& THf -V £ UMAftlK % Digitized by the Internet Archive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/studiesfromsaranOOkrau STUDIES THE SARANAC ■ LABORATORY FOR THE STUDY OF TUBERCULOSIS /, STUDIES IN IMMUNITY TO TUBERCULOSIS I. Experimental studies on the cutaneous reaction to tuberculo-protein First Paper: Factors governing the reaction II. The anaphylactic state in its relation to resistance to tuberculous infection and tuberculous disease (An experimental study) III. Concerning the general tuberculin reaction A. K. KRAUSE, M.D. PUBLICATION OFFICE OF THE JOURNAL OF MEDICAL RESEARCH 240 LONGWOOD AVENUE, BOSTON, MASS., U.S.A. J// MS Studies in Immunity to Tuberculosis.* experimental studies on the cutaneous reaction to tuberculo-protein. First Paper. FACTORS GOVERNING THE REACTION. A. K. Krause, M.D. (From the Saranac Laboratory for the Study of Tuberculosis.) It is remarkable that the phenomenon of cutaneous hyper- sensitiveness has not attracted more of the careful and earnest experimental work that it deserves. Hailed upon its dis- covery as of diagnostic importance in tuberculosis, the appli- cation of the skin test upon human beings by a multitude of observers soon disclosed that in diagnosis and prognosis the test was of very limited value. It is now ten years since von Pirquet and Schick's first publication and an enormous literature has accumulated, outlining studies as they have been made on skin hypersensitiveness on man. Out of this mass of contradictory reports has come comparatively little definite information, for the reason that the work was done on man under conditions that were not easy of comparison and control. It is at once plain that a study of this biologi- cal test as correlated with anatomical conditions could be made only rarely and in isolated instances. Again, it was only in individual and exceptional cases that prolonged observation before infection, at the time of infection, and * Received for publication June 10, 1916. 2 KRAUSE. during the course of disease could be carried out. As a rule the only clinical opportunity was to correlate the results of skin tests with symptoms and physical signs ; and it is notorious how mere symptoms and physical signs can lead us astray when we attempt from them to visualize what may be going on pathologically in a condition like chronic phthisis. However, observation on man has contributed several important facts. It has taught us that a positive cutaneous test means infection or the presence of tubercle, although the disease tuberculosis does not of necessity exist. It has shown that fever and intercurrent infections may on occasion blunt or abolish the capacity of the skin to react. As an index of infection it has greatly advanced our knowledge concerning the epidemiology of tuberculosis. But most of the underlying causes of the skin reactions and the reasons for their fluctuations have apparently eluded the clinical observer. It would seem only natural, in view of what I have just said, that students of tuberculosis would resort to animal experimentation to acquire data under controllable condi- tions. Here the infecting agent could be varied at will, the time of infection could be definitely fixed, the sensi- tiveness of the animal at any period after infection could be noted, and the condition of the animal at any time could be exactly determined by immediate autopsy. At first there was good reason why animal experiment was not fruitful, for it was found that the skin of the ordinary laboratory animals that are susceptible to tuberculosis {e.g., the guinea-pig and the rabbit) would not react to the v. Pirquet test (the method of scarification). Rut this obsta- cle no longer interfered after Mendel, 1 and Moussu and Mantoux 2 described their method of the intracutaneous introduction of tuberculin to which a tuberculous guinea-pig will react very sharply. Notwithstanding the ease and applicability of the Moussu- Mantoux method the literature has remained singularly free from experimental studies on skin sensitiveness. It is no STUDIES IN IMMUNITY TO TUBERCULOSIS. 3 exaggeration to say that their number can be counted on the fingers of one hand. Of these Romer's 3 and Baldwin's 4 stand out most prominently. Baldwin was concerned with the conditions under which skin sensitiveness might occur in guinea-pigs. He injected a number of animals with various products of the tubercle bacillus and of tuberculous foci, but could obtain no positive reactions unless he produced tubercles by the inoculation of either living or dead bacilli. Therefore, so far as he went, he concluded that only the presence of anatomic tubercle endows an animal with the capacity to react cutaneously to tuberculo-protein. Romer was interested more in the meaning of the various degrees of reactive power of the animals. He concludes that the intensity of the reaction varies directly with the extent and advance of the tuberculous process. As our knowledge of the processes of immunity to tuber- culous infection has grown, the desirability of a better insight into the various phases of the skin reaction has become more and more apparent. At the Saranac Labora- tory we have been practising the skin test on animals for about seven years. As time went on I became more and more impressed with the fact that at many points the factors that governed many features of the cutaneous test approached or resembled those that we have come to learn as being associated with the laws of immunity to tuberculous infec- tion. About three years ago I therefore resolved to under- take a more or less exhaustive study on animals from several points of view: first, to study the details of the mere occurrence or non-occurrence of the skin reaction (its rela- tion to time of infection and extent of infection, its modifica- tion by other reactions, etc.) ; and secondly, to study the possible bearing that the inflammatory skin reaction may have on immunity to cutaneous re-infection (see Romer 5 on cutaneous re-infection). The present paper is the first of a series in which I hope to communicate the results of these studies, and shall deal as closely as possible only with the first of the problems outlined above. 4 KRAUSE. A description of the points of resemblance and difference between Romer's method of approach and my own is here necessary. The culture used for infection. — In all his experiments Romer inoculated his animals with a comparatively virulent strain of living human or bovine bacilli. He introduces these in amounts varying from .1 to .0000025 milligram, or, when he uses an emulsion of Bovovaccine, from .01 to .OOOOOOOl cubic centimeter. All but a very few of the animals became infected ; in several which received exces- sively high dilutions, he could find no lesion at autopsy. Nearly all of his animals, too, if allowed to live long enough, showed progressive tuberculosis. At any given time the comparative extent of tuberculosis in a given series varied, depending on the number of bacilli which the individuals received. Practically all of his animals which came to autopsy exhibited macroscopic visceral tuberculosis. He practised intracutaneous tests on these animals at various periods after infection and studied the changing intensity -of reaction. In the work which will be presented in this paper a study was made, not of the variations in reaction as these occurred in animals that had been infected with different quantities of virulent strains, but of the reactions induced in animals infected by two different strains of bacilli which produced two very different and easily distinguished effects. I have worked constantly with both these strains of human tubercle bacilli, Strain Ri and Strain H37, for seven years, and am thoroughly familiar with their invasiveness under many different conditions. During all this time their virulence has remained fairly definite and constant. Strain H37 is of fair virulence and in the doses which we employ will always bring about progressive tuberculosis in guinea-pigs, involv- ing the viscera and leading ultimately to the death of the animal. Strain Ri will always produce regional lymphatic gland tuberculosis after subcutaneous inoculation, but will hardly ever cause gross visceral lesions. Animals after Ri infection always remain in good constitutional condition and STUDIES IN IMMUNITY TO TUBERCULOSIS. 5 after a while glands which were once enlarged tend to return to their former size. Two years after infection the viscera are completely free from any trace of tuberculosis and the regional glands are apparently almost normal. Microscopically these glands at this time are often found to be completely sclerotic, although in a few animals I have seen histological tubercle walled off by dense fibrous tissue. This is a very different picture from that which one sees for several months after Ri infection when the regional glands are usually partially caseous. The time at which the Ri infection stops its march along the lymphatic chain and begins to retrogress is a point that is very difficult to deter- mine with even approximate accuracy. But my experience has taught me that this change begins to disclose itself at from three or four to six months after infection. I judge that this has occurred only by paying attention to the char- acter of the regional glands, when I find that their size is diminishing and their consistence is beginning to be less hard. It must be remembered that in very mild, very slow, and very chronic tuberculosis changes occur with amazing slowness; they are likely to be appreciable only after inter- vals of months. While the Ri bacillus always brings about local glandular changes in guinea-pigs, it has never in my observation produced lesions in them sufficient to cause death. The use of these two different strains of tubercle bacilli provided the opportunity to study the reaction under two very distinct conditions : the one in the presence of advanc- ing tuberculosis ; the other in animals with stationary, self- limited, or healing infection. The conditions were some- what simpler and more sharply differentiated than those in Romer's animals. Method of infection. — Romer infected most of his guinea- pigs intraperitoneally and some subcutaneously. All of the animals in these experiments were infected subcutaneously, so that track could be kept of the glandular changes. 6 KRAUSE. The antigen used. — Romer used old tuberculin in testing his animals. As is well known this is a very complex product consisting of a boiled, filtered, condensed, glyceri- nated extract of tubercle bacilli and contains a high propor- tion of peptone and common salt. He used .02 cubic centi- meter of tuberculin always in a bulk of .1 cubic centimeter of fluid (a dilution of I to 5 in physiological salt solution). In our own experiments we always used a water extract filtrate of dried, pulverized, human (H37) bacilli, made iso- tonic by the addition of common salt, and in this work desig- nated as W.E. 179. The dried residue of W.E. 179 weighed .02 gram per cubic centimeter of extract. This W.E. was always diluted with physiological salt solution in equal pro- portions and a total bulk of .2 cubic centimeter injected into the skin. Every dose represented therefore .002 gram of dried substance. The periods of testing. — In his tables Romer shows no tests that were made earlier than twenty- one days after infec- tion and none that were continued more than eight months after infection. In the following experiments tests were begun at eleven days after infection, or as soon as palpable changes in the regional lymphatic glands were apparent, and continued at various intervals for six hundred and sixty-nine days, or almost two years, after infection. Estimation of the intensity of the reaction. — As is well known a reaction is positive when inflammation appears within twenty-four or forty-eight hours at the site of the injection of the tuberculo-protein. This inflammation can be of all grades, from mere hyperemia to hemorrhagic necrosis. The scale of reaction that one generally meets with runs somewhat as follows: Hyperemia; hyperemia and induration; hyperemia, induration, and necrotic center; hyperemia, induration, and necrotic-hemorrhagic center. An atypical form of reaction is induration with very slight hyperemia or induration with pallid, blanched skin. Any STUDIES IN IMMUNITY TO TUBERCULOSIS. 7 marked hyperemia after twenty-four hours without indura- tion is a rarity. We consider that the reacting area must measure at least ten millimeters in its longest diameter to be noted as positive. To gain all the information possible read- ings should be made both at twenty-four and forty-eight hours after tuberculin injection. And if the reaction is of the delayed type, not reaching its height until after twenty- four hours have elapsed, a reading at seventy-two hours is not superfluous. As a rule the twenty-four-hour reading shows greater intensity than at forty-eight hours. In the following experiments the reactions will be expressed in terms of the extent of hyperemia (redness) and induration and of the amount of central necrosis and hemorrhage, if these are present. Experiments. i. Infection of animals: On June 20, 1913. thirty-one guinea-pigs were infected with living Ri tubercle bacilli as follows: Nos. 1 to 31 received subcutaneous inoculations in the right groin, each .1 cubic centi- meter of a centrifugalized emulsion. On June 28, 1913, thirty-one guinea-pigs were infected with living H37 tubercle bacilli as follows : Nos. 32 to 62 received subcutaneous inocula- tions in the right groin, each .1 cubic centimeter of a centrifugalized emulsion. The Ri emulsion was a heavier emulsion than the H37, so the Ri animals were inoculated with greater numbers of bacilli, but of a low virulent type. 2. Intracutaneous tests : Eleven days after infection : On Ri animals: July I, 1913 : Nos. 1 to 44 injected intracutaneously, each with .2 cubic centimeter diluted W.E. 179 (1 cc. = .01 gm .). July 2 (24 hours) : No reaction in any ; Superficial inguinal glands are apparently normal in all. July 3 (48 hours) : No reactions. On H37 animals: July 9, 1913: The glands of Nos. 32, 33, and 34 are apparently slightly enlarged ; those of 35 are normal ; Nos. 32 to 35 injected intracutaneously, each with .2 cubic centimeter diluted W.E. 179 (1 cc. = .01 gm.). July 11 (48 hours): No. 32, Redness and induration, 25 x 25 milli- meters; Necrotic center, 6 x 6 millimeters. No. 33, Questionable. No. 34, Negative. No. 35, Negative. 