JOURNAL OF THE NATIONAL CANCER INSTITUTE CER Mucosal Injury in Cancer Patients: 2001 I T] TE New Strategies for Research and Treatment Number 29 — Contents Editorial 1 Stephen T. Sonis, Douglas E. Peterson, Deborah B. McGuire, David A. Williams Executive Summary 3 Douglas E. Peterson, Stephen T. Sonis Introduction 6 Douglas E. Peterson, Stephen T. Sonis Biology of Oral Mucosa and Esophagus 7 Christopher A. Squier, Mary J. Kremer Protection Against Mucosal Injury By Growth Factors and Cytokines 16 Dawn Booth, Christopher S. Potten Transgenic Mouse Model of Intestine-Specific Mucosal Injury and Repair 21 Leo Lefrancois, Vaiva Vezys Inflammatory Cytokines and Mucosal Injury 26 David A. Williams Infection and Mucosal Injury in Cancer Treatment 31 Shahab A. Khan, John R. Wingard Biology of Mucosal Pain 37 Christine Miaskowski Mucosal Drug Delivery 41 Vincent H. L. Lee Mucositis in Head and Neck Cancer: Economic and Quality-of-Life Outcomes 45 Amy Peterman, David Cella, Gerald Glandon, Deborah Dobrez, Susan Yount Plenary Session, May 25, 2000 52 Douglas E. Peterson, Stephen T. Sonis I Visit the Journal’s Web site at http://jnci.oupjournals.org/ I EUBLIC HEALTH Liam BERKELEN / LIBRARY " UNIVERS'" o, MONO GRAPHS JOURNAL OF THE NATIONAL CANCER INSTITUTE Number 29 ISSN 0027-8874 ISBN 0—]9—85/539-1 Barnett S. Kramer Editor-in-Chiqf J. Gordon McVie Euro/mun Editor J. Paul Van Nevel Newt [fl/[tor Frederic J. Kaye Douglas L. Weed Reviews Edi/0m Martin L. Brown Economies Editor Hans—Olov Adalni Harry B. Burke Pelayo Correa Ethan Dmitrovsky Merrill J. ligorin Richard Fagerstrom Soldano Ferrone Isaiah J. Fidler Curtis C. Harris Emily L. Harris EDITORIAL BOARD ASS OCIA TE EDITORS Wendy Atkin Frank M. Balis Mariano Barbacid William J. Blot Peter M. Blumherg John D. Boicc. Jr. Louise A. Brinton Ross C. Donehower Susan S. Ellenberg Ellen G. Feigal Suzanne W. Fletcher Michael A. Friedman Patricia A. Gan]. Edward L. Giovannucci Paul Godley John K. Gohagan Michael M. Gottesman Peter Greenwald Jean L. Grem Kathy J. Helllsouer Donald E. Henson Colin R. Jel‘eoatc Patrick G. Johnston Lisa A. Kachnic Frederic J. Kaye Hynda K. Kleinman STA TIS TI CAL EDITORS Janet W. Andersen Per Kragh Andersen Jacques Benichou Stuart G. Baker Donald A. Berry Barry W. Brown Timothy R. Church Bernard F. Cole Kathleen A. Cronin Susan S. Ellenberg Scott S. Emerson Ruth Etzioni Mary A. Foulkes Mitchell H. Gail Edmund A. Gehan Barry I. Graubard Sylvan B. Green Susan G. Groshen Bernard Levin W. Marston Linehan Marc E. Lippman Scott M. Lippman Darrell T. Liu Dan L. Longo Reuben Lotan Douglas R. Lowy Maria Elena Martinez Mads Melhyc Susan G. Nayt‘ield David L. Nelson Ugo Pastorino Daniel F. Heitjan Li Hsu Marek Kimmel Victor Kipnis James R. Murphy J. Jack Lee Richard A. Olshen Philip C. Prorok Timothy R. Rebbeck EDITORIAL ADVISORY BOARD Maureen M. Henderson Gloria H. Heppner Allan Hildesheim Waun Ki Hong William J. Hosk‘ins Peter J. Houghton David H. Johnson V. Craig Jordan [Ian R. Kirsch Kenneth H. Kraemer Alexandra M. Levine Martha S. Linet Henry T. Lynch Pamela M. Marcus Frank L. Meyskens Anthony B. Miller Malcolm S. Mitchell David L. Page Kenneth J. Pienta David Schottenfeld 200/ Gloria M. Petersen David G. Poplack Ross L. Prentice Alan S. Rabson Harvey A. Risch Edward A. Sausville Robert H. Shoemaker Richard M. Simon Michael B. Sporn lan Tannock J. Paul Van Nevel Douglas L. Weed Philip S. Rosenberg Harland N. Sather Daniel J. Schaid Mark R. Sega] Richard M. Simon Donald M. Stablein Robert E. Taronc Sholom Wacholder Richard K. Severson Jerry W. Shay Debra T. Silverman Sandra M. Swain G. Marie Swanson Mario S/.nol Raymond Taetle Peter R. Twentyman Regina G. Ziegler BERKtlEY 513;; A Riv } EDITORIAL STAFF NIVE < r ,kCALtpROLJIA ’ 1/ Scientific Editors: Stephen T. Sonis. D.M.D.. D.M.Sc. \"“’l Douglas E. Peterson. D.M.D., PhD. Deborah B. McGuire, Ph.D., R.N.. F.A.A.N. David A. Williams, M.D. Journal Managing Editor: W. Mark Leader Senior Production Editor: Anne L. Wenzel Manuscript Editors: Elaine Price Beck Joan O’Brien Rodriguez Production Editors: Amanda Maguire Susan L. Sweeney Peer Review Supervisor: Joseph A. McCullough Editorial Assistants: Andrew J. Prugar Stacey Taylor MARKETING STAFF Marketing: Patricia Hudson Press Contact: Dan Eckstein:k EDITORIAL POLICY: Manuscripts from key conferences dealing with cancer and closely related research fields. or a related group of papers on specific subjects of importance to cancer research. are considered for publication. with the understanding that they have not been published previously and are submitted exclusively to the Journal oft/w National Cancer Institute Monographs. All material submitted for consideration will be subject to review. when appropriate. by at least one outside reviewer and one member of the Editorial Board of the Journal oft/1e National Cancer Institute. Opinions expressed by the authors are not necessarily those of the publisher or the editors. Proposals for monographs should be submitted to the Editor-in-Chiel'. Journal of the National Cancer Institute. Oxford University Press. PO. Box 31 110. Bethesda. MD 208244 I 10. Journal of the National Cancer Institute Monographs are available on request to members of the National Cancer Institute Information Associates Program (one copy of each monograph per member). Monographs are also available from Oxford University Press. Journals Subscription Department. 2001 Evans Road. Cary. NC 27513 USA. Telephone toll—free in the US. and Canada, 1—800—8527323. or 919-677-0977. Fax: 919-677-l7l4. E—mail: jnlorders@oup-usa.org. © Oxford University Press *Contact by telephone at 301—986—1891. ext. 1 12. Mucosal Injury in Cancer Patients: New Strategies for Research and Treatment Proceedings of a Multidisciplinary Conference Held in Bethesda, Maryland May 24—25, 2000 Acknowledgments We gratefully acknowledge the following individuals and companies, whose support was pivotal to the success of the conference: National Institutes of Health National Institute of Dental and Craniofacial Research Dr. Ann L. Sandberg National Oral Health Information Clearinghouse Ms. Patricia G. Sheridan Journal ()fthe National Cancer Institute Dr. Barnett S. Kramer, Editor-in-Chief Mr. W. Mark Leader, Managing Editor Ms. Elizabeth Horowitz, Monograph Coordinator Ms. Anne L. Wenzel, Senior Production Editor IntraBiotics Pharmaceuticals, Inc. McNeil Consumer Healthcare Genetics Institute Amgen, Inc. Human Genome Sciences, Inc. We are also indebted to the following colleagues who worked hard and well to coordinate this international conference: Brigham and Women’s Hospital Ms. Maryanne Knasas Ms. Mara McMillan School of Dental Medicine, University of Connecticut Health Center Ms. Kathryn L. Damato Their accomplishments were outstanding. In addition, we thank the speakers for their expert scientific presentations and effective leadership of the often-spirited work- group discussions. Their roles in framing the science and future research directions were invaluable. We hope the results of the conference become viewed as an important milestone relative to research for mucosa] injury in cancer patients. As a result of such research, future cancer pa- tients could likely be spared the often—serious mucosa] toxic effects currently encountered in the clinical setting. STEPHEN T. 50le. D.M.D. D.M.Sc. DOUGLAS E. PETERSON. D.M.D.. PhD. DEBORAH B. MCGUIRE, PhD. RN, F.A.A.N. DAVID A. WILLIAMS, M.D. Mucosal Injury in Cancer Patients: New Strategies for Research and Treatment Contents Editorial Stephen T. Sonis, Douglas E. Peterson. Deborah B. McGuire. David A. Williams Executive Summary Douglas E. Peterson. Stephen T. Sonis Introduction Douglas E. Peterson. Stephen T. Sonis Biology of Oral Mucosa and Esophagus Christopher A. Squier. Mary J. Kremer Protection Against Mucosal Injury By Growth Factors and Cytokines Dawn Booth, Christopher S. Potten Transgenic Mouse Model of Intestine-Specific Mucosal Injury and Repair Leo Lefrangois. Vaiva Vezys Inflammatory Cytokines and Mucosal Injury David A. Williams Infection and Mucosal Injury in Cancer Treatment Shahab A. Khan. John R. Wingard Biology of Mucosal Pain Christine Miaskowski Mucosal Drug Delivery Vincent H. L. Lee Mucositis in Head and Neck Cancer: Economic and Quality-of-Life Outcomes Amy Peterman. David Cella. Gerald Glandon. Deborah Dobrez, Susan Yount Plenary Session, May 25, 2000 Douglas E. Peterson. Stephen T. Sonis I Visit the Journal’s Web Site at http://jnci.oupjournals.org/ l 2001 Number 29 21 26 31 41 45 Editorial Stephen T. Sonis, Douglas E. Peterson, Deborah B. McGuire, David A. Williams Prevention of mucositis in cancer patients is the ultimate goal of the research addressed in this conference. As the proceedings demonstrate, regimen—related mucosal injury is a multifaceted entity that crosses many disciplines at both scientific and clinical levels. Consequently, it is likely that effectively understanding the etiology and biology of mucositis, as well as developing successful interventions, will depend on collaboration among a diverse group of scientists and clinicians. As with most clinical problems, three major areas of inves- tigational activity for mucositis have evolved to date. These areas are 1) discovery of mechanisms contributing to its under— lying pathogenesis, 2) translational studies in models to evaluate potential efficacy of new interventions, and 3) clinical trials to bring new treatments to patients or to evaluate the effect on economic issues and quality of life (QOL). It is clear from the conference outcomes that mucositis as a biological entity is complex. Research has progressed from a collection of observational and descriptive studies to examina- tion of the condition at cellular and molecular levels. There are thus vast opportunities to study the biology of regimen—related mucosal injury. For example, investigators have preliminary in— formation at present regarding changes that occur in tissue fol- lowing challenge with radiation or chemotherapeutic agents. However, knowledge levels remain comparatively limited rela- tive to identifying specifically which cellular and molecular changes occur in various mucosal tissues. For example, what, when, and which receptors are expressed? How do the answers to these questions change over time? These are important issues. Research has demonstrated that a relationship exists between mucositis and venoocclusive disease and that endothelial changes occur in both. What is the contribution of endothelial injury, expression of intercellular adhesion molecules, and changes in permeability to mucosal injury? Morphologic changes in connective tissue suggest that this tissue may serve as a trigger to epithelial damage. How does that occur? What is the sequence? Why is gut epithelium so susceptible? Why are se— lected tnucosal sites more prone to injury than other mucosal sites? These are but some of the questions that require further systematic pursuit. Genetic models may provide a key to fruitful investigation. For example, the role of transcription factors such as NF-KB in triggering cellular injury may be important for understanding initial events that lead to tissue damage. While it appears that certain proteins such as proinflammatory cytokines may have a role in the pathophysiology of the condition, studies have not yet reported how intracellular metabolism or structure is affected by antineoplastic agents. The clinical observation that there is di— versity of mucosal response among patients receiving identical therapy for similar tumors suggests that there may be a genetic predisposition for some forms of mucosal injury. Increasing use of pharmacogenetics as well as assessment by genetic determinants of susceptibility will play substantial roles in understanding patient-to-patient variability relative to occur- rence and severity of mucositis. In addition, identification of Journal of the National Cancer Institute Monographs No. 29, 2001 genetic bases of mucosal cell growth, survival, and differentia— tion will allow future refinement of animal models. Thus, ge— netic determinants will become increasingly important in under- standing both pathophysiology and therapeutics. A different but ultimately related approach seeks to define the effect of new antineoplastic therapies as initiators, promoters, or effectors of mucosal damage. There is obviously a fine line between developing an optimal antitumor agent and minimizing its deleterious impact on normal tissue. Defining biologic dif- ferences between healthy and malignant tissue holds potential benefit for new approaches to antitumor treatment. This strategy also, at the same time, promises to minimize “collateral damage” to normal bystander cells and tissue. Definition of mechanisms will lead to opportunities for new interventions. While current animal models provide an important translational tool for drug development, they are time consuming and relatively expensive. Thus, the models do not provide an opportunity for rapid evaluation of test compounds, and they thereby impair the ability to screen large numbers of potential agents in the context of drug development. Opportunities exist for development of culture—based assays as a means of initially evaluating potential interventions for mucosal injury. On a fun- damental level, manipulated cultures also could serve as the basis for targeted mechanistic studies. Mucosal drug delivery to prevent or treat mucositis offers potential advantages over systemic forms of therapy. Transmu— cosal therapy, especially with molecules of significant size, is a challenge. Studies on formulation, use of permeability enhanc- ers, and techniques to provide locally concentrated levels of drug are all opportunities for future study. Mucositis is also a unique and excellent model for pain stud- ies. The relationship between tissue injury and both activation and sensitization of primary afferent nociceptors provides an important link between objective and subjective findings among patients with mucositis. As each component of mucositis tissue injury is defined, its particular relationship to pain can. in turn, be established and evaluated. Consequently, an opportunity ex- ists for development of directed therapy for pain control. An- other area ripe for potential investigation relies on animal mod- eling of the mechanisms for mucositis—induced pain. Clinical opportunities for pain studies also abound in this arena, both from the standpoint of deriving new techniques to evaluate and accurately record pain as well as from that of developing new forms of pain control. Mucositis carries with it a substantial cost in QOL and eco- Afii’liutions of run/mm: S. T. Sonis, Brigham and Women's Hospital and Dana-Father Cancer Institute. Boston. MA, and Harvard School of Dental Medi- cine, Boston: D. E. Peterson, School of Dental Medicine. Department of Oral Diagnosis, University of Connecticut Health Center, Farmington: D. B. Mc— Guire. University of Pennsylvania School of Nursing. Philadelphia: D. A. Wil— liams. Indiana University School of Medicine and Howard Hughes Medical institute. lndianapolis, IN. Corres/mmlem'e to: Stephen T. Sonis, D.M.D., D.M.Sc.. Brigham and Wom- en‘s Hospital. 75 Francis St.. Boston. MA 021 I5 (e-mail: ss()nis@partners.org). © Oxford University Press nomics. Both of these content areas are potentially fertile areas for future research. As a common and representative regimen— related toxicity, mucositis is a prototypical example of how the “tail can wag the dog" with respect to QOL and cost of care. For example, while the patient’s tumor may be responding well to treatment, toxicity can be sufficiently severe to preclude the patient’s compliance with cancer therapy. The obvious solution to this problem rests with an adequate treatment for mucositis. In the meantime. adequate definition of QOL issues related to mu- cositis may provide opportunity for interim strategies. It is im- portant that these issues become potential endpoints for clinical trials of mucositis interventions. While work to date has initially defined economic and health care costs of mucositis, there is considerable need for additional study. In particular, economic modeling for mucositis associated with a range of protocols has yet to be completed. Scientists, health care providers. and many patients live in a dot—com society. yet there has been relatively little use of the power of the computer to archive. analyze, or model the fre— quency and basis for mucosal toxic effects. Use of multiagent therapies has complicated risk prediction for mucosal toxic ef- fects. However. there is little sharing of data. short of case re- ports. A centralized database is an opportunity waiting to happen. Clearly. the scope of mucosal injury provides an exciting area for innovative, imaginative research in a multiprofessional set— ting. We earnestly hope that this conference has provided a useful context in which to pursue these investigations. NOTE Editor's/10M: D. A. Williams receives royalty payments from Children's Hos- pital. Boston. MA. based on milestone agreements between Children‘s Hospital and Genetics Institute related to the licensing of interleukin 1 1. Journal of the National Cancer Institute Monographs No. 29. 