mrs. M. L., aged sixty. Until August, 1897, patient was well. Her first symjjtom was itching of the skin, which came on suddenly and soon became se- vere. Patient says jaundice did not appear for nearly a month after the onset of the itching. Before the appearance of jaundice diarrhea set in, and there were six or seven stools a day, which were watery and clay-colored. Patient has had no chills, no fever, and no sweating. With the onset of the jaundice she noticed shortness of breath and an occasional swelling of the feet and legs. About the first of January, 1898, she had persistent bleeding of the gums for three days, following the extraction of a tooth. At times the hemorrhage was profuse. Two months ago a tumor was noticed in the region of the gall-bladder. This tumor does not seem to the patient to have increased in size, and has never been tender. In March, 1897, she had several attacks of severe pain in the epigastrium. These attacks were not accompanied by vomiting, fever, or sweating. A few weeks later she had a second but milder attack. The stools were light in color for two or three days at the beginning of these attacks, but patient recalls no change in the color of the urine or the skin at that time. The daughter of the patient states that these attacks of pain were very severe, and that her mother seemed very ill. "Examination February 1.'^, 1S9S. — Patient somewhat emaciated, but fairly well nourished. Mucous membranes pale. Heart and lungs normal. There is a distinct prominence on the right side, the highest point of which is midway between the umbilicus and anterior superior spine. The prominence descends markedly with inspiration. On palpation, the prominent area proves to be pear- shaped and distinctly fluctuating. The border of the liver, wliich reaches almost to the crest of the ilium, can be distinctly felt. "February IJf, 1S9S, Operation. — Vertical incision through rectus muscle. A greatly dilated but not especially dense gall- voL. I — 24 370 WILLIAM J. MAYO bladder presented no adhesions. Liver projects 5 cm. below costal margin. Four silk sutures placed in fundus of gall-bladder with French needles. Small aspirator introduced in center, between su- tures; syringeful of clear fluid mthdrawn. Gall-bladder opened; contents evacuated. In the latter part of the fluid were many fine, sand-like, hard, greenish, round particles, suggesting miniature gall-stones. Common and cystic ducts were dilated to the size of one's thumb. A longitudinal opening, 2 cm. long, was made in the common duct. The same colorless fluid escaped from this in- cision. Duct explored with probe and finger. What seems to be a small, very hard stone, is felt at site of ampulla. To determine the nature of this body an incision was made through the wall of the duodenum. No glandular metastases discoverable. The stone-like body proved to be, as was feared, a carcinoma of the papilla. "Excision of the Cancerous Growth. — To give the growth a wide margin, a large piece of duodenum was excised, a wedge-shaped piece with the apex at the mesenteric border of the intestine. About three-quarters of an inch of the common duct and a shorter piece of the pancreatic duct were excised. The wound in the duo- denum was closed in the usual way with mattress sutures. This was practically an end-to-end anastomosis of the duodenum. The common duct and pancreatic duct were transplanted into the duo- denum along the line of suture. A linear incision into the com- mon duct, which had been made for diagnostic purposes, was closed over a hammer. The gall-bladder was sutured to the peri- toneum. "Abdominal wound closed in the usual way: the peritoneum, with a running silk suture; the muscles and fascia, with buried silver sutures, and the skin, with a continuous subcuticular silver suture. Bismuth gauze inserted to protect the suture of the in- testine and common duct. Drainage-tube surrounded by bis- muth gauze, and gutta-percha tissue inserted into gall-bladder and held in place by a purse-string suture of catgut. Wound dressed with silver-foil. Gutta-percha tissue placed between the raw edges of the skin and the gauze packing. Operation lasted three hours and ten minutes. Patient experienced apparently no shock from the operation, "May 5th, Second Operation. — Cholecystoduodenostomy or cysticoduodenostomy. Suture of fundus of gall-bladder. Com- plete closure of abdominal wound except for drainage. Incision alongside of old cicatrix, circumscribing fistula. Gall-bladder quite small — no larger than one's thumb. Liver about normal CANCER OF THE COMMON BILE-DUCT 371 in size. Many fine adhesions about gall-bladder, which were easily separated. Gall-bladder and ducts thoroughly exposed. The line of suture of common duct at previous operation was readily distinguishable by black silk stitches, bvit it was almost impossible to find any trace of the duodenal suture. Common duct incised at site of old suture. Probe cannot be passed into the duodenum, but there is no positive evidence of the recurrence of the cancer. Unsuccessful attempts had been made before the operation to pass a probe from the gall-bladder through the com- mon duct into the duodenum. Opening into the common duct closed in the usual way with mattress sutures over hammer. An anastomosis between duodenum and the gall-bladder or cystic duct was effected without much difficulty, although the parts to be sutured were very deeply situated and inaccessible. The duo- denum was probably a little less freely movable than at the pre- vious operation, and the gall-bladder was so much reduced in size that we were compelled to pass some of the stitches into what seemed to be the cystic duct; in any event, the neck of the gall- bladder had to be used for the anastomosis. A bougie a boule passed into the gall-bladder was used as a darning-ball to assist in the placing of the sutures. All the sutures were passed (none of them tied) before the openings into the neck of the gall-bladder and duodenum were made, the method employed being that which I described some years ago for intestinal anastomosis. The open- ing in the fundus of the gall-bladder was closed except for protec- tive wicks, wdiich were passed through this fine of suture into the gall-bladder. What seemed to be an enlarged gland was palpated during the operation, but not removed. Patient suffered little or no shock from the operation. In the early autumn of 1898 this patient returned to the hospital too ill for operative interference, and in a few weeks died. During the summer I had corresponded with her, urging her to return to the hospital, for it was clear from her letters that the fistulous communication between the gall- bladder and the duodenum was not working well. At the autopsy it was found that the carcinoma had recurred in the head of the pancreas and duodenum, closing the common duct and inter- fering with the perfect action of the cholecysto-enterostomy or cystico-enterostomy. The anastomosis, as Ave had supposed, had been made between the dilated cystic duct and the duodenum; the fistula was still perfectly pervious, and should have acted nicely except for the interference, a little twisting or bending, created by the new-growth." A STUDY OF 328 OPERATIONS UPON THE GALL- BLADDER AND BILE-PASSAGES* WILLIAM J. MAYO From June 24, 1891, to February 28, 1902, 328 cases of gall- stone or other disease involving the gall-bladder and bihary pas- sages were operated upon at St. Mary's Hospital. This number includes all the cases of this description which were admitted to the hospital during this period. A study of these cases brings out some general features of interest. Three hundred and eleven of the number were of benign origin. The mortality following opera- tion was about 23^ per cent. Seventeen of the operations were for malignant disease, with 3 deaths, a mortality of nearly 18 per cent. Location of the Stones. — In 214 of the cases the stones were located in the gall-bladder or cystic duct or both. In this group there were 2 deaths. In about 10 per cent, of these cases there was obstruction of the cystic duct by a stone or stones, in either case requiring considerable effort to dislodge them. The after-history of many of these cases in which the cystic duct was involved and simply cholecystostomy performed was not wholly favorable. The hospital records do not show the condition of these patients, with the exception of those readmitted for secondary operation. But the number of these cases known to the writer as having had unpleasant symptoms subsequent to the operation leads to the belief that for cases in which the cystic duct has been obstructed, or in which stones have been lodged in the duct for a length of time, cholecystostomy is insufficient, and the gall-bladder should be extirpated at the primary operation if the patient is otherwise in good condition. A large percentage of cases in which the cystic * Reprinted from "Annals of Surgery," June, 1902. OI'KUATIONS OX CALL-BLADDER AND IJILE- PASS AGES 373 (Iiicl lias l)('(Mi iii\(»l\('(l leads to such H disturbance of its mechanism, by stricture, val\(' I'oriiiation, or other unfavorable condition, that it may not furnisli adctiuute drainage to tlie gall-bladder. In some extreme cases the gall-bladder becomes filled with mucus, which is expelled through the duct only by such vigorous contractions as to cause an occasional colic, or a mucous fistula is left, with inter- mittent external discharge. In other cases, at the secondary opera- tion, the gall-bladder is found filled with bile and mucus, develop- ing a condition in which exit to the cystic contents is less easy than entrance of bile. In many instances the discomfort is slight and passes away in time, but there is a large number of cases in which this interference with drainage is sufficent to give symptoms more or less permanent in character. The nature of the difficulty can be aptly compared to stricture of the lacrimal duct or urethra. Nearly one-half of the cholecystectomies performed were secondary to this condition, and only after extirpation of the gall-bladder did a permanent cure result. Stones in the cystic duct are often more easily removed with the gall-bladder than without it. If the peritoneum binding it to the liver be divided on each side, the connective tissue between can be easily separated with the finger; by using the gall-bladder as a tractor and, if necessary, dividing the peritoneal and muscular coats just above the cystic duct, the mucous tube of the latter will strip out readil3% bringing the stone with it. The mucous coat about the neck of the gall-bladder is thick and separates easily from the outer coats, while the fixation by adhesions is to the outer coats alone. At the fundus the mucous membrane is less easy to separate, and a combination of amputation of the fundus with re- moval of the mucous coat from the lower portion of the gall-bladder and cystic duct makes cholecystectomy a safe operation. The drains should be tied to the stump with fine catgut to prevent dis- placement, and a strip of rubber tissue placed between the drains and the stomach to prevent adhesions. The catgut fixation sutures are absorbed before it is necessary to remove the drains. In most cases the whole gall-l)ladder can ])e so easily removed as to render this stri])ping of the mucous membrane unnecessary, but in j)rimary 374 WILLIAM J. MAYO stone impactions and secondary operations for stricture it serves a good purpose; and, as it leaves a pouch composed of the outer coats, into which a tube drain can be securely fastened, the cystic duct can be left open for drainage of the hepatic ducts in cases in which an infective cholangitis is present. Ligation of the cystic duct would prevent this imperative indication (Davis). To leave the cystic duct open in the abdomen for this purpose without direct and complete drainage would be attended with more danger- The bile itself would drain safely to the surface with ordinary care in placing the drains, but not so the infected material from the ducts. If there is no infection of the hepatic and common ducts, — and usually there is none, — drainage of the bile to the surface is unnecessary, and the cystic duct can be closed by ligature. Out of 33 cholecystectomies we had but 1 death, and this was due to ligation of the cystic duct in a case in which the hepatic ducts should have been drained through it. Cholecystectomy will rapidly gain in favor and will undoubtedly supersede cholecystotomy in a large group of cases. Stones Outside the Bile Tract: 13 Cases, no Deaths.-^-In 13 cases stones were found outside of the gall-bladder and biliary ducts. In some the calculi were encapsulated in the adjacent liver border, forming hard nodules from which, upon incision, they could be enucleated. In others a mass about the fundus would contain a number of stones, with perhaps a little mucopuru- lent fluid. Further dissection toward the cystic duct would open a functionating organ of small size, with every evidence that it was but the remains of the gall-bladder. In several cases we have opened a pocket composed of the re- mains of the fundus, but slightly separated by ulceration and con- nective-tissue formation from the neck of the organ. The gall- bladder was perforated, and the extruded stones in a mass of ad- hesions communicated freely with the fundus, or perhaps there were several such pockets lined with granulation tissue and more or less separated from each other. In other cases the stones were found lying in a pocket outside the gall-bladder, with adhesions to the intestine, but communicat- opp:kations on gall-uladdeu and hilk-I'ahsaoes 375 ing with neither. The fiiiidus was contracted to a mass of scar tissue. In three cases we found stones outside the gall-bladder without communication with it, but with a fi.stulous opening into I lie intestine — twice to the duodenum and once to the transverse colon. Removal of the stone in these cases made it necessary to close the fistulous tract connected with the bowel. In the two cases in which the duodenum was involved the friable nature of the infected tissues and the deep seat of the area to be sutured made this a matter of considerable difficulty, and one case formed an intestinal fistula which was very troublesome, but later healed. In both of these cases there were stones in the gall- bladder, but the cystic duct was totally obstructed. It is prol)able that in both cases the stones had been impacted in the cystic duct before the ulceration took place. A study of these cases leads to the belief that stones passing by ulceration and perforation from the gall-bladder and cystic duct to the intestine do so slowly, and that often, if not usually, cicatrization takes place behind before the extrusion into the intestine is accomplished. The next most common direction for stones to travel is toward the surface of the body. The gall-bladder becomes obstructed at the cj^stic duct and its contents infected. Adhesions form to the parietal peritoneum, and eventually, by ulceration, work to the surface as a subcutan- eous abscess. This was met with twice in this series of cases, and one case was admitted with a fistulous opening following spon- taneous rupture. In a considerable experience in the operative treatment of gall-stone disease in private houses and local hospitals we have found this latter condition relatively more frequent than in, St. Mary's Hospital, as the local peritonitis which marks these cases prevents their transportation, while extensive changes at- tending the extrusion of the stones into the intestine may give little symptomatic evidence of trouble. Cholecystitis. — In this group were 34 cases with 5 deaths. This mortality calls attention to the serious nature of the infections. All the cases in which an acute suppurative condition existed at the time of the operation, with or without stones, and all cases in which the gall-bladder was found thickened and containing more 376 WILLIAM J. MAYO or less ropy mucus and bile or sand-like sediment, ^^dthout stones, were classified at the time of operation as cholecystitis. It would seem that the difJerence between these two conditions was marked enough to render a double classification necessary, and that the first should be called "suppurative" and the second "catarrhal" cholecystitis. It was noted in the group which might be termed catarrhal that cholangitis was more frequently an accompaniment, although usually of a mild and irregular type, and that, after the operation, as there was no obstruction by stone or other"\;\ase at the cystic duct, an extension of the inflammatory process manifested itself in these three cases and death resulted. In the suppurative form the gall-bladder was comparable to a closed cavity containing pus, which so thoroughly blocked at the cystic duct as to prevent progressive infection. In all but four of the acute empyemas the stone was removed at the primary operation. In three cases the gall-bladder was also shelled out. Two patients with acute em- pyema in which the stone was removed after great difficulty de- veloped a fatal suppurative cholangitis after cholecystotomy. One of these patients also had a profound jaundice, with purpura hsemorrhagica, and death was probably as much, or more, a result of the hemorrhage as from the progressive infection. The stone in this case was impacted at the juncture between the common and cystic ducts, obstructing both. The other case was typical — the removal of the impacted stone allowed the septic material to pene- trate the ducts. In the four cases in which the gall-bladder was drained, and on account of the serious condition of the patient, no attempt was made to remove the obstruction, each one recovered promptly, and the stone was removed at a secondary operation with the gall-bladder. Cholecystitis, with or without obstruction at the cj^stic duct, is the most dangerous condition for which we are called upon to operate, and although the patient may be apparently in good general condition, progressive infection of the ducts is liable to supervene. In acute infections little manipulation should be made and quick drainage established. If a stone obstructs the cystic duct, it is safer to leave it for a second operation, or, as we OPKUATIONS ON (iALL-HLADDKU AM) I1ILK-P.\>S.SAGES 377 have clone of late, remove the entire ^ail-hhulder to a healthy point proximal to the stone. With the excej)tion of these four cases, all the stones have been removed at tlie primary operation or were discharged through the fistula later in some of the earlier cases. In no case was there a reformation of stone, so far as is known. Gall-bladders which have become cystic from stone obstructing the cystic duct, and in which, after the clear mucus is dra^^•n off, some purulent looking fluid comes up having the physical appear- ance of pus, are not included in this group. These cases are com- mon, and are classed with the ordinary obstructions at the cystic duct ill whicli the stone should be removed at the primary of)era- tion. Tiic author has long held the view that the dependent fundus is an important mechanical factor in that it favors stone formation in cases in which stagnation of the bile, infection of the gall-bladder, and some interference with drainage through the cystic duct are the other factors: that is to say, if the cystic duct were at the bottom, the sediment would pass out first. For this reason it seems that cholecystitis might be more liable to exist without stones in the cases in which the fundus was above the level of the cystic duct. It is possible that the permanent elevation of the fundus, produced by the adhesion to the abdominal incision, may be one cause of the non-formation of new stones after cholecystotomy. In two cases acute suppurative cholecystitis followed typhoid fever, in each instance developing suddenly — one case during the third week and the other during the fifth week after the beginning of the fever. At the time of operation the typhoid bacillus was found in pure culture, and the patient's blood gave the Widal re- action. In both cases stones were present in the gall-bladder, but, on going into the history, it could be shown almost beyond a doubt that the gall-stones existed before the advent of the typhoid, and merely determined a lowered resistance. In taking the histories of the cases of gall-stones operated upon at St. Mary's Hospital, only a very small percentage had had typlioid fever at any time. It would seem that the etiologic importance of typhoid fever in the causation of gall-stones had been overestimated. 