8 KRAUSE. Eleven days after infection, when no anatomic changes are palpable in the Ri animals and slight changes have appar- ently developed in the H37 animals, the Ri animals are still insensitive, while the H37 animals are becoming sensitive. Cutaneous hypersensitiveness therefore apparently begins to develop with the development of tubercle. 3. Intracutaneous tests: Eighteen days after infection: On Ri animals : July 8, 1913 : Injected Nos. 1 to 8 intracutaneously in the usual way. July 9 (24 hours) : No. 1, Negative; Gland, slightly +. No. 2, Ques- tionable; Gland, normal. No. 3, Redness and induration, 20 x 20 milli- meters; Gland, normal. No. 4, Redness, 12 x 12 millimeters ; Gland, normal. No. 5, Slight redness and induration, 15x15 millimeters; Gland, +. No. 6, Questionable; Gland, O. No. 7, Questionable; Gland, O. No. 8, Induration and pale pink areola, 25 x 20 millimeters; Necrotic center, 3 x 3 millimeters ; Gland, definitely +. At eighteen days sensitiveness and glandular changes are beginning in the Ri animals. On H37 animals: July 16, 1913 : Injected Nos. 32 to 39 intracutane- ously in the usual way. July 17 (24 hours): No 32, Hemorrhagic center, 7x7 millimeters; Area of redness, 30 x 25 millimeters ; Induration, + + + ; Gland, + +. No ^^, Redness and induration, 30x25; Induration, -\ — (- ; Glands, + H — K No. 34, Redness and induation, 30 x 20, pale center; Indura- tion, + +; Glands, -f- +. No. 35, Redness and induration, 20x20; Induration, -+- -j-; Glands, + + +. No. 36, Redness and induration, 30 x 20; Induration, + +; Glands, + +. No 37, Redness and indura- tion, 30 x 30 ; Hemorrhagic center, 5x5; Induration, + +; Glands, -f-. No. 38, Redness and induration, 25x25; Induration, ++; Glands, -) — K No. 39, Redness and induration, 25x20; Induration, -\ — hi Glands, slightly +• July 18 (48 hours) : No. 32, Necrotic center, 7x7] Redness, 30 x 25 ; Induration, + + +• No. ^3^ Redness and induration, 15x15; Indura- tion, H — (-• No. 34, Redness and induration, 25 x 15 ; Induration, -\ — (- ; Necrotic center, 2x2. No. 35, Redness and induration, 15x15; Indu- ration, + +. No. 36, Redness and induration, 20x20; Induration, + -f-. No. 37, Redness and induration. 30 x 20 ; Necrotic center, 3 x ^^ Induration, -\ — K No. 38, Redness and induration, 20 x 15 ; Induration, + +. No. 39, Redness and induration, 20 x 15 ; Induration, + +. At eighteen days the H37 animals are uniformly sensitive, and much more so than the Ri guinea-pigs. The glandular reaction of the former to infection is much more marked STUDIES IN IMMUNITY TO TUBERCULOSIS. 9 than that in the latter. From these results there are already suggestions that the intensity of the reaction runs parallel with the amount of tuberculous involvement and conse- quently with the amount of absorption from the foci of disease. From this time on, once the establishment of infection and disease in both lots of animals was assured, the' tests were not always made on both Ri and H37 animals at exactly the same intervals after infection. Hundreds of inoculations on guinea-pigs with H37 had already proved for us that the course of H37 infection is very regular. After subcutaneous inoculation the spleen generally becomes involved about the fourth week, the lungs and liver about the fifth and sixth weeks, and death usually follows from the sixth week on. There was no reason to believe that the results of infection would be any different in these animals. 4. Intracutaneous tests : Twenty-five days after inoculation : On Ri animals: July 15, 1913 : Injected Nos. 1 to 12 intracutaneously in the usual way. July 16 (24 hours) : No. 1, Negative ; Gland, -f-. No. 2, Redness and induration, 15x15; Induration, slight ; Central necrosis and hemorrhage, 4x4; Gland, ?. No. 3, Redness (pale), 30x25 ; Induration,-!-; Gland. + . No. 4, Redness (pale), 20 x 15 ; Induration, slight ; Gland, -(-• No 5, Redness, 25 x 20, pale center; Induration, slight; Gland, -| — (-• No. 6, Redness, 25 x 20, pale center ; Induration, +; Gland, + +. No. 7, Red- ness (pale), 30 x 15 ; Induration, + ; Gland, +. No. 8, Redness, 20 x 20, pale center; Induration, + ; Gland, +. No. 9, Redness, 25 x 20, pale center; Induration, -f- ; Gland, +. No. 10, Pallid reaction, 15 x 15; Induration, + ; Gland, + +. No. 11, Necrotic hemorrhagic center, 4x4; Redness, 25x25; Induration, +; Glands, +• No. 12, Redness, 25 x 20 ; Induration, -f- + ; Glands, + +. July 17 (48 hours) : No. 1, Negative. No. 2, Central necrosis, 6x6; Redness, 10 x 10; Induration, -(-. No. 3, Redness and induration, 15x15; not so pale this morning. No. 4, Almost completely faded. No 5, Redness and induration, 15 x 15. No 6, Redness and induration, 15x15. No. 7, Almost completely faded; Slight induration. No. 8, Redness and induration, 15 x 15. No. 9, Necrotic center, 2x2; Red- ness, 20 x 20 ; Induration, +. No. 10, Necrotic center, 7 xj ; Redness, 10 x 10; Induration, +. No. 11, Necrotic center, 4x4; Redness, 10 x 10; Induration, +. No. 12, Necrotic center, 5x5; Redness, 10 x 10; Induration, +. IO KRAUSE. At twenty-five days the reactions in the Ri animals are increasing in intensity, but are not so marked as those in the H37 animals eighteen days after infection. 5. Intracutaneous tests : Thirty days after infection ; On H37 animals: July 28, 1913 : Injected Nos. 32 to 39 intracutane- ous^ in the usual way. July 29 (24 hours) : No. 32, Redness and induration, 40 x 30; Hem- orrhagic center, 20 x 15 ; Induration, + + -f- ; Glands, + + +. No. 23, Redness and induration, 40 x 30 ; Hemorrhagic center, 5x5; Induration, + + +; Glands, + -\ — K No. 34, Redness and induration, 30 x 30 ; Induration, + -j-; Glands, + +. No. 35, Redness and induration, 35 x 40 ; Induration, -f + + ; Glands, ++-(-. No. 36, Redness and indura- tion, 20 x 25 ; Induration, + + + ; Hemorrhagic center, 5x5; Glands, + + +• No. 37, Redness and induration, 35 x 35 ; Induration, -f- + + ; Hemorrhagic center, 5x5; Glands, + + +. No 38, Redness and induration, 30 x 30 (pale) ; Induration, -f- + ; Hemorrhagic center, 3x3; Glands, -\ — (- +. No. 39, Redness and induration, 20 x 20 (pale) ; Induration, -\ — (- ; Hemorrhagic center, 3 x 3 ; Glands, +. July 30 (48 hours): No. 32, Redness and induration, 30x30; Hemorrhagic, necrotic center, 20 x 20. No. 33, Redness and induration, 25 x 20; Necrotic center, 5x5. No. 34, Redness and induration, 25 X20 ; Necrotic center, 5x5. No. 35, Redness and induration, 20 x 20 ; Ne- crotic center, 5x5. No. 36, Redness and induration, 15 x 15; Necrotic center, 5x5. No. 37, Redness and induration, 35 x 35 ; Necrotic center, ro x 10. No. 38, Redness and induration, 30x30; Necrotic center, 5x5. No. 39, Redness and induration, 20 x 20; Necrotic center, 5x5. With the development and extension of the disease hyper- sensitiveness is increasing. 6. Intracutaneous tests : Thirty-eight days after infection: On Ri animals: July 28, 1913 : Injected Nos. 1 to 16 intracuta- neous^ in the usual way. July 29 (24 hours) : No. 1, Redness and induration, 20 x 20 ; Hemor- rhagic center, 5x5; Glands, -f- +. No. 2, Redness and induration, 30x20; Hemorrhagic center, 3x3; Glands, -j-. No. 3, Redness and induration, 30 x 20 ; Induration, -f- + ; Pale center, 15 x 15 ; Glands, +• No. 4, Redness and induration, 25 x 25 ; Pale center, 5x5; Glands, +■ No. 5, Redness and induration, 25 x 25 ; Induration, -\ — (- ; Glands, -\ — (-• No. 6, Redness and induration, 25x30; Induration, ++; Glands, -j-. No. 7, Redness and induration, 20 x 20 ; Glands, -(-• No. 8, Red- ness and induration, 25 x 25 ; Induration, + + ; Hemorrhagic center, 5x5; Glands, + + +• No. 9, Redness and induration, 25 x 30; Hem- orrhagic center, 8x8; Induration, -\ — | — (-; Glands, slightly +. No. 10, Redness and induration, 25 x 25 ; Hemorrhagic center, 5x5; Induration, STUDIES IN IMMUNITY TO TUBERCULOSIS. I I + + +; Glands, + +. No. n, Redness and induration, 30x35; Hemorrhagic center, 10 x 10; Induration, + + +; Glands, +. No. 12, Redness and induration, 30 x 25 ; Hemorrhagic center, 10 x 10; Indura- tion, + +; Glands, + +. No 13, Redness and induration, 25x20; Induration, + ; Glands, +. No. 14, Redness and induration, 30 x 25 ; Induration, +; Glands, +.• No. 15, Redness and induration, 30x30; Induration, + -f- ; Glands, + +• No. 16, Redness and induration, 30 x 35 ; Induration, + + ; Glands, + +. July 30 (48 hours) : No. 1, Redness and induration, 15 x 15 ; Necrotic center, 3x3. No. 2, Faded; Redness and induration, 8x8. No. 3, Redness and induration, 20 x 25 ; Necrotic center, 3x3. No. 4, Redness and induration, 15x15; Necrotic center, 3x3. No. 5, Redness and induration, 15 x 15 ; Necrotic center, 3x3. No. 6, Redness and indura- tion, 15 x 15 ; Necrotic center, 5x5. No. 7, Slight redness and indura- tion; Faded. No. 8, Redness and induration, 20 x 20; Necrotic center, 5 x 5. No. 9, Redness and induration, 15 x 20; Necrotic center, 2x2. No. 10, Redness and induration, 20 x 20 ; Necrotic center, 5x5. No. 1 1 , Redness and induration, 25 x 30 ; Necrotic center, 3x3. No. 12, Red- ness and induration, 20 x 20; Necrotic center, 7x7. No. 13, Redness and induration, 15 x 15 ; Necrotic center, 4x4. No. 14, Faded; Slight redness and induration. No. 15, Redness and induration, 20x20; Necrotic center, 4x5. No. 16, Redness and induration, 25 x 20 ; Necrotic center, 2x2. It will be seen that the reactions in the Ri animals are steadily increasing in severity and at thirty-eight days resemble those noted in the H37 animals at thirty days. They should be compared with the following results obtained on H37 guinea-pigs, also at thirty-eight days: On H37 animals: August 5, 1913 : Nos. 40 to 43 injected intracuta- neous^ in the usual way. August 6 (24 hours) : No. 40, Redness and induration, 40 x 30 ; Induration, + + +; Hemorrhagic necrotic center, 10x10; Glands, + -+-; Ulcerated at site of infection. No. 41, Redness and induration, 40x40; Induration, + + +; Hemorrhagic necrotic center, 15x15; Glands, + +. No. 42, Redness and induration. 40x40; Induration, + + + ; Hemorrhagic necrotic center, 10 x 10 ; Glands, + + ; Ulcerated at site of infection. No. 43, Redness and induration, 40x40; Indura- tion, + + +; Hemorrhagic necrotic center, 15 x 15; Glands, + + +• August 7 (48 hours) : No. 40, Redness and induration, 30 x 30 ; Induration, + + + ; Necrosis, 10x10. No. 41, Redness and indura- tion, 35 x 35 ; Induration, + + +; Necrosis, 15 x 15. No. 42, Redness and induration, 25x25; Induration, + + +; Necrosis, 15x15. No. 43, Redness and induration, 20x20; Induration, + + ; Necrosis, 10 x 10. 12 KRAUSE. A glance at the above results shows at once that the reactions in the guinea-pigs suffering from progressive tuberculosis are still increasing in severity. They are much more intense than those in the Ri animals at the same time. From this time on the Ri guinea-pigs were not tested until the one hundred and twenty-third day after infection, and were subsequently tested at intervals until almost two years had elapsed since their infection. It was thought wisest not to test them too often, but to conserve them and study the reactions while healing and arrest of the tuberculous process were going on. Before they were killed the H37 animals were tested once more, namely, on the forty-sixth day after infection, as follows: 7. Intracutaneous tests: Forty-six days after infection : On H37 animals: August 13, 1913 : Nos. 44 to 62 were injected intracutaneous!)' in the usual way. August 14 (24 hours): No. 44, Redness and induration, 35x35; Induration, + -j — \-\ Hemorrhagic necrosis, 15x15; Glands, -+-■ No. 45, Pallid type of reaction, 25x25; Induration, + ; Glands, -f- +. No. 46, Redness and induration, 30x30; Induration, -| — \-\ Hemor- rhagic center, 15x10. No. 47, Redness and induration, 35x30; Induration, -| — I — f- ; Glands, + + +. No. 48, Redness and induration, 50 x 40 ; Induration, + -f- -f- + ; Glands, + + + +. No. 49, Redness and induration, 35x30; Induration, + H — h; Hemorrhagic center, 10x10; Glands, H — | — K No. 50, Redness and induration, 35x35; Induration, -\ — (- -f- ; Hemorrhagic center, 20x10; Glands, + + +• No 51, Redness and induration, 35x35; Induration, -f- 4- ; Hemor- rhagic center, 15 x 10; Glands, -)--)-. No. 52, Redness and induration, 30x30; Induration, +; Glands, + +. No. 53, Pallid type, 20x20; Induration, slight; Hemorrhagic center, 10x10; Glands, + + +• No. 54, Redness and induration, 35 x 30; Induration, + + +; Hemor- rhagic center, 20x20; Glands, O. No. 55, Redness and induration, 30x30; Induration, + +; Glands, + -\ — K No. 56, Pallid type, 25 x 25 ; Induration, H — h| Hemorrhagic center, 10 x 10; Glands, -f- +• No. 57, Redness and induration, 30 x 30 ; Induration, -J- + +; Glands, + + -(-• No. 58, Redness and induration, 30 x 30 ; Induration, + -f- -f- ; Hemorrhagic center, 10 x 10; Glands, -\ — \- -j-. No. 59, Redness and induration, 30 x 30 ; Induration, + -f- -f- ; Glands, + H — I — K No. 60, Redness and induration, 25 x 25 ; Induration, -| — (- + ; Hemorrhagic center, 10x10; Glands, -\ — | — (-• No. 6r, Redness and induration, 25x25; Induration, -)- -| — (-; Glands, -) — | — (-■ No. 62, Redness and induration, 25 x 25 ; Induration, + -f-; Glands, +. August 15 (48 hours): No. 44, Redness and induration, 25x25; STUDIES IN IMMUNITY TO TUBERCULOSIS. 