200l Executive Summary Douglas E. Peterson, Stephen T. Sonis BACKGROUND AND SIGNIFICANCE This conference was designed to generate innovative ideas that will ultimately lead to enhanced understanding of mucosal injury and strategically improved therapies for cancer patients. There has been an impressive recent history relative to pub— lications about and funding directed toward mucosal injury in cancer patients. In addition, professional organizations, includ- ing the International Society for Oral Oncology and the Multi- national Association of Supportive Care in Cancer, have targeted mucositis as a major toxicity of cancer treatment for which new research and standards of care are needed. In this context, con— ference participants critically evaluated the current status of sci- ence relative to mucosal injury in cancer patients and delineated future research directions that could ultimately lead to new man— agement strategies. The conference format consisted of a lecture series. multiple workgroup discussions. and a summary plenary session. The research directions that emerged are summarized below. Suc— cessful pursuit of these research themes could lead to clinically important advances in the amelioration of cancer therapy- associated mucositis as well as to enhanced quality of life (QOL) for patients. The research could also potentially permit use of new, more aggressive cytoreductive cancer therapy that results in more durable remissions and improved long-term patient sur— vival rates. FUTURE RESEARCH DIRECTIONS The following two principles were identified as the founda- tion for establishing new research directions: 1) The etiology, progression, and resolution of cancer therapy- associated mucositis are multifactorial in nature. 2) The best research model is ultimately the human model. Specific research themes follow. Models for the Study of Mucosal Biology, Injury, and Repair Mucositis is an important model for the study of mucosal biology, injury, and repair. The discovery of potent agents that might protect or promote healing of the mucosal lining could lead to therapeutic, curative approaches rather than to palliation in cancer patients. Most therapeutic molecules identified to date are cytokines that affect epithelial proliferation. Basic, transla- tional, and applied research involving the most promising mol- ecules should be pursued. In addition to cytokines, there is a critical need to identify and characterize other molecular interventions with similar effects. For example, current research supports the concept that signifi— cant reductions in mucositis can be achieved by appropriate manipulation of stem cell sensitivity by use of growth factors. With the identification of regulatory factors specific for the gastrointestinal tract, it is possible that the stem cells also might be more effectively regulated. New studies are required to assess Journal of the National Cancer Institute Monographs No. 29. 2001 the most efficacious doses and delivery protocols. including combined and sequential use of different cellular and molecular factors. Immune—mediated mucosal injury and repair can be investi- gated in substantial detail with use of in viva model systems that control for antigen expression and the immune response directed toward that antigen. The following are examples: 1) Use of transgenic and gene-targeted mice will continue to define important mechanisms of mucosal injury, including analysis of the stages of epithelial cell damage and repair. These studies thus provide logical and relevant targets for future pharmacologic intervention. New studies are needed to further elucidate basic mechanisms of mucosal cell growth and differentiation. This research should translate these find- ings into patient-oriented research for treatment of inflam- matory bowel disease, in addition to mucositis and other gastrointestinal complications of cancer therapies. Mediators of cell death produced by CD8 T cells that act on intraepithelial lymphocytes can be evaluated via gene knock- out mice and blocking antibodies. Moreover, mechanisms by which tolerance versus autoimmunity is induced are readily testable in this well—defined and tissue—specific system. Fu- ture studies should focus on interaction of CD4 T cells with intraepithelial cell-expressed antigen so that new mecha— nisms relative to mucosal tolerance and immunity can be defined. 3) The gene knockout murine model also presents a unique system in which factors influencing mucosal repair can be studied. This research could directly influence the under- standing of repair of epithelial damage inherent to cancer therapy. Intraepithelial cell damage can be selectively in— duced in enterocytes and is regulated by antigen levels and perhaps other factors such as T—cell number and viral dose. Thus. the system may be manipulated to examine the factors, immune or otherwise, involved in the repair of mucosal tis— sue. A further level of control can be attained by using other mucosal-specific promoters with distinct expression patterns. 2 \_/ Relationships between oral and gastrointestinal mucositis should be further defined, including the potential role of surro— gate markers and the patient-related risk factors, including the possible role of genomics in defining risk profiles for mucositis. Affiliations ()fuullmrx: D. E. Peterson. School of Dental Medicine. Depart~ ment of Oral Diagnosis. University of Connecticut Health Center. Farmington: S. T. Sonis. Brigham and Women‘s Hospital and Dana-Farber Cancer Institute. Divisions of Oral Medicine, Oral and Maxillot‘acial Surgery, and Dentistry. Boston. MA. C0rrus'pnndeuce to: Douglas E. Peterson. D.M.D.. Ph.D.. School of Dental Medicine. Department of Oral Diagnosis. University of Connecticut Health Cen— ter. 263 Farmington Ave.. Farmington. CT 06030—1605 (e—mail: Peterson@ NSO.UCHC.EDU). © Oxford University Press Infection and Mucosal Injury Intact mucosa is an important host defense against systemic infection in neutropenic patients. Conversely. mucosal injury is a significant and identifiable risk factor for localized and sys— temic infections, including those lesions caused by bacteria and fungi. Distinguishing between infectious-related versus regimen— related tissue damage is crucial to maintaining optima] delivery of cytoreductive cancer therapy. These principles collectively provide a basis for future research directed to several concepts, including the following: 1) Understanding the early steps in pathogenesis of infection at damaged mucosal sites could lead to improvements of over— all outcome of cancer patients by reducing morbidity and mortality associated with both mucositis and infection. 2) No molecular intervention has yet been definitively proven to be effective for either prevention or treatment of mucosal injury secondary to cytotoxic cancer therapy. Furthermore. the specific mechanisms by which colonizing pathogens may amplify the severity of pre-existing mucosal damage require further study. It may be possible to bridge these scientific gaps by delineating novel, anti—infective approaches that re- duce overall severity of mucosal toxicity in cancer patients by inhibiting deleterious effects of pathogenic flora at local- ized mucosal sites. Mucosal Pain Basic and clinical studies are needed to characterize the bi- ology of pain associated with mucosal injury. Studies might include the following: 1) Detailed epidemiologic trials to determine patterns and se- verity of acute and chronic pain, as well as related side ef- fects associated with various stomatotoxic chemotherapy and radiotherapy regimens. 2) Characterization of types of pain that result from oral muco— sal injury as well as mucosa at other gastrointestinal tract sites. 3) Development of new animal models that permit evaluation of the anatomy and physiology of nociceptive processes in both normal and inflamed mucosal tissues. Emphasis should be placed on determining which inflammatory mediators acti- vate and sensitize primary afferent nociceptors during muco- sal injury. 4) Delineation of new clinical assessment tools for mucosal pain. including pain arising from nonoral intestinal injury. Knowledge collectively gained from these innovative ap- proaches can be used to develop novel therapies to decrease significant clinical problems associated with pain and its sequel— ae in cancer patients. Mucosal Drug Delivery A major challenge in formulating topical agents for the oral cavity is the need for both adhesion to moist mucosal surfaces and the maintenance of resistance to physical removal by saliva. Strategies to eliminate these research barriers should be pursued, since maximizing drug retention time at localized mucosal sites is important for improving clinical effectiveness. Use of a bio- adhesive gel, for example, may reduce the frequency of appli- cation and amount of drug administered; thus. patient compli— ance is enhanced. In addition, lubrication and physical protection by the bioadhesive gel often lead to reduced discom— fort associated with mucositis. Scientific findings currently exist regarding the effectiveness of transport machinery in facilitating absorption of a diverse array of therapeutic molecules into intestinal epithelial cells. In contrast, further research is needed to understand better the ca- pacity of comparable transport processes in oral epithelial cells that are altered because of oral mucositis. Study of how best to use chemoprotective drugs to mitigate this subcellular injury is also important. The profound effect of selected cytokines on cell proliferation requires that they be delivered locally to mucosa so as to not promote tumor growth in the patient receiving cytoreductive cancer therapy. To exert a mucoprotective effect after topical application, such compounds must transit a surface permeability barrier to reach the proliferative compartment of the epithelium. New research is needed relative to (I) preservation of high local concentrations at the mucosal surface so as to maintain a con— centration gradient and (2) use of permeabilizers to ensure pen- etration of large molecules across the epithelial permeability barrier. QOL and Economic Outcomes Understanding of the effect of mucositis on QOL would be enhanced by a prospective, comprehensive. and longitudinal evaluation of mucositis severity and symptoms in relation to global and specific QOL outcomes. Such research would permit exploration of the potentially complex relationships between physician-graded mucosal injury, patient-reported specific symptom severity, and the multiple domains of QOL. Evaluation of patient preferences for the potentially different acute and long-term consequences of increasingly aggressive cancer treatment protocols is necessary. Precise explanation of QOL implications of different therapeutic regimens may en- hance treatment decision making by the patient, family, and health professionals. This may be particularly valid when there is an absence of clear survival advantage associated with the various treatment modalities under consideration. Systematic, prospective evaluation of economic costs associ- ated with management of mucositis is important. Cost- effectiveness and cost-benefit analyses could be conducted on the basis of the knowledge of true costs of mucositis manage— ment in relation to costs and efficacy of the preventive and therapeutic agents. NEXT STEPS It is essential that ongoing communication occur across rel— evant groups to strategically advance this research agenda. In addition to this Journal of the National Cancer Institute publi- cation, specific next steps, include the following: 1) Posting of conference material on the Web site of the Na- tional Cancer Institute with links to the National Institute of Dental and Craniofacial Research through its National Oral Health Information Clearinghouse as well as Web sites for other relevant National Institutes of Health agencies. 2) Coordination of conference outcomes with the January 2002 clinical consensus conference being developed by the Mul- Journal of the National Cancer Institute Monographs No. 29, 200] tinational Association of Supportive Care in Cancer and the International Society for Oral Oncology. 3) Development of a listserve of conference participants. Continued efforts should be directed to health professional groups and patients to clarify the nomenclature for mucositis across health professional groups. This is essential for the de- termination of precise outcomes. Journal of the National Cancer Institute Monographs No. 29. 2()()I Effective integration of objective and patient—oriented out- comes of interventional clinical trials relative to federal regula- tory mandates is critical. It is important to coordinate this rela- tionship among various user groups, including academic health center investigators, clinicians, industry representatives, govern- ment officials (including those from the Food and Drug Admin- istration and the National Institutes of Health), and patient ad— vocate groups. Introduction Douglas E. Peterson. Stephen T. Sam's A review of mucositis publications during the past 15 years reveals a number of trends. For example. there has been a quan- titative leap relative to the number of publications citing muco- sitis in cancer models; slightly more than l00 studies appeared in the literature in 1986 in contrast to well over 400 studies in 1998. It is likely this escalating pace has been sparked by three factors: I) recognition of mucositis as an important cancer therapy dose—limiting toxicity, 2) high incidence of mucositis in relation to optimal regimens of tumoricidal therapy. and 3) bio— logic complexity Of the condition. An increasing proportion of studies in recent years has evalu— ated mechanistic aspects of mucosal injury. Consequently. un- derstanding of the pathophysiology of mucositis has been stra- tegically advanced. It now seems clear that mucositis represents the endpoint of a process that includes virtually all cell and tissue types within mucosa and that is subject to alteration by local environment and genetic predisposition. Ironically and despite the number of recent studies reporting interventional clinical trials. an effective treatment for mucositis has. to date. been elusive. Thus. mucositis remains an important clinical toxicity for which novel management approaches are needed. Its diverse biologic and clinical nature lends itself to a collaborative effort to ameliorate the condition. The proceedings reported in this monograph derive from the Conference on Mucosal Injury in Cancer Patients: New Strate— gies for Research and Treatment. The symposium was held in Bethesda. MD. May 24—25. 2000. and attracted an eclectic mix of clinicians and scientists that reflected a broad constituency of those interested in mucositis. Approximately one half (51%) of the 120 attendees were from either hospitals or medical or dental schools, 33% were from industry. and 16% traveled the short physical distance from the National Institutes of Health or the Food and Drug Administration. The professional training. re— search backgrounds. and clinical interests of participants were diverse; basic and translational scientists. radiation and medical oncologists. oral medicine specialists. nurses. general dentists. and dental hygienists all were in attendance. Geographic diver— sity was also a hallmark. While the majority of participants came from throughout the United States. other countries and conti- nents including Canada. Europe. Australia. New Zealand. and Asia were also well represented. The National Institutes of Health play a key role in determin— ing the national agenda for biomedical research. The substantial scientific and financial support of the conference by the National Institute of Dental and Craniofacial Research reflected recogni— tion of the importance of mucositis as both a clinical and scien- tific problem. Because of the unmet clinical needs in patients. mucosa] in— jury has become an important niche area for pharmaceutical and biotechnological development. Thus. industry has also exhibited a leading role in driving the science that has enhanced under- standing of the pathophysiology of the condition. In reality. industry will likely convert basic discovery in this area to clini- cally successful therapy. Industrial support. evidenced by atten- dance at the meeting as well as provision of financial resources. was a clear demonstration of a collective corporate commitment both to advance the field and to develop efficacious products for patients at risk for mucosa] injury. The agenda for the conference was designed to combine structured presentations in conjunction with workgroups and broad discussion. This approach. in turn. was directed to the ultimate goal of defining substantive and fruitful areas for future investigation. Text for the formal presentations establishes the basis for this monograph. A summary of the plenary session that evolved based on workgroup proceedings also provides important per— spectives for the future research directions and is included as well. We are pleased that the conference proceedings are being published in the monograph series of the Journal oft/re National Cancer Institute (JNCI). and we thank the JNCI Editorial Board and monograph production team for their support. Access to the monograph will be available through the Web site of the Na- tional Cancer Institute (http://www.nci.nih.gov) and in turn from the National Oral Health Information Clearinghouse Web site (http://www.nohic.nidcr.nih.g0v). Affiliulinns Q/‘uullzm's: D. E. Peterson. School of Dental Medicine. Depart- ment of Oral Diagnosis. University of Connecticut Health Center. Farmington: S. T. Sonis. Brigham and Women's Hospital and Dana-Farber Cancer Institute. Divisions of Oral Medicine. Oral and Maxillofacial Surgery. and Dentistry. Boston. MA. Corn's'pomlmn’i’ to: Douglas E. Peterson. D.M.D.. Ph.D.. School of Dental Medicine. Department of Oral Diagnosis. University of Connecticut Health Cen- ter. 263 Farmington Ave. Farmington. CT 06030—1605 (e-mail: Pctcrson@NS(). UCHC.EDU). © Oxford Univcrsity Press Journal of the National Cancer Institute Monographs No. 29. 