378 WILLIAM J. MAYO • Cholecystotomy has been made by introducing into the gall- bladder a rubber tube, the size of a lead-pencil, wrapped in gauze, then covered with rubber tissue. A catgut purse-string suture is then placed below the incision in the fundus, and the ragged edge of the opening in the gall-bladder inverted into its cavity (Sum- mers). The suture is then pulled taut, compressing the packing about the tube and making a tight joint; the drain is held in place by a catgut suture. If the gall-bladder is too short to reach to the parietal peritoneum for fixation, a few strips of gauze are tacked to it with catgut and form an extension to the surface. In a consider- able number of cases the drains have been carried out through a stab wound and the operative incision completely closed. In the course of other operations, if gall-stones coexist, a stab wound prop- erly placed enables the fundus of the gall-bladder to be drawn out of the opening, and the stones can be removed and drainage established by the aid of the hand inside of the abdomen. Unless it is necessary to remove the gall-bladder, it is not wise to break up adhesions beyond a point necessary to explore the ducts and manipulate the fundus. Time spent in separating adhesions un- necessarily, which must reform, not only prolongs the operation, but breaks down a valuable barrier to the extension of the in- flammatory process and opens up new avenues for infection. Stones in the Common Duct: 31 Cases, One Death. — Stones were found in the common duct in 31 cases, and in only one case was it possible to remove the stone through the cystic duct by dilating it. This was a lucky accident, as I am convinced, from frequent failures, that attempts of this kind are a loss of time. In 29 cases the duct was incised and the stones removed. In 5 cases this was accomplished by separating the gall-bladder from the liver and incising the free surface down to and along the cystic duct to the common duct, the latter being incised at the juncture. In 2 of these 5 cases the cystic duct tore completely loose from the common duct, leaving an irregular opening, which was closed by a plastic operation upon the duct, using the gall-bladder de- nuded of the mucous membrane, except at one point, where enough was left to fill the gap. The remainder of the outer coats OPKUATIONS ON' GALL-IJLADDKH AM) HI LK-PASSAGES 370 of the gall-bladder was trimmed to a convenient sized flap, wrapped about the common duct and held by a few catgut sutures and a light gauze pack. The biliary leakage was very slight in either case, and lasted but a few days. In the large majority of cases of stones in the common duct the stones were movable, and in two-thirds of the cases more than one stone was present in the duct (in one case 27 stones). The typical ball-valve stone of Fenger was met with 7 times. As a rule, where more than one stone was present, the duct was suffi- ciently dilated to enable the introduction of the finger for purposes of exploration. In no other way could we be sure that we had removed all the stones. In 5 cases stones were present in the hepatic ducts also, but were movable, and with varying difficulty were brought to the incision in the common duct for removal. In 2 cases energetic attempts to remove all the stones from the lower end of the duct or a diverticulum from it resulted in forcing the finger well into the duodenum, probably at an ulcerated point, rather than at the site of the papilla. Fitz has shown that large stones, as a rule, pass into the bowel by ulceration rather than by dilating the papilla. In these 2 cases the contents of the duodenum escaped from the drainage-tubes for a number of days, causing rapid emaciation. One patient recovered completely; the second left the hospital after seven weeks in bad condition, and eventually died at her home from inanition. This was the only death in this group. In a number of these cases the head of the pancreas was en- larged, and in 6 cases more or less pancreatic secretion came out with the bile, excoriating the skin, and causing a peculiar odor to the discharge which seems to characterize it. One of these cases had a general acute eczema involving the entire body ; all the cases recovered. Jaundice. — Jaundice in connection with stones in the common duct was a most variable feature. In many cases it was so slight as not to attract especial attention, and finding stones in the com- mon duct was a surprise. In the majority of cases, however, the jaundice was marked. Courvoisier long ago called attention to the 380 willia:vi j. ^l\to fact that jaundice from stone in the common duct was accompanied by a contracted gall-bladder, which could not be palpated ex- ternally in 80 per cent, of the cases. This was true of all but 3 of our cases, in which the gall-bladder was filled with stones, prevent- ing the usual contraction. Jaundice as a Cause of Postoperative Hemorrhage. — In 3 cases capillary oozing was a most serious postoperative comphcation. One case was in a precarious condition for twelve days from this cause. Robson has called attention to the value of chlorid of calcium as a prophylactic in these cases. We have used this for about one year. I am uncertain as to its value, but we have had no deaths from hemorrhage since. One case of empyema of the gall-bladder complicated vdih extreme jaundice from a stone im- pacted in the cystic duct at its juncture with the common duct, and 3 cases of jaundice from malignant disease, died from post- operative capillary oozing. In all these cases there were sub- cutaneous ecchymotic spots, looking like purpura hsemorrhagica, before operation. Every patient with jaundice having this compli- cation died after operation upon the gall-bladder. Tests as to the coagulability of the blood have been rather uncertain, but this clinical means of differentiating the operable from the non-operable cases has been impressed upon our minds. After removing stones from the common duct the incision is closed by a continuous catgut suture, providing the duct is in good condition and no fragments of stone or other detritus are left behind; otherwise the duct is partly closed, leaving a gap for drainage. If the patient be in bad condition, drainage is employed without suture. The suturing is done -u-ith a single row, and, if there be much difficulty in doing this, only enough of a running suture is placed to direct the coaptation, drainage being provided for by fastening gauze wdcks covered with rubber tissue in position with the catgut suture, to prevent displacement. In 26 cases cholecystostomy was made for drainage. In 2 cases cholecystec- tomy was performed, the cystic duct being left open for drainage. In 1 case the duodenum was incised to remove a stone from the ampulla of Vater. OIMOUATIONS ON GALL-IJLAODKR AXD BILPJ-PASSAGES 381 Cholecystenterostomy was pcrfcMiiicd .'] limes for flironic pancrcjilitis and 3 limes for malignaiil disease. 'J'lie aiiasUjiiiosis was made to the duodenum twice and to tlie transverse colon 4 times. So far as we could judge, the anastomosis with the colon answered every purpose. One benign case lived six years in good health and died from other causes, and a second is alive and well now, two years after the operation. While the duodenum is the proper place for the anastomotic opening, it sometimes happens, by reason of adhesions, that this site cannot be secured. The transverse colon is close at hand, and with its appendices — epi- ploica and omentum — furnishes a secure situation for the opening, and the operation itself may in this way be easily accomplished. There are many theoretic objections to it, and a loop of jejunum would seem a more desirable point for the entrance of the bile; however, the fact remains that in most of the reported cases of anastomosis between the gall-bladder and colon the results have been good. The Murphy button was used in making the anas- tomoses. In 12 cases an exploration showed an error in diagnosis. This, however, includes only the cases in which the abdominal wall was incised independently for this purpose, and does not fairly repre- sent the mistakes. In some of the earlier cases a small gall-bladder, with thickened walls extensively adherent, was found, and we contented ourselves with loosening the adhesions. Recovery followed the operation in each instance, and the symptoms were usually relieved. In a few cases, however, there was no abate- ment of the previous pain. In reoperating upon one case a ball- valve stone of small size was found in the common duct, yet so little jaundice was present as seemingly to preclude the possibility of its presence in this locality. In two cases since that we have found a rolling stone in the common duct under preciseh' similar circumstances. Adhesions about a small gall-bladder should lead to a careful exploration of the common duct before deciding that the adhesions alone are the cause of the symptoms. In the cases in which the gall-bladder was explored negatively the real difficulty was usually an old appendicitis or ulcer of the stomach In one 382 WILLIAM J. MAYO case a stone in the right ureter, and once a small ovarian dermoid with a long twisted pedicle was found to be the source of trouble. The abdominal incision for work upon the biliary tract which we have found most useful has been the straight one through the rectus muscle, enlarged, if necessary, either at the top or bottom, after the method of Bevan, with the modification suggested by Robert Weir, incising the sheath of the rectus muscle and the deeper muscles obliquely and retracting the rectus itself rather than severing it. Great difficulty in exposing the gall-bladder, especially if small and under the liver, may be experienced. By dividing the peri- toneum binding the gall-bladder to the liver and separating the cellular space between, the parts can usually be mobilized without dividing the rib cartilages. The venous hemorrhage is quite free for a short time, but stops after temporary gauze packing, and in our cases has never been a serious source of trouble. Of the 8 deaths in the benign cases, 4 were due to progressive infection of the liver ducts with late kidney complications, 1 from the same cause with capillary hemorrhage, and 1 sudden death due to myocarditis, which was recognized previous to operation, but the danger of which was not fully appreciated. Two patients died suddenly on the fourth day. The symptoms after the operation consisted of a peculiar nervous unrest, pulse 110 to 120, tempera- ture 100° to 102° F., gastro-intestinal disturbance not marked, but some tympanitic distention shortly before death, which took place unexpectedly. The postmortem did not show adequate cause for the result. The condition seems to correspond with that described as hepatargia (Eisendrath) , and due to cessation of liver action. The 2 cases belong to the group of cholecystitis without stones. In not a single case was peritonitis a cause of death. Malignant disease involving the bile tract was found 17 times; the results were very discouraging, with a single exception; the palliation secured was of doubtful character, and death followed immediately in nearly 18 per cent, of the cases. The deaths were due to capillary hemorrhage, and all these cases had purpura OPERATIONS ON GAI-L-BLADDER AND BILE-PASSAGES 'iH'.i hsemorrhagica. Stones were also present in all the malignant cases in which the gall-bladder and dnc-ts wore explored. RfiSUMfi Operations for Nnii-malvjnaiil Pinca.ic of the Call-hladdcr atiil Bile-passagea occur- ring in Si. Mary's Ilo.spilal from June 2',, 1001, to February 20. 1002 No. Operated. Recov. Dii.d. CholcfVstosloMiv. Sloncs in f,';ill-bl;i(lgativc 12 12 311 304 7 Operations for Malignant Disease No. Operated. Recov. Died. Ciiolecystostomy. Obstruction of common duct .4 2 2 Cholccvstectomv and partial hepatectomy. Can- cer df gall-i)ladder 1 1 Duodenocholedocliotomy. Cancer in ampulla of Vater 1 1 Cholecystenterostomy. Malignant obstruction of conuiion duct 3 2 1 Exploratory. Inoperable cancer 8 8 17 14 3 *One case died after leaving the iiospital two months after llie operation. MALIGNANT DISEASE INVOLVING THE GALL- BLADDER* WILLIAM J. MAYO From June 24, 1891, to September 23, 1902, 405 operations were performed for all causes upon the gaU-bladder and biliary passages in St. Mary's Hospital. Of this number, 20, or about 5 per cent, of the operations, were for malignant disease. While this does not make mahgnant disease a common malady, it is of sufficient frequency to merit careful attention. The true propor- tion would perhaps be somewhat higher than 5 per cent., as cancer of the gaU-bladder, Hke mahgnant disease of most of the internal organs, is slow to be recognized, and is often, if not usually, so far advanced as to render even an exploration unnecessary, the char- acter of the trouble being only too manifest on physical examina- tion. It is a question whether the relative number of cases in which cancers of the gall-bladder and bile-passages are found and not subjected to operation would be more than 5 per cent, of the num- ber of patients with gaU-stones in which operation is refused or the true nature of the condition is not recognized. After investigation of the subject, Schroder says that 14 per cent, of these patients suffer at some time from cancer of the biliary apparatus. This brings up the query as to whether there is an etiologic re- lationship between gall-stone disease and cancer of the gall-bladder. Cour^^oisier found that 74 out of 84 cases of malignant disease of the gaU-bladder had gall-stones. Siegert states that in 95 per cent, of all cases of primary cancer of the gall-bladder gall-stones *Reprinted from "The Medical News," December 13, 1902. 384 MALIGNANT DISEASE IN'VOLVING THE GAI.L-HLADDEU 385 are present, and adds the significant fact that calculi are found in only 15 per cent, of secondary malignant disease of this organ. In the London Cancer IIos])ilal Jicadles found that in 4 cases of primary cancer of the liver gall-stones were present in all, and in 36 secondary cancers of the liver gall-stones were not detected in a single instance. Musser gives the percentage of gall-stones in the cases of primary cancer of the gall-bladder which he found re- corded as 69 per cent., but says that in many of the instances in which no stones were discovered it was probable that the calculi had passed. In all the explored cases of cancer of the gall-bladder in our own series gall-stones were present. In an exhaustive ex- amination of the subject of cancer of the gall-bladder and bile- ducts Kelynack placed the proportion of primary cancer of the gall-bladder at 75 per cent, of the whole, giving 25 per cent, as the number of primary cancers of the bile-ducts. Kelynack called attention to another noteworthy fact: that cancer of the gall- bladder is at least three times as frequent in women as in men, and that this proportion is also true of gall-stone disease. Siegert found 79 females to 14< males. Cancer of the gall- bladder is most common between the ages of fifty and sixt^', which is the period of greatest frequency of gall-stones. Bland-Sutton states that columnar epithelium from the mucous membrane is the type of the carcinomatous process, but says that in most instances it could not be determined whether its origin was in the mucous glands or in the epithelium. It is to be noted that the cholesterin which forms the chief con- stituent of gall-stones is also a product of the mucous membrane. Butlin says that the cancerous ulcer is the most common form. Bland-Sutton has found general carcinomatous infiltration of the gall-bladder forming a hard, thick- walled tumor, with a small cavity containing stones, to be the more common variety, and with this latter group could be placed the cases which we have recorded. It will be observed that these two varieties which compose the ma- jority of the reported specimens not only contained stones, but the primary lesions would suggest this probable source of irritation. Other gross forms of malignant disease of the gall-bladder are also VOL. I — i5 386 WILLIAM J. MAYO found, particularly that type in which a tumor projects into the cystic cavity. The site of origin of cancer in the gall-bladder is usually near the fundus, although a considerable number have been discovered near the opening of the cystic duct. Perforation into the peritoneal cavity and general peritoneal infection have rarely taken place. The viscera may be found studded with little cancerous growths, causing ascites, or fistulous openings between the gall-bladder and the hollow viscera may occur. The most common method of extension is to the adjacent liver, either directly or through the blood-vessels, and to the lymphatic glands lying in the hepatic fissure. So far as we could discover, there has been no recorded case of operation for primary sarcoma of the gall- bladder, although a few postmortem specimens have been exhibited. Musser found three such cases recorded in the literature. The following indisputable facts should attract attention: First, gall-stones are almost constantly present in primary malig- nant disease of the gall-bladder and rarely in secondary; second, the relative proportion of gall-stone and malignant disease of the gall-bladder in women and men is practically identical; third, the pathologic lesions found are best explainable on this hypothesis; and, fourth, the similarity in age frequency. We are certainly warranted in concluding that gall-stones are the most important etiologic factor in malignant disease of the gall-bladder. Since the relative proportion of cancerous disease involving the gall-bladder and bile tract and simple gall-stone disease was one in 20 in the 405 operations upon the gall-bladder and bile-passages, the question assumes practical importance, and while I would not say that for this reason alone gall-stones should be removed, it certainly aids in deciding that early removal of active gall-stones, other things being equal, is sound surgery, particularly since nearly all the mortality-giving complications are the result of delay. In over 250 uncomplicated gall-stone operations the mortality was less than 1 per cent. The diagnosis of primary cancer of the gall-bladder may be easy. As a rule, a hard tumor is to be detected in the region of the gall-bladder, which is not very tender to touch, and unless there MALIGNANT DISKASE IN\()L\"IN(; TIIK (JALL-HLADDKIt 887 is a peritoneal iiivcjheMiciit, rigidity oi tlie overlying muscles is not marked. There is j)ro^ressive loss of flesh, and later a ca- chexia is dcvclopclI-l)(Mii<,' of flic liosL The active gall-stone, so loii;,' as it has {)erio(ls of latency, is a (icl)atal)le sul)ject, and can l)e claimed })y either the internist or the snrj^eon. It is the ease of chronic and relapsing appendicitis over again. The age and general condition of the patient will, however, be a more dominant factor in this class of cases than in disease of the appendix. The question for consid- eration is whether, in view of the possil)ility, nay, i)rol)al)ility, of further extension of the trouble, it is not wise to remove active gall- stones early. The results of early operation for gall-stone disease are remarkably good. Up to December 11, 190'-2, we had about 250 cases of this kind out of a total of 454 operations, with a death-rate of less than 1 per cent. Robson, Kehr, Ochsner, Murphy, Richard- son, and {)ractically all surgeons with a large experience give sta- tistics which tell the same story. In over !2000 operations of this kind, in the hands of six surgeons, there was not a single instance of reformation of the gall-stones. In this field of surgery delay breeds misfortune. Complications are due to changes in the wall of the gall-bladder or involvement of the bile-ducts, and the calculi may become but an incident in the pathologic process which they initiated. Re- peated infections with prolonged interference with drainage cause the walls of the viscus to become infiltrated with inflammatory products, and the connective tissue formed interferes with its elasticity under pressure and limits its power of contraction. This introduces the element of tension, and results in more marked symptoms, as pointed out by Berg. The struggle is now to a finish, and either the obstructing stone is forced through the cystic duct into the common duct, to remain there, or to pass out into the intestine, or it may become encysted in a thick-walled pouch com- posed of the remains of the gall-bladder, causing recurring attacks of inflammation. The variety of changes found in the gall-bladder remind one of a chronic appendicitis, in which deformed and par- tially obliterated forms, with stones encapsulated ofttimes in the surrounding tissues, are not uncommon. Adhesions to surround- ing viscera may prevent fatal perforation, or divert it into a neigh- boring viscus with spontaneous discharge of the infected fluid, and, 396 WILLIAM J. INIAYO incidentally, of the calculi. Even if all the stones are so discharged, which does not often happen, the adhesions remaining may be a prolonged source of distress to the patient. An operation in this stage is fraught with some danger, and the gall-bladder can no longer be expected to return to the normal. It may do so, but in a number of cases interference with drainage, due to some permanent change in the cystic duct, gave rise to colics or pain. It is in this class of cases that cancer of the gall-bladder is most liable to supervene. In 454 operated cases we found cancer of the gall-bladder or bile-ducts in 21 (5 per cent.), and in nearly all a distinct history of previous colics was elicited. All the cases in which the gall-bladder was examined contained stones. The chronic irritation of calculi in a gall-bladder in which inflammatory changes have taken place seems to be the usual precancerous con- dition. This is certainly a serious factor in deciding operation — a l-to-20 chance is not to be risked lightly. The possibility that all the stones may be expelled through the ducts is alluring to both the physician and the patient. This may happen, but we have never operated upon a case in which the patient brought us stones detected in the feces that the gall- bladder did not contain more. In 49 of our cases stones were found in the common and hepatic ducts, ^ath or without jaundice, but in any event necessitating a serious operation. Robson found stones lodged in the common duct in 20 per cent, of his cases. Evidence that a stone has success- fully passed into the intestine does not contraindicate operation; it is altogether probable that more stones remain, and the next "labor" may end prematurely, leaving one or more stones in the duct. Occasionally one finds a case in which such stones have re- mained for years in the common or hepatic ducts with compara- tively little trouble; but such instances are rare, and the rule is that infections of the liver-ducts, and occasionally those of the pancreas, render surgical intervention a necessity, but with the jaundice which usually coexists the operation is far from safe. Infection of the liver-ducts introduces an element of uncer- tainty in the prognosis, and to this cause the majority of deaths STATUS OF Sf'UGEUY OF (; ALL-HLADDKK AND JHLK-lJl ( TS 'V.)7 afler operation may be traced. In these cases the condition de- scribed as "hcpatarjila," or cessation of liver function from degeneralron of the Uver parenchyma, may be present. (Quincke found that chohingitis, secondary to colon infection of the gall- bladder, was more liable to cause acute liver degeneration than when due to the ordinary pus microorganisms, although this bacillus has a shorter life history. Talma demonstrated that either colon or typhoid bacilli injected into the gall-bladder would not infrociucntly