1 3 Induration, + + +; Necrosis, 15 x 15. No. 45, Redness and indura- tion, 20 x 15 ; Induration, -f- ; Fading. No. 46, Redness and induration, 25x25; Induration, -f + + ; Hemorrhagic necrosis, 15 x 10. No. 47, Redness and induration, 25 x 20 ; Induration, + + + ; Necrosis, 7x7. No. 48, Redness and induration, 25 x 20 ; Induration, + + ; Necrosis, 5x5. No. 49, Redness and induration, 25 x 25 ; Induration, + + + ; Necrosis, 15x15. No. 50, Redness and induration, 30x20; Indura- tion, + + +; Necrosis, 20x20. No. 51, Redness and induration, 25x20; Induration, ++; Necrosis, 10x10. No. 52, Redness and induration, 25x25; Induration, + +; Necrosis, 10x10. No. 53, Redness and induration, 25 x 20 ; Induration, -\ — |- ; Hemorrhagic necrosis, 10 x 10. No. 54, Redness and induration, 25 x 25 ; Induration, + +; Necrosis, 10x10. No. 55, Redness and induration, 25x25; Induration, -\ — h ; Necrosis, 10 x 10. No. 56, Redness and induration, 25x30; Induration, + -| — (-; Scab, 5x10. No. 57, Redness and induration, 25 x 25 ; Induration, + + +. No. 58, Redness and indura- tion, 30 x 20 ; Induration, -| — | — (- ; Necrosis, 20 x 10. No. 59, Redness and induration, 20 x 20 ; Induration, -\ — (- -f- ; Necrosis, 12 x 12. No. 60, Redness and induration, 25 x 25 ; Induration, -\ — f- + ; Hemorrhagic necrosis, 10 x 10. No. 61, Redness and induration, 20 x 20 ; Induration, H — h; Necrosis, 5x5. No. 62, Redness and induration, 25x25; Induration, -\ — | — (-• These were the last tests that were made on members of the H37 series — those with progressive tuberculosis. To summarize: they displayed increasing cutaneous hyper- sensitiveness from eleven days after infection, when the first regional glandular changes were noted, up to forty-six days, when the disease was well advanced. At every stage, too, their reactions were more intense than those in guinea-pigs suffering from Ri, self-limited tuberculosis. On August 16 and 17 autopsies were performed on various individuals of the H37 group, and advanced tuberculosis of the viscera, lungs, livers, and spleens was found in all. 8. Intracutaneous tests : One hundred and twenty-three days after infec- tion : On Ri animals : October 21, 1913 : Injected Nos. 1 to 8 intracuta- neously in the usual way. October 22 (24 hours): No. 1, Redness and induration, 20 x 20 ; Induration, -| — (- ; Glands, -\ — K No. 2, Pallid type; Induration, 15 x 15 ; Induration, -f- ; Glands, + -f-. No. 3, Redness and induration, 30 x 25 ; Induration, -\ — (- ; Glands, +. No. 4, Redness and indura- tion, 15 x 15; Induration, -| — (- ; Glands, -\ — (-■ No. 5, Redness and 14 KRAUSE. induration, 20 x 20; Induration, ++; Glands, + +• No. 6, Redness and induration, 30 x 20; Induration, -\ — h ; Glands, -\ — (-■ No. 7, Red- ness and induration, 30 x 20 ; Induration, -\ — (- ; Glands, +. No. 8, Redness and induration, 30 x 30; Induration, -f- +; Glands, -\ — I — K October 23 (48 hours) : No. 1, Subsiding ; Red papule, 8x8. No. 2, Subsiding; Red papule, 10 x 10. No. 3, Redness and induration, 15 x 15. No. 4, Subsiding; Red papule, 8x8. No. 5, Redness and induration, 25 x 25. No. 6, Red papule, 10 x 10. No. 7, Subsiding; Small papule. No. 8, Red papule, 12 x 12. The reactions are distinctly less intense than in the same group two months ago. The disease is evidently retrogress- ing. On October 23 autopsies were performed on Nos. 1 to 8. No animal showed changes that extended farther than the retroperitoneal glands. As a rule the regional glands are not opaque: they are slightly to moderately enlarged, but none is caseous. There is no trace of visceral disease. 9. Intracutaneous tests : One hundred and sixty-three days after infec- tion : On Ri animals: December 1, 1913: Injected Nos. 9 to 16 intrai_u- taneously in the usual way. December 2 (24 hours) : No. 9, Redness and induration, 30 x 30 ; Induration, -+- + +• No. 10, Redness and induration, 20 x 20 ; Indura- tion, slight. No. 11, Redness and induration, 30 x 25 ; Induration, + + +• No. 12, Redness and induration, 30 x 30; Induration, + + +• No. 13, Redness and induration, 30 x 20 ; Induration, + +. No. 14, Redness and induration, 30 x 30; Induration, + + +. No. 15, Redness and induration, 30 x 20 ; Induration, + + +■ No. 16, Redness and induration, 20 x 20 ; Induration, +• December 3 (48 hours) : No. 9, Redness and induration, 20 x 20 ; Induration, + +• No. 10, Considerably faded; Slight redness and indu- ration. No. 11, Considerably faded; Slight redness and induration, 10 x 10. No. 12, Considerably faded; Slight redness and induration, 10 x 10. No. 13, Considerably faded; Slight redness and induration, 10x10; Slight necrosis. No. 14, Like No. 13. No. 15, Like No. 13. No. 16, Redness and induration, 20 x 15 ; Induration, + -J-. On December 3 autopsies were performed on Nos. 9 to 16. The lesions are about the same in all. There is no vis- ceral disease. All animals are very fat. The regional glands are enlarged and are firmer than normal. It is noteworthy that as the disease heals the reaction tends to fade more rapidly. STUDIES IN IMMUNITY TO TUBERCULOSIS. 1 5 10. Intracutaneous tests : Two hundred and fifty-six days after infection : On Ri animals: March 3, 1914: Injected Nos. 17 to 20 intracu- taneously in the usual way. March 4 (24 hours) : No. 17, Slight induration and redness, 12 x 12 ; Glands, + and rather hard. No. 18, Pale red area, 20 x 20 ; Induration very slight; Glands, + and hard. No. 19, Induration, +; Redness not marked; Glands, O. No. 20, Pale red area, 20 x 20; Induration, + +; Glands, -f- + , hard. March 5 (48 hours) : No. 17, Faded; No longer positive. No. 18, Slight induration. No. 19, Faded; Slight induration. No. 20, Redness and induration, 15x15; Induration, -| — (-• 11. Intracutaneous tests; Two hundred and eighty-four days after infection : On Ri animals: April 1, 1914: Nos. 17 to 23 were given intracu- taneous tests with .2 cubic centimeter diluted W.E. 179 (1 cubic centi- meter = .01 gram.). At the same time they were inoculated intracu- taneous^ with living H37 and Ri bacilli on the opposite side of the body. Only a record of the intracutaneous tests will be given here. April 2 (24 hours) : No. 17, Redness and induration, 30 x 20. No. 18, Redness and induration, 35 x 35. No. 19, Redness and induration, 25x25. No. 20, Redness and induration, 25x25. No. 21, Redness and induration, 20 x 20. No. 22, Redness and induration, 10 x 10. No. 23, Redness and induration, 25 x 20. April 3 (48 hours) : No. 17, Redness and induration, 15 x 15. No. 18, Redness and induration, 20 x 20. No. 19, No induration ; Redness, 15 x 15. No. 20, Redness and induration, 15 x 15. No. 21, Redness and induration, 10 x 10. No. 22, Redness and induration, 15 x 15. No. 23, Redness and induration, 20 x 20. It will be remembered that Nos. 1 7 to 20 were tested twenty-eight days before, on March 3. They now reacted better than they did then and, in general, better than Nos. 21 to 23, which have not been tested until now. 12. Intracutaneous tests : Four hundred and twenty-eight days after infection : On Ri animals: August 21, 1914: Injected Nos. 25 to 27 intracu- taneous^ in the usual way and at the same time with virulent living H37 bacilli in the skin of the opposite side. August 22 (24 hours) : No. 25, Redness and induration, both slight; Area measures 25 x 20. No. 26, Pallid reaction, 15 x 15. No. 27, Slight redness and induration, 20 x 20. 1 6 KRAUSE. 13. Intractdaneous tests: Six hundred and sixty-nine days after infec- tion: On Ri animals: April 20, 1915 : Injected Nos. 28, 30, and 31 intra- cutaneous^ in the usual manner and at the same time with living virulent H37 bacilli in the skin of the opposite side. April 21 (24 hours) : No. 28, Pale pink and indurated, 20 x 20; Indu- ration, + +. No. 30, Pale pink and slightly indurated, 30x20. No. 31, Redness and induration, -j — (-» 20 x 20. April 22 (48 hours) : No. 28, Redness and induration, 25 x 25 ; Indu- ration, + + +. No. 30, Pink, 10 x 10; Induration very slight. No. 31, Redness and induration, 15 x 15 ; Induration, -\ — |— April 23 (72 hours) : No. 28, Redness and induration, 20 x 20 ; Indu- ration, + + +. No. 30, Subsided. No. 31, Redness and induration, 10 x 10; Induration, -| — (-• Discussion. — The skin reactions have herein been studied in two very different lots of tuberculous guinea-pigs: in one series, the Ri animals, which were suffering from a self- limited tuberculosis that after a while showed a tendency to heal; in another series, the H37 animals, which had pro- gressive tuberculosis which ultimately would have caused the death of the animals. In both series the skin reaction . became positive with the palpable establishment of tubercle : at eleven days in the H37 animals; at about eighteen days in the Ri animals. In those with progressive disease the reactions at any given time were always more intense than in those with disease well under control. In the animals with self-limited disease the vigor of the reaction followed the course of the disease. As disease developed, the reaction increased in severity. After it had reached a standstill and began to heal, the reaction became milder. The ability to react never entirely disappeared, and this raises the question as to whether an infection like tuberculosis, once it gains anatomic (not mere bacillary) foothold, is ever entirely wiped out. It is unfortunate that in the above series I have micro- scopic sections of the glands of only one animal; namely, of No. 28 (Ri), killed six hundred and seventy-nine days after infection. In this animal the right and left inguinal glands and the retroperitoneal glands were all enlarged, — much more than usual so long after infection. The bronchial STUDIES IN IMMUNITY TO TUBERCULOSIS. I J glands were not enlarged and their consistence and appear- ance were normal. There was no visceral disease. The inguinal glands were hard and sclerotic, but showed no case- ating points to the naked eye. A histological examination of these glands, however, revealed minute tubercles (giant cells and fine patches of necrosis set in the center of dense masses of fibrous tissue). Now I have often examined the glands of other Ri guinea- pigs long after infection. Many were normal to the naked eye. Some which appeared almost normal showed no anatomical tubercle, yet revealed a cross section made up wholly of scar tissue. Only recently I examined micro- scopically sections of glands of guinea-pigs that had been infected with Ri six hundred and ninety-seven days previ- ously. They showed no anatomical tubercle. There was nothing in fact except dense, hyaline, scar tissue. But the possibility always exists that somewhere tubercle, however slight, may be lurking, sufficient to set up at least a mild skin reaction. It is a remarkable phenomenon, and one that has often been noted in the human body, that although the body can keep tubercle under perfect control, it cannot eradicate it entirely. I have a very small number of observations in which I studied the effects of re-infection on a decreasing power to react cutaneously. 14. The effects of re-infection on cutaneous hypersensi- tiveness. — Numbers 28, 30, and 31 reacted six hundred and sixty-nine days after infection, as described in Section 13. At that time, April 20, 191 5 , they were given inoculations of virulent H37 bacilli in the skin. Ten days later, April 30, Nos. 30 and 31 were again tested intracutaneously with the following results : Twenty-four hours : No. 30, Redness and induration, 30x30; Indu- ration, -| — I — K No. 31, Redness and induration, 25 x 25 ; Induration, + + +• 48 hours : No. 30, Redness and induration, 20 x 20 ; Induration, H — \-. No. 31, Redness and induration, 15x15; Induration, -\ — ( — (-. 1 8 KRAUSE. It will be noted that re-infection definitely increases the capacity to react once this is diminishing. I have also noticed this feature in studies made on animals that were inoculated with living non-pathogenic acid-fast bacilli. Such bacilli are non-pathogenic, inasmuch as they do not cause progressive disease in animals. But they will give rise to abscesses at the point of inoculation and to enlargements of the regional glands. The results of inoculation are apparent much earlier than when tubercle bacilli are inoculated. Tumor (going on to abscess) appears at the site as early as two, three, or four days after inoculation. But the effects are much more transitory than when real tubercle exists. On June 17 I inoculated four white guinea-pigs subcu- taneously in the right groin with one cubic centimeter of a very heavy emulsion of living acid-fast Timothy bacilli. On June 24, seven days later, there were large swellings in the groins of all, and on this day I injected each animal intracutaneously with .2 cubic centimeter W.E. 179 (1 cubic centimeter = .01 gram). The results were as follows: June 25 (24 hours) : No. 1, Redness and induration, 20 x 20. No. 2, Redness and induration, 25 x 20. No. 3, Redness and induration, 20 x 15. No. 4, Redness and induration, 25 x 20. June 26 (48 hours) : No. 1, Redness and induration, 12 x 12. No. 2, Redness and induration, 6x6. No. 3, Redness and induration, 6x6. No. 4, Redness and induration, 10 x 10. On July 8, twenty-two days after inoculation, there were no palpable signs of tumor in the right groin, except that the glands were slightly enlarged. On this day I again injected each animal intracutaneously with .2 cubic centimeter W.E. 179 ( 1 cubic centimeter = .01 gram) and obtained the fol- lowing results : July 9 (24 hours) : No. 1, Pallid type of reaction, 20 x 20 ; Induration slight. No. 2, Negative. No. 3, Pallid type, 30 x 30 ; Slight induration. No. 4, Pallid type, 35 x 20 ; Slight induration. July 10 (48 hours) : Nos. 1, 2, 3, 4, All reactions completely faded. In this experiment the noteworthy points are: 1, the early appearance of cutaneous hypersensitiveness (coincident STUDIES IN IMMUNITY TO TUBERCULOSIS. 