200] Biology of Oral Mucosa and Esophagus Christopher A. Squier, Mary J. Kremer The mucosal lining of the oral cavity and esophagus func- tions to protect the underlying tissue from mechanical dam- age and from the entry of microorganisms and toxic mate- rials that may be present in the oropharynx. In different regions, the mucosa shows adaptation to differing mechani- cal demands: Masticatory mucosa consists of a stratified squamous keratinized epithelium tightly attached to the un- derlying tissues by a collagenous connective tissue, whereas lining mucosa comprises a nonkeratinized epithelium sup- ported by a more elastic and flexible connective tissue. The epithelium is constantly replaced by cell division in the deeper layers, and turnover is faster in the lining than in the masticatory regions. Chemotherapeutic agents and radiation limit proliferation of the epithelium so that it becomes thin or ulcerated; this will first occur in the lining regions. The principal patterns of epithelial differentiation are repre- sented by keratinization and nonkeratinization. As keratino- cytes enter into differentiation, they become larger and begin to flatten and to accumulate cytokeratin filaments. In addi- tion to the keratins, the differentiating keratinocytes synthe- size and retain a number of specific proteins, including pro- filaggrin, involucrin, and other precursors of the thickening of the cell envelope in the most superficial layers. The con- cept of epithelial homeostasis implies that cell production in the deeper layers will be balanced by loss of cells from the surface. There is a rapid clearance of surface cells, which acts as a protective mechanism by limiting colonization and invasion of microorganisms adherent to the mucosal surface. [J Natl Cancer Inst Monogr 2001;29:7—15] INTRODUCTION The oral cavity has sometimes been described as a mirror that reflects the health of the individual. Changes indicative of dis— ease are seen as alterations in the oral mucosa lining the mouth, which can reveal systemic conditions, such as diabetes or vita- min deficiency, or the local effects of chronic tobacco or alcohol use. Modern anticancer therapy represents a significant chal— lenge to the integrity of the oral mucosa. Chemotherapeutic agents and radiation therapy limit the proliferative ability of the epithelium so that it becomes thin or ulcerated. This is manifest first in the more rapidly proliferating tissues, such as gastroin— testinal and oral lining mucosae. There may also be indirect effects, such as damage to the salivary glands, that will reduce salivary production and impair barrier efficiency and a reduction in immunocompetence as a result of myeloablative therapy. This will increase the risk of local infection from oral organisms. This article will first describe the organization of the oral mucosa and esophagus, then examine important functional as- pects of the covering epithelium, including epithelial prolifera— tion, differentiation, turnover, and barrier function, all of which have important implications for the maintenance of the integrity of this tissue in the face of anticancer therapy. Finally. since Journal of the National Cancer Institute Monographs NO. 29. 200] most cancer is a disease of the elderly, there will be a brief consideration of changes caused by the aging of the tissue. ORGANIZATION AND FUNCTION OF THE ORAL AND ESOPHAGEAL MUCOSA The mucosa of the mouth and esophagus may appear to differ little from the rest of the moist lining of the gastrointestinal tract, with which it is continuous. In fact, with the notable exception of the uterine cervix, this tissue is remarkably different from other mucosae of the body and has more in common with skin, with which it forms ajunction at the lips, than with the intestinal mucosa. The soft tissues of the human oral cavity and esophagus are covered everywhere by a stratifying squamous epithelium (I). In regions subject to mechanical forces associated with mastication (i.e., the gingiva and hard palate) there is a keratinizing epithe- lium resembling that of the epidermis covering the skin. In these masticatory mucosae, the keratinized epithelium is tightly at— tached to the underlying tissues by a collagenous connective tissue, or lamina propria. The floor of the mouth, buccal regions, and esophagus, which require flexibility to accommodate chew— ing, speech, or swallowing of a bolus, are covered with a non— keratinizing epithelium. The connective tissue of lining mucosae is more elastic and flexible than the connective tissue in the masticatory mucosa. The dorsum of the tongue is covered by a specialized epithelium, which can be represented as a mosaic Of keratinized and nonkeratinized epithelium. This epithelium is attached tightly to the muscle of the tongue. Fig. 1 illustrates diagrammatically the distribution of the dif— ferent types of mucosa within the oral cavity (2). From mea— surements made by Collins and Dawes (3), it can be calculated that the masticatory mucosa represents approximately 25%, the specialized mucosa (dorsum of tongue) approximately 15%. and the lining mucosa approximately 60% of the total surface area of the oral lining. The esophagus extends from the upper esophageal sphincter, which delineates it from the oropharynx, to the lower esophageal sphincter, representing the junction with the gastric mucosa (4). The organization of the tissues reflects their function—that of transporting ingested food from the oral cavity to the stomach. The process of peristalsis, which is initiated by swallowing and involves rhythmic contractions of the muscular walls, accom— plishes this transportation. The extensibility and motility of the mucosal lining are reflected in the presence of a nonkeratinized mucosal surface resembling that of the oral lining mucosa (Fig. Affiliation ol'unrlmrs.‘ Dows Institute for Dental Research, College of Den- tistry. University of Iowa, Iowa City. ('m-ravpmnlmnr to: Christopher A. Squier. Ph.D., D.Sc.. F.R.C.Path. N419 DSB. College of Dentistry. University of Iowa. Iowa City. IA 52242 (e-mail: christopher—squier@uiowacdu). See “Note” following “References." 0 Oxford University Press Fig. 1. Diagram to show the anatomic location and extent of inasticatory. lin- ing. and specialized mucosa in the oral cavity. lModified from reference (2).] Upper Lip Underside of Tongue Alveolar Mucosa Hard Palate Gingiva Soft Palate Cheek Floor of Mouth Tongue Lower Lip I Masticatory Mucosa [j Lining Mucosa D Specialized Mucosa 2). This surface is separated from the submucosa by a muscularis mucosa, consisting of a smooth muscle and elastic fiber layer. which may serve to reduce the excursion of the luminal lining mucosa as a result of the contractions of the external esophageal muscle. consisting ofcircular and transverse layers of striated or smooth muscle. The primary function of oral and esophageal epithelium is the protection of the underlying tissue (1). In the masticatory re— gions, the mechanically tough stratum corneum serves to dissiv pate shearing forces, and in the lining areas. including the esophagus, there is a distensible and flexible surface layer. In both regions, lipid-based permeability barriers in the outer epi— thelial layers protect the underlying tissues against fluid loss and against the ingress of a range of potentially harmful environ- mental agents. These include microbial toxins and enzymes and antigens and carcinogens from foods and beverages. STRUCTURE OF THE ORAL AND ESOPHAGEAL MUCOSA All covering and lining tissues of the body consist of a sur— face epithelium supported by a fibrous connective tissue. Epi— thelium. by virtue of the close packing and constant turnover of cells. is well adapted to protect underlying tissues and organs against mechanical and chemical insult, whereas the connective tissue, consisting of relatively few cells in an extensive matrix. provides mechanical support and nutrients for the epithelium. In comparing the structure of skin and oral mucosa to the gastro- intestinal tract, a major difference emerges in the organization of the epithelium, which reflects the different functions of these regions. The lining of the stomach and small and large intestine consists of a simple epithelium composed of only a single layer of cells, which facilitates absorption across the tissue. Skin, oral mucosa. and esophagus are covered by a stratified epithelium (Fig. 3) composed of multiple layers of cells that show various patterns of differentiation (or maturation) between the deepest cell layer and the surface. Features that distinguish the oral and esophageal mucosa from skin are its moist surface and the absence of appendages. The skin contains numerous hair follicles. sebaceous glands, and sweat glands, whereas the glandular component of oral and esophageal mucosa is represented primarily by the minor sali— vary glands. These glands are concentrated in the submucosa. and the secretions reach the mucosal surface via small ducts. The salivary glands have an important role in maintaining a moist surface containing mucins and a variety of antimicrobial sub stances as well as epidermal growth factor (EGF). In the esopha— gus, the minor salivary glands can produce a secretion with high bicarbonate concentration to neutralize refluxing stomach acid (5). Sebaceous glands are present in the upper lip and buccal mucosa in about three quarters of adults. Unlike the esophagus. longitudinal smooth muscle circular submucous coat squamous epithelium lamina propria of the mucous membrane Fig. 2. The organization of the tissues ofthe human esophageal lining. [Modified from reference (4).] Journal of the National Cancer Institute Monographs No. 29. 200] Layer Membranmcoaung- Granules Layer Fig. 3. Principal ultrastructural features of differentia- a tion in (a) keratiniled oral epithelium and (b) nonkera— linized oral and esophageal epithelium. [Modified from reference (48“ Membrane-coating Granules Granular ’ PrickIe-cell the oral mucosa has no muscularis mucosae, and. consequently, it is difficult to identify clearly the boundary between it and the underlying tissues. In many regions. such as the cheeks, the lips, and parts of the hard palate, a layer of loose fatty or glandular connective tissue containing the major blood vessels and nerves supplying the mucosa separates the oral mucosa from underlying bone or muscle. This represents the submucosa in the oral cavity, and its composition determines the flexibility of the attachment of the oral mucosa to underlying structures. A similar organiza- tion is seen in the esophagus. In regions of the oral mucosa. such as the gingiva and parts of the hard palate, the oral mucosa is attached directly to the periosteum of underlying bone with no intervening submucosa. This arrangement is called a mucoperi- osteum and provides a firm, inelastic attachment. In several regions of the oral cavity, there are nodules of lymphoid tissue Journal of the National Cancer Institute Monographs No. 29. 200! consisting of crypts formed by invagination of the epithelium into the lamina propria. These areas are extensively infiltrated by lymphocytes and plasma cells. Because of their ability to mount immunologic reactions, such cells play an important role in com- bating infections of the oral regions. The mucosa] lamina propria consists of cells, blood vessels, neural elements, and fibers embedded in an amorphous ground substance. The lamina propria shows regional variation in the proportions of its constituent elements. particularly in the con- centration and organization of the fibers. Cancer therapies will tend to lower cell proliferation and turnover in connective tissue; ionizing radiation has a direct effect on large molecules that make up the ground substance, so that depolymerization occurs, vascular permeability increases, and there will be tissue edema and an inflammatory infiltrate (6). Damage to fibroblasts will result in cell loss and the appearance of abnormal cells leading to fibrosis after about 6 months (7). Similarly, damage to blood vessels will lead to hypovascularity and tissue ischaemia (6). Together, these changes will reduce the ability of the tissue to heal and resist infection (8). CELLULAR AND MOLECULAR EVENTS IN DIFFERENTIATION IN ORAL AND EsoPHAGEAL EPITHELIUM The effects of cancer therapy primarily manifest in the oral and esophageal mucosae as changes in the epithelium that reflect damage to proliferating and differentiating cells. This section will describe keratinocyte structure and function in normal tis- sue. Chemotherapeutic agents and radiation therapy limit the proliferative ability of the epithelium so that it becomes thin or ulcerated. Basal keratinocytes are cuboidal or columnar cells with a bounding plasma membrane and a full complement of the normal intracellular organelles (Fig. 3). These cells are capable of division so as to maintain a constant epithelial population as cells are shed from the surface. Tissue homeostasis requires differentiation and desquamation at the epithelial surface to be matched by cell division. Many factors, including aging and disease, can alter this balance so that an epithelium may become thicker (hyperplastic) or thinner (atrophic) than normal. The progenitor cells are situated in the basal layer in thin epithelia, such as the floor of the mouth, and in the lower two to three cell layers in thicker epithelia, such as the cheek, esopha- gus, and palate. Dividing cells tend to occur in clusters so that more are seen at the bottom of epithelial ridges than at the top. The progenitor compartment is not homogeneous but consists of two functionally distinct subpopulations of cells. A small popu- lation of progenitor cells cycles very slowly and is considered to represent stem cells whose function is to produce basal cells and retain the proliferative potential of the tissue (9—11). Because it divides infrequently, the epithelial stem cell may be important in preserving the genetic information of the tissue, since DNA is most vulnerable to damage during mitosis. While the position of stem cells can be related to anatomic structure in some tissues, such as intestine. tongue papillae, and hair follicles, the cells are not morphologically identifiable in most areas of skin and oral mucosa. There have been many attempts to develop specific stem-cell markers, including the presence of adhesion mol- ecules. such as the Bl—integrins, B—catenin, and cytokeratins 15 and 19 which some have claimed can be used to identify these cells in skin and oral mucosa (l2—15). The larger portion of the progenitor compartment is composed of amplifying cells whose function is to increase the number of cells available for subse- quent maturation by entering into mitosis. The control of epithelial proliferation and maturation is the subject of extensive research, and there are a large number of biologically active substances, most of which are peptide growth factors that are collectively termed cytokines and that may stimulate or suppress epithelial cell proliferation. Those that stimulate keratinocyte proliferation include epidermal growth factor (EGF), transforming growth factor—0t (TGF—OI), platelet- derived growth factor (PDGF), and interleukin 1 (IL-1) (16—18). The rate of proliferation is the result of interaction between positive and negative regulators, which act via a complex control system involving the binding of peptide factors to cell surface receptors, a cascade of cytoplasmic elements regulated by the activities of kinases and phosphatases, and transcriptional activ— 10 ity in the nucleus leading to expression of proteins involved in cell cycle regulation (18.19). Mitotic activity can also be affected by a number of factors, such as time of day, stress, and inflammation. For example, the presence of a slight subepithelial inflammatory cell infiltrate stimulates mitosis, while severe inflammation causes a marked reduction in proliferative activity. It has recently been demon- strated that, for buccal epithelium, there is a clear circadian rhythm, with most cells being in the mitotic (M) phase at 2100 hours ( [9). Since the M phase represents one of the most radio- sensitive stages of the cell cycle, radiation therapy involving the oral mucosa should optimally be administered in the morning. The use of different techniques has led to a wide range of estimates of the rate of cell proliferation in the various epithelia, but, in general, the rate is highest for cells in the thin nonkera- tinized regions, such as floor of mouth and underside of tongue, than for the thicker keratinized regions, such as palate and gin- giva (20) (see Table 1). Apart from measuring the number of cells in division, it is also possible to estimate the time necessary to replace all of the cells in the epithelium. This is known as the turnover time of the epithelium and is derived from knowledge of the time it takes for a cell to divide and pass through the entire epithelium. Published human data for turnover times range from a median value of 34 days for epidermis to 4 days for the small intestine, with the values for oral and esophageal epithelium falling between (2/,22) (see Table 1). The regional differences in the patterns of epithelial maturation appear to be associated with different turnover rates; for example, nonkeratinized buccal epithelium turns over faster than keratinized gingival epithelium. Such differences can have important implications for healing and for the rate of recovery of the tissue from damage, which is of particular relevance in considering the effects of cancer therapy on