travel up the hepatic ducts and cause extensive patho- logic changes in the liver. Lavastine recently reported six autop- sies in cases of acute liver insufficiency in which these changes were most marked. In our experience this has been the cause of death in nearly one-half of the cases. The symptoms of this condition are chiefly nervous — usually sudden in onset and rapid in their course. The only safeguard in operating upon cases with infective cholangitis is free drainage of the bile to the surface. In 1897 we operated upon two cases in which the diagnosis of gall-stones seemed certain, and in neither case were calculi found. The gall-bladder was not opened, although its walls were thickened and lacked that peculiar bluish-green color characteristic of it when filled normally with bile. The operation was thought to be a mistake in diagnosis. About this time we had several cases in which we could not be sure whether or not the unopened gall- bladder contained calculi. In 2 instances the gall-bladder was incised; no stones were found, but it contained thick mucus and bile. These patients were cured by drainage. During the next two years we had a small number of such cases, and finally re- operated upon the two original patients, in whom the symptoms had continued, and obtained a cure in each. In 454 cases we have had 26 of this description, — evidently a chronic cholecystitis without stones, — but with a similar symp- tomatology; the colics were undoubtedly due to plugging of the bile-passages with the tarry material with which the gall-bladder is ordinarily filled. In nearly all these cases we have, before open- ing and draining the gall-bladder, carefully examined the pylorus, the appendix, and the right kidney to complete a differential diag- 398 WLLLLLM J. MATO nosis. In all the cases the gall-bladder was evidently the affected organ, its walls being thickened and very often exceedingly ad- herent to neighboring structures. There was no evidence that stones had ever been present. I have recently written to each of these patients to find out the present condition, and from the 23 letters received 19 replies; of these, 15 write that they are well, 2 improved, and 2 unimproved. Three died as a result of the operation, from hepatargia. This death-rate alone demonstrates that the infection is often more active than in gall-stone disease. A few words in regard to the principles underlying operations upon the gall-bladder and bile-ducts. The most important ques- tion is that of infection, and as to whether it is acute or chronic, and especially whether or not it involves the liver-ducts. Every gall-bladder containing calculi can be looked upon as infected, although with the free circulation of bile and nearly normal gall-bladder found in slumbering stones it is slight, and the "ideal" operation — ^that is, complete closure of the gall- bladder and abdominal incision — may be performed; but it will occasionally, even then, lead to disaster — to say nothing of the danger of overlooking stones — which may and does happen to the most expert. The ideal operation has been performed largely upon these latent cases found during other operations, and the conclusions dra^m cannot be safely applied to gall-stones in a state of activity, which of itself argues an infection of a more acute type. Cholecystotomy and drainage is the operation of choice in these cases. Patients suffering from active gall-stones with periods of latency, in a fairly normal gall-bladder with patent ducts, are nearly always cured by this method. Drainage is con- tinued until the bile is normal. This is a safe operation, as the gall-bladder is fastened to the abdominal incision, the fundus is permanently elevated, and gravity drainage through the cystic duct is even better than in the normal state. Gall-bladders with thickened walls, and especially if the cystic duct has been ob- structed, are hable to give trouble after cholecystotomy, and, if possible, the organ should be removed. If the liver-ducts are STATUS OF SUUCiKUY OF (JALL-IlLADUKIt AM) 111LK-DL( TS 'JIM) entirely free from iiivoivenieiit, then, and only llien, sliouM the cystic duet be tied and the pdl-hladder excised without drainage of the bile from the liver-ducts; but if the hepatic ducts are in- volved, free drainage of the bile to the surface must be provided for by leaving the cystic duct open. Kehr says that the hei)atic ducts require drainage in 37 per cent, of cases, and, with increased ex- perience, we believe that this is not too high a percentage. It was in the management of such cases that we recommended the removal of the nmcous membrane — the peritoneal, as much as may be of the muscular coats being left. The cystic duct is cut across and left open in the bottom of the pouch, the outer margins of which are ])rought to the surface and drained in the same manner as after cholecystotomy. It gives all the advantage of temporary hepatic drainage with the permanent benefits of chole- cystectomy. To cut the gall-bladder away, leaving the cystic duct open, spouting bile into the abdominal cavity, requires a quantity of drainage, and does not compare in ease and safety with the method we have described and practised. The raucous membrane of the fundus of the gall-bladder does not separate easily, but in the vicinity of the cystic duct it can be readily removed. For this reason we usually amputate the fundus and remove the mucous membrane from the lower portion. We can recommend this with much confidence, as in a considerable experience we had no deaths and no relapses. In stones in the common duct drainage of the hepatic duct is essential. This usually has been done by suturing the common duct and draining through the gall-bladder by cholecystotomy. W. E. B. Davis first recommended leaving the common duct open after removing stones, for drainage of the hepatic ducts, and as these cases are usually jaundiced and bad subjects for operation^ this time-saving practice has often proved to be of the greatest value. In 49 choledochotomies we had but 3 deaths, and the method of Davis, in whole or in part, has been the one of choice, the gall- bladder being also drained at the same time. In comparing the surgery of the gall-bladder with that of the 400 WILLIAM J. MAYO appendix, this essential point of difference exists : in removing the appendix every effort is made to prevent leakage from its attach- ment to the cecum, while in the gall-bladder the necessity of thor- ough drainage from the liver-ducts out of the external incision is, in many cases, necessary to success. Of the 454 cases up to December 11, 1902, upon which this paper is based, 438 are from the records of St. Mary's Hospital and 16 occurred in the Minne- sota State Hospitals for the Insane. MALIGNANT DISEASE OF THE COMMON BILE DUCT* WILLIAM J. MAYO Primary carcinoma of the common duct is rare. In 4578 autopsies Kelynack found 8 cases of primary cancer of the gall- l)Iadder, only 2 of which had their origin in the duct. Musser collected 100 cases of carcinoma of the gall-bladder and 18 of the hile-ducts. The site of the neoplasm in the common duct is usually either at the juncture of the hepatic and cystic ducts or near its duodenal termination. The 18 cases collected by Musser showed 3 in the hepatic duct and 14 in the common duct, and of the latter, 9 were at or near the papilla. In 17 cases of cancer of the ducts Rolleston found 15 in the common duct, and of these 10 were at or near the papilla. In 511 operations upon the gall- bladder and bile-passages performed in St. Mary's Hospital up to March 21, 1903, 22 were for malignant disease, and of this number it was thought that G began in the ducts. In several cases the exploration did not reveal the exact site of origin. As to the etiology of carcinoma of the common duct, there is some question. It must be conceded that gall-stones are the most common cause of cancer of the gall-bladder. In Musser's 100 cases of gall-bladder carcinoma 69 contained gall-stones and good evi- dence that calculi had at one time been present in the majority of the remainder. Primary cancer of the gall-bladder and gall-stone disease are more common in females than in males, and in about the same pro- portion. This is not true of duct carcinoma. Malignant disease of the common duct is equally frequent in the male antl female, * Reprint from " Xorlhwestern Medicine," 1903, vol. i, No. 4. VOL. I— ;2G 401 402 WILLIAM J. MAYO which does not favor the behef that gall-stones are the cause of the malignant process in the duct. In 22 collected cases of cancer at or near the papilla of the com- mon duct Edes found gall-stones in but 4, and 3 of these were in the gall-bladder. In 36 cases of cancer of the common duct Rolleston found gall-stones in less than half. The extensive experience of Mayo Robson, however, entitles his opinion to great weight, and he states his belief that gall-stones are the most common cause of malignant neoplasms in the biliary passages, although the calculi have not remained in situ. The histologic variety of carcinoma of the ducts is always of the columnar-cell type, although Robson says that secondary degenera- tion of papillomata occurs. Systemic infection is rare ; the growth usually progresses by contiguity, and sooner or later the lymph- glands of the gastrohepatic omentum are involved. In some of the cases reported the growth was very small at autopsy, notably the case reported by Edes, in which it was not larger than a bean, even after a year or more of marked symptoms. Death usually occurs from debility, the result of the jaundice and infection of the biliary ducts. The symptoms are not distinctive, and the diagnosis can- not often be made. The chronic jaundice and cachexia are not dissimilar to malignant growths of the head of the pancreas, and the occurrence of glycosuria and fatty stools is not sufficiently common in the latter disease to aid differentiation. In primary carcinoma of the common duct pain is not usually severe, in this respect differing from stones in the same situation; but as so many cases have at one time had stones, the diagnostic importance of the pain symptom is not great. There is usually no tumor present, although, not infrequently, the distended gall- bladder can be palpated below the margin of the liver, but not the distinct hard tumor of cancer of the gall-bladder. The question of jaundice is an important one. In the beginning it is often inter- mittent — very much like stone in the common duct. Occasionally this will lead to an infection, with the fever and chills of cholangitis. In the later stages jaundice is complete. Maury reports a recent case with the typical symptoms of Hanot and Rendu, early inter- MALIGNANT DISEASE OF THE COMMON HII.K-DUCT 403 mittent jaundice, insidious onset, and diarrhea, with, later, coni- plete obstruction of the common duct. An ex[)Ioration of tlic ducts in doubtful cases is the only way a positive diagnosis can be es- tablished. McBurney first called attention to the case with whicli the duo- denum could 'be opened for the pur])ose of removing stones im- pacted in the diverticulum of \'ater, and was the first surgeon to perform the operation. We have several times successfully opened the duodenum for this purpose, and a large number of such opera- tions are now on record. Carle strongly urges incision of the duo- denum for removal of stones or growths from the duodenal end of the common duct, and cites cases in which stones were formed in the duct which later might give rise to carcinoma. Cancer higher up in the duct would necessitate union between the remaining frag- ment of the duct and the duodenum, as Halsted succeeded in doing in his case. As a palliation, cholecystenterostomy is the indicated procedure. The anastomosis may be made either between the gall- bladder and duodenum, or, if the latter is involved, with the trans- verse colon or jejunum. We have joined the gall-bladder to the transverse colon for inoperable obstruction of the common duct or chronic pancreatitis five times, and these cases did fully as well in every respect as five cases in which the duodenum was used as a receptacle for the biliary discharge. One case, supposed to be malignant, proved not to be so by living in good health six years after uniting the gall-bladder and colon. Judging from our own experience, I see no reason why the transverse colon may not serve as well as the duodenum, provided the latter, more favorable, situa- tion is not practicable. The proximity of the large bowel and the nature of its coats render anastomosis with the gall-bladder easy, and in the palliation of malignant disease it is, perhaps, almost as good for the purpose as the duodenum. In the original work of Winiwarter it was the chosen method. A few cases of enormous distention of the common duct have been reported. Robson details an instance in which he had been able to suture such a cystic formation to the surface of the body. Summers, in a most in- teresting case, united the common duct to the duodenum, with a 404 WILLIAM J. MAYO successful result. In these cases the obstruction was, however, non-malignant. The following case came under our observation: Carcinoma, Ampulla of Vater. — M. K., female, age fifty-nine, German. Admitted to St. Mary's Hospital November 1, 1900. History. — For many years patient has suffered from sudden attacks of pain arising in the epigastric and extending to the right hypochondriac region. The suffering has been severe, lasting from two to six hours, and ending with an attack of vomiting accom- panied by prostration. At times she has been somewhat jaundiced after these attacks. About one year ago her appetite began to fail, and distress in the stomach became more constant but less severe. There has been a progressive loss of weight — over 40 pounds in all. Family and personal history good. Examination. — Patient somewhat emaciated ; there was a marked cachexia with a moderate jaundice. Pulse, temperature, and respiration normal. Liver could be outlined just below the free margin of the ribs; gall-bladder could not be felt. There were tenderness and some rigidity of the muscles in this region, other- wise examination negative. Urine had a trace of albumin and much bile. Test-meal developed free hydrochloric acid, and, on distention with air, the outlines of the stomach were found to be normal. Stools contained traces of bile, but were light colored. The history was clear as to the presence of gall-stones, but the patient had a distinctly cachectic look. Diagnosis. — Either gall-stones in the common duct or malig- nant disease. Operation November 3, 1900. Incision through the right rectus muscle. Liver somewhat larger than normal; gall-bladder en- larged, containing bile mixed with ropy mucus and a single non- faceted, dark-colored stone, the size and shape of a small pea. The cystic and common ducts were moderately dilated, but no stone nor other obstruction could be detected on most careful ex- ploration. Gall-bladder drained to the surface after attaching to the parietal peritoneum. The findings were unsatisfactory, and did not account for the condition of the patient. For forty-eight hours drainage of bile was free, but gradually increased in quantity up to two or more pints a day; the skin became greatly irritated from the discharge, and examination showed that a large part, if not all, the pancreatic secretion was being discharged, with all of the bile, to the surface. Stools now contained bile. A self-retaining female catheter was inserted into the gall-bladder MALIGNANT DISEASE OF THE COMMON BILE-DUCT 405 through the fistulous opening, and in this way the drainage was directed into a receptach* without <-«)nta(t with the skin. It was evident {\n\l there was an obstruction which liad been overlookecl at tlie duodenal extremity of the duct. Patient was in a very feeble condition, and on November 'iOth was allowed to return home. Even with the continuous drainage, she imi)roved somewhat. The jaundice disappeared, and on Jan- uary '-20, 1{)01, she was readmitted to the hosjjital. On January 31, 1901, an incision 4 inches in length was made to the inner side of the fistula. The adhesions were separated, and the common duct and duodenum thoroughly exposed. At the extreme end of the connnon duct a hard body could be felt through the wall of the duodenum, the size of a filbert, and was supposed to be a stone lodged in the ampulla of ^'ater. An incision was made 2 inches in length in the anterior wall of the duodenum, exposing a grayish- white mass which was strictly localized to the site of the papilla of the common duct. Its size did not exceed the end phalanx of the forefinger. About one-third of its length projected into the free lumen of the duodenum, and two-thirds posterior to the intestinal wall. The tumor was excised, exposing the free end of the common duct. The removal was made partly with a knife and partly with the Pacpielin cautery, and finally the whole raw surface was seared with the cautery. The common duct was otherwise free from ob- struction. The incision in the duodenum was sutured. No en- larged lymphatics could be detected. A small drainage wick was inserted and the wound closed. The attachment of the gall- bladder to the skin was left undisturbed. The discharge from the fistula diminished rapidly, and in three weeks had com- pletely ceased. Stools became normal in color and the gain in weight and general appearance was rapid. For nearly a year and a half the patient remained well. About the middle of June, lOO'-i, she began to notice some pain of a boring character in the epi- gastrium, soon followed by jaundice, which slowly progressed, and on July 7th the abdomen was opened in the region of the former incisions. Gall-bladder found distended. On opening the duo- denum it was discovered that the growth had returned at the site of former operation. Enlarged lymphatics were present and also deep attachments to the pancreas. Cholecystduodenostomy was performed with the Murphy button, with recovery. The specimen was examined by Le Count, and Ids report is as follows : 406 WILLIAM J. iL^YO "The tissue is from the duodenal wall, and some sections show portions of Brunner's glands. Lieberkiihn's glands may be traced to lower depths than normal through a very inflammatory mucosa that contains a few small lymph-nodes and small areas of hemor- rhage. Certain of these glands are directly continuous vdih groups of epithelial cells that lie deeply within the mucosa and the mus- cular coats. The epithelium in these invasions are altered as follows: they lose their colunmar shape, become possessed of larger and more deeply stained nuclei, possess karyokinetic nuclei in many instances, and do not retain their characteristic grouping, being, instead, arranged in disorderly clumps and bunches that vary in size ; these deeply lying collections of epithelial cells always possess an irregular cavity that simulates a gland of the simple tubular type or a gland-duct. One must conclude that this is the tissue from a cylindrocellular carcinoma." So far as the writer can ascertain, the only case in which a carcinoma of the common duct has been excised previous to the one herewith reported was that of Wm. S. Halsted. The report of the case was pubhshed in the "Boston Medical and Surgical Journal," December 28, 1899, No. 26, vol. cxh, p. 645. A few cases have been reported since then, but the total number is small. A STUDY OF 534 OPERATIONS UPON THE GALL-BLADDER AND BILP:-PASSA(;ES, WITH TABULATED REPORT OF 547 OPER- ATED CASES* WILLIAM J. MAYO In nature's defense against infection within the abdominal cavity there are three weak situations — the Fallopian tube, the appendix, and the gall-bladder. The first to gain an accepted surgical position was the infective lesions of the tube. The ease of diagnosis and the remarkable results of operative interference contributed largely to this result. The appendix, after much dis- cussion, has also reached an assured place in surgery; but the gall- bladder has been slow to receive the attention from the medical profession which its importance deserves. The patient is usually along in years, and often, by reason of degenerative lesions or adipose tissue, a poor subject for operative interference; and, again, death does not frequently come with that tragic sudden- ness which oftentimes marks appendicitis. Like the appendix, the gall-bladder is a dependent organ with a limited outlet, connected, although less directly, with the in- testinal tract; but, fortunately, at a point in which the micro- organisms normally present are not so active. Both organs are liable to stone formation, but infections of the gall-bladder are usually due to less virulent bacteria, and its better blood-supply and distensibility equalize that tension which is so fatal a char- acteristic of the inflamed appendix. Perforation and sudden death, therefore, are less frequent in diseases of the gall-bladder. * Presented May 13, 1903, at the Sixth Triennial Congress of American Phy- sicians and Surgeons. Reprinted from "Boston Med. and Surg. Jour.," May il, 1903. 