1 9 with early manifestation of lesions) ; and, 2, the transitory- character of the hypersensitiveness. On July ill again inoculated Nos. I to 4 subcutaneously in the left groin, each with 1 cubic centimeter of a heavy emulsion of Timothy bacillus. Three days later, on July 14, all received intracutaneous injections of .2 cubic centimeter W.E. 1 79 ( 1 cubic centi- meter = .01 gram) with the following results: July 15 (24 hours) : All animals show good skin reactions. Indura- tion and redness are marked. In all guinea-pigs they measure 25 x 25. These reactions are the best obtained thus far. From these experiments it would appear that re-infection tends to increase the capacity of the skin to react. 15. The influence of the general tuberculin reaction on cutaneous hypersensitiveness. — The cutaneous reaction, the focal reaction (and consequent general tuberculin reaction), and the anaphylactic reaction to tuberculo-protein are gen- erally looked upon as being different phases of hypersensi- tiveness to tuberculo-protein, but whether they are depend- ent on the same fundamental causes is not understood. In the following experiments the effect of the general tubercu- lin reaction on cutaneous hypersensitiveness has been studied. In Section 6 was given the record of the skin reactions obtained on H37 guinea-pigs on August 5, 1913, thirty-eight days after infection. On August 7 the animals, Nos. 40 to 43, were injected intraperitoneally, each with one cubic centi- meter W.E. [79. The next day, August 8, they were notice- ably ill and were then injected intracutaneously with .2 cubic centimeter W.E. 179 (1 cubic centimeter = .01 gram). The following results were noted : August 9 (24 hours) : No. 40, Pallid reaction, 15x15; Induration, slight. No. 41, Pallid reaction, 15 x r5 ; Induration, very slight. No. 42, Negative. No. 43, Pink area, 12 x 12; No induration. August 10 (48 hours) : No. 40, Negative. No. 41, Slight redness and induration. No. 42, Negative. No 43, Negative. 20 KRAUSE. In Section 7 are detailed the effects of intracutaneous tests on Nos. 44 to 62, made on August 13, forty-six days after infection. On August 15 these guinea-pigs, Nos. 44 to 62, each received I cubic centimeter W.E. 179 intraperitoneally. During the night Nos. 50, 5 I and 53 died and several became very ill. The next day, August 16, several of the stronger were selected for intracutaneous test, and reacted as follows : August 16 (24 hours): No. 44, Redness and induration, 25x25; Induration, -f- -\ — f- ; Necrosis, 5x5. No. 46, Redness and induration, 20x20; Induration, H — K No. 49, Redness and induration, 30x30; Induration, -\ — | — (- ; Necrosis, 10 x 10. No. 54, No reaction; Very weak. August 17 (48 hours) : No. 44, Faded. No. 46, Faded. No. 49, Faded. No. 54, Dying; Killed. It would seem that there is no doubt that the general reac- tion blunts skin hypersensitiveness considerably. We usually conceive of the general reaction as being due to the absorp- tion of focal products (focal reaction). The result here is comparable to what we often find during the height of a recrudescence of tuberculous disease. Then, during the febrile period, when more or less absorption from the focus is taking place, the cutaneous hypersensitiveness is frequently in complete abeyance. General discussion. — The reader who is familiar with the general principles of immunity to tuberculous re-infection must have already noted the many points at which the fac- tors underlying cutaneous hypersensitiveness approach those of immunity to re-infection. Some of these general princi- ples of immunity may be summarized as follows: 1 . There is no immunity without a preexisting tuberculous focus. 2. Immunity begins to manifest itself with the develop- ment of the focus. 3. It decreases with the healing of the focus. 4. It disappears with the enucleation of the focus. 5. The degree of immunization varies directly with the virulence of the immunizing microorganism. STUDIES IN IMMUNITY TO TUBERCULOSIS. 2 1 From the above work, as well as from the researches of Baldwin 6 and Romer, 7 it would appear that : 1. There is no cutaneous hypersensitiveness without a focus (tubercle). 2. This hypersensitiveness appears coincident with the establishment of the focus. 3. It diminishes with the healing of the focus. 4. It varies directly with the intensity of the disease, which in its turn is dependent on the virulence of the invading bacillus. Whether immunity to re-infection and cutaneous hyper- sensitiveness are both merely expressions of focal activity, having no interrelationship, or whether there is a causal rela- tionship between the two phenomena are problems that remain to be investigated. At the present time we are engaged on work the aim of which is to discover whether immunity to re-infection is a function of hypersensitiveness. The expression of cutaneous hypersensitiveness, as it mani- fests itself to the observer, is the ability of the skin of the affected animal to react in an inflammatory manner — a capa- city which the non-infected organism does not possess ; and our immediate point of attack is the meaning and nature of this inflammatory reaction. Of more immediate importance is why the numerous cli- nicians who have investigated the cutaneous reaction on man have arrived at such hopelessly contradictory results as regards the interpretation of the reaction. The answer does not seem particularly difficult. We must remember that as pulmonary tuberculosis presents itself for clinical observation it comes to the clinician as a disease really grounded on a very complex anatomical basis. We may take for granted that at bottom the simplest case is a mixture of very varied anatomical and physiological conditions; a jumble of foci in all stages of proliferative, exudative, and degenerative pro- cesses, continually changing one into the other, and thereby altering the amount and the time element of give and take between foci and host. Again, unless violent alterations are 22 KRAUSE. taking place in the diseased areas, the net changes take unconscionably long periods to express themselves and to make themselves apparent to the observer. It is also evident that it is very difficult to measure an inflammation or the effect by which cutaneous hypersensitiveness manifests itself. It is not easy to tabulate differences of induration and differ- ent degrees of hyperemia. It is only comparatively gross changes that are appreciable by us. To compare the results in one patient with those in another brings us no useful knowledge. At any particular time the same intensity of reaction in two or three different patients might mean two or three entirely different things. But some information might be gained by a prolonged and intensive study of the fluctu- ations of hypersensitiveness of each individual patient, by correlating observations of the reaction with those obtained by every diagnostic adjuvant. Holmes 8 has attempted some- thing of this kind. He has studied individuals and has com- pared their reactions, one with another; and he has arrived at results that resemble what we have obtained from animal experimentation. He concludes that " individuals with pul- monary tuberculosis present marked tuberculin hypersensi- tiveness (cutaneous) ; this diminishes if they improve, fluctu- ates, or remains stationary, if they do not, rises if they fail." It has also seemed to me that Hamman 9 displayed unusual acumen when he wrote : " We believe that tuberculin hyper- sensitiveness in relation to tuberculous disease runs, roughly, somewhat as follows : Since nearly all adults are infected with tuberculosis, we assume a low grade of tuberculin- hypersensitiveness to begin with. Should there be a fresh invasion of the body from within or from without, the tuber- culin-hypersensitiveness rapidly rises. If the disease sub- sides and the individual recovers the hypersensitiveness gradually falls to a lower level, perhaps to the original low level; if the disease remains active, the high level of hyper- sensitiveness persists and lasts until the body is overwhelmed and its resistance broken down completely by the disease, when hypersensitiveness disappears." STUDIES IN IMMUNITY TO TUBERCULOSIS. 23 SUMMARY. 1. Cutaneous hypersensitiveness to tuberculo-protein is inaugurated by the establishment of infection and the development of the initial focus. 2. It increases with progressive disease. 3. It varies directly with the extent and intensity of the disease. 4. It diminishes with the healing of the disease. 5. It is probably never entirely lost (except in the presence of intercurrent disease, pregnancy, etc.). 6. It is increased by re-infection. 7. It is diminished or completely wiped out during the period of the general tuberculin reaction. 8. It is suggested that tissue hypersensitiveness may be a function of immunity to re-infection. [This work was made possible through the generosity of Mrs. A. A. Anderson of New York City.] BIBLIOGRAPHY. 1. Mendel. Med. Klinik, 1908, 12. 2. Mantoux and Moussu. Acad, des Sc, Sept. 14, 1908. Mantoux. La Presse Med., 1910, No. 2, 10. 3. Romer. Beitr. z. Klin. f. Tuberc, 1909, xiv, 1. 4. Baldwin. Trans. Nat. Assoc. Study and Prev. of Tuberc, 7th Ann. Meeting, 191 1, 352. 5. Romer. Beitr. z Klin. f. Tuberc, 1910, xvii, 287. 6. Baldwin. Loc cit. 7. Romer. Beitr. z. Klin. f. Tuberc, 1909, xiv, 1. 8. Holmes. Trans. Nat. Assoc. Study and Prev. of Tuberc, 10th Ann. Meeting, 1914, 133. 9. Gelien and Hamman. Bull. Johns Hopkins Hosp., 1913, xxiv, No. 268, 180. Studies in Immunity to Tuberculosis.* the anaphylactic state in its relation to resist- ance to tuberculous infection and tuberculous DISEASE. AN EXPERIMENTAL STUDY. A. K. Krause, M.D. {From the Saranac Laboratory for the Study of Tuberculosis.) The one outstanding development of our hard-won knowl- edge of tuberculosis during the last ten years is the rapidly growing belief that tuberculous infection and the disease tuberculosis are two entirely different phases of an inter- dependent situation. It required the widespread application of the cutaneous tests and the further discovery of the fact that the positive cutaneous reactions occurred only in the presence of tubercle, to clinch the pathologists' contention that a large proportion of civilized mankind is tainted with tubercle, yet does not suffer from tuberculosis. (Throughout this paper the term " tubercle " is used to denote merely anatomic change, while the term " tuberculosis " is used to denote clinical disease.) It followed further, as a natural and logical conclusion, that tuberculous infection or the initial development of tubercle does not carry with it the necessity of the immediate development of clinical tubercu- losis. The great majority of people who have tubercle pass through a long life reacting positively to infection tests; yet with bacilli in some way or other under control, they never show symptoms of specific illness. In the same way, but to a lesser degree, they are in the same condition as the man who once had clinical tuberculosis, whose disease has become arrested and who lives through a shorter or longer interval — perhaps through the rest of a long life — without ever again experiencing a symptom of tuberculosis. We know that such a man harbors hidden sources of danger to himself, we * Received for publication June 10, 1916. (25) 26 KRAUSE. know too that he will always react to tuberculin tests. We cannot guarantee that the old, once active foci will never again flare up. We can only tell him that so long as his foci of disease and the bacilli which they often undoubtedly contain remain well walled off and encysted, just so long he is likely to be free from tuberculosis. It follows, too, that the tuberculosis of adolescents and adults in many cases develops from benign tubercle that has been implanted earlier in life. In other words, tuberculosis often originates from a preexistent benign focus. Although there are a great many facts to support this view, I shall content myself here with merely making the assertion as a matter of fact. I would remind the reader that bone and joint, adrenal, and meningeal tuberculosis must develop from some primary focus, whether this primary focus is demon- strable or not. And I would further remind him that such tuberculosis frequently develops in those who have hitherto enjoyed perfect health. If now we grant that tubercle is very commonly implanted in early life, that tuberculosis is usually not an immediate consequence, that the majority of people go through life with tubercle, but without tuberculosis, and that in a great many cases tuberculosis develops from preexisting tubercle, it becomes of the utmost importance that we acquire some well-controlled information concerning the conversion of tubercle into tuberculosis. In the light of our newer knowl- edge, too, the conditions that influence infection and the further dissemination of tubercle need much more study than they have thus far received. Once tuberculous infection is established in an animal it is obvious that numerous mechanical factors, alone or working with other forces, will determine whether the initial focus will remain localized or whether dissemination from it will take place. As example of such influences I need only mention the site of the initial focus, its degree of fibrous investment or the character of its center, whether more or less caseous, more or less fibrous or calcareous, variations in the blood and lymph flow to and from the focus as these are STUDIES IN IMMUNITY TO TUBERCULOSIS. 