these regions. Clinically, these differences are re— flected both in the more rapid appearance of therapy—induced mucositis than in dermatitis and in the prevalence of damage to nonkeratinized rather than to keratinized surfaces. After cell division, each daughter cell either recycles in the progenitor population or enters the maturing compartment. The switch between proliferation and differentiation is modulated by the presence of factors, such as extracellular calcium, phorbol esters, retinoic acid, and vitamin D3 (23). Cells in the basal layer are attached by integrin-containing focal adhesions, and differ- entiation involves migration with a loss of integrin expression and an increase in cadherin-mediated adhesion via close inter— cellularjunctions or desmosomes. There are also changes in the Table l. Epithelial cell proliferation and turnover in selected tissues Mean labeling index. ‘7(* Tissue region Median turnover time, days‘i' Small intestine — 4 Floor of mouth 12.3 20 Labial mucosa 1 1.8 — Buccal mucosa 10.2 14 Ventral tongue 10.1 — Esophagus 21 :1: Gingiva 9.1 w Hard palate 7.2 24 Dorsal tongue 4.3 — Skin _ 27 *Reference (/9). +Refercncc (20). :Reference (2]). Journal of the National Cancer Institute Monographs No. 29, 2001 composition of intracellular proteins, termed cytokeratins, and in the development of new ones, including involucrin, loricrin, and filaggrin (24,25). The principal patterns of differentiation are represented by keratinized and nonkeratinized epithelia. Differentiation in ke- ratinized epithelia (Fig. 3, a) leads to production of the stratum corneum. The cornified cells making up this layer are flat and hexagonal in shape (26), filled with a compact array of con- densed cytokeratin filaments (27), bounded by a thickened cell envelope (28), and surrounded by an external lipid matrix (29,30). As cells leave the basal layer and enter into differentiation, they become larger and begin to flatten and accumulate cyto- plasmic protein filaments, representing the cytokeratins. Kera- tins represent 30 different proteins of differing molecular weights; those with the lowest molecular weight (40 kd), such as keratins 8 and 18, are found in glandular and simple epithelia; keratins of intermediate molecular weight are found in stratified epithelia; and the largest keratins (approximately 67 kd) are found in keratinized stratified epithelium. All stratified oral epi— thelia possess keratins 5 and 14 in the undifferentiated basal cells, but differences emerge in the suprabasal layers with dif- ferentiation. Ortho-keratinized oral epithelium, such as the pal- ate, contains keratins l and 10, whereas gingiva and parakera- tinized palatal epithelium contains keratins l and 10 or keratins 4 and 13. Nonkeratinized epithelium, including esophagus. con- tains keratins 4 and 13 (31,32). As the cells enter the prickle cell layer, small organelles known as membrane-coating granules or lamellar granules rep— resenting accumulating lipid become evident (Fig. 3, a) (33). In addition to the accumulation of lipids and keratins, the differ- entiating keratinocytes synthesize and retain a number of spe- cific proteins, including profilaggrin (34,35), involucrin (36), and other precursors of the thickening of the cell envelope (37). At the boundary between the granular and cornified layers. the membrane-coating granules migrate to the superficial (apical) aspect of the keratinocyte, where the bounding membrane of the organelle fuses with the cell plasma membrane so that the lipid lamellae are extruded into the extracellular spaces of the surface layer (28,29). Thus, the membrane-coating granules are believed to be responsible for the formation of a superficial, intercellular, permeability barrier in stratified squamous epithelium. After the granules are extruded. the interior of the cell becomes filled with aggregated cytokeratin filaments, and involucrin. loricrin, and other proteins are deposited on the inner aspect of the plasma membrane as a thick band of protein that becomes covalently cross-linked (24,25). In keratinized oral epithelium, about 50% of the intercellular space of the stratum corneum is occupied by desmosomes (38), and the interdesmosomal regions are frequently dilated. Al- though the extruded membrane-coating-granule contents fuse to form multiple broad lipid sheets in the intercellular spaces of the stratum corneum of this tissue, the number of individual lamel- lae in oral tissue is less than that observed in epidermis. In nonkeratinizing epithelia (Fig. 3, b), the accumulation of lipids and of cytokeratins in the keratinocytes is less evident and the change in morphology is far less marked than in keratinizing epithelia. The mature cells in the outer portion of nonkeratinized epithelia become large and flat and possess a cross—linked pro— tein envelope, but they retain nuclei and other organelles, and the cytokeratins do not aggregate to form bundles of filaments, as Journal of the National Cancer Institute Monographs No. 29. 2001 seen in keratinizing epithelia. As cells reach the upper one third to one quarter of the epithelium, membrane-coating granules become evident at the superficial aspect of the cells and appear to fuse with the plasma membrane so as to extrude their contents into the intercellular space. The membrane—coating granules found in nonkeratinizing epithelia are spheric in shape and mem- brane bounded and measure about 0.2 am in diameter (39). They have often been referred to as cored granules because of their appearance in transmission electron micrographs. Such granules have been observed in a variety of human nonkera— tinized epithelia. including oral mucosa (4042), esophagus (43), and uterine cervix (44). Studies employing ruthenium te— troxide as a postfixative have indicated that a small proportion of the granules in nonkeratinized epithelium do contain lamellae, which may be the source of short stacks of lamellar lipid scat— tered throughout the intercellular spaces in the outer portion of the epithelium (45). In contrast to the appearance of the inter— cellular spaces of the surface layer of keratinized epithelia, those of the superficial layer of nonkeratinizing epithelia contain elec- tron lucent material, which may represent nonlamellar phase lipid, with only occasional short stacks of lipid lamellae. It is the absence of organized lipid lamellae in the intercellular spaces that accounts for the greater permeability of this tissue. The concept of epithelial homeostasis implies that cell pro— duction in the deeper layers will be balanced by loss of cells from the surface. While there has been much focus on pro- grammed cell maturation and death (e.g., apoptosis) in other systems, comparatively little is known about the events deter— mining desquamation in skin and mucosa. The available evi- dence suggests a programmed breakdown of cell adhesion mol- ecules, involving both lipids and proteins, probably by intercellular enzymes that might originate in the extruded mem— brane-coating granules (46). Regardless of the nature of the pro— cess, the rate at which cells leave the surface represents a de- fense mechanism by rapidly clearing the substrate to which many microorganisms adhere so that they are unable to produce toxic effects or to invade. Data for murine oral mucosa from Kvidera and Mackenzie (47) suggest a clearance of surface cells in 2—4 hours, depending on the region. While these rates are likely to be lower in humans, the process will clearly limit colo- nization and invasion. NONKERATINOCYTES IN ORAL AND ESOPHAGEAL EPITHELIUM Many histologic sections of oral and esophageal epithelium contain cells that differ in appearance from the other epithelial cells, and it is obvious from ultrastructural and immunochemical studies that they represent a variety of different cell types, in- cluding pigment—producing cells (melanocytes), Langerhans’ cells, Merkel cells, and inflammatory cells such as lymphocytes, which together can make up as much as 10% of the cell popu- lation in the oral epithelium (48). All of these cells except Merkel cells lack desmosomal attachments to adjacent cells, so that during histologic processing, the cytoplasm shrinks around the nucleus to produce the clear halo. None of these cells contain the large numbers of tonofilaments and desmosomes seen in the epithelial keratinocytes nor do they participate in the process of maturation seen in oral epithelia; therefore, they are often col- lectively called nonkeratinocytes. Melanocyte and Pigmentation The endogenous pigments most commonly contributing to the color of the oral mucosa are melanin and the hemoglobin in the blood. Melanin is produced by the specialized pigment cells called melanocytes. which are situated in the basal layer of the oral epithelium and the epidermis. Melanocytes lack desmo— somes and tonofilaments but possess long dendritic processes that extend between the keratinocytes. often passing through several layers of cells. Melanin pigment is synthesized within the melanocytes as small structures called melanosomes. These are inoculated or injected into the cytoplasm of adjacent kera- tinocytes by the dendritic process of the melanocyte. Similar cells have been described in the esophageal epithelium and can give rise to melanotic lesions (49). Another type of dendritic cell sometimes seen in the su- prabasal layers of epidermis and oral and esophageal epithelium is the Langerhans’ cell (48,50). It is usually demonstrated by specific immunochemical reactions that stain cell surface anti— gens. Langerhans' cells may be capable of limited division within the epithelium, but it is clear both that they can move in and out of the epithelium and that the source of these cells is the bone marrow. This is in accord with evidence suggesting that they have an immunologic function, recognizing and processing antigenic material that enters the epithelium from the external environment and presenting it to helper T lymphocytes. It also seems likely that Langerhans‘ cells can migrate from epithelium to regional lymph nodes. The Merkel cell is situated in the basal layer of the oral and esophageal epithelium and epidermis (48,51). It possesses kera- tin tonofilaments and occasional desmosomes linking it to adja— cent cells, but the characteristic feature of the Merkel cell is the presence of small. membrane-bound vesicles in the cytoplasm, sometimes situated adjacent to a nerve fiber associated with the cell. These granules may liberate a transmitter substance across the synapselike junction between the Merkel cell and the nerve fiber and. thus, trigger an impulse. This arrangement is in accord with neurophysiologic evidence suggesting that the Merkel cell is a sensory cell responding to touch. Merkel cells may arise from division of an epithelial cell (keratinocyte). Inflammatory Cells When sections of epithelium taken from clinically normal areas of mucosa are examined microscopically. a number of inflammatory cells can often be seen in the nucleated cell layers. These cells are transient, and the cell most frequently seen is the lymphocyte. although the presence of polymorphonuclear leukocytes and mast cells is not uncommon. Lymphocytes are often associated with Langerhans’ cells. which are able to acti- vate T lymphocytes. It is becoming evident that the association between nonkera— tinocytes and keratinocytes in skin and oral mucosa represents a subtle and finely balanced relationship in which cytokines represent the controlling factors (16). Thus, keratinocytes pro- duce interleukins (l. 6. 7. 8. 10, 11, and I2). colony-stimulating factors (GM, G. and M), and tumor necrosis factor-0t, all of which modulate the function of Langerhans' cells. In turn, Lang- erhans‘ cells produce lL-l. which can activate T lymphocytes. which secrete lL—2. thus bringing about proliferation of T cells capable of responding to antigenic challenge. lL-l also increases the number of receptors to melanocyte—stimulating hormone in 12 melanocytes and so can affect pigmentation. The influence of keratinocytes extends to the adjacent connective tissue where cytokines produced in the epithelium can influence fibroblast growth and the formation of fibrils and matrix. EPITHELIAL SURFACE BARRIER In a variety of stratified squamous epithelia. there is an ef- fective permeability barrier in the tissue. For example. present in the oral mucosa and esophagus is an abundant flora contain- ing many opportunistic organisms, yet inflammatory lesions are relatively infrequent, except around the teeth. The location of this barrier in the superficial layers of the epithelium has been confirmed by experiments that demonstrate an increase in per- meability when the surface layers are removed by stripping (52). Studies with microscopically visible tracers. such as small pro— teins (53) and dextrans (54), suggest that the major pathway across stratified epithelium of large molecules is via the inter- cellular spaces and that there is a barrier to penetration as a result of modifications to the intercellular substance in the superficial layers. described in the previous section. However. it is clear from measurements of permeability that different compounds may penetrate an epithelium at different rates, depending on the chemical nature of the molecule and the type of tissue being traversed. This has led to the suggestion that materials with different chemical properties cross the barrier region by different routes, some crossing the cell membrane and entering the cell (transcellular or intracellular route) and others passing between the cells (intercellular route). For oral mucosa, Squier and Lesch (55) have used light and electron microscopic autoradiography to show the route taken by isotopically labeled compounds ap- plied to the surface of the tissue. Compounds ranging from water to cholesterol. applied to either keratinized or nonkeratinized oral epithelium, could be subsequently localized in the intercel— lular regions of the superficial layer of the tissues. suggesting that this compartment is the predominant route for compounds moving across the barrier layer of oral epithelium. However, Zhang and Robinson (56) have pointed out that the pH depen— dency that is evident in absorption of ionizable compounds reflects their partitioning into the epithelial cell membrane, so it is likely that such compounds will tend to penetrate transcel— lularly. Finally, from the point of view of delivering bioactive peptides that might protect the epithelium during cancer therapy. it is worth noting that the superficial layers of the tissue may act as a reservoir for topically applied compounds. Although this phenomenon has been inferred from kinetic studies in oral mu- cosa (57,58), it is poorly understood. As we have already men- tioned. the permeability barrier in nonkeratinized epithelia con— sists of groups of lipid lamellae located in the intercellular spaces of the superficial epithelial layer (45,59). These limit the penetration of nonpolar compounds, which may become trapped in a nonlipid or fluid lipid intercellular compartment of the bar— rier layer. Thus, the surface layer of the epithelium may take up a compound relatively rapidly (depending on its lipophilicity and the nature of the vehicle). Once saturated. this layer cannot adsorb any more material. regardless of the duration of expo- sure. Subsequently. the adsorbed material diffuses into the deeper layers of the tissue at a fairly constant rate that is more dependent on the capacity (or loading) of the reservoir than on the duration of surface exposure. The constancy of the oral environment is ensured to a large extent by the continual secretion of saliva into the oral cavity Journal of the National Cancer Institute Monographs No. 29. 200! from the three major salivary glands and numerous minor sali- vary glands located in. or beneath. the mucosa. Saliva. by con- tinually bathing the surface of the oral mucosa, maintains a moist atmosphere and a stable. but slightly acidic. pH. Corn- pared with the secretions of the gastrointestinal tract. saliva is a relatively mobile fluid with less mucin. limited enzymatic activity. and virtually no proteases. The mucosal surface has a salivary coating that has been estimated to be 70 mm thick (3) and that may act as an unstirred fluid layer. Several independent lines of evidence suggest that saliva and salivary mucin contrib- ute to the barrier properties of oral mucosa (60). Within saliva there is a high-molecular—weight mucin named MG] (6/) that can bind to the surface of the oral mucosa so as to maintain hydration. provide lubrication, concentrate protective molecules such as secretory immunoglobulins. and limit the attachment of microorganisms. Histatins are small salivary-derived histidine- rich polypeptides with marked antifungal activity (62). These may be augmented by the activity of a recently discovered class of antimicrobial peptides. the defensins. that are expressed by oral epithelium (63). AGING or ORAL MUCOSA Skin shows well—documented changes in structure and func— tion with age. most of which arise from chronic exposure to UV radiation (i.e., photoaging). The oral mucosa. being protected from such environmental effects. shows few changes that can be unambiguously ascribed to aging. In some regions. there is a slight thinning of the epithelium with a concomitant flattening of the epithelial—connective tissue interface (64). Despite claims of a reduction in the rate of cell proliferation with age, there are no clear data to support this for human tissue. although there may be some increase in turnover time (65). The limited information available on the permeability of oral mucosa indicates that there is a trend toward decreased perme- ability to water with age. which is statistically significant for floor-of—mouth mucosa from females (66). It is of interest to note that. in skin. where the mOIphologic changes with age are more marked than in oral mucosa. there have been a number of reports that demonstrate a significant decrease in permeability with age: Squier et al. (66) discussed the reasons for this. Among the age changes evident in the lamina propria are those affecting the vascular system. Although there is some evi- dence for a reduction in the number of individual vessels with active flow (67), it is not known whether this reduction affects overall blood flow and perfusion. Systemic conditions encoun- tered in the elderly that can affect the oral vasculature include diabetes and atherosclerosis (68,69). In a study of blood flow in atherosclerotic monkeys. Goodman and Squier (70) reported a 50% reduction in flow in the oral mucosa. However. given the ample blood supply to the oral tissues. it appears that perfusion is still sufficient to tissue viability. even in the presence of these vascular alterations (7/). RELATIONSHIP TO MUCOSAL INJURY IN CANCER Anticancer therapy represents a significant challenge to the integrity of mucosal tissues. Chemotherapeutic agents and ra- diation limit proliferative ability so that the overlying epithelium becomes thin or ulcerated. This effect is first seen in the more rapidly proliferating tissues, such as gastrointestinal and oral lining mucosa. where atrophy and ulceration can represent a dose—limiting and potentially serious complication of treatment. Journal of the National Cancer Institute Monographs No. 29. 