407 408 WILLIAM J. MAYO As a cause of chronic distress and disability in adult life, how- ever, diseases of the gall-bladder nearly equal in frequency those of the appendix, while in later years the gall-bladder undoubtedly takes first place. The inflammatory affections of the Fallopian tubes, which gave such an impetus to pelvic and abdominal sur- gery, have dropped to third place, with lesions of the stomach a close competitor. In St. Mary's Hospital in 1902 the relative proportion of these operations was: Appendix, 345; gall-bladder, 143; ovaries and tubes, 98; stomach, 77. The 534 operations upon the gall-bladder and bile-passages which form the basis of this report were performed upon 518 pa- tients, with 19 deaths — a mortality of 3.5 per cent. Of the total number, 510 were for gall-stone disease, with a mortality of 3 per cent. Considering stones in the gall-bladder as uncompli- cated, there were 208 cases with 2 deaths — a mortality of less than 1 per cent. On the other hand, grouping as complicated stones in the cystic duct stones in the common duct, infections with and without stones, and malignant disease, there were 326 cases with 16 deaths — a mortality of 5 per cent. This is a most significant fact, and offers a strong argument in favor of early operation. It has been said that of the 5 to 10 per cent, of adults who have gall- stones, the large majority do not have symptoms. I am convinced that many times physicians do not recognize and attribute symp- toms really present to their proper source in the gall-bladder; we too often make an indefinite diagnosis of dyspepsia or indigestion. The standard of measure in the diagnosis of gall-stones is the "colic," yet this is but a small part of the clinical picture, and is readily diagnosticated. Not so the chronic distress and gastric disturbance which the evolution of the pathologic process so often develops. In the latter case the victim goes the round of the reputable practitioners of medicine, and then, unrelieved, falls into the hands of the charlatan or patent medicine vender, until com- plications develop or a condition of encapsulation is established, subject to occasional attacks of regional inflammation. That the large majority of adults with gall-stones never suffer is true, yet these stones but "slumber," with the possibility of a OPERATIONS UPON THK C;ALL-HLAI)DP:R AM) HILE-PAR8AGE8 400 I)aiiifiil awakciiiii/^. Once ";uli\('," I lie cliaiicc of j>crmanently regaining a condilioii of "rrst" is nol good, altlioiigli the inlcrval may last for years. Uiuler normal (;onciitions tlie })il(' lias hccii supposed to be sterile; but it has been demonstrated that if col- lected in sufficiently large (juantitics, a few bacteria can be de- tected, and, as shown by Lartigau, it is possible that the necessary attenuation of the infective agents is produced by passage through the liver, and that stagnation of the infected bile in the gall- bladder is the usual cause of the "stone-building catarrh," rather than an ascending inflammation from the intestines through the ducts. A gall-bladder containing stones is always infected, and when this organ once becomes restless from the irritation of its host, the chance of a complete reconciliation is improbable. In 19 cases we have found and removed unsuspected gall-stones during an abdominal operation performed for another purpose, and in the light of the findings an inquiry properly directed demon- strated that the majority of these patients had suffered at times, but that the trouble was attributed to another cause. In some of these cases marked contraction of the gall-bladder or dense ad- hesions indicated that there had been at one time an active disease present. The truth is that there are varying degrees of gall-stone activity, of which the form characterized by "colics" is the most noticeable, for obvious reasons. Not only is the mortality in the uncomplicated cases low, but the operation, as a rule, is of the simplest character — opening the gall-bladder, removal of the stones, and attachment of the fundus of the organ to the abdominal wall, with temporary drainage of the bile to the surface. Chole- cystostomy not only effectually cures the condition, but by per- manently elevating the fundus, the cystic duct is brought to the bottom of the cavity, and gravity drainage ensues in the future. It is probable that even in these cases cholecystectomy will be the indicated operation in the near future, as this operation can be performed with almost equal safety. The "ideal" operation, that is, complete closure of the gall-bladder incision, has been success- ful in many cases of slumbering gall-stones accidentally found during an abdominal operation performed for other purposes. This 410 WILLIAM J. MAYO is a dangerous practice in active gall-stone disease, as activity means an increased infection. As long as the stones do not become lodged either in the pelvis of the gall-bladder (Hart- mann's pouch) or in the cystic duct, the usual short colic is the chief symptom, and cholecystostomy is sufficient. If the stones are impacted in either situation, the colic is replaced by a constant distress, less acute but more continuous. The gall-bladder be- comes distended, and its walls infiltrated with inflammatory prod- ucts. In such cases, as a rule, the stones can be dislodged and brought up and out through the gall-bladder, and an ordinary cholecystostomy performed; but occasionally an incision through the wall of the duct is necessary for relief of the impaction, followed by suture of the duct and cholecystostomy or ablation of the organ itself. Operation in 183 cases of stones impacted in the pelvis or in the cystic duct resulted in 6 deaths — a mortality of 3.25 per cent. This included, however, a number of cases of severe in- fections. Should the walls of the gall-bladder have undergone marked changes, or angulation and stricture of the cystic duct, resulting in mucous fistula, seem a possible outcome, cholecystec- tomy is more certain to afford permanent relief. If the cystic duct is completely obstructed, so that the gall-bladder contains no bile, it is a simple operation to detach the organ from the liver and ligate with catgut at the base; but if the gall-bladder participates in the biliary circulation in spite of the obstruction, it is not always wise to ligate the cystic duct, especially if there is a cholangitis present. Under such circumstances, if it is decided to remove the organ, the cystic duct should be left open for drainage of the bile to the surface. It is in this class of cases that we have, where possible, removed the fundus and the entire mucous membrane, the peritoneum and outer layers of the base of the gall-bladder being retained to form a pouch into which the end of a drainage- tube can be securely fastened for the purpose of safely conducting the bile to the surface. In 75 cholecystectomies, either complete or with the modifications outlined, there were but 2 deaths. This does not include 23 cholecystectomies made in the course of other operations. OPERATIONS UPON THE GALL-BLADDKU AM) HILE-FASSAGES 411 Kehr, Robson, and all surgeons of large experience in gall- stone surgery unite in declaring that tliey have not known gall- stones to reform after operation, and this has been our experience. In our early work it sometimes happened that stones would be discharged from the fistula subsequent to the operation, or occa- sionally the outer opening would cicatrize before all the calculi were discharged, and those retained gave rise to future colics, requiring secondary operations, not because the stones had reformed, but because the primary operation had been incomplete. It is the fond hope of the patient that the stones will pass down and out through the common duct, and not infrequently this is the case, but there are usually more behind. We have never operated upon a patient who has passed calculi that we have not found more in the gall-bladder. The passing of a calculus, instead of being a reason for delay, is an indication for operation, as the next " labor" may miscarry and lodge the stone in the common duct, in which situation the results of operative interference are no longer certain on account of the attendant jaundice and infection of the common and hepatic ducts. Surgery of the common duct has become a fairly safe operation, the mortality depending more upon the condition of the patient than upon any difficulties in the technic. This, of course, has some marked exceptions. Davis, both experimentally and clin- ically, called attention to the necessity of freely draining the bile to the surface after common-duct operation, and demonstrated that suture of the duct was usually unnecessary and occasionally harmful. This saved much time in a class of cases with general debility from jaundice and infections of the ducts, in which time is an element of great value. If the cystic duct is open, enabling the bile to flow outward through the gall-bladder to the surface by means of cholecystostomy, suture of the common duct may be harmless, but not otherwise. To Robson we owe great improvement in the technic of this operation, especially the sand-bag under the back at the level of the liver, to tilt the costal margin outward and upward on the prin- ( iple of a reverse Trendelenburg position. The incision of Bevan 412 WILLIAM J, ^L\YO through the abdominal wall, as modified by Robson, in dispensing with the lower angle and carrying the upper limb close to the costal margin, enables one to draw the liver downward and outward, and straightens the angle of the cystic duct at the common duct. In this way removal of common-duct stones is rendered easy. In many cases the common duct is so dilated by stones impacted in the terminal portion that the finger can be easily introduced, and by aiding the extraction with counterpressure with the other hand over the duodenum, the stone may be coaxed backward and out of the duct incision. The finger can be used in the same way in the hepatic ducts. In no other way in some cases can one be certain the ducts are free. In a few cases incision of the duodenum and direct extraction, as advised by McBurney, may be necessary. We had two such cases in our earlier experience. The "ball- valve" stone, so graphically described by Fenger, was met with 9 times. Kehr has extended the usefulness of hepatic drainage by direct tubage of the hepatic duct, and has shown us how to cure some cases heretofore considered hopeless. Out of the 534 cases, we had 58 of stones in the common duct, with 3 deaths — 5.5 per cent. The relative proportion of common-duct cases was 11 per cent, of the whole — considerably less than in either the Kehr or Robson series, who report a percentage of nearly 20 per cent. Even 11 per cent, is too high. Gall-stones should, in the large majority of instances, be diagnosticated, and the patient sent to the surgeon before common duct symptoms clinch the diagnosis and force operation upon the patient. Chronic pancreatitis, shown by enlargement of the head of the pancreas, was met with in connection with gall-stone disease 18 times. In 6 of these patients cholecystduodenostomy was per- formed and all recovered. In the remaining 12 cases no special treatment beyond the removal of the gall-stones and establishment of drainage was adopted. One case of acute pancreatitis and fat necrosis, due to a cholecystitis from one large gall-stone in the gall-bladder, recovered after operation, as did one subacute hem- orrhagic cyst of the pancreas from the same cause. It has been noted experimentally that high grades of infection OPERATIONS UPON THE GALL-Ur,.\I)I)KH AM) HILPI-PASSAGES 413 seldom cause stones, although these infections are most liaMe to occur ill j)aticnts who already suffer from cahuli. In the acutely infected cases gangrene or perforation of the gall-l)ladder may occur. In the more chronic infections of this type the symptoms are almost identical with gall-stones. In 534 cases we had 27 of this chronic character, with 3 deaths. The mortality at once demonstrates tliat the infection was more virulent than in gall-stone disease. The deaths were usually due to "hepatargia," or cessation of liver function, which the experimental work of Adami helps us to appre- ciate. These cases of chronic cholecystitis are very interesting because of the liability to overlook the condition at operation, on account of the absence of gall-stones, although the gall-bladder shows evidences of trouble. In the benign series it is to be noted that 16 cases came to secondary operation, 14 were cured by the second operation, 2 required several operations before cure resulted. The average stay in the hospital was slightly less than three weeks, the attempt being to remove all the stones at the primary operation and make the drainage more efficient by accurate placing and retention with catgut sutures, which are absorbed before it is necessary to remove the drains. This reduces the quantity of packing to a minimum and enables better closure of the wound without marked hernia liability. The sutures closing the abdominal incision are left in place two weeks, which makes it quite safe to allow the patient to get about early — an important factor in the recovery of old people. If the patient is cholemic, we give chlorid of calcium as a prophy- lactic against hemorrhage. Otherwise we have not found a special preparatory treatment to be of value beyond the ordinary prepa- ration for laparotomy. The after-care is very simple. If possible, the bile is conducted into a bottle, and the drainage packing is not disturbed until the end of the first week, and after removal of the drains, repacking is rarely necessary. In severe cases bile drain- age to the surface is essential, and for this purpose the cystic duct, if unobstructed, offers an easy and safe avenue of escape from the hepatic ducts. In I>ut few cases have we found direct incision 414 WILLIAM J. MAYO and tubage of the common duct, as recommended by Kehr, ad- vantageous. Malignant disease of the gall-bladder and bile-ducts was met with 24 times, or 4 per cent, of the total number of cases. The proportion was about 4 times in the gall-bladder to once in the ducts. In some cases the exact origin could not be determined by a reasonable exploration. Five times the gall-bladder was removed for cancer. In 2 of these cases a considerable portion of the liver was also excised. Three patients are alive, one nearly two years, but, unfortunately, with recent recurrence and ob- structive jaundice. In one case a carcinoma involving the ter- minal end of the common duct was excised by the duodenal route; recurrence after one and one-half years necessitated cholecyst- duodenostomy. As the presence of gall-stones occurs in only 15 per cent, of secondary cancers of the gall-bladder and in over 90 per cent, of primary cancers, we must conclude that they are the chief etio- logic factors in the production of mahgnant disease of this organ. In nearly all our cases gall-stones were present, and a clear history of active trouble could be elicited, although there may have been years of quiet between the cohcs and the development of the cancer. While the possibility of cancer could not be of itseK considered an indication for the removal of gall-stones, it is worth considering, as about 1 case in 25 coming to the operating-table in this series had malignant disease. The usual precancerous condition would seem to be a gall- bladder with thickened walls, due to chronic infection and calculi. It is just this class of gall-bladders which are now being subjected to excision rather than drainage, and it is probable that in the future many cases of cancer will be removed in an early stage with permanent cure. Palliative operations for malignant disease of the gall-bladder and bile-ducts are not of great service. In a few cases of common- duct obstructions cholecystenterostomy gives relief for a time, and if the duodenum cannot be easily reached, the transverse colon serves as well for the purpose of anastomosis. OPERATIONS UPON TIIK OALL-ULADDKIl AM) MILK- PASSAGES 415 In 5 cholecystenterostomies for m.-ili^Mumt disease 4 were attached to the transverse colon, with 1 death, and in 1 to the duodenum. In non-niah'gnant disease we attach to the duodenum, if possible, the Murphy button being used in all. A statistical table is attached herewith: A TABLE OF 547 OPERATIONS UPON THE GALL-BLADDER AND BILE- PASSAGES (MORE THAN ONE OPERATION PERFORMED AT ONE TIME, ONLY THE MAJOR IS TABULATED;. OCCLRRJNC; IN ST. MARY'S HOSPITAL OF R0CHESTE1{, MINN., FROM JUNE !24, 1891, TO MAY 13, 1903 St. Mary's Hospital reports, 527; Minnesota State Hospital and private practice, 20. Operations fob Benign Disease Total. Recov'd. Died. Cholecystostomy: stones in gall-bladder, cystic duct, or both' 299 296 8 Cholecystostomy: polypus in gall-bladder 1 1 Cholecystostomy: gall-bladder stone with acute pancreatitis and fat necrosis 1 1 Cholecystostomy: cholecystitis with and without stones 51 46 5 Choledochotomy: stones in common duct 59 56 S Cholecystectomy: gall-stone disease 56 55 1 Cholecystectomy: cholecystitis 9 8 1 Cholecystectomy: cyst of gall-bladder containing 10 quarts, supposed to be ovarian 1 1 Cholecystenterostomy: chronic pancreatitis and jaundice, 4 with gall-stones, 1 without 5 5 Perforation of calculus, al)scess and general peri- tonitis 2 . . 2 Division of adhesions 16 16 Duodenocholedochotomy: stone in ampulla of Vater 1 1 Exploratory: negative 21 21 522 507 15 Operations for Malign.^nt Dise.\se Total. Recov'd. Died. Cholecystectomy 4 3 1 Cholecystostomy: obstruction common duct. .. . 5 3 2 Cholecystectomy and partial hejiatcctomy: can- cer of the gall-bladder 1 1 Duodcnocholcdocliotomv: cancer in ampulla of Vater ' _ 1 1 Cholecystenterostomy: malignant obstruction of common duct 4 3 1 Exploratory: inoperable cancer 10 Malignant, total 25 11 4 May 13, 1903 547 518 19 SOME CAUSES OF FAILURE OF OPERATION TO CURE GALL-STONE DISEASE* WILLIAM J. MAYO Surgery of the gall-bladder and bile-passages is one of the most satisfactory branches of our art. The relief following operation is perfect and rapid, leaving little to be desired. The death-rate, taking the cases as they come, is hardly more than 3 or 4 per cent., and in uncompKcated cases less than 1 per cent., depending to a large extent on the condition of the patient. Including all causes of failure to cure, either complete or partial, and such late sequelae as adhesions and hernia, the number of instances is small. In 580 operations upon the gall-bladder and bile-passages we had but 17 cases, or 3 per cent., which required a secondary operation. Dur- ing this period, however, we have, on a number of occasions, oper- ated a second time for symptoms arising after an operation per- formed elsewhere. It is fair, therefore, to presume that some of our cases have, unknown to us, been operated upon at a later period by other surgeons, and that failures to establish a complete cure have been more numerous than this percentage would indi- cate. It must be taken into consideration also that many pa- tients have symptoms referable to uncured lesions which are not sufficiently^ serious to demand operation, and these may be ac- counted as partial or temporary failure ; but looked at even in this light, gall-stone surgery is wonderfully successful. Practically all the patients are benefited, and few would exchange their post- operative for their previous condition. Poor results usually occurred in our earlier work, and meeting *Read at the Thirty-fifth Annual Meeting of the Minnesota State Medical Society at St. Paul, June 17, 1903. Reprinted from the "St. Paul Medical Journal," August, 1903. 416 FAILURE OF OPERATION TO TURE OALL-STONE DISEASE 417 with such cases has gradually enabled us to overcome the causes which lead to the suh.scc|uerit tn)ul)k'.s. Of course, in rare in- stances the condition of the patient may not warrant a comjjlete procedure at one operation, and a second operation is deliberately elected. With a single exception, to he referred to later, all the cases in which results were less perfect than was desirable occurred in complicated cases, and it can be laid down as an axiom that de- lay in seeking surgical relief was the direct cause of the complica- tions. It is the experience of surgeons generally that compli- cated cases have usually had symptoms long enough to have made a diagnosis possible before the development of serious lesions, and that an operation at that time would have been safer and cure more certain. I would call attention to the clinical fact that a small number of patients who have had a cholecystostomy performed will have a colic or two following operation, and sometimes the colic accom- panied by transient jaundice. We have observed this most often during the first month or two after discharge from the hospital. In the large majority of instances the colics do not recur and the patient remains well. The temporary trouble is probably due to a crippled gall-bladder becoming filled, and by reason of recent ad- hesions not emptying properly, so that a single spell of pain shortly after closure of the fistula does not indicate a second operation unless there are other evidences of trouble. The most common cause of later symptoms is incomplete re- moval of stones. Tait advocated cholecystostomy and drainage based upon the frequency of overlooking stones, as it enabled spontaneous discharge. In one case in our early experience 55 calculi worked out of the fistula during the first two weeks. There is, however, little excuse for leaving stones in the gall-bladder, as by using the finger as a guide, even a small calculus will rarely be overlooked. Stones in the cystic duct frequently escape attention, and it was only after several such misfortunes that we began to exercise greater care in exploring the cystic duct. The parts are deeply situated, and as these patients are often obese, it was not easy to VOL. I — 27 418 WILLIAM J. MAYO locate such a calculus previous to the development of the Robson technic, that is, the sand-bag under the back, the high incision, and dislocation of the liver downward and outward, which exposes the cystic and common ducts perfectly. In most of these cases cholecystectomy is indicated. If the stone completely obstructs the cystic duct, the duct and cystic vessels are caught with curved forceps just beneath the impacted stone. The duct is then cut across and the gall-bladder and duct with the stone quickly re- moved from below upward, almost by traction alone, with an occasional division of some firm adhesion to the liver. Twice we have reoperated upon cases in which the gall-bladder had been removed distal to the stone, leaving it in the duct to cause future trouble. This is more apt to be the case when the gall-bladder is dissected out from above downward. The deep field is obscured by the blood running downward, and the same vessels are cut over and over again. Stones are often overlooked in the common duct, as they may lie quiescent for years. The jaundice may be very slight, and in some cases not noticeable. The gall-bladder in the mean time may become obstructed at the cystic duct, so that this organ may be enlarged and cystic, with calculus at the neck and nothing to call attention