27 determined by a multiplicity of physical conditions. If for the time being we leave out of consideration these well- recognized mechanical factors, we then assume that infection and the further extension of tubercle are also modified by qualities that appertain to one or both of the two organisms concerned ; namely, the invader and the host, the tubercle bacillus and the infected individual. It is generally held that the issue of tuberculous infection turns on the one hand upon the number and virulence of the bacilli and the poisons elaborated by them or inherent in them, and, on the other hand, upon the susceptibility or resistance of the body. Some at least of these attributes need elucidation. It is becoming more and more generally accepted to-day that the tubercle bacillus contains or elabo- rates no specific poison or toxin, but that the various symp- toms of intoxication met with in tuberculosis are due to bacillary protein sensitization and intoxication in their various forms. Therefore, what was formerly looked upon as being due purely to the effects of a bacillary attribute, is now viewed as being dependent in part on the changed reactions of the body which help to make up that complex and elusive condition that we are pleased to call resistance. There is no doubt that ceteris paribus, the number of bacilli concerned at any time does have a marked influence on the degree and extent of infection. This fact can be easily established by simple experiment and can be dismissed in a word. But whether, as is often assumed, the virulence of tubercle bacilli, as they occur in disease, is so very variable, is not altogether so clear. There is no doubt that strains that have been isolated from the animal body fre- quently decrease in virulence once they are trained to a vegetative test-tube existence. This every observer of any experience must have noted. But it has been the experience of the Saranac Laboratory that this lost virulence is never regained, though a variety of well-known methods have been tried in the effort to bring back original virulence to a culture. 28 KRAUSE. Variation in bacillary virulence has often been brought in to explain the multiplicity of clinical and anatomical types of tuberculosis. But to attempt in this way to explain too much of the variety of clinical finding in tuberculosis would soon involve one in endless contradiction. As a matter of fact, most of the studies that have aimed to differentiate virulence in bacillary strains as they have been isolated from tuberculous disease of many varieties and in many locations have not been very fruitful (Vagedes, 1 Krompecher and Zimmermann, 2 C. Frankel and Baumann, 3 Moller, 4 Burnet, 5 English Commission, etc.). They have at least not brought to light the variations in virulence that many writers have too often assumed. Perhaps our technical methods have heretofore been too coarse. At any rate, satisfactory information on this point is much to be desired and it should be the subject of prolonged and carefully controlled research, for its importance can hardly be exaggerated. Although a great deal of laborious work has failed to bring forward much definite knowledge of the variations in virulence, as these arise in tuberculous man and animals, the most casual observer has noted that once infection is estab- lished its subsequent course is largely dependent on fluctua- tions in the resistance or susceptibility of the individual — in other words, in the way in which at any given time the indi- vidual reacts to the invader. We see the individual reacting now exudatively, now proliferatively to his infection. By the application of the cutaneous test we can discover very prettily that the reaction of at least one of his tissues, the skin, can be a very changeable one. What underlies these changes we do not know : the fact remains that they are there. It is very probable that man has very little native resist- ance to tuberculous infection. Evidence is continually accu- mulating that if exposed to the living tubercle bacillus, every man will become infected. How else explain our autopsy and cutaneous test results which show tubercle to be such a universal taint? But there is a very real and a very impor- tant difference in the way in which individuals react to this STUDIES IN IMMUNITY TO TUBERCULOSIS. 29 infection. This observation is too patent to need demonstra- tion. This is the difference between what we call resistance on the one hand, and, on the other, lack of resistance. The term resistance and its opposite are vague terms, yet their manifestations are none the less real. We know that we can diminish or increase resistance by certain hygienic measures: one of Dr. Trudeau's 6 first experimental works showed this. Murphy's 7 recent work on the effects of radiation, and Arloing's, s and Lewis and Margot's 9 on those of splenectomy, also illustrate the point. The conditions of the experiments of Lewis and Murphy are, however, such as can hardly occur in natural infection and are only valuable in so far as they give us an insight into the fundamental causes of immunity in tuberculosis. In the experiments on resistance and virulence which I shall detail I have had in mind several of the features to which I have referred. For the purposes of infection I have used several strains of tubercle bacilli with which I have worked constantly for seven years, and whose constancy of virulence under various circumstances I am familiar with and fairly certain of. With one of them, Strain Ri, I aimed to produce a non-progressive and self-healing tuberculosis in guinea-pigs, such as we see approximated in the non-pro- gressive or quiescent infection of man. I then sought for a medium which might presumably reduce the resistance of my animals to infection and further invasion, a medium which I desired to be comparable to what we may meet with in man under natural conditions. By the application of this medium I tried to convert a non-progressive or self-limited tuberculosis into a progressive and spreading infection. My endeavor was to study some small possible element in resist- ance, some element that was sharp and definite and easily controlled. The medium which I selected was non-lethal anaphylactic shock. It is a common clinical experience to find tubercu- losis developing soon after some other acute infection, such as measles, tonsilitis or catarrhal affections of the respiratory tract. It is conceivable, and indeed probable, that during 30 KRAUSE. the course of the measles or the influenza, local circulatory conditions (congestion, etc.) are set up around an already existing focus of tuberculous infection, that favor the mobil- ization and dissemination of bacilli from what was hitherto a benign tubercle. But that this is not the whole story is sug- gested by the fact that during an attack of measles or during the febrile period of other infections, a patient who formerly reacted positively to the cutaneous tuberculin tests often loses his skin sensitiveness. This loss of skin sensitiveness is also not infrequently met with during an acute febrile exacer- bation of tuberculosis in the tuberculous patient himself. We must assume, therefore, that during intercurrent infection or renewed activity of tuberculosis, the specific immunity reaction to tuberculin that is represented by cutaneous hypersensitiveness may be in abeyance. Now both the cutaneous tuberculin reaction and the gen- eral anaphylactic reaction are phases of hypersensitiveness to protein. Furthermore, the idea developed by Vaughan that many of the constitutional symptoms of infection, of which fever is such a prominent one, are due to intoxication by dissociated proteins, which in turn depends on the hyper- sensitiveness of an organism that has become allergic to the specific protein under consideration, is gaining more and more adherents. It was therefore but a step to reason that the resistance of the body to tuberculous infection and exten- sion might be reduced by protein intoxication or anaphy- laxis. Several years ago I 10 had done a few experiments that led me to believe that such might be the case. At that time I studied the relation between hypersensitiveness to tuberculo- protein and immunity to tuberculous infection. If my non- tuberculous animals were first artificially sensitized to tuberculo-protein and then inoculated with living tubercle bacilli, after a proper length of time I could not determine that sensitized animals reacted to infection any differently from non-sensitized normal controls. If, however, a few of the sensitized were given a non-lethal toxic injection of tuberculo-protein, and if they developed symptoms of acute STUDIES IN IMMUNITY TO TUBERCULOSIS. 3 I anaphylaxis and were then very shortly afterwards inoculated with living bacilli it struck me that the disease extended further than was the case in either the merely sensitized animals or in the normal controls. As a result of these experiments I published the following conclusions: (1) " Sensitization of non-tuberculous guinea-pigs with tuber- culo-protein does not alter their resistance to experimental tuberculous infection; " and (2) " Resistance to infection is markedly lowered during the period that a sensitized animal is suffering from symptoms of anaphylactic shock." In these experiments of 191 1 the microorganism used for infection was our strain, H37 tubercle bacillus, which is a virulent human bacillus and produces progressive tubercu- losis in guinea-pigs. I could interpret my results only by comparing the extent of tuberculous involvement in the different sets of animals at a given time after inoculation. In any experiments of this kind, unless the differences were sharp and constant and the number of animals was large, any interpretation would always be open to justifiable criti- cism. As a matter of fact, the differences in these particular experiments were fairly marked and uniform, but the number of animals used never satisfied me as being sufficient to overcome all sources of error in interpretation. I therefore determined to repeat the experiment under conditions that could be more easily controlled and more sharply differenti- ated. The microorganism used for infection in the following experiments was our strain, Ri human bacillus. This germ was isolated by Dr. Trudeau in 1891. After a few years of incubator existence (about 1893) its virulence for guinea- pigs dropped to a level where it has remained constant ever since. Subcutaneous inoculation into the groin of a guinea- pig will bring about hyperplasia and caseating tuberculosis of the regional glands, as well as of the iliac glands, but macroscopic tuberculosis of any of the abdominal and thoracic viscera occurs infrequently. Only under excep- tional conditions, pregnancy, intercurrent disease, excessive or repeated dosage, will macroscopic tubercle develop in the 32 KRAUSE. spleen, or with extreme rarity in the lungs or liver. Six months or a year after inoculation the superficial inguinal glands, which were once enlarged, begin to return to their normal size and gradually approach their normal consistence. I have carried some guinea-pigs that I inoculated subcu- taneously with Ri for two years and more. At the end of this time I found all viscera clean and normal to the naked eye, the iliac, and deep inguinal glands apparently normal, and with the exception of slight enlargement and radi- ating, translucent grayish lines and here and there an opaque yellowish-white point, the superficial inguinal glands were almost normal in appearance. Cutaneous tests before death, however, proved such animals to be still hypersensitive to tuberculo-protein, and histological examination revealed microscopic tuberculosis in the superficial inguinal glands. One and two years after inoculation the animals are always sleek, fat, lively, and in apparent perfect health, though they harbor a localized tuberculosis and react to tuberculo-protein. We have, therefore, always looked upon Ri infected animals as affected with self-limited, healing tuberculosis, and more or less comparable to those human beings who are infected, but whose infection shows no tendency to progress. In the first series of our new experiments one set of animals was first sensitized with tuberculo-protein, then anaphylactized, and finally infected with Ri. Another set was first sensi- tized, "then infected with Ri and finally anaphylactized. Both sets were then allowed to live for a certain length of time, after which the effects of anaphylactic shock on infec- tion and the spread of the disease were studied. STUDIES IN IMMUNITY TO TUBERCULOSIS. 