200] In the oral mucosa. lesions first appear on the soft palate. tongue. and cheeks; as they enlarge, they lead to extreme pain and dys— phagia. As a consequence. there may be dehydration, a compro- mised nutritional status because of painful chewing, and a de- creased quality of life. The effect of cancer therapies is not limited to epithelia and will tend to lower cell proliferation and turnover in connective tissue; ionizing radiation has a direct effect on the tissue matrix leading to an increase in vascular permeability. tissue edema, and an infiltration of inflammatory cells. Damage to fibroblasts will result in cell loss and fibrosis; similarly. damage to blood vessels will lead to hypovascularity and tissue ischemia. To- gether. these changes will reduce the ability of the tissue to heal and resist infection. There may also be indirect effects, such as damage to the salivary glands, which will reduce salivary pro- duction and impair barrier efficiency, and a reduction in immu- nocompetence as a result of myeloablative therapy. This will increase the risk of local infection from oral organisms. FUTURE RESEARCH DIRECTIONS The treatment of mucosal injury in cancer patients has tended. in the past. to focus on palliation. Apart from effective combinations of antimicrobials. local anesthetics, and. possibly. anti-inflammatory agents. nonirritative vehicles that will coat the mucosa to enhance lubrication and provide some degree of oc- clusion so as to relieve acute symptoms have also been included. The discovery of potent agents that might protect or promote healing of the mucosal lining leads to the possibility of therapy rather than palliation. Most of the candidates are cytokines. with an effect on epithelial proliferation that has already been men- tioned (“Cellular and Molecular Events in Differentiation in Oral and Esophageal Epithelium.” above). While intuitively, it might be assumed that agents that increase epithelial prolifera- tion would protect the epithelium during anticancer therapy, ani— mal studies suggest the opposite effect—~increasing proliferation sensitizes epithelial cells to the effects of chemotherapy and results in increased mucositis. This discovery has led to in- creased interest in cytokines that act by arresting epithelial cell division. thus sparing the cells from the effects of anticancer therapy. After release from arrest, there is rapid proliferation and repopulation of the tissue so as to restore normal mucosal func— tion. Studies to identify other compounds with these effects and to characterize their behavior will be critical if management of mucosal injury is to progress from palliation to therapy. The profound effect of cytokines on cell proliferation makes it essential that they be delivered locally to the mucosa, so as not to interfere with anticancer therapy. In practice, this demands topical application. To exert an effect after topical application. such compounds must pass across a surface permeability barrier (described in “Cellular and Molecular Events in Differentiation in Oral and Esophageal Epithelium,” above) to reach the prolif- erative (basal) compartment of the epithelium. These require- ments demand the maintenance of high local concentrations at the mucosal surface. so as to maintain a concentration gradient. and the presence of permeabilizers to ensure penetration of large molecules across the epithelial permeability barrier. A major challenge in formulating topical agents for the oral cavity is the need for adhesion to the moist surface of the mucosa and the need to resist the flushing action of saliva. The use of bioadhesive gels reduces the frequency of application and the amount of drug administered and can also improve patient com— 13 pliance and acceptance. Optimizing the retention time of the drug is important in improving its clinical effectiveness. Finally. for the mucositis patient, the occlusion and lubrication of a bio- adhesive gel reduce the discomfort of the lesion. REFERENCES (I) Squier CA. Hill MW. Oral mucosa. In: Ten Cate AR. editor. Oral histol— ogy: development. structure and function. St. Louis (MO): CV Mosby: 1989. p. 341—81. (2) Roed-Petersen B. Renstrup G. A topographical classification of the oral mucosa suitable for electronic data processing. Its application to 560 lett— koplakias. Acta Odontol Scand 1969;27:681—95. Collins LM. Dawes C. The surface area of the adult human mouth and the thickness of the salivary film covering the teeth and oral mucosa. J Dent Res 1987;66:1300—2. Gavaghan M. Anatomy and physiology of the esophagus. AORN J 1999: 69:372—86. (5) Long JD. Orlando RC. Esophageal subtnucosal glands: structure and func- tion. Am J Gastroenterol 1999;94:2818—24. (6) Baker DG. The radiobiological basis for tissue reactions in the oral cavity following therapeutic x-irradiation. Arch Otolaryngol 1982;108:21—4. (7) Sonis ST. Epidemiology. frequency. distribution. mechanisms. and histo- pathology. In: Peterson DE. Sonis ST. editors. Oral complications of cancer chemotherapy. The Hague (The Netherlands): M. Nijhof; 1983. p. 1—12. (8) Squier CA. Mucosal alterations. NCI Monogr l99():9:169—72. (9) Hume WJ. Potten CS. Advances in epithelial kinetics—an oral view. J Oral Pathol 1979:823—22. (10) Hume WJ. Potten CS. Proliferativc units in stratified squamous epithelium. Clin Exp Dermatol 1983182954106. (ll) Bickenbach JR. Identification and behaviour of label-retaining cells in oral mucosa and skin. J Dent Res 1981;60:1620—2. ([2) Watt FM. Epidermal stem cells: markers. patterning and the control of stem cell fate. Philos Trans R Soc Lond B Bio Sci 1998:353:831—7. ([3) Jones PH. Epithelial stem cells. Bioessays 1997;19:683490. (l4) lzumi K. Terashi H. Marcelo CL. Feinberg SE. Development and charac— terization of a tissue~engineered human oral mucosa equivalent produced in a serum—free culture system. J Dent Res 2000;79:798—805. (15) Slack JM. Stem cells in epithelial tissues. Science 2000;287:1431—3. (l6) Feliciani C. Gupta AK. Sauder DN. Keratinocytes and cytokine/growth factors. Crit Rev Oral Biol Med 1996:7:300—18. (l7) Marcinkiewicz M. Grabowska SZ. C/yzewska E. Role ofepidermal growth factor (EGF) in oesophageal mucosal integrity. Curr Med Res Opin 1998; 14:145—53. (18) Kamer AR. Krebs L. Hoghooghi SA. Liebow C. Proliferative and apoptotic responses in cancers with special reference to oral cancers. Crit Rev Oral Biol Med 1999;10:58—78. (19) Bjarnason GA. Jordan RC. Sothern RB. Circadian variation in the expres— sion of cell-cycle proteins in human oral epithelium. Am J Pathol 1999: 154:613—22. (20) Thomson PJ. Potten CS. Appleton DR. Mapping dynamic epithelial cell proliferative activity within the oral cavity of man: a new insight into carcinogenesis? BrJ Oral Maxillofacial Surgery 1999;37:377—831. (21) Squier CA. Johnson NW. Hopps RM. Human oral mucosa: development. structure and function. Oxford (UK): Blackwell Scientific: 1976. (22) Lipkin M. Proliferation and differentiation of gastrointestinal cells. Physiol Rev 1973;53:8917915. (23) Dotto GP. Signal transduction pathways controlling the switch between keratinocyte growth and differentiation. Crit Rev Oral Biol Med 1999:10: 442—57. (24) Mehrel T. Hohl D. Rothnagel JA. Longley MA. Bundman D. Cheng C. et al. Identification of a major keratinocyte cell envelope protein. loricrin. Cell 1990;61:1103—12. Candi E. Melino G. Mei G. Tarcsa E. Chung SI. Marekov LN. et al. Biochemical. structural. and transglutaminase substrate properties of hu- man loricrin. the major epidermal cornified cell envelope protein. J Biol Chem 1995;270:26382—90. (26) Kligman AM. The biology of the stratum corneum. In: Montagna W. Lobitz WC. editors. The epidermis. New York (NY): Academic Press: 1964. p. 387—433. (3 \_ (4 \— (25 V (27) Steinert PM. Cantieri JS. Epidermal keratins. In: Goldsmith LA. editor. Biochemistry and physiology of the skin. Vol. 1. New York (NY): Oxford University Press: 1983. (28) Matoltsy AG. Keratinization. J Invest Dermatol 1976;67:20—5. (29) Elias PM. Friend DS. The permeability barrier in mammalian epidermis. J Cell Biol 1975;65:180—91. (30) Madison KC. Swartxendruber DC. Wertz PW. Downing DT. Presence of intact intercellular lipid latnellae in the upper layers of the stratum cor- neum. J Investig Dermatol 1987;88:714—8. (3/) Tomakidi P. Breitkreutz D. Fusenig NE. Zoller J. Kohl A. Komposch G. Establishment of oral mucosa phenotype in vitro in correlation to epithelial anchorage. Cell Tissue Res 1998;292:355—66. (32 Barrett AW. Selvarajah S. Franey S. Wills KA. Berkovity. BK. Interspecies variations in oral epithelial cytokeratin expression. J Anat 19981193: 185—93. (33) Martinez 1R. Peters A. Membrane—coating granules and membrane modi— fications in keratini/ing epithelia. Am J Anat 19711130293. (34) Dale BA. Ling SY. Evidence of a precursor form of stratum corneum basic protein in rat epidermis. Biochemistry 1979;18:3539—46. (35) Dale BA. Holbrook KA. Steinert PM. Assembly of stratum corneum basic protein and keratin filaments in macrofibrils. Nature 1978;276:729—31. (36) Rice RH. Green H. The cornified envelope of terminally differentiated human epidermal keratinocytes consists of cross-linked protein. Cell 1977; 1 12417—22. (37) Simon M. Green H. Participation of membrane-associated proteins in the formation of the cross-linked envelope of the keratinocyte. Cell 1984;36: 327—34. (38) Schroeder HE. Differentiation of human oral stratified epithelium. Basel (Switzerland): Karger: 1981. p. 33—156. (39) Squier CA. Membrane coating granules in nonkeratini/ing oral epithelium. J Ultrastruct Res 1977;60:212—20. (40) Hashimoto K. Dibella RJ. Shklar G. Electron microscopic studies of the normal human buccal mucosa. J Investig Dermatol 1966;47:512. (4/) Schenk P. Wersall J. The ultrastructure of papillae—fillil'ormes of human tongue. Arch Dermatol Forsch 1975;252:91. (42) Hayward AF. Membrane-coating granules. Int Rev Cytol 1979;59:97. (43) Hopwood D. Logan KR. Bouchier IA. The electron microscopy of normal human oesophageal epithelium. Virchows Arch B Cell Path 1978;26: 345—58. (44) Grubb C. Hacketnann M. Hill KR. Small granules and plasma membrane thickening in human cervical squamous epithelium. J Ultrastruct Res 1968; 22:258. (45) Wertz PW. Swartzendruber DC. Squier CA. Regional variation in the struc- ture and permeability of oral mucosa and skin. Adv Drug Del Rev 1993; 12:1—12. (46) Madison KC. Sando GN. Howard E]. True CA. Gilbert D. Swartxendruber DC. et al. Lamellar granule biogenesis: a role for ceramide glucosyltran- ferase. lysosomal enzyme transport. and the Golgi. J Investig Dermatol Symp Proc 1998:3280—6. (47) Kvidera A. Mackenzie IC. Rates of clearance of the epithelial surfaces of mouse oral mucosa and skin. Epithelial Cell Biol 1994:3:175—80. (48) Squier CA. Finkelstein MW. Oral mucosa. In: Oral histology. devel— opment. structure and function. St. Louis (MO): CV Mosby; 1998. p. 345—85. (49) Yamazaki K. Ohmori T. Kumagai Y. Makuuchi H. Eyden B. Ultrastructure of oesophageal melanocytosis. Virchows Arch A Pathol Anat Histopathol 1991;418:515922. (5(1) Terris B. Potct F. Structure and role of Langerhans‘ cells in the human oesophageal epithelium. Digestion 1995;56(Suppl l):9714. (5/) Harmese JL. Carey FA, Baird AR. Craig SR. Christie KN. Hopwood D, et al. Merkel cells in the human oesophagus. J Pathol 1999;189:176—9. (52) Squier CA. Hall BK. Iii-vim) permeability of porcine oral mucosa after epithelial separation. stripping and hydration. Archs Oral Biol 1985:30: 485—91. (53) Squier CA. The permeability of keratinized and non-keratinized oral epi— thelium to horseradish peroxidase. J Ultrastruct Res 1973;43:160v77. (54) Hoogstraate AJ. Bodde HE. Methods for assessing the bttccal mucosa as a route of drug delivery. Adv Drttg Del Rev 1993;12:99—125. (55) Squier CA. Lesch CA. Penetration pathways of different compounds through epidermis and oral epithelia. J Oral Pathol 1988;17:512—6. \, Journal of the National Cancer Institute Monographs No. 29, 2001 (56) Zhang H. Robinson JR. Routes of drug transport across oral mucosa. In: Rathbone MJ. editor. Oral mucosal drug delivery. New York (NY): Marcel Dekker; I996. p. 5l—63. (57) Ho NH. Biophysical kinetic modeling of buccal absorption. Adv Drug Del Rev I993;12:6I—97. (58) Du X. Kremer M. Wertz PW. Squier CA. Oral epithelial flux and reservoir function following brief exposure to ethanol [abstract]. J Dent Res 1997; 76:362. (59) Law S. Wertz PW. Swan/,endruber DC. Squier CA. Regional variation in content. composition and organization of porcine epithelial barrier lipids revealed by thin-layer chromatography and transmission electron micros- copy. Arch Oral Biol 1995;40:1085—9l. (60) Schenkels L. Gururaja TL. Levine MJ. Salivary mucins: their role in oral mucosal barrier function and delivery. In: Rathbone MJ. editor. Oral mu- cosal drug delivery. New York (NY): Marcel Dekker; 1996. p. 191—220. (6/) Levine M]. Jones PC. Loomis RE. Reddy MS. AI-Hashimi I. Bergey EJ. et al. Functions of human saliva and salivary mucins: an overview. In: Mackenzie IC. Squier CA, Dabelsteen E. editors. Oral mucosal diseases: biology. etiology and therapy. Copenhagen (Denmark): Laegeforeningens Forlag: I987. p. 24—27. (62) Tsai H. Bobek LA. Human salivary histatins: promising anti-fungal thera— peutic agents. Crit Rev Oral Biol Med l998;9:48()—97. (63) Weinberg A. Krisanaprakornkit S. Dale BA. Epithelial antimicrobial pep— tides: review and significance for oral applications. Crit Rev Oral Biol Med 1998;92399—414. (64) Williams DM. Cruchley AT. Structural aspects of aging in the oral mucosa. In: Squier CA. Hill MW. editors. The effect of aging in oral mucosa and skin. Boca Raton (FL): CRC Press: 1994. p. 65—74. Journal of the National Cancer Institute Monographs N0. 29. 2001 (65) Hill MW. Epithelial proliferation and turnover in oral epithelia and epi— dermis with age. In: Squier CA. Hill MW. editors. The effect of aging in oral mucosa and skin. Boca Raton (FL): CRC Press; I994. p. 75—83. (66) Squier CA. Wertz PW. Williams DM, Cruchley AT. Permeability of oral mucosa and skin with age. In: Squier CA. Hill MW. editors. The effect of aging in oral mucosa and skin. Boca Raton (FL): CRC Press: I994. p. 91—98. (67) Johnson D. Roth G. Richardson D. Matheny J. Effect of age on the dy— namics of human gingival microcirculation [abstract]. J Dent Res 1992;712 I780. (68) Murrah VA. Diabetes mellitus and associated oral manifestations: a review. J Oral Pathol 1985;14:271—81. (69) Semba I. A histometrical analysis of age changes in the human lingual artery. Arch Oral Biol l989;34:483—9. (70) Goodman CH. Squier CA. Blood flow in the oral mucosa of normal and atherosclerotic rhesus monkeys. J Oral Pathol 1988;17:34—8. (7/) Johnson GK. Effects of aging on the microvasculature and microcirculation in skin and oral mucosa. In: Squier CA. Hill MW. editors. The effect of aging in oral mucosa and skin. Boca Raton (FL): CRC Press; 1994. p. 99—105. NOTE Supported by Public Health Service grant RZIDE13778 from the National Institute of Dental and Craniofacial Research. National Institutes of Health. Department of Health and Human Services: and by North Atlantic Treaty Or- ganization Collaborative Linkage grant 975870. 