to the common-duct stone. This is so contrary to the usual condition of contracted gall-bladder and open cystic duct in common-duct stone as to lead to error. The ducts should be explored with the fingers in every case before open- ing the gall-bladder. After opening the gall-bladder the relief of tension prevents moving the stone in the dilated duct, and escape of the cystic contents is apt to soil the field. If the gall-bladder is distended, it is well to explore a second time after tapping, but before opening the gall-bladder with the attendant possibility of infecting the deep parts. If stones are found in the common duct, it will usually be suflSciently dilated to introduce the finger into the duct for exploration. In many cases in no other way can we be sure the common and hepatic ducts are clear. One source of failure of cholecystostomy to cure is from secon- dary obstruction of the cystic duct preventing free drainage of the gall-bladder down through the passages. This may eventuate in a mucous fistula or repeated attacks of colics as the gall-bladder FAILURE OF OPERATION TO CURE GALL-STONE DISEASE 411) secretions are periodically forced through the stricturcd duct. In some cases the gall-hhidder will distend and finally rupture through the scar, discharging bile and mucus. In practically all these cases the cystic duct has been obstructed by stone, causing ulceration, the healing of which induces a stricture, or kinking of the channel may occur. Other things being equal, it is better to excise the gall-bladder in all cases in which the cystic duct is in- volved. In this way we have of late eliminated the most common cause of secondary trouble. In septic cases drainage is necessary for a long period of time, and if the fistula be allowed to close too quickly, severe symptoms may ensue. On two occasions we have had to reopen and re- establish drainage in septic cholecystitis; both cases were colon infections. This is less liable to happen in ordinary empyema of the gall-bladder, in which a fistula will usually remain until ster- ilization has been accomplished by natural processes. In colon infections tubage should be continued until the bile becomes sterile. Cancer can also be said to be secondary to stone formation, and may take place after cholecystostomy or, being present, may be mistaken for inflammatory disease. All thick-walled gall- bladders should be looked upon as suspicious, and as they are functionally useless, cholecystectomy should be done rather than cholecystostomy; in this way many early cancers will be cured. Chronic pancreatitis may exist at the time of operation, and to obtain a good result drainage should be long continued. We have twice allowed the fistula after cholecystostomy to heal too quickly. The secondary symptoms were marked by attacks of slight jaundice, occasionally fever and chills, and rather persistent stomach trouble. At the second operation the only cause found for the condition lay in a chronic pancreatitis, and both cases were cured by cholecystenterostomy. We should examine every case as to the condition of the pancreas at the time we explore the ducts, and if this disease is present, either drain the gall-bladder for a long time or do a cholecystenterostomy with the Murphy button in addition to removing the stones. In the beginning of this paper I spoke of having seen only one uncomplicated case which re- 420 WILLIAM J. MAYO quired a secondary operation. This was a cholecystostomy with a very large gall-bladder; the stones were easily removed and the ducts were free. For weeks after operation bile escaped occasion- ally from the fistula, not much, but troublesome. On dissecting out the gall-bladder it was found that by a low attachment to the abdominal incision this viscus had formed a channel along which, in certain positions of the body, bile would gravitate outward. After cholecystostomy the gall-bladder should be attached as high up in the incision as is convenient. Persistent biliary fistula us- ually means obstruction of the common duct. I take it that every one understands the importance of not attaching the gall-bladder to the skin. In the early days persistent bile fistula was usually due to a mucocutaneous suture, the evils of which obsolete prac- tice I do not need to point out. Turning in the margins of the incision in the gall-bladder and drawing a purse-string suture closely about the drainage-tube in a similar manner to a Kader gastrostomy enables healing of the fis- tula to take place promptly. There are two rather common causes of failure to effect a perfect cure which can not always be avoided. Postoperative adhesions are liable to cripple the movements of the viscera in this neighborhood. Adhesions to the stomach and duo- denum are the most annoying. Secondary separation may be necessary, with the use of Cargile membrane. As a rule, like the pain of old pleuritic adhesions, in time relief comes as bands stretch out. We make it a rule not to allow gauze drains to come in con- tact with the stomach and duodenum on account of the develop- ment of adhesions. We always interpose a piece of rubber tissue and leave it from six to eight days, until the adhesive film surround- ing the drains becomes organized. Hernia following operations for gall-stone disease is not usually troublesome, but long incisions in obese people may give rise to serious hernial protrusion requiring secondary operation. If practicable, in such cases we now make a second opening outside the working incision, close to the ribs. Through this the drainage material may be brought out and en- able us carefully to close the full length of the original incision. The necessity for this is increased if the opening has been extended downward for the purpose of removing the appendix. SOME OBSERVATIONS ON THE SURGERY OF THE COMMON DUCT OF THE LIVER* WILLIAM J. MAYO That the gall-bladder is infected through the bile in the large majority of cases, and not by way of the common duct and duo- denum, is at least an interesting theory. Undoubtedly this latter route of bacterial invasion takes place, especially in the more serious grades of infections, and often without stones. The early experiments, which showed the normal bile to be sterile, have been proved untrue, and we must look upon the bili- ary secretion as containing a few bacteria which can be detected if a sufficiently large quantity be examined. In the formation of stones the necessary attenuation of the infective agents is obtained by passage through the liver under conditions which have proved most difficult to duplicate in an artificial manner, hence the few examples of true stone formation which have followed experi- mentation. The result is usually a more severe and non-calculous form of infection. A gall-bladder once infected remains infected, needing only a disturbing element to reproduce its original intensity. Given an infected gall-bladder containing stones which cause mechanical irritation, and we have the most favorable conditions for frequent bacterial excursions into the common duct, and should this be accompanied by calculi which fail to pass the duodenal orifice, we have all the conditions favorable for the development of cholangi- tis, pancreatitis, etc. At least 90 per cent, of all the diseases of the common duct upon which we are called to operate arise in this way. *Rcprinted from the "Medical Record," April 30, 1904. 4:il 422 WILLIAM J. MAYO The primary importance of the gall-bladder must be recognized in the etiology of these infections, and operation should be per- formed at a time when the infection and stones are still in their originating viscus. The cystic duct and its patency are most important elements in the problem, and the question of treatment is to a large extent based upon its condition. If the cystic duct is closed mechanic- ally, either by stone, kink, or stricture, a diseased gall-bladder results, which is better out than in. The infection has confined its ravages to the organ itself, and we find a cystic gall-bladder with impacted stones, thick walls, and, above all, with an obstructed cystic duct. Such organs are functionless, and if drained and left, they will be a possible source of future troubles, among which may be mentioned reinfection, mucous fistula, cancer, or colics due to failure to drain its secretions down through the common duct. The keynote to the diagnosis of this condition is found in the fact that the gall-bladder contains no bile. Not all cystic gall-bladders require removal, as a large stone in the pelvis of the gall-bladder may block without injury to the duct — after removal of the stone, bile at once appears. In these cases, if there is no evidence of a cholangitis, there is no need of bile drainage to the surface, and the duct can be per- manently occluded by ligation. Cholecystectomy is easily per- formed from below upward; the cystic duct and vessels are caught with forceps; a second forceps on the gall-bladder side prevents leakage and enables division of the cystic duct and vessels. The gall-bladder can be readily separated from below upward without hemorrhage or trouble. When the dissection has progressed a short distance, the forceps on the deep end of the duct and cystic vessels are removed after ligation with catgut, to which latter a light gauze drain is attached to prevent displacement. This guard against leakage should always be taken. The above method of removal enables careful ligation of the deep parts and bloodless separation from the liver, whereas if the gall-bladder is removed from above downward, the same vessels are cut over and over again, obscuring the deep field with blood. SURGERY OF COMMON DUCT OF THK LIVER 423 If the gall-bladder contains hile, we have an entirely different condition to deal with. Free communication through an open cystic duct means a possible infeetion of the common duct, and if the infection is at all marked, provision must be made for bile drainage to the surface. This can be accomplished best by chole- cystostomy and draining through natural channels. Should it be deemed wise to remove the gall-bladder in such cases, we have practised in many instances amputation of the fundus, retaining the base to form a pocket from which the mucous membrane is removed, and into which the drainage-tube can be securely fastened to the cut end of the cystic duct. This gives temporary bile drainage, with the permanent benefits of a cholecystectomy. The foregoing procedure is also very useful in some cases of cholecystectomy in which deep adhesions render the typical opera- tion fraught with peril to the liver, duodenum, and deep vessels. There have been some purely theoretic objections advanced to this device, and it is not ideal; but the fact remains that in a very large number of cases it has proved both safe and curative. Many excellent surgeons believe that the gall-bladder should be removed in every case. In simple cases of stones with latent infections and the gall-bladder in good condition, no harm results from ligation of the cystic duct, even if patent, which is shown by the presence of bile; but in such cases cholecystostomy gives per- manent results, which fully equal cholecystectomy, and has the advantage of less manipulation. It also leaves the gall-bladder, and while it is only a vestigial remnant and probably of no value, we are by no means sure that it should be sacrificed unnecessarily. Time, however, may prove the radical course the proper one to pursue in all cases. In our experience, cholecystectomy has had a rapidly growing field of usefulness. We hear a great deal about hepaticus drainage — that is, open- ing the common duct and introduction of a tube drain into the hepatic duct. We should not forget that the cystic duct is an offshoot of the common duct, and, if patent, furnishes natural drainage to the surface. In many cases, however, direct tube drainage from the main ducts is life-saving. Bile drainage through 424 WILLIAM J. MAYO an open cystic duct amounts to identically the same thing, differ- ing only in degree, and the question to be decided is when it will not be sufficient. We have usually found it efficient, and unless the common duct has been opened for the removal of stones, nat- ural drainage by way of the cystic duct and gall-bladder furnishes results which, when compared with indiscriminate opening and tubage of the common and hepatic ducts, proves the former method both safe and easy, and greatly lessens the disability. In a few cases the cystic duct has not been sufficient for this purpose. We have then removed the gall-bladder and split the cystic duct down into the common duct and drained the deep ducts. Stones in the common and hepatic ducts require removal and drainage. In a considerable number of cases stones in the deep ducts may be latent, giving no symptoms, and especially without jaundice. Therefore it behooves us to examine carefully the ducts for stones in all cases, even if we have no special reasons to antici- pate their presence beyond the fact that the gall-bladder contains calculi. For all operations on the ducts Robson's technic should be em- ployed : the sand-bag under the back, longitudinal incision of the right rectus muscle, with division of its inner half at the curved margin of the costal arch up to the ensiform cartilage (Bevan). In two of our cases, at the extreme upper limit of this incision the anterior mediastinum was opened, allowing air to suck back and forth. This incident is rather startling, and makes one think the pleura may have been accidentally injured. A few catgut sutures close the rent and no harm results. If the common duct contains stones, one of them is seized be- tween the left forefinger and thumb, and, using the stone as a guide (Elliot), two lateral mattress sutures are placed, leaving a free space between for longitudinal incision of the duct in its visible portion between the cystic duct entrance and the duo- denum. These two threads with long ends act as tractors, and after removal of stones may be crossed to unite the duct margins, always leaving a little chance for drainage at the ends of the in- cision. In the majority of cases, unless the cystic duct is freely Fig. 41. — Cholecystectomy, showing clamps applied to cystic duct and cystic vessels. ^iC^ Fig. 42. — Cholecystectomy, showing cystic duct and vessels ligated. Gall-bladder paitially separated and sutured in position to cover tie exposed liver substance. SURGERY OF COMMON DUCT OF TIIK LIVER 4^3 open for drainage through a cholecystostomy, it is Ixllcr practice to leave the incision open, either completely or in part, and use the threads to fasten the drains in position. Fixing the drains prevents floating hy biliary escape, or tlerangenient hy diaphragmatic action exerted upon the liver. The gauze is surrounded by rub- ber tissue on its inner and lower surface, to prevent peritoneal a3 l.(t()k(>(l ;il from I lie slaiidjioiiiL of iiioii;ilily, (•liolc<-y>lo.stoiiiy is tlic safcsl, and should 1k« considered tlic iioiiual, operation for the averagecil.se. As we lia()0 o|)(>ralions in wliicli r, this cannot he taken as a valid objection to leaving it in sitii. There are some conditions in which, after cholecystostomy, future trouble may be expected. First, in all those cases in which the cystic duct is obstructed by a stone and the gall-bladder takes no part in the biliary circulation (contains no bile), other things being equal, it should l)e removed, as in this condition we lia\e oc- casionally had to remove it secondarily for the relief of mucous fistula or colics due to obstructions to drainage from kinking or stricture. Second, thick-walled gall-bladders which have become functionless lead to a suspicion of malignant disease and should be excised. Several times we have unexpectedly removed what proved to be an early carcinoma of the gall-bladder. One such patient is now alive — more than three years. In connection with common-duct surgery it is not wise to re- move a functionating gall-bladder unless for direct indication. This is particularly true if cholangitis exists, as common-duct cases more often require a secondary operation than any other, and the gall-bladder not only affords easy drainage and enables cholecyst- enterostomy should there be future contraction and obstruction of the common duct, but it is also a safe guide to the deep ducts if future trouble should arise. As to permanency of cure: Patients upon whom we have per- formed cholecystostomies have remained well, expect in a few instances of bad selection in our early experience in which cholecys- tectomy would have been the better operation. The operative disability after cholecystostomy was brief. A short incision witli separation of the fibers of the rectus muscle rendered early union without liernia almost a certainty. By turn- ing in the cut margins of the gall-lihulder about the tube (Summers) in a similar manner to the Stamm-Kader gastrostomy the bile dis- charge stopped promptly, since on removal of the tube at the end 454 ^TLLIA^SI J. MAYO of the week the peritoneal surfaces agglutinated. The average patient was up in twelve days and left the hospital within two weeks. Cholectstectoimt There was a total of 319 cholecystectomies, with a mortality of 3.13 per cent. In the cholecystectomies in the last series of 500 cases the mortahty was 1.62 per cent. Cholecystectomy has an increasing field of usefulness, but its increase of mortality, although sHght, is for one reason or another fairly certain, and prevents it from replacing cholecystostomy. At the same time, where the circumstances permit easy removal of the gall-bladder and the disease is confined entirely to this organ, it is the operation we most commonly perform even in cases in which cholecystostomy would answer the purpose. But if the patient is very obese and the gall-bladder has a broad attachment to the Kver, necessitating prolongation of the incision or increased manipulation, cholecystectomy is the more difficult and dangerous operation. The permanence of cure after cholecystectomy is, of course, absolute when the disease is confined to the gall-bladder. In the majority of cases the incision was made nearly if not quite as short as for cholecystostomy. The period of convalescence was, therefore, about the same. In a few cases a longer incision was re- quired, adding several days to the disability. Verj^ rarely was a patient in the hospital for more than fourteen days. Operations upon the Cosevion Duct — 207 Cases The operations upon the common duct, so far as the mortalit}^ is concerned, can be di^'ided into four groups, although this ar- rangement is more or less artificial, since some of the cases are hard to classify. Group 1: 105 patients, with 3 deaths, — 2.9 per cent., — con- sisting of those patients in whom gall-stones were present in the common duct, but without immediate active symptoms. Jaun- dice was moderate or not present. If it was present, the obstruc- tion was incomplete or intermittent and permitted the escape of a 1500 OPERATIONS ON fJALL-ULADUKIl AND I5ILE-PASSAGES 455 certain umouiiL of bile into llie intestine. There was comparatively little infection of the ducts, and except for the presence of mucus, the bile' was normal. The operation under such circumstances was simple and the eonvaleseenoe short, the patients usually being able to leave the liospital within fifteen days. The cures have been permanent. Group 2: 61 patients, with 10 deaths — 10 per cent. A series of cases in which there was active infection not only in the common duct, but also involving the ducts of the liver. Stones were usually present. The patients not only had jaundice, but suffered from Charcot's fever (malarial type, irregular chills, followed by a temperature of from 103° to 107° F., passing off in a few hours with sweating), pain intermittent and most marked just previous to the active symptoms; during the remissions a little bile passed the obstruction, relieving the liver. Among the older writers this was called "remittent bilious fever." The added infection at once introduced an element of grave danger, not only from the opera- tion, but from the production of certain complications which caused death within two months. It is in this group of cases that hepatic duct stones may form ; we have seen 7 examples of this condition. The infection and interfer- ence with drainage from a stone formed in the gall-bladder, but which had passed into and become lodged in the common duct, furnishes the necessary conditions for their formation. The cholangitis may subside and the stones reach a more or less cjuiescent state, but after removing the calculi from the common duct others which have formed in the hepatic ducts may pass into the common duct, to cause future trouble. Coincident enlargements in the head of the pancreas or changes in the duct-wall may lead to secondary stone formation. In four instances under such circumstances we have seen stones reform in the common duct requiring second operations after periods of from one to five years. In two the gall-bladder had been removed at the primary operation, and the stones were too large to come down from the hepatic ducts. The possibility that these stones had as their nuclei hepatic duct calculi cannot be denied in one case, but 456 WILLIAM J. MAYO it does not seem possible that this was true in the other three. In this group of cases inflammatory diseases of the pancreas were often found associated. As a rule, these patients were in the hospital from three to four weeks. Group 3: Complete obstruction of the common duct; 29 cases and 10 deaths; 34 per cent. It is hardly necessary to call attention to the fact that formation of bile is only one of the functions of the liver, and that a patient may live for a great length of time with nearly if not quite complete obstruction of the common duct, the necessary amount of bile being absorbed by the blood and elim- inated with the urine, perspiration, etc. In Group 1 we found the bile comparatively healthy, containing only a moderate amount of mucus. In Group 2 the bile was darker, containing a large amount of mucus and often showing colon bacillus on culture. The third group showed almost no