33 Experiment i. 1 . Sensitization of guinea-pigs : May 12 and 13: Nos. 1 to 17, each injected intraperitoneally with 2 cubic centimeters W.E.* (water-extract) 179. 2. Toxic injection of animals : June 17 and 18: Nos. 1 to 17 injected post-orbitally with E.E.* (extract-emulsion) 179, in doses varying from .2 to .5 cubic centimeter. Results: Nos. 5 and 15 died of anaphylactic shock. Nos. 1, 9, and 10, no symptoms. Nos. 4 and 8, symptoms, +; recovered. Nos. 3, 7, 12, 13, 14 and 17, symptoms, -\ — (- ; recovered. Nos. 2, 6, 11 and 16, symptoms, + -\ — (- ; recovered. 3 Inoculation of animals with Ri : June 18: Emulsion used: 140 milligrams of a glycerin-beef-serum culture of Ri, planted June 1 and showing good growth, were ground in a mortar with a few drops of .85 per cent physiological salt solution. Ten cubic centimeters of physiological salt solution were then added to the mush and the emulsion was centrifugated until all large clumps were thrown down. The supernatant fluid, which was milky, was then used for inoculation. Animals inoculated: Set A: Nos. 1 to 4, sensitized May 12; intoxi- cated, June 17, twenty-four hours before inoculation. Set B: Nos. 6 to 10, 12 to 14, 16 and 17, sensitized May 12 and 13; intoxicated, June 18, one to three hours before inoculation. Set C : Nos 41 to 48, Originally normal non-sensitized, but on June 18 given post-orbital injection of .25 cubic centimeter E.E. 179, to control the effects of the post- orbital injection on the infection. Injection of E.E. given one hour before inoculation. Set D : Nos. 49 to 56, Normal controls ; no preliminary preparation before inoculation. Inoculation : Each of the above-named animals was inoculated with .25 cubic centimeter of the above emulsion, subcutaneously in the right groin. 4. Results : On August 8, fifty-one days after Rr inoculation, all the infected guinea- pigs were killed and autopsied. The findings at section were not very conclusive. At the end of my notes for the day I have a summary in which I say, "Differences between the various sets of animals are too * W.E. 179 : A water extract of Bac. Tub. H37, containing bacillary protein. Prepared as follows : 122 gms. dried H37 (from glycerin-broth cultures), pulverized for one week in a mill ; powder extracted with Aq. dest., 1,830 cc, at 50° C. for 48 hrs. ; let stand 3 days in the dark ; filtered through Berkefeld filter. The filtrate is the W.E. 179. 1 cc. of dried residue = .02 gm. E.E. 179 : Is a small portion of the above extract withdrawn before filtration and allowed to settle out in the dark. 1 cc. dried = .022 gm. 34 KRAUSE. slight to conclude that nqn-lethal anaphylactic shock affected the animals 1 resistance. The two most advanced lesions are in Nos. 13 and 14 (anaphylactized animals), which had the added burden of pregnancy." However, careful analysis of the autopsy details of each animal as they are set down reveals the fact that the gross tuberculous involvement was more advanced in more anaphylactized animals than in those of Sets C and D. There were no anaphylactized animals which showed as little involvement as the best of the controls {i.e., Nos. 41 to 56), seven of which had no visceral disease whatever. In five of the anaphylactized animals the lungs showed miliary foci. This, a very rare event after an ordinary Ri infection, occurred in only one control. In only one anaphylactized animal, No. 17, was the disease localized strictly to the glandular (lymphatic) system. In seven of the controls it was so confined. If, in all but a few of the anaphylactized animals the infection had escaped from the glandular system, while in most of the controls it had been limited to the lymphatics, I should have no hesitancy in concluding that if original infection (and perhaps, too, dissemination from an already existing focus) takes place during the phase of anaphylactic shock, it occurs at a time when resistance to infection is lowered. In view of the results of this experiment I can do no more than suggest that this is probably the case. It must be remembered that these animals were killed fifty-one days after infection. We have found by experience that Rr infection reaches its maximum and maintains it at from thirty or forty days to three or four months. From this time on, under ordinary conditions, the disease tends to retrogress. Enlarged and hard glands tend to diminish in size and regain some of their normal elasticity, and the occasional nodule that is seen in the spleen tends to disappear, so that after a year or two the naked eye can detect little except enlarged and partially sclerotic, super- ficial inguinal glands. It is impossible to say now whether these various sets of animals would have shown more striking variations later in the disease. I hope to study this matter further. Experiment 2. {Plan : To sensitize guinea-pigs with tuberculo-protein ; after sensitive- ness is established to infect them with Ri ; after disease is established to give the originally sensitized ones non-lethal anaphylaxis and study the effects of anaphylaxis on the disease already present.) 1 . Sensitization of guinea-pigs : May 15 : Nos. 31 to 40, each injected intraperitoneal^ with 2 cubic centimeters W E. 179. 2. Inoculation of animals with Ri : August 9: Emulsion used: A spadeful of Ri from glycerin-beef- serum culture, good growth, planted July 10, rubbed up in mortar with a STUDIES IN IMMUNITY TO TUBERCULOSIS. 35 few drops of physiological salt solution. Emulsified in salt solution. Not filtered. Emulsion opalescent. Animals inoculated : Set A : Nos. 31 to 40, sensitized May 15. Set B : Nos. 67 to 76, non-sensitized normal controls. 3. Tests of sensitiveness of animals at time of inoculation : August 11 : Nos. 18 to 20, from same original series as Nos. 31 to 40, and sensitized on May 13, were to-day given post-orbital injections of E.E. 179. They proved to be very sensitive. 4 . Toxic injections of animals : September 30: Fifty-two days after inoculation and 138 days after sensitization : Nos. 31 to 40 injected post-orbitally with E.E. 179, in doses of from .2 to .25 cubic centimeter. All had -\ — (- or + + + symptoms of anaphylactic shock except No. 37, which had no symptoms. All recovered. 5 . Results : On December 3, sixty-four days after anaphylactic shock and 116 days after inoculation with Ri, all the infected guinea-pigs were killed and autopsied. There was no appreciable difference between the two lots. None of the animals had visceral disease. All showed practically healed processes, with slightly enlarged and firm regional glands the only evidences of infection. The animals' resistance to disease was therefore apparently not lowered by the anaphylactic shock which they had experienced some time before. Other phases of this experiment will be taken up in the general discussion at the end of this paper. Experiment 3. (Plan: Exactly the same as that of Experiment 1, except that for purposes of infection Moeller's Mist Bacillus was used instead of the Ri human tubercle bacillus. As is well known the Mist Bacillus is one of the so-called saprophytic or non-pathogenic acid-fast bacilli. It possesses practically no parasitism for mammals. If inoculated subcutaneously into guinea-pigs, even in enormous doses, it produces an abscess at the site of inoculation, but does not cause lesions in the viscera. The regional glands also undergo some hyperplasia. Speaking broadly, we may say that the local effects differ from those brought about by the tubercle bacillus, in that it takes relatively enormous doses of the Mist Bacillus to produce any appreciable change (abscess), in that the changes are evident in from two to three days, and in the fact that they are very transient and non-progressive as compared to what is found in real tuber- culosis. As in R 1 infection the aim of this experiment was to discover whether anaphylactic shock would so reduce the resistance of the body as to allow a saprophytic microorganism, the Mist Bacillus, to " take hold." ) 36 KRAUSE. i . Technic of the experiment : The steps followed were exactly the same as those outlined in Experi- ment I, except that the time intervals were somewhat different. The animals were intoxicated with E.E. 179, forty-three days after sensitization with W.E. 179. Inoculation with enormous doses of an emulsion of Bac. Mist was made immediately after intoxication on six guinea-pigs that survived the toxic dose and on controls. 2 . Results : August 12. The animals were killed forty-seven days after inoculation. The different sets of animals revealed no differences of involvement. A few of each set had slightly enlarged regional glands. The rest were entirely normal. Careful search of stained smears of some of the enlarged glands revealed no acid-fast bacilli. Anaphylactic intoxication shortly before inoculation apparently did not alter the animals 1 response to the parasitism of the Mist Bacillus. Experiment 4. At the time that we were engaged on some of the experi- ments described above, Thiele and Embleton 11 published a work entitled, " The Pathogenicity and Virulence of Bacteria." They reported that they had been able to change non-patho- genic into pathogenic bacteria by a process which in its gen- eral principles (except the time element) resembled the one which I have already described. Their usual procedure was as follows: They would sensitize normal guinea-pigs with large doses of a non-pathogenic microorganism such as the Timothy bacillus and the Smegma bacillus. From five to fifteen days later they would then inoculate these animals with smaller doses of the living organism of the same type and find that this second inoculation caused disease and death. They reported further that the virulence of the originally non-pathogenic organism was so enhanced that bits of tissue from the animals thus infected produced disease in animals to which they were transferred. " The optimum time for the inoculation was from five to ten days after the sensitizing dose," they report. They detail one experiment on one animal with the tuber- cle bacillus which I shall quote : " A rabbit of 2,500 grams was inoculated with .05 milligram of human tubercle bacilli, fourteen days after sensitization with 50 milligrams of STUDIES IN IMMUNITY TO TUBERCULOSIS. 37 killed pulverized tubercle bacilli (Meister, Lucius, and Briining). The rabbit died in one week, having the post-mortem appearance of acute mili- ary tuberculosis, the only difference being large necrotic areas in the spleen. "The control rabbit of 2,300 grams received only a dose of .05 milli- gram of the same tubercle culture as was used for the previous animal. Six months later it was alive and well, and after killing it a post-mortem examination revealed no signs of tuberculosis." Now the reports of a number of Thiele and Embleton's experiments were at total variance with what had been our experience at the Saranac I aboratory. I had often made repeated inoculations, spaced at various intervals, of the vari- ous saprophytic acid-fast bacilli, and had never noted any effect on the animal that would compare with what Thiele and Embleton found. I had also frequently practised the procedure outlined by them in the above experiment with tubercle bacilli and had not arrived at a similar result. It is true that my technic had not been exactly similar to theirs : my experiments had been done in work performed from very different points of view. But their results were so surprising in view of what I had already been taught by my own work that I resolved to repeat systematically some of the experi- ments, using the tubercle bacillus for the infecting agent. I aimed, too, to reproduce their technic in its essential features. I therefore outlined the following experiment: Inocula- tion of guinea-pigs with very large quantities of dead virulent tubercle bacilli; re-inoculation ten days afterwards with large quantities of living tubercle bacilli of low virulence ; compare the effects on the re-inoculated animals as against those on controls. 1. Material used for primary inoculation. — For my pri- mary inoculation I desired tubercle bacilli that were dead and which nevertheless retained their integrity as much as possible — which, in other words, had not been altered by heating or by chemicals. I therefore selected the following material : In the Saranac Laboratory we are constantly col- lecting the bacilli from broth cultures, washing them, and 38 KRAUSE. drying them in the dark in vacuo over sulphuric acid. As these accumulate we put them aside in the dark in a dried state. They have been subjected to no other agencies than washing and drying and their original constitution has pre- sumably not been altered. At the time at which I began this experiment (June, 191 5), we had in one bottle a collec- tion of originally virulent H37 strain which had been washed, dried, and assembled between January, 1914, and March, 19 14 At this distance, fifteen to seventeen months, they were presumably dead and I resolved to use these for my sensitizing inoculation. 