15 Protection Against Mucosal Injury By Growth Factors and Cytokines Dawn Boot/1, Christopher S. Patten This article provides an overview of published studies in which growth factors and cytokines were used to modify the sensitivity of intestinal stem cells to a dose of radiation. In these experiments, growth factors were used to manipulate the sensitivity of stem cells in the gastrointestinal tract to reduce the severity of gastrointestinal mucositis in cancer therapy patients. Transforming growth factor BS, interleu- kin 11, and keratinocyte growth factor were used. All three agents, given according to appropriate protocols, can result in a threefold to fourfold increase in the number of intestinal stem cells that survive a dose of radiation therapy. This re- sult was assessed by using the crypt microcolony assay of stem cell functional capacity. The changes in stem cell sur- vival that were observed resulted in increased animal sur- vival. This increased survival was taken as a surrogate for improvement in patient well-being. The severity of diarrhea, a marker of functional impairment, was concomitantly re- duced. [J Natl Cancer Inst Monogr 2001;29:16—20] Cancer therapy involves a fine balance between the use of a large enough dose of a drug to kill tumor cells and the prevention of damage to the normal tissues of the body, notably the oral mucosa, the gastrointestinal tract. and the hematopoietic system. Damage caused during cancer therapy to the epithelial layer lining the mouth causes the lining to become depleted and ul— ceration of the mouth to occur (oral mucositis). This makes it difficult for the patient to eat. swallow. or speak and causes pain and susceptibility to infection. A combination of patient discom— fort and possible infection can lead to treatment delay or dose reduction. which can result in a less favorable outcome for the patient. Another major dose—limiting tissue is the gastrointestinal tract, whose rapid cell cycle makes it more susceptible to the effects of cytotoxic exposure and to a rapid expression of dam— age. This damage can be seen in the small intestine within a few days and within a slightly longer period of time in the large bowel. This article will concentrate on the specific problems associated with damage to the normal small intestine. The small intestine is a constantly renewing tissue. continu— ously replacing cells that are lost in the lumen of the intestine. This renewal is achieved by the production of new cells in the crypts by the stem cells and their progeny. arranged in an am— plifying transit lineage of six to eight generations. Under steady- state situations, there are thought to be between four and 16 actual stem cells that are located near the base of the crypt. However. if the system is damaged (6."., by radiation) an acute response occurs. causing some cells in the lower region of the crypt. possibly the stem cells, to die via apoptosis. The stem cells. probably along with some early transit generation cells. will recognize the damage and undergo rapid cell division to regenerate the crypt and. hence, the tissue by clonal growth. The number of clonogenic cells per crypt is thought to depend on the level of damage induced. but it could be as many as 30 cells per 16 crypt (I). Some crypts will become reproductively sterile. will be unable to initiate this regeneration response, and will disap- pear within approximately 48 hours. The time sequence for this response is as follows (Fig. 1). In the first 2—3 days following a dose of radiation (e.g.. a 14-Gy dose). some crypts can be seen to be regenerating (forming microcolonies) alongside other crypts that have been reproductively sterilized. On day 4 following radiation therapy, those crypts that have survived the radiation damage will be approximately 1.5 to two times bigger than a normal crypt. and the sterilized crypts will have disappeared. These large. regen- erating crypts will then start to split or bud into several crypts (2). Approximately 14 days after irradiation. the regenerating crypts and foci will have grown visible to the naked eye as macrocolonies (3,4). In certain cases, budding will continue until the whole of the intestine has been regenerated and the normal architecture restored. There are advantages in protecting the clonogenic stem cells frotn damage. since they are the key to the survival of an in- dividual crypt. The number of crypts that survive following cy- totoxic damage determines how intact the intestinal mucosa is and. hence. how well an animal or a patient can survive the damage. The number of surviving crypts plays a pivotal role in the competitive race between depopulation and ulceration and regeneration. Alternatively. a stimulatory factor given be- fore cytotoxic exposure could increase the number of stem cells per crypt that are subjected to the cytotoxic insult. A possible method of protecting the clonogenic stem cells might be to ma— nipulate them by using growth factors. This could be achieved by using a growth factor that will take the stem cells out of the cycle. making them more resistant to the cytotoxic damage. and ensuring the survival of more stem cells capable of regenerating any damaged tissue. Finally, giving, after cytotoxic insult, a growth factor that could increase the rate of proliferation or initiate regeneration earlier may help in speeding up the regen- eration process. A combination of various protocols could give maximum protection for the epithelium. Various experiments to manipulate clonogenic stem cells to protect against cytotoxic damage have been carried out. but this review will outline work that has used transforming growth fac— tor B3 (TGFvB3). interleukin 1] (IL—11). and keratinocyte growth factor (KGF). using lO—IZ-week-old (C57BL/6 x DBA/ 2)FI (BDFI) mice. All experiments were performed within the regulations of the UK. Scientific Procedures Act (1986). Alfiliulion of‘auI/im‘r: Cancer Research Campaign Epithelial Biology Group. Paterson Institute for Cancer Research. Christie Hospital. Manchester, UK. Cm‘rmpmulwn'v II): C. S. Potten. B.Sc,. M.Sc. I’h.D.. D,Sc.. Epistem Ltd.. Incubator Building. Grafton St., Manchester M13 ‘)XX. U.K. (c—mail: cpotten@ epistemcouk). © Oxford University Press Journal of the National Cancer Institute Monographs No. 29. 2001 Fig. 1. Radiation effects on small intestine. (a) (Original magnifica— tion Xl00 ). The earliest response is an increase in apoptosis 4.5 hours following 8—Gy irradia- tion. Arrows indicate apoptotic bodies. (b) (Original magnifica— tion x50). A selected area of in- testine demonstrating a single large regenerating crypt 4 days after l4-Gy irradiation. Vincris- tine helps identification by arrest- ing cells in metaphase. Arrow- heads : mitosis. Larger arrows indicate two reproductively ster— ilised crypts that are dying and disappearing. (c) (Original mag- nification x40). Day 6 following l4-Gy irradiation. Many areas of intestine are completely devoid of crypts and villi with only a few epithelial cells remaining. ((1) (Original magnification x40.) Section through a normal (control) small intestine showing crypts and villi. (e) (Original magnification x40). By day 8 following l4-Gy irradiation the few regenerating crypts (see Fig. l. e) have split (budded) and are starting to form new crypts and villi. (f) (Original magnification x40.) By day 30 following l4—Gy irradiation the epithelium in the surviving animals is restored to its normal small intestinal architecture, as seen in Fig. l, d. TRANSFORMING GROWTH FACTOR [33 TOP-[33 is a known inhibitor of some epithelial cell prolif- eration (5,6) by preventing cell cycle progression and accumu- lating cells in GI or Go. For clonogenic stem cells in the intes- tine. this might render them more resistant to the cytotoxic damage, leaving more clonogenic stem cells to start the regen— eration process. To assess the capacity of TGF-B3 to protect the intestine from radiation damage, we used the crypt microcolony assay as de— scribed by Withers and Elkind (7). This is an accepted test for crypt stem cell functional capacity. A standard dose of 2.5 mg TOP-[33 was administered intraperitoneally to male BDF| mice 24, 8. and 4 hours before and then once immediately after irra- diation. Animals were culled and samples of small intestine were taken 4 days after irradiation. The number of surviving crypts per circumference (a unit length) was counted for 10 circumfer— ences per mouse. Data were analysed by using the DRFIT pro- gram (8), which allows curves to be fitted and tests the statistical difference between treated and vehicle control groups by using a variance-ratio F test. As shown in Fig. 2, a, four injections of TOP-83 administered once immediately after and again at 4, 8, and at 24 hours before radiation therapy compared with the vehicle control shows a statistically significant shift to the right of the treated curve (9) (P<.001). This indicates that there are four times more clonogenic stem cells surviving at 15 Gy in the TGF—BB-treated group than in the control group. It is interesting that if an additional dose of TOP—[33 was administered 4 hours after irradiation (—4. —8, —24, 0, and +4 hours), then the level of protection afforded is reduced (data not shown). TGF-B3 ad— ministered intraperitoneally after radiation therapy caused crypts to become sensitized, with only approximately one third of the crypts surviving. in comparison with the vehicle control (data not shown). This finding demonstrates the importance of the use of a growth factor in the correct protocol. These studies demonstrated that the small intestinal clono- genic stem cells can be protected from radiation damage. How does this finding relate, however, to the patient’s well—being in Journal of the National Cancer Institute Monographs No. 29, 2001 cancer therapy? To answer this question, we conducted a series of experiments to study animal survival over time after a dose of radiation. Groups of 20 animals were pretreated with TGF-B3 or vehicle, then irradiated with a 15.8-Gy x—ray and observed for 30 days. Animals had their heads, thoraxes. and forelimbs shielded to reduce the complications of oral and hematopoietic damage. Almost 100% of animals pretreated with TGF-B3 sur— vived to 30 days, but only approximately 35% of the vehicle control animals survived (10) (Fig. 3, 3). During the 30 days postirradiation, animals were checked twice daily for diarrhea (verified by evidence of wet feces on the fur of the anal region). The duration of diarrhea can be seen on individual animal “life- lines” as shown in Fig. 4. Administration of TGF-B3 before the radiation dose reduces the total number of days during which diarrhea was recorded, suggesting not only an improvement in life expectancy within this experiment but also an improvement in quality of life and a reduction in diarrhea, a surrogate marker of intestinal dysfunction (10). The histology of the intestine of the animals surviving to 30 days is indistinguishable from that of control subjects (Fig. 1). This indicates truly remarkable regen— erative capacity, especially when the appearance of the intestine at days 3—5 (Fig. l) is considered. INTERLEUKIN 11 A pleiotrophic cytokine that affects many different systems, recombinant human IL-11 was originally isolated and cloned from an immortalized primate bone marrow stromal cell line (1]). Similar experiments to the TGF—BS study were carried out by using IL-1 1. A number of protocols were tested with use of the crypt microcolony assay, and these were carried out as follows: pro- tocol A: lL-l 1 administered subcutaneously at 9 AM and at 9 PM, for a total of five injections before irradiation; protocol B: IL-1 1 administered subcutaneously at 9 AM and at 9 PM, for a total of seven injections postirradiation; and protocol C: IL-ll adminis- tered subcutaneously at 9 AM and at 9 PM, for a total of three injections prior to irradiation and six injections following irra- diation. Fig. 2. Intestinal crypt survival curves comparing treated groups 3 against respective vehicle con» trols. All vehicle groups are rep— resented as open circles and all treated groups as closed symbols. (2) Animals were given vehicle or 2.5 mg transforming growth factor B3 (TGF—B3) intraperito— neally at 24. 8. and 4 hours before irradiation and once immediately after irradiation. Values defining the TGF-B3 survival curve are (mean 1 SE) D” = 1.26 10.10. N : 5664 1 4698 and the values for ' TGFB-3 0.1~ Surviving Fraction 0.01. ' IL-11 ° Saline - KGF o Saline the vehicle curve are D” : 1.12 1 156 o sallve . . 0.83. N = 5294 1 4492. There is 0 5 10 15 a statistically significant differ— ence between the two curves (P<.001). D0 is the mean lethal dose. the reciprocal of the slope on the exponential position of the curve#it is a measure of the ra- diosensitivity; N is the back ex» trapolate to [em dose of the ex- ponential portion of the curve and a measure of the size ofthe shoul— der. (b) Animals were given ve- 10 12 14 hiele or 2.5 mg interleukin 11 (IL-1 1) subcutaneously 1 day be- Protection Factor 0 —* N (.0 «L‘- 01 0') O 5 10 15 10 12 14 16 18 10 12 14 16 18 Dose(Gy) fore, at the time of irradiation. and continuously throughout the postirradiation regeneration phase. All injections were given at 9:00 AM and 9:00 PM. for a total of three injections before irradiation and six injections after irradiation. Values for the lL-ll survival curves are (mean 1 SE) D0 : 2.301 0.94. N = l 105 1 23.7. and the values for the vehicle group are D() = 1.84 1 0.67. N 2 232.5 1 52.9. There is a significant difference between the two curves (P<.001). (c) Animals were given vehicle or 12.5 mg keratinocytc growth factor (KGF) subcutaneously once a day for 3 days before irradiation]. Values for the KGF curve are D0 = 1.911 0.12. and N = 173 1 65. and the values for the vehicle control curve are D“ = 1.43 1 0.06. and N = 886 1 294. There is a statistically significant difference between the two curves (P<.001). In all cases, IL-11 protected the clonogenic stem cells from radiation damage, but to varying degrees. Protocol A gave good protection with a statistical significance of P<.001, and protocol B was only modestly protective (P = .003), while the best protection was afforded by administering IL-1 1 subcutaneously both before and after radiation therapy (protocol C; P<.001) (Fig. 2, b) (12), when up to 3.5 times more crypts survived a dose of l6-Gy irradiation in the IL-1 1 treated groups. Studies examining lL—l 1 and animal survival over time after radiation therapy did not take into consideration the damage caused to the hematopoietic system. Protocols A and C were carried out with a dose of 12-Gy x-ray irradiation. Animals were observed over the following 30 days. As a result of damage to the hematopoietic and gastrointestinal systems, all of the animals treated with a dose of 12—Gy irradiation died during the first 1 1 days; however, the lL—l l—treated groups survived an average of 2 days longer than the controls (Fig. 3, b) (13). It is unclear by what mechanisms IL-ll protects against cy— totoxic damage, but lL—l 1 given before radiation therapy alters the sensitivity of clonogenic stem cells, making them more re- sistant; hence, more of them survive. It is uncertain whether IL-ll alters cell cycle progression or has an indirect effect by altering other growth factors. IL—1 1 protects best when it is given both before and after irradiation, and it may act via different mechanisms before and after irradiation. IL—1 1 is known to syn- ergize with other factors [e.g., stem cell factor and steel factor (14)] and also may enhance the effects of IL-3 ([5). 18 KERATINOCYTE GROWTH FACTOR KGF was identified originally as a factor that stimulated ep- ithelial cells in vitro (I6). Further studies demonstrated that it also stimulates both proliferation and differentiation in a number of epithelial cell types when given to healthy animals. This is very noticeable in the gastrointestinal tract (17). It has been observed that administration of KGF has trophic effects on the gastrointestinal tract, increasing crypt depth and villus length (18). KGF also has a particularly pronounced trophic effect on oral mucosa (19). Crypt survival studies with the use of KGF as a protective agent against radiation damage have shown that KGF given for 3 days before irradiation showed a statistically significant pro- tective effect over a range of doses (P<.001) (Fig. 2, c), with three times more crypts surviving a dose of l6-Gy irradiation in the KGF treated group and 3.5 times more crypts surviving after a dose of l4—Gy irradiation than controls (vehicle treated) ( [8). Use of KGF postirradiation was not found to improve crypt survival. Animal survival studies have also been performed with KGF and 12 Gy of radiation by using postirradiation bone marrow transplantation to protect the hematopoietic system (Fig. 3, c). This experiment showed that 90% of mice pretreated with KGF before receiving 12 Gy of irradiation and a bone marrow trans- plant survived a 16-day observation period, whereas mice that received only 12 Gy of irradiation and a bone marrow transplant all died during the first 8 days. Journal of the National Cancer Institute Monographs No. 29, 2001 100 "“- '- so 60 40 o—o—o—o—o—o—o-o—o—o-o—o—o—o—o—o—o—o 20 —'- TGFa-a -°- 15.BGy PB ob 5 i0 i5 éo 25 30 ._ B 100 ”R 5580 ' $60 > E40 3 (020 --—|L-11 ‘IO‘12Gy/WB O0 5 10 15 20 25 30 100 80 60 40 20 o._n_l_n_l_n_l_._l_._l 0-. '- X-I-I-I-I-I-I-I-I C --— KGF+1ZGy+BMT —°— 12Gy WB+BMT 5 1 0 1 5 20 25 30 Days Fig. 3. Survival time of animals exposed to 12~Gy (experiment with interleukin- l l [IL—l l]) or 15.8—Gy ( experiment with transforming growth factor B3[TGF- B3!) xarays delivered whole body (IL—1 1) or abdomen only (TGF~BS). For the keratinocyte growth factor (KGF) experiment. l2-Gy Cs'J7 was delivered whole body, followed by a bone marrow transplant. All vehicle groups are represented as open symbols. and all treated groups are represented as closed symbols. A) Animals were given vehicle or 2.5 mg TGF—B3 24. 