bile in the ducts, and the little present was thin and of a dark, spinach-green color, or in the worst cases a condition of complete acholia was manifest, the ducts being filled with a clear, colorless, mucoid secretion. The patients' general condition was extremely poor, pulse feeble and rapid, and in the long-standing cases edema of the feet and free, bile-stained fluid in the peritoneal cavity. Albumin and casts in the urine and other evidences of extreme toxemia were usually manifest. The operative mortality in this group during the period of complete obstruction was very high — 34 per cent. This included deaths from early and late complications. Acute obstructions of this type, when accompanied by evidences of infection, were es- pecially fatal, and since acute obstruction from stone is seldom permanent, it is often wise to wait for a period of remission before operation. It frequently happens that the duct will dilate suffi- ciently in the early stages to permit some relief of the symptoms, and this is the time to interfere. Later the inflammatory products in the duct-wall may contract down upon the stone, giving rise to permanent obstruction. In a few instances of complete obstruc- tion which came on suddenly and which remained without tem- porary remission of symptoms, spontaneous cure by sloughing of 1500 OPERATIONS OX (JAr-L-HLADDKIt AND UILP>PAS.SAGES 457 the stone into the intestine took jilace. We have seen 4 such ex- amples. In each, after years of typical gall-stone symptoms, there was sudden and complete obstructive jaundice. In 2 there was a steady temperature, and in all 4 there was a {)eculiar rigidity of the upper abdomen. After from six to twelve weeks of acute and severe symptoms the patient suddenly became relieved, the jaun- dice disappeared, and a large gall-stone was found in the stool. Three of these patients were subjected to operation subsequently. In all, one or more stones were found in the gall-l)ladder or in the adjacent liver border, the center of a cicatricial mass, but without communication with the ])ile-tract, the common duct being densely adherent to the duodenum at the site of perforation. The most common causes of death after operation in this group were exhaustion from cholemia, with or without capillar^' hemor- rhage, and sudden cessation of the liver function. All the patients who recovered remained well. The hospital disability averaged a little over three weeks. Group 4-" This group concerned malignant disease — 12 cases, 4 deaths — 333^ per cent, mortality. Cancer of or involving the common duct occurs in two forms: First, the primary tumor of the common duct or papillie, a small, hard, grayish-white mass, with a tendency to remain localized until a late stage. We have seen several examples and have had two primarily successful excisions, but none of the patients lived beyond three years. Second, com- mon-duct obstructions from carcinoma extending downward from the gall-bladder and cystic duct, or from cancer of the head of the pancreas. These cases are, of course, inoperable, and even an exploration proved fatal in several instances. Relation to Pancreatitis One of the most interesting problems in connection with surgery of the bile-tract concerns coincident inflammations of the pancreas. In a total of 86 out of the 1500 cases the pancreas was involved to an extent noticeable on examination. Four of these cases were acute, 2 of wliich recovered and 2 died. Six were subacute; 2 of these had hemorrhagic cysts; 5 recovered and 1 died; 9 cancer, 5 458 "U'lLLIAM J. MAYO deaths; 67 had chronic pancreatitis; the evidences usually con- sisted of hard nodules, most marked in the head of the pancreas and near the common duct. Four cases, supposed to be common- duct obstruction from chronic pancreatitis alone, were shown by subsequent operation to have had an undiscovered stone in the ampulla. In a few cases the pancreatic disease apparently was not secondary to the bile-tract. That the acute forms had a deleterious effect upon the patient is unquestioned, but I have been unable to separate the harm done by the chronic inflammations from the essential condition in the bile-tract, and I do not believe that unless it was obstructive it had a decided influence on the prognosis. In summing up the causes of the 66 deaths, 10, or 15 percent., were accidental and could be eliminated. The largest number were due to cessation of liver function, usually the result of infections. Microscopic examination showing destruction of the epithelial elements of the liver and often fatty degeneration. Next came ex- haustion from blood changes due to chronic cholemia. The mortality and the complications of delay placed the early operation for appendicitis on a sound surgical footing. To remove the disease while still in the appendix and before its rupture in- volved the abdominal cavity was the logical conclusion. The same reasons apply, and with equal force, to the early operation for gaU-stone disease. Remove the disease while it is still in the gall-bladder, and a mortality of from 1.47 per cent, (cholecystostomy) to 1.62 per cent, (cholecystectomy) is the result. This includes death from accidental causes, acute perforation, and gross infections. Excluding these cases, a mortality of less than 1 per cent, can be shown. With the passage of the stone into the common duct we no longer have a localized disease, but one fraught wdth grave dangers from liver infection and cholemia, and in this condition nearly 1 in 7 of our cases came to operation, while 1 in 25 developed malignant disease of the gall-bladder or bile-tract, and in most of these cases gall-stones were present. In other words, 1 patient in 6 had al- lowed the favorable time for operation to go by, although the very large majority had ample warning in the early and safe stage. PANCREAS CASE OF ACUTE PANCREATITIS WITH FAT NECROSIS— OPERATION: RECOVERY* WILLIAM J. MAYO J. C, male, aged fifty-nine, American, was admitted to St. Mary's Hospital on June 11, 1901, with the following history: For two years he had sufifered from attacks of indigestion and at times had refrained from eating for twenty-four hours or more in order to obtain relief, ^yhile able to attend to his professional duties during this time he lost '25 pounds in weight. On June -ith he was suddenly seized with agonizing pain in the epigastrium, accompanied by marked symptoms of collapse, re- cjuiring anodynes and vigorous stimulation. Vomiting and retch- ing were frequent. The abdomen became greatly distended, and symptoms of acute obstruction of the bowels developed. June 5th and 6th the condition remained about the same, temperature ranging from 100° to 10^2° F., pulse from 06 to 120. Hiccup be- came most distressing. The bowels acted slightly as a result of purgation and enemata, but wathout relief to the abdominal distention. Beginning on June 6th the stomach was washed out and rectal feeding instituted, with some relief to the acuteness of the symptoms, although the main features were practically unchanged. On this date an in- definite tumor of irregular outline could be detected under the right rectus above the umbilicus, in the region of the gall-bladder. On June 7th and 8th his condition was practically unchanged, the hiccup and extreme nervous unrest being most marked. The physical examination on admission to the hospital re- vealed the following: A large, heavy man, of splendid physique, vital organs in good condition except for a trace of albumin in the urine. The ab- domen was verj' tympanitic. To the right of tlie umlnlicus, and above it, was an irregular, indefinite mass, apparently the size of a *Reprinted from "Jour. Amer. Med. Assoc," January 11, 190^. 401 462 WILLIAM J. MAYO large fist. The temperature was 101° to 102° F., pulse, 120, and of poor quality. He was very restless, hiccuping at intervals, and having every appearance of extreme illness. There was slight jaundice. Diagnosis: Gangrenous cholecystitis with probable perforation. The patient was operated on at once. The abdomen was opened through the right upper rectus, coming directly upon a greatly thickened and adherent omentum, which was infiltrated with little white or brownish spots from the size of a hempseed to that of a pea or larger. On loosening the adhesions some bloody fluid es- caped from the peritoneal cavity. It was now noticed that the mesentery was infiltrated in a similar manner. The peritoneum, while reddened, was unaffected. The diagnosis of fat necrosis was e\adent. Raising the omentum and transverse colon, the greatly enlarged pancreas could be felt like a pudding in a tight sac. With a small aspirating needle this was aspirated in several places, withdrawang only bloody fluid. The rectus was severed laterally and a search instituted for the gall-bladder, which was found far to the right and wholly unconnected with the tumor previously detected. The gall- bladder was greatly thickened and contained one enormous stone, the size of a small hen's egg, also some mucopurulent material. The stone was removed, and a large rubber drain inserted and sutured to the opening in the gall-bladder with a catgut suture, in purse-string fashion, drawn tightly to prevent leakage. The drain was brought out of a stab wound in the right groin. A large wick of gauze was placed along with the tube into the right kidney pouch. The whole of the anterior wound was closed. Time of operation was forty-five minutes. The patient was placed in bed in extreme shock, with restless- ness, muscle twitching, cold perspiration, etc. Temporary delir- ium developed after the anesthesia had passed away. Atropin, as advised by Crile in this form of shock, was found most eflaca- cious. Strychnin, rectal exhibition of saline solutions, etc., were also resorted to. At the end of eighteen hours an immense drain- age through the rubber gall-bladder drain commenced. It was a bloody, serous fluid, -udth little evidence of bile. This discharge was very irritating, and on examination showed pancreatic fluid and bile. The quantity was so great as to saturate a large dressing every four hours. In two weeks this irritating discharge was gradually replaced by bile of a more normal appearance, and at the end of four weeks the fistula closed. The patient, while in a most CASE OF ACUTE PANCREATITIS WITH FAT NECROSIS 463 critical condition for a week, slowly regained his health, leaving the hospital in seven weeks, and he is now in perfect health, up to his usual weight, and can eat and digest normally. It is evident in this case that the gall-stone had been the cause of a cholangitis which extended to the pancreatic ducts, with resultant acute pancreatitis and fat necrosis. The free drainage and relief of tension following the opening of the gall-bladder checked the process short of abscess formation. The fine constitution and previous exemplary habits of the patient were great factors in the recovery. PANCREATIC CYST* WILLIAM J. MAYO The pathology of this rare affection has advanced but little since 1885, when Senn ^^Tote his classic paper upon the "Surgery of the Pancreas." However, workers in many fields are reporting cases and making comments upon the prominent features which characterize the disease, and out of the growing material at our disposal more exact pancreatic surgery will evolve. The following is a report of a typical case of pancreatic cyst: Mrs. J. S., aged twenty-eight years; married; mother of one child, aged eighteen months. Admitted to St. Mary's Hospital September 30, 1893, with the following history: One month after the birth of her child she first noticed a tumor just to the left of the midline, between the umbilicus and the mar- gin of the left costal cartiLage. The enlargement gradually in- creased in si^e, gi^'ing rise to a feehng of pressure and weight; during the past four months it has grown rapidly, is more painful, and cohcs or cehac neuralgias are of frequent occurrence. There is marked digestive disturbance, partly due, no doubt, to the fact that she is two and one-half months pregnant at the present time. Her general health is faihng, and she has lost 15 pounds in weight. Physical examination reveals a smooth, rounded, somewhat elastic tumor in the left hypochondriac region, the size of a child's head, slightly movable in a direction toward the umbilicus. Tumor of the spleen was easily excluded, as there was an intestinal percussion-note to be obtained anterior and exter- nal to the enlargement, showing it to be retroperitoneal in origin. Renal cyst from the upper part of left kidney was ruled out with more difficulty. The urine was normal, and there was no indica- tion of urinary disturbance whatever. Insufflation of the colon ♦Reprinted from "Medical Record," February 10, 1894, pp. 168, 169. 464 PANCREATIC CYST 465 witli air by the rectum sliowcd tliat the (Icsccndin^ colon passed external to the tumor, hetwecn it and the h)in, and that the trans- verse colon lay just below it, while the stonuich percussion-note was immediately above. A diagnosis of jjancreatic cyst was made by exclusion, admitting, however, of a possibility of cyst of the u{)per part of the mesentery or hydatid. On October 15, 18I)'5, an incision was made in the median line above the umbilicus. The tumor was found to lay between the stomach and the colon, under the omentum, but too far to the left to be fixed in the opening. A second incision was made in the left semilunar line, from the ninth costal cartilage downward; the gastrocolic omentum under which the tumor presented was stitched to the parietal peritoneum and to the tumor-wall, and the cyst opened and drained of about one quart of straw-colored pan- creatic fluid of the consistence of syrup. Tubular drainage was established, with definite healing in five weeks. The course of the pregnancy was uninterrupted. VOL. I — 30 THE SURGICAL ASPECTS OF PANCREATITIS* WILLIAM J. MAYO In 1879 Balzer described acute pancreatitis in association with fat necrosis. Little attention was attracted to the subject, how- ever, and it was not until ten years later, when Fitz wrote his papers on the subject, that the medical world really became aware of the inflammatory diseases of the pancreas. Fitz soon after pointed out the fact that many supposed cysts of the pancreas due to traumatism were really accumulations of fluid in the lesser cavity of the peritoneum and in the omental bursse. A proper understanding of chronic pancreatitis is largely due to Robson, who first noted the condition in connection with his operative work upon the biliary tract. In fact, the surgical study of the inflammatory diseases of the pancreas may be said to be the result of an inquiry into the causation of some of the complications of gall-stone disease. The reason for this depends upon certain anatomic facts. Brewer, discussing the question from this stand- point, calls attention to the embryologic development of the organ. The pancreas is a racemose gland without a firm capsule, its pro- tected situation defending it from injury. It is originally formed of two offshoots from the intestinal tube, each having a duct com- municating with the intestine. The two buds soon coalesce, the upper channel, known as Santorini's duct, becomes obsolete, and the inferior, or duct of Wirsung, carries on the function. Later investigations show that the duct of Santorini is not usually com- pletely closed, although it is ordinarily functionless. Schirmer *Read at the Fifty-third Annual Meeting of the American Medical Association, in the Section on Surgery and Anatomy, and approved for publication by the Executive Committee: Drs. H. O. \Yalker, A. J. Ochsner, and DeForest Willard. Reprinted from "Jour. Amer. Med. Assoc," October 4, 1902. 466 THE SURGICAL ASPECTS OF PANCREATITIS 407 found, ill only two out of !().> cadavers, that tliis duct was al)sont, and in 53 it was capal)le of carrying the entire secretion of the ^datid tJirough its intestinal orifice, which is situated al)Out 13^ inches al)ovc tile oi)ening of the pancreatic duct of Wirsung. It is to be noted that as the pancreas develops it is completely surrounded by peritoneum, the posterior layer in later life becoming infiltrated with fat, changing its character, and aiding to account for the direc- tion of diffusion of material escaping from the pancreas. The common duct of the liver passes around the head of the pancreas, between it and the second part of the duodenum. For V/2 inches the common duct lies fused or in close contact with the main pan- creatic duct, and in the sulimucous tissue joins with it, forming the ampulla of Vater. The joint opening on the duodenal surface is situated 4 inches below the pj'lorus. It is easily seen that a gall-stone may obstruct the common intestinal orifice without blocking the pancreatic duct, or at any point in the lower 13^ inches of the common duct may exert in- jurious pressure upon the pancreatic duct, or a stone may occlude it entirely at the diverticulum of Vater, thus explaining the etio- logic relationship existing between pancreatitis and gall-stone disease. This brings us to the effect of tliis pressure on the pancreas. Hildebrand experimentally ligated the pancreatic duct, and the result was acute pancreatitis, usually of the hemorrhagic type, w ith or without fat necrosis. Oj^ie repeated Hildebrand's experiments with the same results. He also found that injecting the pancreas with irritating fluids through the duct would produce acute pan- creatitis, and in Halsted's case the postmortem examination re- vealed the fact that a stone at the papilla had obstructed the open- ing so that bile had passed throughout the pancreatic ducts, caus- ing the fatal pancreatitis. Evans, in one case of fat necrosis, at autopsy found the pancreas stained with bile, even in the tail. Flexner, in experimental work, demonstrated that acute pan- creatitis resulted from the injection into the ducts of many sub- stances, such as dilute gastric juices and so forth, without an in- fection. 468 WILLIAM J. MAYO He also produced the disease with various bacterial cultures which he injected into the ducts. In this connection Brewer also points out the effect of the passage of large stones through the papilla in relaxing the sphincter and permitting infections from the duodenum. Robson, in his operative work, found 20 cases or more in which chronic pancreatitis resulted from partial obstructions or infections of the duct, interfering with free drainage. The patho- logic condition present was a chronic interstitial pancreatitis, with enlargement and hardening of the gland, much resembling malig- nant disease. This enlargement of the head of the pancreas com- presses the common duct and causes jaundice. It can be seen that a gall-stone obstructing the papilla may cause bile to pass through the duct of Wirsung, and in some cases out of the duct of Santorini, resulting in pancreatitis, or the main pancreatic duct may be obstructed and yet allow the pancreas escape by reason of the accessory duct of Santorini. Considering the frequency of gall-stones and infections of the liver-ducts which must affect the pancreatic ducts more or less, biliary calculi occupy the most important position as to the causation of pancreatitis, and the form which the disease takes depends, to a large extent, on the amount of obstruction and the degree of infection or chem- ical irritation which takes place. In 13 recently reported cases of acute pancreatitis operated on gall-stones were found in 10. Pan- creatitis also occurs independent of gall-stone disease through in- fections from the gastro-intestinal tract, as during the course of a duodenitis, gastric ulcer, or cancer; and at times its origin is in- trinsic, having no external evidences as to its causation. Pancreatitis has been divided by Robson into acute, subacute, and chronic. The knowledge of acute pancreatitis has been gained so largely from postmortem examination that we recognize only the more severe grades of the disease. The few surgical observa- tions with recovery, and the accumulations of fluid of pancreatic origin due to traumatisms, occurring in the omental bursa in con- nection with certain heretofore unexplainable complications found during operations on the biliary tract, lead to the belief that many cases of acute pancreatitis are not recognized because recovery has Tin: SURGICAL ASPECTS OF PAN'CREATITIS 4G!> taken i)lacc. It is alto^'ctlier prohahlc that this fliscase is not so fatal as our limited knowledge of the subject wouhl lead us to be- lieve. The course of the acute form is so often marked by hemor- rhage into and about the j)an(reas that the condition has been called acute hemorrhagic pancreatitis without regard to whether or not hemorrhage was a prominent factor, and in other cases fat necrosis was a feature so important that no reference was made to the pancreas at all. Acute pancreatitis often follows an injury, as in one of Robson's cases, in which a servant fell against the corner of a table, striking the al)domen over the pancreas, and died in forty-eight hours. The disease is ushered in by the most acute and sudden pain in the epigastric region, followed with profound collapse. The abdomen becomes distended at once, and this, with the nausea or vomiting, may lead to the diagnosis of intestinal obstruction. Among the special symptoms noted are "nervous unrest" (Halsted and Opie), "repeated attacks of collapse" (Fowler), "lividity" (Evans), and "hiccup" (Mayo). In the hemorrhagic form death usually follows in a few hours or days. The indications at this time are to relieve the shock and sustain life. If the patient is tided over these acute symptoms and a subacute pancreatitis is established, death may result later from infection with formation of abscess. It is in this condition that incision and drainage may aid recovery, as in cases reported by Fowler and others. The drainage may be anterior, through a tube ])rotected by gauze packing, or a large posterior incision at the left costovertebral angle, as recommended by Robson. In the few successful cases reported the drainage was prolonged for weeks or months. Fat necrosis is the most interesting phenomenon connected with acute pancreatitis. It occurs in a large percentage of cases, and may coexist with the hemorrhagic form, or in some of the cases which end in recovery without hemorrhage or infection, as in the case which we reported.