2. Primary inoculation. — June 21. Two grams of the dried H37 tubercle bacilli were ground by hand in an agate mortar for one and one-half hours. Twenty-five cubic centi- meters of physiological salt solution were then added and the bacilli were carefully emulsified. One cubic centimeter of the emulsion therefore represented eighty milligrams of the dried bacilli. On the same day twenty guinea-pigs were inoculated sub- cutaneously in the right groin as follows : They were divided into two lots of ten each: lot No. 1, Nos. 1 to 10, and lot No. 2, Nos. n to 20. Nos. I and 2 and 10 to 13 received .5 cubic centimeter each (40 mg. dried bacilli) ; Nos. 3 to 5 and 14 and 15, .25 cubic centimeter each (20 mg. dried bacilli) ; Nos. 6 to 10 and 16 to 20, .1 cubic centi- meter each (8 mg. dried bacilli). 3. Re-inoculation. — July I. Ten days after primary inoculation. Re-inoculation was made with Ri, described in the foregoing pages. A heavy, milky, non-centrifugalized emulsion of Ri was prepared. Of the above animals, Nos. 1 to 10 were inoculated subcutaneously, each with .25 cubic centimeter of the Ri emulsion in the left groin. Nos. 11 to 20 were not re-inoculated with Ri emulsion, but were left to control the effects of the original, primary inoculation with old, dried H37. At the same time, five new animals Nos. 21 to 25, which were normal, untreated animals, were STUDIES IN IMMUNITY TO TUBERCULOSIS. 39 inoculated like Nos. 1 to 10, each with .25 cubic centimeter of the Ri emulsion to control the effects produced by Ri. We therefore had three sets of animals : Set 1 : Nos. 1 to 10. Inoculated June 21 with old, dried H37 and re-inoculated July I with living Ri. Set 2: Nos. 11 to 20. Inoculated June 21 with old, dried H37. Set 3: Nos. 21 to 25. Inoculated July I with living Ri. With Set 2 we could study the effects of a single dried H37 inoculation; with Set 3, those of a single living Ri inoculation ; and we could compare both with those pro- duced in the doubly inoculated animals of Set I. 4. Results. — August 9. Forty days after inoculation with Ri. All the animals, Nos. 1 to 25, were killed and' autopsied. Nos. 1 to 10: All have large caseous glands and abscesses in the right groin (site of dried H37 inoculation) and not quite such large, caseating glands in the left groin (Ri inoculation). The iliac glands are considerably enlarged (right and left) and have caseous spots in all of them. In no instance has the disease got beyond the iliac glands ; there are no visible lesions in the spleens, livers, or lungs of any of the animals. Nos. 11 to 20 : Those animals that had only an inocula- tion of old, dried H37 on June 21. In general, the effects are the same as those in Nos. 1 to 10, except that there is no enlargement or involvement of the left superficial and deep inguinal glands. The iliac glands on both sides are enlarged. In No. 14 the spleen is slightly enlarged with numerous small, pin-head, raised, grayish-yellow nodules resembling tubercles. In No. 15 the spleen is slightly enlarged and shows one raised, yellowish nodule, five millimeters in diameter. In No. 17 the spleen is almost of normal size, but has one small, raised, yellowish- nodule, two millimeters in diameter. The spleens of all the 40 KRAUSE. other animals seem normal. The livers and lungs of all the animals are apparently normal. Nos. 21 to 25 : Those animals that had only an inocula- tion of living Ri on July 1. There is involvement of the superficial and deep inguinal and iliac glands of the left side in all. All spleens, livers, and lungs seem normal. 5. Discussion of Experiment 4. — It will be seen that those animals of Set I that were treated according to Thiele and Embleton's method, i.e., those that had received a double inoculation during a ten-day interval, did not differ materially from those of Set 3 that had only the Ri or control inocu- lation. But in Set 2 which had only the old, dried H37 inoculation, and which was used to control this inoculation, several interesting and surprising results occurred. In Nos. 14, 15, and 17 the spleens were found to be involved with what looked like tubercle. These nodules were accordingly enucleated and ground up, each in a separate mortar. Smears were prepared from them and stained. After prolonged search, one acid-fast bacillus was found in the smear made from the spleen of No. 15 ; none was found in those from the spleens of Nos. 14 and 17. Each emulsion of spleen juice was therefore re-inoculated separately into a normal guinea-pig on August 9. On September 25 these three guinea-pigs were killed and autopsied, and all showed a tuberculous involvement that corresponded with what we usually obtain when we inoculate living H37 under ordinary conditions. In all three diagnostic animals the spleens were considerably enlarged and contained caseous nodules. It will be remembered that Nos. 14, 15, and 17 were inoculated only with old H37 bacilli, which had been thoroughly dried from fifteen to seventeen months before. It was astonishing to us that under such conditions these microorganisms should maintain their viability and their parasitism Yet such was undoubtedly the case. Transference of the spleen nodules of Nos. 14, 15, and 17 showed that under the above mentioned conditions H37 had lost none of its virulence as STUDIES IN IMMUNITY TO TUBERCULOSIS. 4 1 we had learned to know it. It would appear that in the great mass of dried organisms that had been preserved over all this time a few out of countless numbers, a few perhaps far in the interior of large clumps, had remained alive during fifteen to seventeen months. The net result of the experiment, however, carefully con- trolled as regards the infectiousness of the two types of tubercle bacilli used, failed to substantiate Thiele and Embleton's view that a sensitization of from five to fifteen days can so diminish the resistance of the animal organism that the parasitism of a bacillary invader may be increased thereby. The result, too, is more or less in harmony with those of the first three experiments detailed in this paper. GENERAL DISCUSSION AND SUMMARY. The foregoing work is merely a first attempt to study the anaphylactic state as it is related to resistance to infection by the tubercle bacillus and to the extension of tuberculous disease. Under the conditions of the experiments, especially experiments Nos. 1 and 2, no relationship has been con- clusively proved. Yet other methods of attack remain and it is to be hoped that these will be applied. It is certainly a plausible view that under the influence of the continuous absorption of bacillary protein and intoxication thereby, resistance may be so lowered that bacillary invasion is a comparatively easy matter. After watching the course of long-continued febrile phthisis one gets this impression. The experiments herein outlined were concerned with the influence on resistance of a single, marked, anaphylactic intoxication. In Experiment 1 it may have had an effect, but this was so obscure that no generalization could be drawn therefrom. In Experiment 2 it had apparently no effect. But here we must remember that the existing disease may perhaps have been so well localized and walled off that no bacilli were mobilized and none were free to exert their influence on freshly anaphylactized soil. If, after infection was once established, an animal could be kept in a more or less continuously intoxicated condition {i.e., anaphy- laxis intoxication) perhaps the result would be different. 42 KRAUSE. Under the experimental condition above outlined we are justified in concluding : i. Anaphylactic shock, experienced by guinea-pigs a short time before infection with tubercle bacilli of low viru- lence, did not reduce their resistance to such an extent that the parasitism of the particular microorganism concerned was markedly increased, although there were suggestions that the extent of disease was probably increased. 2. If tuberculous disease of low grade is once established in guinea-pigs a single attack of anaphylaxis does not bring about conditions that favor the extension of the disease. 3. Anaphylactic shock suffered just before the inoculation of a non-pathogenic acid-fast organism, the Mist Bacillus, does not lay the body open to progressive invasion by this germ. 4. No success attended the efforts to enhance the viru- lence of a strain of the tubercle bacillus (Ri) when the method of Thiele and Embleton was followed. 5. Evidence is herewith submitted that tubercle bacilli can preserve their viability and their original virulence after being kept in a dried state for as long a time as from fifteen to seventeen months. [This work was made possible through the generosity of Mrs. A. A. Anderson of New York City.] BIBLIOGRAPHY. 1. Vagedes. Zeitschr. f. Hyg., 1898, xxviii, 276. 2. Krompecher and Zimmermann. Centralbl. f. Bakter., Orig., 1903, xxxiii, 580. 3. Frankel and Baumann. Zeitschr. f. Hyg., 1906, liv. 4. Moller. Ibid., 1906, lv, 506. 5. Burnet. Ann. de l'lnst. Past., 1912, xxvi, 868. 6. Trudeau. Amer. Jour. Med. Sci., 1SS7, No - I §7> New Series, 11S. 7. Murphy and Ellis. Journ. Exp. Med., 1914, xx, 397. 8. Arloing. Compt. Rend. Soc. de Biol., 1905, 58, 261. 9. Lewis and Margot. Journ. Exp. Med., 1914, xix, 187. 10. Krause. Journ. Med. Research, 191 1, xxiv, 399. 11. Thiele and Embleton. Zeitschr f. Imm. 'frschung, Orig., 1913, xix, 643. Studies in Immunity to Tuberculosis.* concerning the general tuberculin reaction. A. K. Krause, M.D. {From the Saranac Laboratory for the Study of Tuberculosis. ,) Several years ago in a discussion of the probable mechan- ism of the general tuberculin reaction I 1 took the ground that this reaction is at bottom dependent upon a focal reac- tion. This was a point of view that had been advanced before and has always had a number of adherents. In elab- orating it I sought to explain the well-known symptoms of intoxication that mark the general reaction and suggested that these were due to the reaction of some absorbed focal products which exerted their effect upon an organism that had already become hypersensitive to them by reason of previous focal absorption. We could then comprehend the series of events that make up this exceedingly complicated reaction as follows : The antigen (tuberculin) injected into the body is absorbed and in due time reaches the tuberculous focus, with which it will react. This focal reaction is mani- fested anatomically by exudation or inflammation, the sever- ity of which is dependent upon a combination of at least two factors; first, the accessibility of the focus, that is, the ease with which the antigen can reach it, a condition which will vary with the degree of fibrous investment of the focus ; and second, the amount of tuberculin which gets to the focus. The acute focal inflammation favors the absorption into the blood or lymph channels of focal products, which is more rapid and quantitatively greater than before the reaction. This sudden absorption of comparatively large amounts of focal material reacts on an already sensitized body, produc- ing an anaphylactic intoxication, the symptoms of which we term the general reaction. Since then I have had no reason to change my belief that the * Received for publication June 10, 1916. (43) 44 KRAUSE. absorption of focal products brings about the general reac- tion. But the results of a series of experiments undertaken to prove the validity of the above hypothesis have led me to modify my view that the symptoms are anaphylactic in origin. My first experiments were made with the idea of determin- ing whether the reacting focus contains a primary toxin, and, if it does, whether this. toxin is present during the late stages of reaction in larger amounts than during the earlier phases. Guinea-pigs with advanced tuberculosis received intraperitoneal injections of varying quantities of a water extract of dried, pulverized tubercle bacilli. As they fell sick they were killed at various times, some at the beginning of symptoms, some after illness was well established, and others when moribund. Their tuberculous organs showing the familiar focal reaction of inflammation were recovered and ground in mortars or their pulp squeezed out in a press. The pulp was taken up in physiological salt solution and the resultant suspension was centrifugated. The supernatant fluid was then used for intravenous injection into normal guinea-pigs. The result was uniformly the same. All animals thus injected became very acutely ill, and generally died in from one to five minutes. The solution of tubercu- lous organ pulp was very dilute; the lethal dosage com- paratively small, as low as one cubic centimeter. Quantitative experiments failed to show that the organs of moribund animals with intense hemorrhagic focal reactions were any more toxic than those with beginning reaction, in which the focal inflammation was not visible to the naked eye. I was next led to test the toxicity of ordinary, non-react- ing foci. Tuberculous organs of guinea-pigs were treated in essentially the same manner as outlined above and the solutions were injected intravenously into normal guinea- pigs. Such animals exhibited the same symptoms and died as acutely as though they had received the products of the reacting focus. Sometimes, if the dose was very small they would almost immediately become very ill and then gradually recover. Sometimes the supernatant diluted tuberculous lung STUDIES IN IMMUNITY TO TUBERCULOSIS. 