8. and 4 hours before irradiation and once immediately after irradiation. B) Animals were given ve- hicle or 2.5 mg lL-ll subcutaneously 1 day before irradiation. at the time of irradiation. and continuously throughout the postirradiation regeneration phase. All injections were given at 9:00 AM and 9:00 PM. C) Animals were given vehicle or 5 mg/kg KGF per day subcutaneously on days 2. l. and 0: that is. before and at the time of (day 0) radiation. PB = partial body. WB 2 whole body. BMT = bone marrow transplant. It seems somewhat puzzling that the best protocol for pro- tection is to give KGF before cytotoxic insult when it is known that KGF is a stimulator. The mechanism by which KGF pro- tects the intestinal system is unknown and could be multifacto- rial. However. KGF has been shown to have trophic effects on the gastrointestinal tract. increasing crypt size. This may suggest that KGF protects the crypts by proportionally increasing the number of clonogenic stem cells per crypt (18). Indeed, cell pro— liferation studies have clearly indicated increased stem cell pro— liferation following KGF treatment. as well as (rather suipris— ingly) a reduction in transit cell proliferation (20). Stimulatory factors. such as KGF and. possibly, IL—l 1. may also act as cell survival factors that prevent apoptosis. Journal of the National Cancer Institute Monographs No. 29, 200] TGFB-3 Mouse no. I Diarrhoea Vehicle Mouse no. 0 10 20 30 Days Fig. 4. Each line represents the life of an animal through a survival experiment, with the black boxes indicating when diarrhea was observed and the termina- tion of a line indicating when an animal died. This is taken from an experiment where a dose of 15.3 Gy partial body x—ray was delivered to the abdomen and the mice were pretreated with transforming growth factor [53 (top panel) or vehicle (lower panel). KGF has also been shown to increase the number of goblet cells in the gastrointestinal tract (17). Goblet cells produce mu- cins that act as a barrier between the epithelium and the luminal and trefoil proteins that act to defend the gut and can also aid in its repair. One difficulty in making comparisons between the effects of these different agents is that different injection regimens have been used. as have, in some cases, different radiation doses and delivery protocols. However. it is very clear that the sensitivity of the potential clonogenic sterr cells in the small intestine of the mouse can be experimentally manipulated by exogenous growth factors or cytokines in an advantageous manner. Such manipu- lations have been shown to afford overall radioprotection to an animal. and this protection manifests itself in potentially dra- matic changes in animal survival and well-being. CONCLUSIONS These studies show that statistically significant and, in some cases, dramatic reductions in mucositis can be effected by ap— propriate manipulation of stem cell sensitivity with the use of growth factors and cytokines. The results presented here are 19 preliminary; extensive additional studies are required to deter— mine the most effective doses and delivery protocols. Many more growth factors and cytokines should be tested. together with combined and sequential use of different factors. With the identification of intestinal—specific regulatory factors, it would seem likely that the sensitivity of the critical stem cells might be even more effectively manipulated. This would reduce the se- verity of gastrointestinal mucositis even further. improving the quality of life of cancer therapy patients and possibly allowing for dose escalation and improvement in cure rates. REFERENCES (1) Hendry JH. Roberts SA. Potten CS. The clonogen content of murine in— testinal crypts: dependence on radiation dose used in its determination. Radiat Res 1992;132:1154). Cairnie AB. Millen BH. Fission of crypts in the small intestine of the irradiated mouse. Cell Tissue Kinet 1975282189796. (2 \7 (3 V mouse intestinal mucosa. Radiology 1968;91:99871000. Withers HR. Elkind M. Radiosensitivity and fractionation response of crypt cells of mouse jejunum. Radiat Res 19691381598413. Kurokawa M. Lynch K. Podolsky DK. Effects of growth factors on an intestinal epithelial cell line: transforming growth factor b inhibits prolif— eration and stimulates differentiation. Biochem Biophys Res Comm 1987: 1422775—82. Barnard JA. Beauchamp RD. Coffey RJ. Moses HL. Regulation of intes- tinal epithelial cell growth by transforming growth factor type beta. Proc Natl Acad Sci U S A 1989;86:1578—82. (7) Withers HR. Elkind MM. Microcolony survival assay for cells of mouse intestinal mucosa exposed to radiation. Int J Radiat Biol Relat Stud Phys Chem Med 197():7:261—2. (8) Roberts SA. DRFlT: a program for fitting radiation survival models. Int J Radiat Biol Relat Stud Phys Chem Med 1990;51:1243—6. (9) Potten CS, Booth D. Haley JD. Pretreatment with transforming growth factor beta-3 protects small intestinal stem cells against radiation damage in viva. Br J Cancer 1997;75:1454—9. (4 V (5 \. (6 \— 20 Withers HR. Elkind M. Dose—survival characteristics of epithelial cells of (IO) Booth D. Haley JD. Bruskin AM. Potten CS. Transforming growth factor— b3 protects murine small intestinal crypt stem cells and animal survival after irradiation, possibly by reducing stem cell cycling. Int J Cancer 2000; 86:53-9 (II) Paul SR. Yang YC. Donahue RE. Goldring S. Williams DA. Stromal cell-associated haematopoiesis: immortalization and characterization of a primate bone marrow—derived stromal cell line. Blood 1991;77: 1723—33. Potten CS. Interleukin-ll protects the clonogenic stem cells in murine small-intestinal crypts from impairment of their reproductive capacity by radiation. Int J Cancer 1995;62:356’61. (/3) Potten CS. Protection of the small intestinal clonogenic stem cells from radiation—induced damage by pretreatment with interleukin 11 also in- creases murine survival time. Stem Cells 1996;14:452—9. (/4) Hiryama F. Shih J. Awgulewitsch A. Warr CW. Clark SC. Ogawa M. Clonal proliferation of murine lymphohemopoietic progenitors in culture. Proc Natl Acad Sci U S A 1992;89:5907—11. (/5) Paul SR. Schendel P. The cloning and biological characterization of recombinant human interleukin 11. Int J Cell Cloning 1992;10:13543. (l6) Rubin JS. Osada H. Finch PW. Taylor WG. Rudikoff S. Aaronson SA. Purification and characterization of a newly identified growth factor spe- cific for epithelial cells. Proc Natl Acad Sci U S A 1989;86:802—6. (/7) Housley RM. Morris CF. Boyle W. Ring B. Biltl. R. Tarpley JE. et al. Keratinocyte growth factor induces proliferation of hepatocytes and epi— thelial cells throughout the rat gastrointestinal tract. J Clin Invest 1994294: 1764—77. (/8) Farrell CL. Bready JV. Rex K. Chen JN. DiPalma CR. Whitcomb KL. et a1 Keratinocyte growth factor protects mice from chemotherapy and radia— tion-induced gastrointestinal injury and mortality. Cancer Res 1998258: 933—39. (/9) Farrell CL. Rex KL. Kaufman SA. Dipalma CR. Chen JN, Scully S. et al Effects of keratinocyte growth factor in the squamous epithelium of the upper aerodigestive tract of normal and irradiated mice. Int J Radiat Biol 1999;75:609—20. (20) Potten CS. O‘Shea JA. Farrell CL. Rex K. Booth C. The effects of repeated doses of keratinocyte growth factor on cell proliferation in the cellular hierarchy of the crypts of the murine small intestine. Cell Growth Differ 2001;12:265—75. —\ \. '9 Journal of the National Cancer Institute Monographs No. 29. 2001 Transgenic Mouse Model of Intestine-Specific Mucosal Injury and Repair Leo Lefrang'ois, Vaiva Vezys Most studies of injury and repair to mucosal tissue have used nonspecific mediators to induce injury. Damage to the mu- cosal epithelium resulting from chemical or radiation treat- ment associated with cancer therapy may fall into this cat- egory of injury. When such treatments are applied, it is generally not possible to predict or control the extent of possible injury. This fact makes analysis of inductive and reparative processes difficult. In addition, the role of the immune system in the etiology and subsequent healing of mucosal tissue following cancer therapy with or without bone marrow transplantation remains unclear. To study tis- sue- and antigen-specific immune damage of intestinal mu- cosal tissue, we generated transgenic mice that express a nominal antigen exclusively in intestinal epithelial cells. The transfer of antigen-specific CD8 T cells with concomitant virus infection resulted in the destruction of intestinal epi- thelial cells and disease. The destructive phase in some cases was followed by complete recovery and tolerance induction. This model will provide a system that can be regulated for analysis of the mediators of mucosa-specific tissue damage and repair. [J Natl Cancer Inst Monogr 2001;29:21—5] The intestinal mucosal immune system is composed of in— ductive and effector sites. Peyer’s patches (PPs) and mesenteric lymph nodes (MLNs) serve as sites for antigen presentation and primary activation of naive T and B cells. After activation in these sites, some lymphocytes then migrate to tertiary tissue sites. including the epithelium, the lamina propria (LP), and the effector sites of the intestinal mucosa (1,2). Numerous activated CD4 and CD8 T cells, as well as B and plasma cells, can be found within the LP (3—7). Plasma cells in the LP mainly pro- duce immunoglobulin A. which is subsequently actively trans~ cytosed across the epithelium by a receptor—mediated process (8). The majority of this antibody is presumed to be specific for bacterial antigens. Similarly. in normal situations, the activated and memory T cells residing in the LP and the epithelium may have been initially primed by antigens derived from normal flora. Indeed. in germ—free mice, the intestinal immune system is poorly developed and contains severely reduced numbers of lymphocytes with reduced effector function (6,9). Thus, coloni— zation of the gut with normal flora is, in part, responsible for the formation of the mucosal immune system tissue. This process is, in effect, a symbiotic relationship between bacteria and the host in which the latter gains considerable benefit through the for- mation of a normal mucosal immune system poised to respond to pathogenic insult. The intestinal epithelium is home to a substantial population of T lymphocytes termed “intraepithelial lymphocytes” (lELs). lELs are a complex population of cells that contain several lym— phocyte subsets (10). Most small intestinal lELs express CD8, while large intestinal lELs contain larger populations of CD4 and CD4-8—T—cell receptor (TCR) (XB subsets (11.12). TCR us- Journal of the National Cancer Institute Monographs No. 29, 2001 age is also distinct for lELs as compared with lymphocytes in secondary lymphoid tissue, such as that found in the spleen and lymph nodes. Small intestinal lELs in most species examined contain an appreciable population of cells that express TCRyB [250% in some mouse strains (13—15)]. This enigmatic popu- lation of lymphocytes has been proposed to be involved in early responses to bacterial infection, perhaps through recognition of antigens presented by nonclassic major histocompatibility com- plex (MHC) molecules. The precise function of TCRyB lELs remains elusive, although these cells have been shown to pro- duce keratinocyte growth factor (KGF) when activated, suggest— ing that they may play a role in homeostasis or repair of the intestinal epithelium (l6). Mucosal epithelial cells are far from being simply a quiescent physical barrier. Rather, intestinal epithelial cells (lECs) are ac- tive participants in innate as well as adaptive immune responses (17.18). lECs produce, as well as respond to, a variety of cyto— kines, including factors involved in lymphocyte development, such as stem-cell factor and interleukin 7 (19,20). lECs also produce chemokines that are likely essential for mucosal lym- phoid organ formation and for mounting effective immune re— sponses (21—23). In inflammatory bowel disease, lECs can be- come targets of a dysregulated immune system (24). lt remains unclear whether lECs are victims of direct antigen—specific im— mune destruction or whether inflammatory cytokines induce damage in a bystander-like fashion. To determine the conse- quences of antigen-specific T-cell interactions with lECs, we established the model presented here. MATERIALS AND METHODS Mice. C57BL/6J (Ly5.l) ‘nice were purchased from The Jackson Laboratory. Bar Harbor, ME. The OT—I mouse line (25) was from W. R. Heath (Walter and Eliza Hall Institute, Parkville, Australia) and F. Carbone (Monash Medical School, Prahan, Victoria, Australia) and was maintained as a C57BL/6-Ly5.2 line on a RAG—/— background. Intestinal fatty acid—binding pro- tein promoter-truncated ovalbumin (IFABP—tOVA) transgenic mice were created by use of a construct containing the long form of the lFABP promoter (nucleotides —1 I78 to +28; a gift of J. 1. Gordon, Washington University School of Medicine, St. Louis, MO) (26.27), tOVA cDNA (encoding amino acids 138—386). which does not include the signal sequence so the protein re- mains cytosolic (28), and human growth hormone (hGH) Affiliation ()fuur/mm: Division of Rheumatic Diseases, Department of Medi- cine. University of Connecticut Healt‘l Center, Farmington. Currz'.\'/mrulclm' to: Leo Lefrancois. Ph.D., Division of Rheumatic Diseases. Department of Medicine. University of Connecticut Health Center. MC1310. 263 Farmington Ave.. Farminglon. CT 06030 (e-mail: llet'ranc@neuron. uchc.cdu). See “Notes” following "References." © Oxford University Press 2] (nucleotides 498—2652; from R. M. Perlmutter, University of Washington, Seattle) (29). An Sail fragment containing these elements was microinjected into C57BL/6—Ly5.1 fertilized eggs by the Transgenic Animal Facility at the University of Connecti- cut Health Center. To detect the transgene. genomic DNA was analyzed by polymerase chain reaction to identify a 608-base— pair band by using an lFABP—specific 5’ primer (5’-GCCATC- ACACTTGACCCTAA—3') and an OVA-specific 3’ primer (5’—TCAGGCAACAGCACCAACAT-3'). Mice were kept in specific pathogen-free housing and were analyzed between 8 and 10 weeks of age. RNA analysis. Total RNA from the indicated tissues was isolated by cell lysis with guanidine isothiocyanate followed by 16 hours of centrifugation at 22 0C over a cesium chloride cushv ion (30). Purification of poly (A) RNA was accomplished by using a Poly(A) Quik messenger RNA (mRNA) Isolation Kit from Stratagene (La Jolla. CA). according to the manufacturer’s instructions. One microgram of poly (A) RNA was dot blotted onto a nylon membrane. which was then hybridized with a 32P-labeled OVA cDNA fragment. The blot was stripped and reprobed with a glyceraldehye—3-phosphate dehydrogenase- specific cDNA probe to allow for mRNA quantification. A Mo- lecular Dynamics (Sunnyvale. CA) Phosphorlmager was used to quantitate hybridization. Isolation of lymphocyte populations and adoptive transfer of OVA-specific CD8 T cells. IEL and LP cells were isolated as described previously (12,13). Lymph nodes (LN) and spleens were removed and single-cell suspensions were prepared. Pe— ripheral LN included brachial. axillary. and superficial inguinal lymph nodes. The resulting preparation was filtered through Ni— tex. and the filtrate was centrifuged at 200g for 5 minutes at 4 °C to pellet the cells. For adoptive transfer. 5 X 105 OT—l/RAG—/—/ Ly5.2 pooled LN cells were injected intravenously into Ly5.1 B6 or transgenic hosts. Where indicated. mice were infected 24 hours later by intravenous (IV) injection of l x 10“ plaque- forming units (pfu) vesicular stomatitis virus (VSV)—encoding OVA (3]). Detection of antigen-specific CD8 T cells with MHC te- tramers. Mice were infected by injection of 1 x 10“ pfu of VSV-OVA. Six days later. lymphocytes were isolated and VSV nucleoprotein (N)-specific or OVA—specific CD8 T cells were detected by using H-2Kb tetramers containing the N protein— derived peptide RGYVYQGL (32) or the OVA—derived peptide SIINFEKL (33). Peptides were purchased from Research Genet— ics. Huntsville. AL. Tetramers were produced essentially as de- scribed previously (34.35). In brief. H—ZKh containing the BirA— dependent biotinylation substrate sequence (the construct was provided by J. Altman, Emory University. Atlanta. GA) was folded in the presence of human [32—microglobulin and the N or OVA peptide. Biotinylation was performed with biotin—protein ligase (Avidity. Denver, CO). Tetramers were then produced from biotinylated high-pressure liquid chromatography—purified monomers by the addition of streptavidin—allophycocyanin (APC) (Molecular Probes. Eugene, OR). Flow cytometric analysis. Lymphocytes were resuspended in 0.2% phosphate—buffered saline (PBS). 