* The symptoms are those of an acute form of pancreatitis with repeated attacks of shock and collapse, *"Jour. Amcr. Med. Assoc," January \i, I90i. 470 WILLIAM J. MAYO probably as new foci are formed by fresh leakage from the pancreas. The fat necrosis is usually limited to the upper abdominal region, and affects the omentum, mesentery, and retroperitoneal fat by preference, although cases have been reported in which the epi- cardial fat has been involved and even the bone-marrow. In Beck's case the peritoneum was attacked, having much the ap- pearance of tuberculosis. The distribution of the areas involved is hard to explain. Is it due to intraperitoneal leakage with ab- sorption through the lymphatics, or is it not possible that the de- velopment of the pancreas as an intraperitoneal organ and the fatty conversion of the posterior layer of the peritoneum 'may be the important factor in the diffusion of the irritant.'^ The ■ appearance of fat necrosis is characteristic: Little areas of opaque spots scattered throughout the fat, of a whitish or brown- ish hue, from the size of a millet-seed to that of a pea, or often very much larger. The pancreas itself may be so slightly affected that the condition has been overlooked upon examination. This is the most reasonable explanation of cases reported without mention of the pancreas. Why does injury or disease of the pancreas cause hemorrhage and fat necrosis? Robson's observation that there is a greater tendency to hemorrhage in cases of jaundice complicated with pancreatic disease has been confirmed by the experience of others, and it is probable that disarrangement of its secretory func- tion may have this effect. The action of the pancreatic secretion upon the fats is to split the fat-globules into fatty acids and glycerin, and it has been thought that the absorption of the glycerin may act in this manner. It would appear, however, that the amount of glycerin absorbed would be too small to account for the phenom- enon. The fatty acids unite with the calcium salts, giving rise to the opaque spots which characterize the disease. As aids to diag- nosis, glycosuria, fatty stools, lipuria, and so forth are of value, but are not often present, and, therefore, if not found, have no negative weight. Mr. Cammidge, in a few cases, discovered a peculiar crystal in the urine which he has described (Robson) . Opie found steapsin in the urine obtained after death in one case. Walker points out that the absence of pancreatic secretion from stools is shown by a pale color of the feces, much like that condition which TIIK SUI«;i(AI, ASI'KCTS OK pan(kf:atitis 471 i> peculiar lo cliolemia. Most cases have, at some time, a rise of leinperaliire, and as a cliolatij^itis is often the source of infection in acute pancreatitis, diills and fever with sweating may l)e expected in this chiss of cases. In 29 cases of chronic pancreatitis which came to necropsy at Johns Hopkins, Opie found the three most com- mon causes to be pancreatic calculi, gall-stones in the terminal portion of tiie duct, and carcinoma, and he notes its frequent ap- pearance as a complication of cirrhosis of the liver. Glycosuria coming on late in the course of liepatic cirrhosis is suspicious of interstitial changes in the pancreas. In the diagnosis of the chronic forms of pancreatitis Thayer says that the feces should be searched for excess of fats and for fragments of the pancreas. Evidence of imperfect digestion of albuminoitls, as shown by Sahli's glutoid capsule, he also believes to be of value. In acute and subacute pancreatitis the surgical indications, be- yond the opening and draining of septic accumulations in or about the pancreas, would be the removal of the gall-stones, if present, and the establishment of free drainage through the gall-bladder, relieving the tension and aiding the secretions to escape. This procedure alone resulted in the recovery of the case reported by Beck and in our case. The most important diagnostic feature of chronic pancreatitis is jaundice, and in thin subjects the enlarged pancreas may be felt. Frequently a distended gall-bladder can be palpated, which is un- usual in stone obstructing the common duct. The slow course aids in differentiating from malignant disease. The treatment of chronic pancreatitis is by drainage, and this is best accomplished by way of the gall-bladder. We have had 7 well- marked cases, 4 treated by cholecystostomy, 3 by cholecystenteros- tomy; the anastomosis was made twice to the transverse colon and once to the duodenum. All the patients recovered, and the ultimate result in the cases in which the transverse colon was used was as good as with the duodenum. It is sometimes difficult to make a satisfactory anastomosis between the gall-bladder and duodenum in these cases, on account of adhesions. Cholecystos- tomy is preferred by Robson, but the prolonged external discharge is a cause of much annoyance to the patients. BIBLIOGRAPHIC INDEX Abbe, 21, 22, 23, 24, 27, 29, 44, 85, 103, 117, 333, 334, 387 Adami, 230, 320, 413 Albu, 199, 220 Allen (Dudley), 216 Andrews, 40, 89, 127, 179, 300 Annandale, 20 Audisten, 201 Audry, 12 Auerbach, 317 Bai.dt, 86, 102 Balzer, 232, 466 Beadles, 385 Beck, 470, 471 Behrend, 164 Berg, 395 Beraays, 25, 28, 37, 38, 47, 72, 75, 90, 105, 110 Berthold, 196, 244 Bettman, 207 Bevan, 101, 228, 247, 350, 382, 392, 411, 424, 449 Beyea, 184, 192, 193 Billroth, 19, 27, 34, 37, 38, 40, 41, 75, 107, 166, 170, 207, 214, 280, 282, 290, 304 Binnie, 100 Bircher, 35, 38, 54, 92 Blake, 258, 263, 275, 297 Bland-Sutton, 72, 97, 100, 385 Bloodgood, 117 Boas, 52, 95 Bobbs, 343 Bond, 311, 321, 323 Bramwell, 271 Brandt, 48 Braun, 56, 76, 112, 138 Brewer, 235, 236, 425, 466, 468 Brigham, 86, 105 Brinton, 97, 196, 200, 224 Bristow, 86, 97, 224 Brunner, 239, 244, 268 Bull, 34, 37, 38, 41, 48, 53, 71 Butlin, 14, 37, 385, 388 Cabot, 142 Cammidge, 470 Cannon, 184, 195, 241, 258, 263, 297, 309, 311, 31-2 Carey, 13 Carle, 111, 138, 156, 250, 365, 403 Carrel, 309 Chlumsky, 108, 118, 147, 150 Clark, 321 Collins, 75 Cone, 12 Conner, 41, 86, 105 Cordier, 154, 192 Cour^-oisier, 232, 379, 384, 387, 443 Cramer, 219, 220, 221 Crile, 251, 309 Cruveilhier, 181 Cuneo, 187, 188, 210, 230, 284, 315 Cunningham, 295, 309 Curtis, 109 Cushing, 82, 115, 117, 122, 213, 309, 319 Czerny, 40, 91, 94, 102, 108, 113, 115, 116, 151, 153, 161, 190, 228, 261 Davis, 374, 399, 411 Dawbarn, 192 Deaver, 275 Debove, 200, 225 Dennis, 60, 109 473 474 BIBLIOGRAPHIC INDEX Dickinson, 88, 126, 322 Dieulafoy, 182, 271 Doyen, 76, 91, 112, 122, 251, 261,297, 319 Dudgeon, 233 Dunn, 201, 209, 225 Duplant, 201 Duret, 92 Edes, 364, 365, 402 Einhorn, 14, 98 Eiselsberg, 156, 319 Eisendrath, 89, 127, 382 Elliot, 344, 345, 353, 424 Evans, 467, 469 Ewald, 13, 52 Fagge, 340 Fantino, 111, 156, 250 Felitzet, 33 Fenger, 20, 23, 38, 39, 45, 51, 53, 68, 91, 100, 109, 113, 114, 165, 332, 339, 345, 349, 361, 379, 412 Ferguson, 288 Ferrier, 39, 156, 390 Fiedler, 196, 224 Finlayson, 101 Finney, 126, 127, 175, 189, 192, 2^7, 242, 249, 254, 255, 256, 260, 270, 273, 276, 304, 306 Fitz, 232, 379, 466 Fletcher, 26 Flexner, 12, 467 Foote, 179 Ford, 320 Fowler, 251, 252, 261, 322, 469 Francine, 267, 268 Frank (Kendall), 13 Frank, 20, 27, 29, 37, 40, 45, 69, 90, 109 Frankel, 394 Frankl-Hochwart, 199 Futterer, 201, 209, 225 Gaston, 334 Gerhardt, 88, 125, 198 Gerster, 16 Goffe, 280 Graham, 52, 64, 73, 163, 172, 201, 208. 209, 225, 277, 285 Graser, 27 Greenough, 274 Greiss, 196 Grunfeld, 268 Guillot, 115, 117 Guinard, 107, 280 Gunzburg, 52 Gussenbauer, 20, 37, 71, 97, 99 Haberkant, 99, 165, 209, 229, 280, 287 Haberlin, 93, 228 Hacker, 13, 14, 23, 37, 39, 45, 109, 111, 137, 257 Hagenback, 27 Hahn, 39, 45 Halsted, 95, 100, 104, 115, 165, 366, 369, 403, 406, 448, 467, 469 Hamilton, 81, 342 Hanot, 362, 402 Harley, 342 Harrington, 319 Harris, 349 Hartley, 71 Hartmann, 21, 175, 176, 187, 188, 199, 201. 211, 213, 214, 217, 224, 276, 284, 410 Haslam, 73. 74 Heidenhain, 110 Heineke, 36, 38, 46, 53, 92, 144, 189. 227, 249, 254, 277 Hektoen, 72. 97 Hemmeter, 94, 95, 97 Heydenreich, 88 Hildebrand, 467 His, 318 Hoegh, 352 Homans, 447 Howard. 267 Huntington. 178, 236, 309 Ingals, 12 Israel, 98 IMIJLIOCUAIMIIC INDKX 413 Jiicobi, 3.'} Jiikscli, 'H Jonu.s, 4K) Jonnesco, .Sl;{ Jnslin, 27 i K.voKU, !)(), 10!), Mi, 45:$ Kjimmcrer, 27, 114 Kirn. Hi), 17!), '.Ui, f5!)() Kclir, D.y.i, ;5!)'i, ;j!):j, ;5!).j, ;5!)!), 41 1, 41^, 414, 4;e!) Keith, ;5()!) Killing, ij.'j, 157, :nfi Kilyiiack, 3G4, 885, 401 Kilhourne, 33G KochtT, 47, 56, 75, 70, 91, 98, 106, 107, 108, 113, 114, 15.3, 166, 214, io.'i, 259, 276, 280, 282, 284, 288, 289, 315, 447 Kiinif,', 11. 2C. 27, 34 Kijrtc, 151, 156 Kraske, 94 Krauso, 3!)4 Kriinlein, 161, 162, 212, 228, 282, 288, 291 Kuhn, 95 Kiimmell, 91, 108, 114, 115 Kussmaul, 34, 52 Kuster, 89, 127 Landerer, 52 Landois, 331 Lange, 50 I-ingcnbcck, 333, 345 Lartigan, 320, 409, 440 Laucnstein, 37 Lavastine, 397 Lcbert, 201, 209 I^ Count, 368, 405 lA'nnander, 440 Leube, 88, 125, 174, 200, 225, 226 Levy, 108 Lindner. 100, 119 Littlefield, 261 Liilnner. 368 Loreta, 36, 38, 46, 55, 74, 81, 92 Lublinski, 12 Lun, ;{()(>. M».'5 of diiodi'iuiin, 17!) of gall-l)Iu(l(li'r. .'Wl cliolecysleclouiy in, '.i'Ai, tiHH cholecystentcrosloiiiy for, ;J81, 414, 415 diagnosis, S86 difrercniiiil, .SH7 fro((ncncy, 414 fj.'ill-stones as cause, '584, 411 involving liver, .'$!)() jaundice in, iid'i, .'5H7 of liver and gall-stones, relation, ;540 of pylorus, 98, !)S frequency, 'M obstruction from, 71. See also Pyloric obstritclion, iintlignanf. of stomach. See (iuslric cancer. Cancerous cachexia, 73 Carcinoma. See Cancer. Cardiac extremity of stomach, 33 orifice of stomach, method of ex- ploring, 85 obstruction, Kader's ojjeration in, 90 Ssabanejew-Frank operatioa in, 90 surgical treatment, 89 Witzel's operation in, 89 Catarrh, stone-building, 393, 409 Catarrhal cholecystitis, 376 jaundice, 3G3 Cecocolic sphincter, 310 Cerebrosj)inal nerves, control of, over stomach, 317, 318 Cholangitis, infective, oi)eration in, 348 Cholecystectomy, 349, 353, 355 for cancer of gall-bladder, 388 for gall-stones, 333, 345. 409, 410 for injuries of gall-bladder, 356 for malignant disease of gall-bladder, 356 for stones in cystic duct, 42'2 indications for, 355 VOL. I — 31 ( liolccN ^trcloiiiy, MiorLility from, 433 4.37, 45 I ((crmanence of (lire after, 451 niiioval of muccjus mi-mbrane of gail-l)la cholecystotomy in, 378 chronic, operation in, 397 gall-stones from, 440 operations in, 375 sui)purative, 370 (holecystostomy. causes of failure, 418 for gall-stones. 409, 410 mortality from, 4.33-437 operati\c disability after, 453 Cholecystotomy, 348 for gall-stones, 333, 342, 348 drainage in, 344 ideal, 343, 350, 398, 409 for gall-stones, 409 in cholecystitis, 378 in gall-stones, 398 Choledochotomy, 349 for gall-stones, 349 mortality from, 433-437 Cholelithiasis. See Gall-stones. Cicatricial stricture of esophagus,ll. See also Esophagus, stricture of, cicatricial. Cirrhosis of liver, jaundice in, 36:2 Cocain anesthesia in operations on stomach, 117 Colic in diagnosis of gall-stones, 237 in gall-stones, 3.38, 408, 409, 441 renal colic and, differentiation, 441 renal, and gall-stone colic, differentia- tion, 441 Collapse after operation for malignant disease in abdomen, 114 in acute pancreatitis, 409 Comnu)n bile-ihict. See Bile-ducts, common. 482 INDEX OF SUBJECTS Contraction of anastomotic opening after button gastro-enterostomy, 115 Cruveilhier's acute round ulcer of stom- ach, 181 Cureting of gastric cancer for diag- nostic purposes, 95 Cylindric-cell cancer of stomach, 97 Cyst of pancreas, 464 Cystic duct, anatomy of, 440 stones in, 338 care in exploring, 417 cholecystectomy for, 422 operations for, 372, 373, 396 removal of, 344 Davis' method of drainage in gall-stone operations, 399 Diet after operations on stomach, 117 before operations on stomach, 116 potato, for foreign bodies in stomach, 35 Dieulafoy's mucous erosion, 182, 271 Dilatation, atonic, of stomach, in neurasthenia, 176, 201, 272 gradual, in cicatricial stricture of esophagus, 16 of stomach, 128 and malignant pyloric obstruction, differentiation, 72 Brandt's operation for, 48 gastro-enterostomy in, 131 gastroplication in, 131 gastrotomy for, 38 method, 123 not of organic origin, 176, 201 operation in, 184 pyloroplasty in, 131 results, 140 surgery, 35, 38 retrograde, and gastrotomy, in im- passable stricture of esophagus, 27 Displacement of stomach, effect of weight of tumors, 34 Distention of stomach before opera- tions on stomach, 34 method, 33 Distention of stomach with air, value of diagnosis, 43 Diverticulum of esophagus and cica- tricial strictm-e of esophagus, differ- entiation, 14 Doyen's operation to prevent re- gurgitant vomiting after gastro- enterostomy, 112 Drainage, gastric, 144, 189 in gastric ulcer, 226 value of, in neurotic cases, 143 hepaticus, 423 in chronic pancreatitis, 471 in operations for removal of gall- stones, 344 of fluids in operations on upper ab- domen, 322 of gall-bladder. 352 Duodenum and common bile-duct, anastomosis between, 446 cancer of, 179 diseases of, associated with gall- bladder disease, 179 operations on, review of cases, 178 perforation of. 238, 239 ulcer of, 178. 223, 244 acute perforating, operation in. 247 and gastric ulcer, relative fre- quency, 268, 269 chronic, 267 from surgical standpoint, 194 gastro-enterostomy in, 242 pain in, 198 surgical treatment, 301 symptoms, 198 classification, 270 clinical, 270 frequency, 244, 268 gastro-enterostomy in, 249, 302 bile in stomach after, 250 hemorrhage in, operative treat- ment, 248 indurated, 270 medical, 270 treatment, results of, 274 non-indurated, 270, 272 operative indications, 247 INDEX OF SUBJECTS 483 DiKuli-num, )il(<'r of, sex froqueiHy, iio suFf^ical Ircatniont. ii)i) symptoms, iiti treatment, surgical, €99 with gall-hladder ami liver compli- cations, operation in, €48 with gastric complications, opera- tion in, €48 Electhicity in non-dilatable cicatricial stricture of esophagus, €6 Emaciation in gastric cancer, 64 Enlarged glands in omentvmi in pyloric obstruction, 53 Entero-anastomosis after gastro-enter- ostomy, cases illustrating, 154, 155 in regurgitant vomiting after gastro- enterostomy, 138 to prevent regurgitant vomiting after gastro-enterostomy, ll!2 Enterocholecystotomy for gall-stones, 334 Erosion in gastric ulcer, 'JTl mucous, of stomach, 181, IS'i Esophagectomy in non-dilatable cica- tricial stricture of esophagus, €0 Esophagoscopy in stricture of esopha- gus, 14 Esophagotomy, external, combined with gastrotomy, for foreign bodies in esophagus, 8 for foreign bodies, 7, 8 in non-dilatal)le cicatricial stricture of esophagus, 19 internal, in non-dilatable cicatricial stricture of esophagus, io Esophagus, 1 diverticulum of, and cicatricial stric- ture of esophagus, diflferentiation, 14 foreign bodies in, 3, (> and foreign bodies in trachea, differentiation, 3 external esophagotomy for. 7, 8 gastrotomy, combined with ex- ternal esophagotomy for, 8 Esophagus, foreign biwlies in, illustrative cases, 8, 9 instrumcntjj for removing, 7 symptoms, 7 treatment, 7 open buckle in, location with R5ntgen rays, 10 removal, 10 stricture of, cicatricial, 11 and esojjhageal diverticula, dif- ferentiation, 14 bougies in, 16 diagnosis, 13 dilatable, treatment, 16 esophagoscopy in, 14 etiology, 11 general character, 11 impassable, gastrotomy and ret- rograde dilatation in, €7 gastrostomy in, €8 treatment of, €6 location, 13 non-dilatable. Abbe's string-saw method in, £1 Billroth's operation in, 19 electricity in, €6 esophagectomy in, 20 Gussenbauer's operation in, €0 internal esophagotomy in, 25 Ochsner's operation in, €1, €4 treatment, 19 prognosis, 14 treatment, 16 congenital, 13 fibrous. \i simple, 12, 13 ulceration of, cicatricial stricture from, 11, 12 Ether anesthesia in radical operation for gastric cancer, 287 Excision in gastric ulcer. 126.176.203.304 Exhaustion after gastro-enterostomy, 148 Exploration in radical operation for gastric cancer, 287 Exploratory operations in diseases of stomach, 84, 123, 124, 141 484 INDEX or SUBJECTS Exploratory operations in gastric cancer, 64, 65, 10-2, 16-2, 163, 17^2 effects of, 65 ulcer, 20-2 in malignant pyloric obstruction, 74 Fat necrosis in acute pancreatitis, 461, 469 Fatal suture angle, 166, 214 Feces, gall-stones in, 332, 444 Feeding, rectal, after operations on stomach, 117 Fenger's ball-valve stone, 361, 412 incision in operations on stomach, 38 method of gastrostomy, 39 modification of Kocher's method of gastro-enterostomy, 113 Fibrous stricture of esophagus, 12 Finney's gastroduodenostomy, 192, 255 in gastric ulcer, 204, 227, 242, 276, 304 Fish-hook pj'lorus, 129 Fissiue ulcer of stomach, 181, 271 Fistula, gastric, surgery of, 35 sutiu-e in, 38 Floating gall-stones, 339 Foregut, organs derived from, 309 Foreign bodies in air-passages, 3 in bronchi, 4 in esophagus, 3, 6 and foreign bodies in trachea, differentiation, 3 external esophagotomy for, 7, 8 gastrotomy combined with ex- ternal esophagotomy for, 8 illustrative cases, 8, 9 instruments for remo\'ing, 7 sj-mptoms, 7 treatment, 7 in stomach, gastrotomy for, 38 potato diet for, 35 surgery, 35 in trachea, 3 and foreign bodies in esophagus, differentiation, 3 Foreign bodies in trachea, diagnosis, 4 illustrative cases, 6 sj-mptoms, 3, 4 prognosis, 4 tracheotomy for, 5 treatment, 4 i Fowler's position in septic peritonitis, I 322 Frank's method of gastrostomy, 40 Functions of stomach, 63 Gall-bladder, 325 anatomy of, 235, 439 aspiration of, blind, dangers of, 333 bacteria in, 321 in gall-stones, 394 Sevan's incision in operations on, 350 blood-supply of, 440 cancer of, 384 cholecystectomy in, 356, 388 cholecystenterostomy for, 381, 414, 415 diagnosis, 386 differential, 387 frequency, 414 gall-stones as cause, 384, 414 involving liver, 390 jaundice in, 362, 387 disease, diseases of duodeniun asso- ciated with, 179 drainage of, 352 ducts of, operations on, report of seven cases, 331, 335 function of, 331 incision in operations on, 350 infection of, from bile, 421 in gall-stones, 332, 412, 413 injuries of, 332 cholecystectomy in, 356 mucous membrane of, removal, in gaU-stones, 399 nerve-supply of, 440 operations on, 372, 407 after-care, 413 classification, 333 disabilitv after, 453 INDEX OF SUBJFXTS 485 (jiill-liladdcr, opcr.it ions cpii, incision for, jiiundice as caiiso of licniorrlia^i' after, 380 inorlalitj' from, 'i.'i'i perrnancncc of cure, i.V.i report of seven eases, SIJl, 335 secondary, 413 I)elvis of, r,i5, UO perforation of, ■i'.W, '23!) rupture of, 333 Smith's incision in o])erations on, 3i2 surgery of, 348 status, 392 suture of, after removal of stones, 333, 334 in wound, dangers, 343 Tait's incision in operations on, 342, 350 Gall-stones, 330 active, operation in, 395 and cancer of liver, relation, 340 as cause of cancer of common bile- duct, 402 of gall-I)lad(ler, 384, 414 bacteria in fiall-hiadder in, 304 ball-valve, 3(il, 412 cholecystectomy for, 333, 345, 409, 410 cholecystenterostomy for, 345 cholecystitis as cause, 440 cholecystostomy for, 409, 410 cholecystotomy for, 333, 342, 348, 398 drainage in, 344 choledochotomy for, 349 chronic pancreatitis in, 412, 419 colic in, 338, 408, 409, 441 diagnosis of, 237 renal colic and, difTenntiation, 441 complications in, 395 dangers of suturing gall-bladder in woinid in removal of, 343 diagnosis of, 3,38, 438 (litTerential. 340 enterocholecystolomy for, 334 etiology of, 336 (jall-sloiics, (ifl li stage, s\(ii[)tonis of, 443 first stage, symptoms of, 4 U floating, 339 fourth stiigc, sympt«mis of, 442 frequency of, 33(i, 392 ideal cholecystotomy for, 409 in feces, 332, 444 in intestine, 341 in women, reasons for frequency, 320 incision for, 342 infection of gall-bladder from, 322, 413 jaundice in, 332, 337, 300, 443, 444 location of, 372 number of, 337 operation for, causes of failure to cure, 410 mortality from, 240, 430 outside; bile-tract, operations for, 374 pain in, 338, 441 pancreatitis in, 412, 419, 457 prognosis, 340 removal of raucous membrane of gall- bladder in, 399 report of two operations for, 327 second stage, symptoms of, 441 slumbering, 392, 408 suture of gall-I)ladder after removal, 333, 334 tension in, 395 third stage, sj-mptoms of, 442 treatment of, 341 medical, 341 surgical, 342 typhoid fever as cause, 377 urine in, 444 Gas pain, 319 Gastrectasia, forms of, 128 Gastrectomy, 40, 41 Rillroth's method, 41 in gastric cancer, 104, 105, 108 in malignant disease of stomach, 8(5 partial, 41 and pylorectomy, in gastric cancer, 106 Gastric cancer, 93, 228 absence of free hydrochloric acid in, 73.94 486 INDEX OF SUBJECTS Gastric cancer, adhesions in, 99 operation in, 165, 209 age in, 93 and gastric ulcer, relation, 163, 201, 209, 225, 277, 285 Bernays' method of cureting, 47 Billroth's operation for, 166 cachexia in, 64 checking of, after exploratory in- cision, 37 collapse after operation for, 114 colloid type, 97 condition of patient in, curability depending on, 101 curability of, 96 ciu-eting of, for diagnostic pur- poses, 95 cylindric-cell, 97 diagnosis, 37, 63, 141, 163, 229, 284 early, 94 diagnostic massage in, 95 disappearance of, after exploratory incision, 37 emaciation in, 64 examination of blood in, 95, 286 exploratory operations in, 64, 65, 102, 162, 163, 172 effects of, 65 extension of, operation in, 165 to surrounding structures, cura- bility depending on, 99 fragments in stomach contents in, 95 frequency, 71 gastrectomy in, 86, 104, 105, 168 gastro-enterostomy in, 65, 110, 133, 145 anterior method, 111 mortality from, 134 posterior method. 