45 juice or spleen juice was divided into two portions, one of which was passed through Berkefeld bougie or cotton or paper. The filtrate was apparently harmless, while the unfiltered fluid was highly poisonous. If the material was injected intraperitoneally instead of intravenously the animal suffered no apparent discomfort, but would of course develop tuberculosis after the proper incubation interval. The net result of this series of experiments was that I could detect no difference between the effects of reacting focal products and those of non-reacting foci. I then compared the foregoing results with those obtained from the injections of similar preparations of normal organs of healthy guinea-pigs. It was found that the juice of normal organs produced exactly the same effects. Wasser- mann and Keyser 2 had reported that the blood serum of normal guinea-pigs was toxic for other normal guinea-pigs, if injected intravenously into the latter. In common with many other authors I was unable to confirm their observa- tions. I usually killed my normal guinea-pigs by bleeding them and used the serum thus obtained for intravenous injections on normal hosts. So far as I could determine such injections never produced the least harmful effect on the recipient, even though the latter received ten cubic centi- meters and more. The juice of organs of the same donors was, however, uniformly toxic. It was therefore plain that the organs were not toxic because of any blood from which they had not been freed. Water extracts of dried spleens and lungs were also harmless. These organs were dried at 50 C. for twenty-four hours, then powdered and extracted in distilled water at 50 C. for twenty-four hours. The emul- sion was made isotonic and centrifugalized. The resultant supernatant fluid produced no symptoms when injected intravenously into guinea-pigs. From these experiments it would seem reasonable to con- clude that any tissue products if thrown rapidly into the cir- culation in unusual amounts will intoxicate an animal. In other words, one's own tissues, whether normal or patholog- ical, may conceivably poison one if they get into the blood 46 KRAUSE. stream in excess. Therefore, if the focal reaction, following a tuberculin injection, establishes a better circulatory give and take between the tuberculous focus and the animal organism, and, as it were, " unlocks" or "opens" the focus, the resultant symptoms of a general reaction may be due merely to substances that are primarily toxic for the organ- ism. The same absorption bringing about symptoms of illness will occur whenever the conditions for focal absorp- tion are enhanced, whether it be the more or less continuous absorption from a poorly invested focus or the occasional absorption following physical strain, intercurrent disease, or drug action, any of which may produce circulatory condi- tions about the focus that favor increased absorption. It is interesting that in a recent paper, " The Cellular Factor in Infectious Diseases," Riesman 3 puts forward a sim- ilar hypothesis to explain the toxemia in certain conditions that are remarkable for cellular proliferation and exudation, such as miliary tuberculosis, carcinomatosis, pneumonia, etc. He conceives that in these and similar other " productive " diseases, the proliferated cells which are present in enormous numbers undergo parenteral digestion and in consequence throw " into the blood products that if not qualitatively are at least quantitatively abnormal," and which may give rise to a large share of the symptoms of such diseases. The fore- going work furnishes some experimental basis for Riesman's view. While it would seem from the above experiments that normal tissues, introduced parenterally, are toxic for the organism, it is difficult to imagine how normal tissues could enter the circulation if an individual is in perfect health. But given a fracture of bone, an extensive burn, or a focus of disease, and conditions are at once established that favor the absorption of cellular elements to which the body may react with symptoms of illness, the severity of which would be proportional to the amount of material taken up in a given time. STUDIES IN IMMUNITY TO TUBERCULOSIS. 47 SUMMARY. 1. An extract of an animal's own normal tissues if intro- duced rapidly into its circulation is toxic. 2. The products of tuberculous foci are primarily toxic if received into the blood stream (or, perhaps, lymph stream). 3. It is suggested that the symptoms of the general tuberculin reaction are due to the primary toxicity of focal products, the absorption of which is favored by the focal reaction that results from the injection of tuberculin. Record of Experiments. A. Injections of Tuberculous Tissue Juices : 1. February 21 : Five guinea-pigs, inoculated intraperitoneally on January 27th with tubercle bacilli, H39, were bled to death by cardio- puncture. In every case the omentum was rolled up, nodular and caseous. All had moderate miliary tuberculosis of the spleen. One liver showed a number of pinhead, caseous tubercles. There was no tuberculosis of the lungs in any. The omenta and spleens of the five animals and a small piece of tuberculous liver were ground together in a meat chopper, mixed with 15 cubic centimeters of physiological salt solution, and pressed out with an oil press. (a) Guinea-pig (550 gms.), normal, injected intravenously with 5 cubic centimeters of the tuberculous organ juice. It immediately went into collapse and died in one and one-half minutes Autopsy : No macro- scopic lesions (b) Guinea-pig (650 gms.), normal, injected intravenously with 2.5 cubic centimeters. Lively for one or two minutes, then collapsed and death in three minutes. Autopsy: No macroscopic lesions. The material was then centrifugalized and the supernatant liquid used for the next injection. (c) Guinea-pig (350 gms.), normal, injected intravenously with 1 cubic centimeter of the supernatant liquid. Lively for one minute, then sudden collapse and convulsions, with death in three minutes. Autopsy: Lungs pale and slightly insufflated, few epicardial hemorrhages. The supernatant fluid was then filtered through paper and the filtrate used for the next injection. (d) Guinea-pig (270 gms.), normal, injected intravenously with .75 cubic centimeter of the filtrate. No symptoms. 2. October 5: Three tuberculous guinea-pigs, inoculated September 9th with tubercle bacilli, H37, bled to death. All spleens had diffuse, caseating miliary tuberculosis. The spleens were crushed with sand in a 48 KRAUSE. mortar, mixed with 10 cubic centimeters of distilled water, and stood in the ice-chest for twenty-four hours. The emulsion was then filtered through cotton. Total filtrate equals 7 cubic centimeters. October 6: Guinea-pig (650 gms.), normal, injected intravenously with 6 of filtrate. No symptoms. B. Injections of Non-tuberculous Tissue Juices: 1. February 23: Five normal guinea-pigs were bled to death. Their omenta and spleens and a small piece of liver of one pig were ground together in a meat chopper, mixed with 10 cubic centimeters of physio- logical salt solution, and pressed out in an oil press. February 24: (a) Guinea-pig (500 gms.), normal, injected intrave- nously with 4 cubic centimeters of normal organ juice. Lively for two minutes, then restless, urinating, bucking, falls over, convulsions, marked inspiratory spasms, death in four minutes. Autopsy: No macroscopic lesions. (b) Guinea-pig (530 gms.), normal, injected intravenously with 2.5 cubic centimeters. Lively for two minutes, then uneasy, urinating, coughing, bucking, falls over, convulsions, unconscious, death in four minutes. Autopsy : Lungs pale, slightly blown out. The juice was then centrifugalized and the supernatant fluid filtered through paper. (c) Guinea-pig (500 gms.), normal, injected intravenously with 5 cubic centimeters of filtrate. No symptoms . 2. December 9: Three non-tuberculous guinea-pigs bled to death. The liver tissue from three animals and the lungs and spleen from two were taken and treated separately. Moist weights; liver, 20 grams; lungs, 7.5 grams; spleen, 1 gram. All the organs were crushed in mortars with sand and mixed with physiological salt solution, 10 cubic centimeters to liver pulp, 10 cubic centimeters to lung pulp, and 2 cubic centimeters to spleen pulp. The different emulsions were centrifugalized and the supernatant fluids pipetted off. These fluids were even in appear- ance and contained no macroscopic particles. (a) Guinea-pig (270 gms.), normal, injected intravenously with 1 cubic centimeter of supernatant fluid from spleen juice. Animal lively for one minute, then twitching, convulsions, opisthotonus, rapid respiration. Animal gradually recovered. (b) Guinea-pig (300 gms.), normal, injected intravenously with 3 cubic centimeters of lung juice fluid. No symptoms for one minute, then spasmodic movements, convulsions, death in four minutes. Respiratory picture at end very much like that of anaphylaxis. Autopsy: Lungs resemble the insufflated lungs of anaphylaxis, but are not so pale and not so distended ; no epicardial hemorrhages. The liver juice fluid was filtered through ordinary filter paper, hard filter paper and cotton, successively, and the filtrate used in the following injection : STUDIES IN IMMUNITY TO TUBERCULOSIS. 49 (c) Guinea-pig (220 gms.), normal, injected intravenously with 2 cubic centimeters of filtrate. Same symptoms as foregoing animal, b ; death in two minutes. Autopsy: Lungs moderately distended. No lesions. This is the only instance in which a filtrate of tissue juice produced intoxication. (d) Guinea-pig (260 gms.), normal, injected intravenously with 5 cubic centimeters of mixed blood serum from the same animals from which the organs were taken. No symptoms. 3. December 15: Material consisted of lungs from five non-tubercu- lous guinea-pigs, four of which had been bled to death. Total moist weight is 17 grams. Ground in mortar with sand and mixed with 25 cubic centimeters of physiological salt solution. Stood in ice-box over night. The next morning centrifugalized and the supernatant fluid divided into two portions, one to undergo Berkefeld filtration. The blood serum was removed from the clots one hour after bleeding, incubated at 2,7° C. for one hour and fifteen minutes, then stood at room temperature for twenty-four hours. December 16: (a) Guinea-pig (300 gms.), normal, injected intra- venously with 7.5 cubic centimeters of blood serum. No symptoms. (b) Guinea pig (300 gms), normal, injected intravenously with less than 1 cubic centimeter of unfiltered lung fluid. Collapse coming on one minute after injection. Slow recovery. (c) Guinea-pig (300 gms.), normal, injected intravenously with 2 cubic centimeters unfiltered lung fluid. Same symptoms as foregoing animal, b. Slow recovery. (d) Guinea-pig (220 gms.), normal, injected intravenously with 1.2 cubic centimeters filtrate. No symptoms. 4. December 19 : Three non-tuberculous guinea-pigs were bled to death. The sera of the three animals were mixed together and part of the mixture was incubated one and one-half hours at 37 C. and stood at room temperature for twenty-four hours. The other part was not incu- bated and stood in the ice-box for twenty-four hours. The livers were crushed with sand in a mortar and mixed with physio- logical salt solution. The mixture was centrifugated and the supernatant fluid used. The lungs and spleen were dried on a water bath at 50 C. for twenty- four hours. They were then powdered and extracted in distilled water for twenty-four hours at 50 C. A few cubic centimeters of physiological salt solution were then added and the emulsion was centrifugated. The super- natant fluid was used for injection. (a) Guinea-pig (250 gms.), normal, injected intravenously with 2.5 cubic centimeters of liver juice fluid. Collapse and death in two minutes. Autopsy: Lungs collapsed, no lesions. December 20: (b) Guinea-pig (350 gms), normal, injected intra- venously with 10 cubic centimeters of the incubated serum mentioned above. No symptoms. 50 KRAUSE. (c) Guinea-pig (300 gms.), normal, injected intravenously with 8 cubic centimeters of the non-incubated serum. No symptoms. December 21: (d) Guinea-pig (270 gms), normal, injected intra- venously with 4 cubic centimeters of the extract of dried spleens. No symptoms. .(e) Guinea-pig (280 gms.), normal, injected intravenously with 3 cubic centimeters of the extract of dried lungs. No symptoms. Discussion. — Kraus and Volk 4 have reported that fil- trates of extracts of tuberculous focal material are primarily- toxic for guinea-pigs, while those of normal organs are not. The above experiments fail to confirm their observations as to the effects of tuberculous organ filtrates. The filtrates of material from both tuberculous and normal organs are usually non-poisonous, whether the filtration is done through a Berkefeld filter, paper, or cotton. The even appearance of the supernatant fluids and their freedom from visible parti- cles dispels the idea that the symptoms are caused by cerebral or pulmonary emboli. The harmlessness of the dried tissue extracts makes it seem as though the symptoms are not due to protein intoxication. It would appear that the cellular elements carry some intoxicating substance which exerts its action when large enough quantities of cells or of material from disintegrated cells are rapidly thrown into the circulation. [This work was made possible through the generosity of Mrs. A. A. Anderson of New York City.] BIBLIOGRAPHY. 1. Krause, A. K. Johns Hopkins Hospital Bulletin, Special Tubercu- losis Number, 1911, xxii, 250-258. 2. Wassermann, M., and Keyser, F. Fol. Serol., 191 1 , vii, 243-251. 3. Riesman, David. J. Am. M. Ass., 1915, lxiv, 649-652. 4. Kraus, R., and Volk. Wien. med. Wchnschr., 1910, xxiii, 289-290. COLUMBIA UNIVERSITY LIBRARIES 0050168916 DATE DUE Demco, Inc 38-293