111 prolongation of life after, 291 regurgitant vomiting after, 112 gastrostomy in, 108, 109 Hacker's gastrostomy in, 109 Hartmann's incision in, opposite page 185 histologic structure, 71 Gastric cancer, histologic structure, cvu-ability depending on, 96 history in, importance of, 63, 73, 163 jejunostomy in, 110 Kader's gastrostomy in, 109 Kocher's operation for, 166 pylorectomy and partial gas- trectomy in, 107 lactic acid in, 74, 94 lines of incision in, opposite page 185 location of, curability depending on, 97 lymphatic involvement in, 99, 119, 210 operation in, 165 Marwedel's gastrostomy in, 109 Mayo's incision in,opposite page 185 medullary, 97 Moynihan's incision in, opposite page 185 operations in, 184 radical, 65 pain in, 64 palliative operations in, 108, 110, 291 percentage of hemoglobin in, 286 progress of, 93, 96 pylorectomy and partial gastrec- tomy in, 106 pylorectomy in, 66, 87, 166, 167 pyloric end, frequency of, 37 radical operation for, 161, 166, 228, 280 after-care, 291 anesthesia in, 287 exploration in, 287 freeing greater curvature, 288 mobilization of lesser curva- ture in, 287 mortality from, 280, 281 Murphy's proctoclysis after, 291 recurrences after, 161 removal of diseased structures, 288 INDKX; OF SUBJECTS 487 Gastric cancer, nulical oix-raliuti fur, restoration of fjastro-inlcstiniil canal, liilirotii No. 2 inctliofi, iiW Kochcr's nictliod. ^H!) results of, IG'i separation of pyloric end of stomach, !288 steps. 18(i, iJ87 suture of gastric stump, iSi) reduced motor power of stomach in, 94 Robson's incision in, opposite page 185 scirrlms type, 97 Ssabanajcw-Fraiik gastrostomy in, 109 Stamm's gastrostomy in, 109 surgery of, 30 status, 177 surgical treatment, iiO, 279 review of, 280 symptoms of, 94 test-meals in, 64 tests of stomacii-contents for, 103, 172 treatment, 03, 0,3 surgical, 279 review of, 280 Tuholske's operation in, 108 urine in, 95 value of test-meal in diagnosis of, 43 vomiting in, 04, 94 von Hacker's gastrostomy in, 109 Witzd's gastrostomy in, 109 diseases, diagnosis, 33 disturbances acting from within stomach, causes, 62, 69 after abdominal operation, cause, 49 from omental hernia, 48, 69, 70 from traction iif adherent omentum, 34 drainage. 144, 189 in gastric ulcer, 226 value of, in neurotic cases, 143 Gastric incjtility, inlcrfiniice with, 273 myasthenia, 130 tetany, 220 in gastric ulcer, 199 ulcer, 124, 221 acute, 173 and duodenal ulcer, relative fre- quency, 268, 269 antl gastric cancer, relation, 103, 201, 209, 225, 277. 285 anemia in, 222 chronic, 173, 267 from surgical standpoint, 194 I)ain in, 197 prognosis of, 225 pyloric stenosis from, 199 Robson's, 183 surgical treatment, 301 symptoms of, 197 classification of, 180, 181, 270 clinical, 270 complications, 226 diagnosis, 36, 125, 141, 224 distortion and contraction after healing, 89 erosion, 271 etiology, 195, 222 excision in, 126, 176, 203, 304 exploratory operation in, 202 Finney's pyloroplasty in, 227, 242, 276, 304 fissure, 181 fissure-like. 271 frequency. 195. 196. 267 gastric drainage in. 226 tetany in, 199 gastro-enterostomy in, 125, 134,145, 175, 203, 204, 227, 276, 302 case illustrating, 151 complications after. 154 hemorrhage from, 224 Andrews and Eiscndrath's oper- ation for, 127 gastro-enterostomy in. 127 treatment, 89, 127 surgical, 300 488 INDEX OF SUBJECTS Gastric ulcer, hour-glass stomach from, treatment, 127 hyperchlorhydria in, 222 indurated, 270 line of incision in, 202 opposite page 183 location of, 184, 196, 224 McGraw ligature operation in, 205 mechanical injury in, 223 medical, 270 treatment, results of, 274 multiple, 224 non-indurated, 270, 272 number of, 196 obstruction from, gastro-enteros- tomy in, 128 pyloroplasty for, 127 treatment, 127 operations in, 180 advisability, 273 pain in, 173, 223 perforation in, 88, 126 results, 36 surgical treatment, 299, 300 treatment, 89, 126 pore-like, 181 prognosis, 200, 225 pyloric obstruction from, 50, 127 spasm in, 273 pyloroplasty in, 203, 227, 277 Rodman's operation in, 203, 227, 242, 277, 305 round, 181 acute, 181 chronic, 181 sentinel lymph-glands in locating, 184, 271 sex frequency, 196, 224 surgical treatment, 36, 88, 124, 125, 173, 299 indications for, 225, 226 symptoms, 173, 223 tests of stomach-contents in, 200 treatment of, medical, 125 results of, 274 surgical, 124, 125, 299 vomiting in, 224 Gastroduodenostomy. See Pyloroplasty. Gastro-enterostomy, 145, 253, 256 anterior method, 137, 146, 150, 257 bile in stomach after, 138 bronchopneumonia after, 149 complications after, 148 detachment of bowel from stomach after, 147, 149, 150 entero-anastomosis after, cases illus- trating, 154, 155 exhaustion after, 148 for gastric ulcer, case illustrating, 151 in benign pyloric obstruction, 136 in chronic duodenal ulcer, 242 in dilatation of stomach, 131 in gastric cancer, 65, 110, 133, 145 anterior method. 111 mortality from, 134 posterior method. 111 prolongation of life after, 291 regurgitant vomiting after, 112 ulcer, 125, 134, 145, 175, 203, 204, 227, 276 complications after, 154 in hemorrhage from gastric ulcer, 127 in hour-glass stomach, 137 in interference with mechanical action of stomach, 46 in malignant pyloric obstruction, 76 cases, 79, 80 in obstruction from gastric ulcer, 128 in pyloric obstruction, 55, 56, 91 cases illustrating, 58, 59, 60 in ulcer of duodenum, 249 bile in stomach after, 250 indications for, 41, 132 Kocher's method, 113 Fenger's modification, 113 length of jejunum loop in, 158 lung complications after, 149 McGraw ligature method, 259 Mikulicz's method, 190 mortality from, 148 Moynihan's method, 261 Murphy button in, 68, 91, 110, 114, 153, 258 INDEX OF SUBJECTS 480 Gustro-cntcrosloiiiy, Miirpliy liultun in, contraclion of iinaslDinolic oixfi- ing afUT, 115 no-loop method, 2G1, 203 steps of, 290 opening in, location of, 1"'<) posterior nu'llioil, l:J7, IKi, I.W, 257 Postnikow's met hod, 41 regurgitant vnmitin;,' .'iflcr, 138, IIH, 2o8 Doyen's oi)eralioii for preven- tion, 112 entero-anastomosis in, 138 Rutkowski's operation for pre- vention, 112 Roux's method, 201 Senn's metiiod, 113 suture method. 08, 91, 110, 114, 153, 258, 2(i() contraction of anastomotic open- ing after, 115 technic of, 293 vicious circle after, 138, 140, 258 AVoIHer's method, 113 Gastroiiepatie omentum, division of, in operations on stoTnaeli, 03 Gastro-intestinal canal, embryology of, 309 Gastrojejunostomy. See Gaslro-entcr- ostomy. Gastroplieation, 131 Gastroptosis, 130, 143 Gastrorriiaphy in gunshot wound of stomach, 09 in pyloric obstruction, 54 in wounds and injuries, 38 Gastrostomy, Andrews' method, 40 Fenger's method, 39 for feeding purposes, 39 Frank's method, 40 Hacker's method, 39 in gastric cancer, 109 Ilahn's method, 39 in gastric cancer, 108, 109 in impassable stricture of esophagus, 28 in mechanical interference with action of stomach, 45 Gastrostomy, Kadcr's, in gastric cancer, 10!) Marwcrlcl's, in gastric cancer, 109 Ssabanajcw-I"'rank, in gastric cancer, 109 Stamm's, in gastric cancer, 109 Witzd's method, 40 in gastric cancer, 109 Gastrotomy and retrograde dilatation in impassable stricture o( esopha- gus, 27 comliined with external esophagot- omy for foreign bodies in eso|)lia- gus, 8 for dilatation of stomach, 38 for foreign bodies, 35, 38 for mechanical interference with action of stomach, 44, 45 Gunshot wound of liver, 340 of stomach, gastrorriiaphy in, 69 Gussenbauer's o[)cration in non-dilata- ble cicatricial stricture of esophagus, 20 Hacker's method of gastrostomy', 39 in gastric cancer, 109 Hahn's method of gastrostomy, 39 Hartmann's incision in gastric cancer, opposite page 185 point of election, 213 pyloric syndrome, 175, 199 Heineke-Mikulicz pyloroplasty. See Pyloro phi.it y. Hemoglobin, percentage of, in gastric cancer, 280 Hemorrhage after operations on gall- bladder and bile-passages, jaundice as cause, 380 control of, in radical operation for cancer of pyloric end of stomach, 212 from gastric ulcer, Andrews and Eisendrath's operation for, 127 gastro-enterostomy in, 127 treatment, 89, 127, 224, 300 490 INDEX OF SUBJECTS Hemorrhage in operation for stones in common bile-duct, 425 in ulcer of duodenum, operative treatment, 248 Hemorrhagic pancreatitis, acute, 469 Hepatargia, 382, 397, 413 Hepaticus drainage, 423 Hernia, omental, as cause of gastric dis- tress, 48 gastric distress from, 69, 70 Hiccup in acute pancreatitis, 469 Hindgut, 311 History in gastric cancer, importance of, 63 Hormones, 316 Hour-glass stomach from gastric ulcer, treatment, 127 gastro-enterostomy in, 137 resection in, 305 treatment, 90 Hydrochloric acid, free, absence of, in gastric cancer, 73, 94 Hyperchlorhydria, 131 in gastric ulcer, 222 Ideal cholecystotomy, 343, 350, 398, 409 Impassable stricture of esophagus, treat- ment, 26 Incision, Bevan's, in operations on gall- bladder, 350 Fenger's, in operations on stomach, 38 Hartmann's, in gastric cancer, oppo- site page 185 in operations for pyloric obstruction, 53 on gall-bladder, 342, 350 on stomach, 84 line of, in gastric cancer, opposite page 185 ulcer, 202 opposite page 183 Maylard's, for exploratory examina- tion of stomach, 124, 141 modification, 141 Mayo's, in gastric cancer, opposite page 185 Incision, Mikulicz's, in gastric cancer, opposite page 185 Moynihan's, in gastric cancer, oppo- site page 185 Robson's, in gastric cancer, opposite page 185 Smith's, in operations on gall-bladder, 342 Tait's, in operations on gall-bladder, 342, 350 Tiffany's, for exploratory examination of stomach, 124, 141 Intestine, bacteria in, 321 detachment of, from stomach, after gastro-enterostomy, 147, 149, 150 embryology of, 309 gall-stones in, 341 large, antiperistalsis in, 311 small, traction weight of, producing funnel shape of stomach, opposite page 142 vestibule of, 234 Irrigation of stomach before operations on stomach, 34 Jaboulat-Braun operation to prevent regurgitant vomiting after gastro- enterostomy, 112 Jaundice as cause of hemorrhage after operations on gall-bladder and bile- passages, 380 catarrhal, 363 in cancer of common bile-duct, 402 of gall-bladder, 362, 387 in chronic pancreatitis, 471 in cirrhosis of liver, 362 in gall-stones, 332, 337, 360, 443, 444 in pancreatitis, 362 in stones in common bile-duct, 379 surgical significance, 360 Jejunostomy in gastric cancer, 110 Jonnesco's pyloric canal, 313 Kader's gastrostomy in gastric cancer, 109 INDEX OF SUBJECTS 491 Kadcr's operation in ol).striKlir)n in cardiac orifice of stomach, 90 Kidney, movable, as cause of pyloric obstruction, 5i Kocher's method of gastro-entcro.stf firctioii, iJli.'J. HH I'liagiK-ytosis, onifntnin in ijcvclopinf,', liii I'Icxus, Aiicrhacli's, :5I7 Mfis.siicr'.s, lUT I'lifunioiiia after oixTatitins on stoni- acli, Ki, IIG I'oro-liki- iili-iT of stoniacli, ISl I'ostnikow's nu'lliod of f,'a.stro-cnlt'r- osloniy, 41 l'«)tato diet for foreign Ijodies in stom- ach, 35 Proctoclysis, Murpliy 's, after radical operation for gastric cancer, -2i)l I'ylcjrectoniy, 41 and gastroduodenostoniy in malig- nant pyloric obstruction, 75 cases, 77. 78 and partial gastrectomy in gastric cancer, 106 IJillroth's method, 41 in gastric cancer, GO, 87, 16C, 167 in malignant pyloric obstruction, 75 cases, 77 in mechanical interference with ac- tion of stomach, 46 in pyloric obstruction, 53 Pyloric antrum, anatomy, 314 canal of Jonnesco. 313 end of stomach, cancer of, adhesions in, 209 extension to other organs, 209 laboratory methods of diag- nosis, 208 lymphatic infection in, 210 radical operation for, 207 after-treatment, 216 control of hemorrhage, 212 shock in, 215 steps, 211-217 significance of palpable tumor, 208 Pyloric ob.stniclioii, 50 Ix-iiign, gastro-eiiterostoiiiy in, 136 operations for, 67 causes, 51) diagnosis, 52 enlarged omental glands in, 53 from gastric ulcer, 50, 127 from movable kidney, 52 gastro-enterostomy in, 55, 56, 01 cases illustrating, 58, 59, 60 gastrorrhaphy in, 54 l>orcta's operation in, 55 malignant, and dilatation of stom- ach, ditferentiation, 72 diagnosis of, 71 exploratory operation in, 74 gastro-enterostomy in, 76 cases, 79, 80 methyl-blue in, 75 pylorectomy and gastroduoden- ostomy in, 75 cases, 77, 78 surgical treatment, 71, 75 treatment of, non-operative, 74 surgical, 71, 75 ^Yblfler's operation in, 76 operations for, 53 incision in, 53 pylorectomy in, 55 pyloroplasty in, 53, 92 cases illustrating, 58, 59 results of, 140 surgical treatment, 90 valve formation in, 50, 51, 90, 128, 142, 273 vomiting in, 64 spasm, 129, 175, 183 diagnosis, 142 in gastric ulcer, 273 physiology of, 318 pyloroplasty in, 130 stenosis from chronic gastric ulcer, 199 syndrome, 175 of Hartmann, 199 Pyloroplasty. 144, 189, 254 Finney's, 192, 255 in gastric ulcer, 204, 276, 304 494 INDEX OF SUBJECTS Pyloroplasty in dilatation of stomach, 131 in gastric ulcer, 203, 227, 277 in obstruction from gastric ulcer, 127 in pyloric obstruction, 53, 92 cases illustrating, 58, 59 spasm, 130 Pylorus, adenoma of, 72 cancer of, 98 frequency, 37 obstruction from, 71. See also Pyloric obstructioyi, malignant. fish-hook, 129 malignant diseases of, 93 obstruction of. See Pyloric obstruction. sarcoma of, 72 spasm of. See Pyloric spasm. stricture of, 50 from gastric ulcer, 50 Rectal feeding after operations on stomach, 117 Regurgitant vomiting after gastro- enterostomy, 112, 138, 148, 258 Doyen's operation for preven- tion, 112 entero-anastomosis in, 112, 138 Rutkowski's operation for pre- vention, 112 Remittent bilious fever, 455 Renal colic and gall-stone colic, differ- entiation, 441 Resection oi stomach, 40 BUlroth's method, 41 Retrograde dilatation and gastrotomy in impassable stricture of esophagus, 27 Reverse mucous currents, 311 Robson's chronic ulcer of stomach, 183 incision in gastric cancer, opposite page 185 technic for operations on bile-ducts, 424 Rodman's operation in gastric ulcer, 203, 227, 242, 277, 305 Rontgen rays, locating open buckle in esophagus with, 10 Round ulcer of stomach, 181 acute, 181 chronic, 181 Roux's method of gastro-enterostomy, 261 Rupture of gall-bladder, 333 Rutkowski's operation to prevent re- gurgitant vomiting after gastro- enterostomy, 112 Saxtoeixi's duct, 236, 466 Sarcoma of pylorus, 72 Scirrhus of stomach, 97 Secretin, 316 Senn's method of gastro-enterostomy, 113 Sentinel enlarged lymph-nodes in locat- ing gastric ulcer, 184, 271 Septic peritonitis. Fowler's position in, 322 staphylococcus albus in, 323 Shock in radical operation for cancer of stomach, 215 Silk sutures, buried, value of, 86, 166 Silver wire sutures, buried, value of, 86, 166 Slumbering gall-stone, 392, 408 Smith's incision in operations on gall- bladder, 342 Spasm, pyloric, 129, 175, 183 diagnosis of, 142 in gastric ulcer, 273 physiology of, 318 pyloroplasty in, 130 Sphincter, cecocolic, 310 Ssabanejew-Frank gastrostomy in gas- tric cancer, 109 operation in obstruction of cardiac orifice of stomach, 90 Stamm's gastrostomy in gastric cancer, 109 Staphylococcus albus in septic periton- itis, 323 Stenosis. See Stricture. Stomach, 31 INDEX OF HUBJECT8 49.5 Slomiuh, adenoc'urciriuiiiii of, !J7 ariiitoiny of, iiVZ, 'Mi and associated viscera, surgery of, j)riiici|)l'cs iiiKlerlyiiij,', .'iOS atcjiiic (iiliitalioii, in nciirasllienia, 176, !eoi, m bacteria in, 122, 319 bile in, after gastro-cnterostoniy, 138, 2!)3 for (iiioilciial ulcer, 250 blood-siipply of, 02, 121 blood-vessels of, 31-t Hristow's \vater-l)ottle, 07 cancer of. See Gastric cancer. cardiac orifice of, 33 method of exploring, 85 obstruction, Kadcr's operation in, 90 Ssabancjcw-Frank operation in, 90 surgical treatment, 89 ^^ itzel's operation in, 89 control of cerebrospinal nerves over, 317, 318 of sympathetic nerves over, 318 dilatation of, 128 and malignant pyloric obstruction, differentiation, 72 Brandt's operation for, 48 gastro-cnterostomy in, 131 gastroplication in, 131 gastrotomy for, 38 method, 123 not of organic origin, 17fi, 201 operation in, 18-t pyloroplasty in. 131 results, 140 surgery, 35, 38 diseases of, diagnosis, C2, 171 exploratory operations in, 84, 123, 124, 141 non-malignant, enlarged lymphatic glands in, 53, C5, 85 surgical treatment, 82, 121 treatment, C2 displacement of, effect of weight of tumors, 34 St<)ina4 value in diagnosis, 43 embryology of, 309 emptying of, at time of operation, 124 examination of, methods, 123 test-meals for, 123 fistula of, surgery, 35 suture in, 38 foreign bodies in, gastrotomy for, 38 potato diet for, 35 surgery, 35 fragments of cancer in, 95 functions of, G3, 122, 140, 194, 236 funnel shape of, traction weight of small bowel producing, opposite page 142 gunshot wound of, gastrorrhaphy in, 69 hour-glass, from gastric ulcer, treat- ment, 127 gastro-enterostomy in, 137 resection in, 305 treatment, 90 injuries of, gastrorrhaphy in, 38 surgery, 35 lavage of, before gastric {)perations, 34, 115 before operations on stomach, 34 lesser curvature, anatomy of, 312 blood-vessels of, 315 lymph-glands of, 85, 100, 185, 210. 284, 315 enlarged, in non-malignant gastric disease, 53, 65. 85 sentinel, in locating gastric ulcer, 184, 271 malignant disease of. See Cancer of stomach. mechanical interference with action, 43 gastro-enterostomy in, 46 gastrostomy in. 45 gastrotomy in, 44 496 INDEX OF SUBJECTS Stomach, mechanical interference with action, methods of diag- nosis, 43 omental hernia as cause, 48 pylorectomy in, 47 treatment, 43 method of emptying contents, 312 of examination, 43, 44, 123 motor power of, reduced, in gastric cancer, 94 mucous erosion of, 181, 182 muscular action of, 219, 312 nerve-supply of, 317 new growths of, surgical treatment, 86 obstruction of, benign, operations for, 67 surgical treatment, 89 operations on, 38, 180 after-care, 117 after-treatment, 34 anesthesia in, 83, 117 aspiration pneumonia after, 83 cocain anesthesia in, 117 diet after, 117 diet before, 116 distention of stomach before, 34 division of gastrohepatic omentum in, 63 Fenger's incision in, 38 incision in, 84 lavage before, 34, 115 lung complications after, 149 preparation of patient before, 77 pneumonia after, 116 preparation for, 115 rectal feeding after, 117 review of cases, 178 outlining of, methods, 34, 123 perforation of, 238, 239 physiologic facts concerning, 262 physiology of, 316 position of, 122 preparation of, for operation, 115 pyloric end, cancer of, adhesions in, 209 extension to other organs, 209 laboratory methods of diagno- sis, 208 Stc»mach, pyloric end, cancer of, lym- phatic infection in, 210 radical operation for, 207 after-treatment, 216 control of hemorrhage, 212 shock in, 215 steps, 211-217 significance of palpable tumor, 208 resection of, 40 Billroth's method, 41 scirrhus of, 97 storage function, 312 supersecretion of, 131 surgery of, 33, 169 principles underlying, 308 problems relating to, 139 tramnatisms of, surgical treatment, 92 ulcer of. See Gastric ulcer. wounds of, gastrorrhaphy in, 38 surgery, 35 Stomach-contents, tests of, for gastric cancer, 163, 172, 200 Stomach-tube, value of, 172, 220 Stone-building catarrh, 393, 409 Storage function of stomach, 312 Stricture, cicatricial, of esophagus, 11 and esophageal diverticula, dif- ferentiation, 14 bougies in, 16 diagnosis, 13 dilatable, treatment, 16 esophagoscopy in, ,14 etiology, 11 general character, 11 impassable, gastrostomy in, 28 gastrotomy and retrograde dilatation in, 27 treatment, 26 location, 13 non-dilatable. Abbe's string-saw method in, 21 Billroth's operation in, 19 electricity in, 26 esophagectomy in, 20 Gussenbauer's operation in, 20 INDEX OF SIBJECTS K)7 Striclurt', (•i<'atri(i;il, iKni-dilal.iljIc, internal csuplmgotoniy in, 25 OcIisiut's operation in, '21, 24 treatment, 19 prognosis, 14 treatment, 15 congenital, of esophagus, 13 fil)rous, of esophagus, 12 of common (kict of Hver, report of operation for, 'i-2H of pylorus. .■>() from gastric ulcer, .'iO, 199 simple, of esophagus, I '2, 13 Subdiaphragmatic abscess from i)er- forated gastric ulcer, 3G Supersecretion of stomach, 131 Suppurative cholecystitis, 376 Suture angle, fatal, IGG in gastric fistula, 38 method in gastro-enterostomy, 08, 91, 110, 114, 153. 258, 260 contraction of anastomotic open- ing after, 115 of gall-bladder after removal of stones, 333, 334 in wound, dangers, 343 of gastric stump after radical opera- tion for cancer, 289 Sutures, silk, buried, value of. 80, 106 silver wire, buried, value of, 80, 166 Sympathetic nerves, control of, over stomach, 318 Tait's incision in operations on gall- bladder, 342, 350 Tension in gall-stones, 395 Test-meals in examination of stomach, 123 in gastric cancer, 64 value of, 34 in diagnosis of gastric cancer, 43 Tetany, gastric, 220 in gastric ulcer, 199 Tiffany's incision for exploratory ex- amination of stouuieh, 124, 111 32 Trachea, foreign bodies in, 3 and foreign bodies in esophagus, ditrerentiation, 3 diagnosis, 4 illustrative cases, prognosis, 4 symptoms, 3, 4 tracheotomy for, 5 treatnu'nt, 4 Tracheotomy for foreign bodies in trachea, 5 Traumatism of stomach, surgical treat- ment, 92 Treitz's ligament, 265 Tuholske's operation in gastric cancer, 108 Tumors, cfTect of weight, in displace- ment of stomach, 34 Typhoid fever as cause of gall-stones, 377 I'lcer, gastric. See Gastric ulcer. of duodenum, 178, 223, 244 acute perforating, operation in, 247 and gastric ulcer, relative fre- quency, 268, 269 chronic, 267 from surgical standpoint, 194 gastro-enterostomy in, 242 pain in, 198 surgical treatment, 301 symptoms, 198 classification, 270 clinical, 270 frequency, 244, 268 gastro-enterostomy in, 249, 302 bile in stonuich after, 250 hemorrhage in, operative treat- ment, 248 indurated, 270 medical, 270 treatment, results of, 274 non-indin-aled, 270, 272 operative indications, 247 sex frefpiency, 245 surgical treatment, 299 498 INDEX OF SUBJECTS Ulcer of duodenum, symptoms, 223 treatment, surgical, 299 with gall-bladder and liver compli- cations, operation in, 248 with gastric complications, opera- tion in, 248 Ulceration of esophagus, cicatricial stricture from, 11, 12 Urine in gall-stones, 444 in gastric cancer, 95 Valve formation in pyloric obstruc- tion, 50, 51, 90, 128, 142, 273 Vestibule of small intestines, 234 Vicious circle after gastro-enterostomy, 138, 146, 258 Vomiting in gastric cancer, 64, 94 ulcer, 224 in pyloric obstruction, 64 of blood as sign of gastric injury, 35 regurgitant, after gastro-enterostomy, 112, 138, 148, 258 Doyen's operation for preven- tion. 112 Vomiting regurgitant, after gastro- enterostomy, entero-anastomo- sis in, 112, 138 Rutkowski's operation to pre- vent, 112 Water, value of, in body, 311 Water-bottle stomach, Bristow's, 97 Wirsung's duct, 235, 466 W^itzel's method of gastrostomy, 40 in gastric cancer, 109 operation in obstruction of cardiac orifice of stomach, 89 Wblfler's method of gastro-enterostomy, 113 operation in malignant pyloric ob- struction, 76 Wounds, gunshot, of liver, 346 of stomach, gastrorrhaphy in, 69 of stomach, gastrorrhaphy in, 38 siirgery, 35 X-RAYS. See Rontgen rays. RD 14M45 C 1 A codec'