0.1% bovine serum albumin (BSA). and NaN3 at a concentration of l x 106 to l X 107 cells/mL followed by incubation at 4 0C for 30 minutes with 100 (LL of properly diluted monoclonal antibody (MAb). The MAbs either were directly labeled with fluorescein isothiocya- nate (FlTC), phycoerythrin (PE), Cy5. APC or were biotinyl- 22 ated. For the latter. avidin—PE—Cy7 (Caltag Laboratories. Burlin- game. CA) was used as a secondary reagent for detection. For tetramer staining. cells were first reacted with PE-labeled anti- CD8 (Caltag Laboratories) and FlTC-labeled anti—CD1 Ia at 4 °C followed by staining for 1 hour at room temperature with APC- coupled MHC tetramers. After staining, the cells were washed twice with PBS/BSA/NaN3 and fixed in 3% parafortnaldehyde in PBS. Relative fluorescence intensities were then measured with a FACSCalibur (Becton—Dickinson. San Jose. CA). Data were analyzed by using WinMDl software (J. Trotter. Scripps Clinic, La Jolla, CA). Histologic analysis. Duodenum. jejunum. and ileum from experimental animals were fixed in 10% formalin (Fisher Sci- entific, Pittsburgh. PA). Paraffin—embedded tissue was sectioned and then stained with hematoxylin—eosin. All images were mag- nified X 200. RESULTS With the goal of testing the impact of T-cell reactivity with an [EC-specific antigen. we generated transgenic mice expressing a chicken tOVA gene under control of an IEC—specific promoter (Fig. l). The lFABP directs protein expression to mature entero— cytes but not to crypt epithelial cells (26,27). In addition. IFABP is expressed primarily in the small intestine and is expressed weakly or not at all in the stomach or the large bowel. The hGH gene that contains introns and exons was used to provide signals for the poly (A) addition and to increase in viva transgene ex- pression (29). When expressed. the OVA lacks a signal se- quence. so that the protein remains cytosolic (28). Because of this, IECs should effectively process and present OVA~derived peptides in the context of MHC class I. The mouse strain used for transgenic production was C57BL/6J, an H-2h haplotype strain. The OVA contains an eight amino acid peptide, SIINFEKL. which has been shown to bind to H-2Kh (33). Thus. MHC class l—restricted CD8 T cells of the appropriate specificity should recognize IECs expressing the tOVA. We examined the OVA mRNA levels in two lines of IFABP— tOVA transgenic mice. There was a striking difference in mRNA levels between the two lines. with the 232—4 line ex- pressing approximately lO-fold more mRNA than the 232—6 line. The expression patterns were similar in the two lines, with the highest concentrations of OVA mRNA present in the ileum. Fig. 2 depicts the relative mRNA levels detected in sections of the intestine of lFABP—tOVA mice. We were unable to detect OVA by immunohistochemistry or western blotting. suggesting that low levels of protein were expressed or the truncated protein was rapidly degraded in the cytoplasm of IECs. We first tested whether CD8 T cells of the transgenic mice were tolerant to Sal I Sal 14260 lFABP tOVA hGH Fig. 1. Construct used in generation of intestinal fatty acid—binding protein promoter-truncated ovalbumin (lFABP-tOVA) transgenic mice. The SAL I frag- ment used for production of C57BL/6J transgenic mice contained the long form of the lFABP promoter. tOVA complementary DNA (cDNA)—encoding amino acids 1387336, and the human growth hormone complementary DNA. Journal of the National Cancer Institute Monographs No. 29. 2001 Expression of the iFABP-tOVA Transgene stomach duodenum proximal jejunum distal jejunum F‘\ i \ \ j .. colon/.1/ (:« ”W‘\{:/. Jv-r _ r.# “D “If, 1- Q ,, 1 high no expression expression Fig. 2. Depiction of the relative concentration of ovalbumin (OVA) messenger RNA in 232 mice along the length of the intestine. The figure is based on data from dot blot analysis of poly(A) RNA from duodenum. proximal jejunum. distal jejunum. and ileum of intestinal fatty acid binding protein promoteHruncated OVA mice by using OVA complementary DNA as a probe. OVA. This test was performed by infecting the mice with a recombinant virus containing the OVA gene (VSV—OVA). In normal mice, a robust OVA-specific CD8 T-cell response can be visualized by use of MHC class I tetramers that contain the SIINFEKL peptide. In contrast, few if any OVA—specific CD8 T cells could be found after VSV—OVA infection of lFABP-tOVA mice (36). This result indicated that [EC—expressed antigen had gained access to the systemic immune system. The mechanism by which this occurred is unknown but may be mediated by dendritic cells carrying IEC-derived proteins to the draining lymph nodes and beyond (37). At that point, cross tolerance would be induced to this “self—antigen” via interaction of T cells with dendritic cells that had not been activated by inflammatory signals (see Fig. 4). Further experiments will be necessary to delineate this important pathway for maintenance of self- tolerance to gut—specific protein. Since the CD8 T—cell compartment of IFABP-tOVA trans- genic mice was apparently tolerant to OVA, we were unable to test the reactivity of endogenous T cells with IECs. To circum- vent this problem, we used an adoptive transfer system in which naive, OVA—specific TCR transgenic CD8 T cells (called OT—I) were transferred to unmanipulated IFABP-tOVA mice (Fig. 3). The transferred cells are trackable by virtue of differences be— tween host and donor Ly5 alleles. Thus, a monoclonal antibody specific for Ly5.2 (expressed by the donor CD8 T cells but not by host cells) can be used in flow cytometry or immunohisto— chemistry to detect the OVA—specific CD8 T cells after transfer to the transgenic mice (31). Small numbers (5 X 105) of OT—l cells were injected by the IV route into host transgenic mice; their presence was detected 4 and 5 days later in secondary lymphoid tissues and in intestinal mucosal effector sites (LP and epithelial lymphocytes, respectively). Four days after transfer to the low antigen-expressing line 232—6, few OT—l cells were pre- sent in peripheral lymph nodes (PLNs), LP, or IELs (Table 1). In contrast, a large expansion of OT-l cells had occurred at this time point in PP and MLN. By day 5 after transfer, the MLN and PP OT-l populations had declined, whereas significant numbers of OT-l cells were now present in LP and IELs. Four days after Journal of the National Cancer institute Monographs No. 29, 2001 MODEL SYSTEM FOR ANALYSIS OF T CELL REACTIVITY WITH INTESTINE-SPECIFIC ANTIGEN TrackCIble (Ly5.2) CD8+ ova-SPECIFIC T CELLS I tvsiii‘ be {-— lmmunlze with virus non-secreted antl en (I’OVA) Transgenic mouse expressln ut not crypt epithe lum in mature enterocyies Monitor Health Histolo ical Analysis Flow Cytometry o isolated mucosal T cells Assess Recovery and Factors involved In Tissue Repair Fig. 3. Model system for analysis of immune-mediated intestinal epithelial cell injury. Table I. Mucosa—specific expansion of antigen-specific CD8 T cells adoptively transferred to lFABP-tOVA transgenic mice* Tissue Mouse line PLN MLN PP LP IELs 232—6 7 ’ 7 Day 4 1.1102 11.3 1 3.5 44.3 1 6.5 1.8 10.6 1.1103 Day 5 0.7 1 0.3 3.8 1 0.8 3011.0 11.4 1 2.7 8.8 11.2 232—1 Day 4 5.2107 28713.5 49.3191 11.6119 11.7113 Day 5 2.6 1 0.4 15.2 1 1.9 57.9 1 4.8 50.7 1 9.6 60.9 1 7.4 *PLN = peripheral lymph nodes; MLN = mesenteric lymph nodes: PP = Peyer‘s patch: LP : lamina propria: IELs = intraepithelial lymphocytes. 5 x 105 Ly5.2 OT—l—RAG—l— T cells were transferred to LyS.l lFABP—tOVA trans— genic mice (232—6 or 232—4). Four or five days later lymphocytes were isolated from the indicated tissues and the presence of CD8+ OT—l cells was determined by fluorescence flow cytometry. Values indicate means and standard errors of at least six determinations per tissue. transfer of OT-l cells to the high antigen-expressing line 232—4, many donor cells were found in MLN and PP. Unlike in the 232—6 mice, OT-I cells were also found in the LP and IELs at this time point. By day 5 after transfer [0 232—4 mice, a massive expansion of OT—l cells had occurred in the LP and [EL com- partments. Overall, these results suggested that [EC-derived OVA or OVA fragments are being transferred to PP and MLN, where antigen presentation to T cells can occur. The extent of activation is dependent on the antigen levels, since less activa— tion occurred in the 232—6 (low antigen) compared with the high antigen 232—4 mice. These findings also indicated that activated CD8 T cells did not enter the intestinal mucosa until after acti- vation in PP or MLN. since there were substantial numbers of activated OT—I cells in PP and MLN at least 1 day before their appearance in the mucosa. These data are in agreement with our previous study (2) showing that (x487 integrins are important for migration of activated OT-I cells into the intestinal mucosa. [J b) Despite the fact that large numbers of antigen-specific CD8 T cells entered the mucosa and apparently responded to antigen expressed by IEC, no overt damage to intestinal tissue was observed by histologic analysis. OT-I cells isolated from the epithelium of these animals exhibited potent lytic activity in a standard chromium release assay in vitro, indicating that they were functional (data not shown). Notwithstanding this fact. the apposition Of cytOlytic T cells with potential target cells did not result in induction of cellular damage. Therefore, func— tional “tolerance“ had been induced in viva but not in virm, at least with regard to lytic activity. To determine whether the inclusion Of inflammatory signals would alter the outcome Of the response. mice were infected with VSV-OVA at the time Of OT—l transfer. Four days after OT-I transfer and con— comitant VSV—OVA infection, donor cells had expanded to a much greater extent in the MLN of IFABP-tOVA mice, as compared with their expansion in normal mice. This result showed that endogenous OVA served tO potentiate the CD8 T- cell response to the VSV-OVA infection. The virus infec— tion, along with OT—I transfer to the transgenic mice. also resulted in a transient weight loss in 232—6 mice and a sus- tained weight loss leading to death in most cases in 232—4 mice (36). In the absence of VSV—OVA infection, OT—I transfer did not cause weight loss (Fig. 4). and virus infection alone also had no effect on the weight Of the mice or on the mucosal tissue. or?” 1 Falled Tolerance . (tntectlons, lmmunosuppression) Fig. 4. Proposed pathway of acquisition of intestine—specific antigen and induc» tion of immune tolerance or autoreactivity. Antigen is acquired by secondary lymphoid antigen—presenting cells from ova-expressing epithelium of 232 trans- genic mice either via absorption from the gut lumen as material from intestinal epithelial cell turnover or from dying epithelial cells and migrate to the Peyer‘s patches and mesenteric lymph nodes. In the secondary lymphoid tissue. in the absence of inflammation. tolerance induction will occur primarily through de— letion for CD8 T cells. When inflammatory signals are present. as in a virus. infection. or dysfunction of regulatory cells. then the response is converted to a productive one that is potentially pathogenic. The photograph on the right shows a hematoxylin—eosin stain Of a section from the ileum of a 232—4 mouse 4 days after OT-l cell transfer and concomitant virus infection. 24 Histologic examination Of intestinal tissue after OT—I transfer and VSV-OVA infection revealed significant damage to the in— testinal epithelium (Fig. 4). In 232—6 mice, destruction Of duo- denal and jejunal tissue was substantial with loss Of epithelial cells. shortening Of villi, and elongation Of crypts (36). Crypt epithelial cells were not damaged. in keeping with the known expression pattern of lFABP. There was much less damage to ileal tissue and no effect on large intestinal tissue. In 232—4 mice. the disease was much more severe compared with that seen in 232—6 mice (Fig. 4) There was extensive destruction Of epithelium in duodenal. jejunal, and ileal sections, although the latter sections were once again the least affected. The majority of these mice died of the disease within 6 days after infection. In those 232—4 mice that survived and in all 232—6 mice, intestinal tissue was histologically normal approximately 2 weeks after the destructive episode. These results clearly demonstrated that CD8 T cells are capable of antigen-specific recognition of IEC and of inducing IEC death. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH The model described here will be invaluable for analysis of the stages Of epithelial cell damage and repair inherent to the intestinal mucosa] immune system. The mediators Of cell death produced by CD8 T cells that act on IEC can be evaluated with the use of gene knockout mice and blocking antibodies. More- over, the mechanisms by which tolerance versus autoimmunity are induced are readily testable in this well—defined and tissue- specific system. Future studies will focus on the interaction of CD4 T cells with IEC—expressed antigens and will provide fur— ther clues to the workings of mucosal tolerance and immunity. This model also presents a unique system in which factors in- fluencing mucosal repair can be studied and that could have a direct impact on the repair of the epithelial damage inherent to cancer therapy. Since IEC damage can be induced selectively in enterocytes and regulated by antigen levels and, perhaps, by other factors such as T-cell number and virus dose. the system can be easily manipulated to examine the factors, immune or Otherwise, of repair of mucosal tissue. By using other mucosa— specific promoters with distinct expression patterns, a further level Of control can be attained. In sum, the study of immune- mediated mucosal injury and repair can be investigated in sig- nificant detail by using in vivo model systems that allow control of antigen expression and the immune response directed toward that antigen. REFERENCES (I) Butcher EC. Picker LJ. Lymphocyte homing and homeostasis. Science 1996;272:60—6. (2) Lefrancois L. Parker CM. Olson S. Schon MP. Muller W. Wagner N, et al. The role of B 7 integrins in CD8 T cell trafficking during an anti—viral immune response. J Exp Med 1999;189:1631—8. (3) Zeitz M. Greene WC. Peffer NJ. James SP. Lymphocytes isolated from the intestinal lamina propria of normal nonhuman primates have increased expression of genes associated with T-cell activation. Gastroenterology [988:942647—55. Lundqvist C. Melgar S. Yeung MM. Hammarstrom S. Hammarstrom ML. lntraepithelial lymphocytes in human gut have lytic potential and cytokine profile that suggest T helper l and cytotoxic functions. J Immunol 1996: [57:1926—34. (4 \. Journal of the National Cancer Institute Monographs No. 29, 2001 (5) Fuss lJ. Neurath M. Boirivant M. Klein JS. de la Motte C. Strong SA. et al. Disparate CD4+ lamina propria (LP) lymphokine secretion profiles in in- flammatory bowel disease. Crohn's disease LP cells manifest increased secretion of IFN gamma. whereas ulcerative colitis LP cells manifest increased secretion of lL-S. J Immunol 1996;157:1261770. Lefrancois L. Goodman T. In viva modulation of cytolytic activity and Thy-1 expression in TCR-y8+ intraepithelial lymphocytes. Science 1989: 243:1716—8. Huleatt JW. Lefrancois L. Antigen—driven induction ofCDl lc on intestinal intraepithelial lymphocytes and CD8+ T cells in vim. J Immunol 1995:154: 5684—93. Mcintyre TM. Strober W. Gut—associated lymphoid tissue — regulation of IgA B—cell development. In: Ogra PL. Mestecky J. Lamm ME. Strober W, Bienenstock J. Mcghee .IR. editors. Mucosal immunology. San Diego (CA): Academic Press; 1999. p. 319—56. lmaoka A. Matsumoto S. Setoyatna H. Okada Y. Umesaki Y. Proliferative recruitment of intestinal intraepithelial lymphocytes after microbial colo— nization of germ—free mice. EurJ lmmunol 1996:262945—8. (IO) Lefrancois L. Phenotypic complexity of intraepithelial lymphocytes of the small intestine. J lmmunol 1991;147:1746—51. (ll) Camerini V. Panwala C. Kronenberg M. Regional specialization of the mucosal immune system—intraepithelial lymphocytes of the large intestine have a different phenotype and function than those of the small intestine. J lmmunol 1993;151:1765—76. (l2) Laky K. Lefrancois L. Puddington L. Age—dependent intestinal lympho» proliferative disorder due to stem cell factor receptor deficiency: param- eters in small and large intestine. J Immunol 1997;158:1417+27. (l3) Goodman T. Lefrancois L. Expression of the 'y8 T—ccll receptor on intes- tinal CD8+ intra—epithelial lymphocytes. Nature 1988;333:855—8. (14) Takagaki Y. DeCloux A. Bonneville M. Tonegawa S. Diversity of y8 T»cell receptors on murine intestinal intraepithelial lymphocytes. Nature 1989;339:712—4. (15) Goodman T. Lefrancois L. Intraepithelial lymphocytes. Anatomical site. not T cell receptor form. dictates phenotype and function. J Exp Med 1989: 170:1569—81. (16) Boismenu R. Havran WL. Modulation of epithelial cell growth by intra- epithelial y?) T cells. Science 1994;266:1253—5. (l7) Kagnoff MF. Eckmann L. Yang SK. Huang G. Jung HC. Reed SL. et a1. Intestinal epithelial cells: an integral component of the mucosal immune system. In: Kagnoff MF. Kiyono H. editors. Essentials of mucosal immu— nology. San Diego (CA): Academic Press; 1996. p. 63+71. (18) Dignass AU. Podolsky DK. Interleukin 2 modulates intestinal epithelial cell function in virro. Exp Cell Res 1996;225:422—9. (l9) Puddington L. Olson S. Lefrancois L. Interactions between stem cell factor and c-Kit are required for intestinal immune system homeostasis. Immunity 1994:1:733—9. (20) Laky K. Lefrancois L. von Freeden-Jeffry U. Murray R. Puddington L. The role of lL-7 in thymic and extrathymic development of TCR y?) cells. J Immunol 1998;161:707—3. (2]) Kraehenbuhl JP. Pringault E. Neutra MR. Review article: intestinal epi- thelia and barrier functions. Aliment Pharmacol Ther 1997:] 123—8. (22) Yang SK. Eckmann L. Panja A. Kagnoff MF. Differential and regulated expression of C—X—C. C-C. and C-chemokines by human colon epithelial cells. Gastroenterology 199721 13:1214—23. (23) Dwinell MB. Eckmann L. Leopard JD. Varki NM. Kagnoff MF. Chemo- (6 (7 \. (8 x, (9 x, Journal of the National Cancer Institute Monographs No. 29. 2001 kine receptor expression by human intestinal epithelial cells. Gastroenter— ology 1999:] 17359417. (24) Strober W. Ehrhardt RO. Chronic intestinal inflammation: an unexpected outcome in cytokine or T cell receptor mutant mice. Cell 1993;75:203—5. (25) Hogquist KA. Jameson SC. Heath WR. Howard JL. Bevan MJ. Carbone FR. T cell receptor antagonistic peptides induce positive selection. Cell 1994:7621747. (26) Sweetser DA. Hauft SM. Hoppe PC. Birkenmeier EH. Gordon J1. Trans- genic mice containing intestinal fatty acid-binding proteimhuman growth hormone fusion gencs exhibit correct regional and cell-specific expression of the reporter gene in their small intestine. Proc Natl Acad Sci U S A 1988: 85:96] 1—5. (27) Green RP. Cohn SM. Sacchettini JC. Jackson KE. Gordon J1. The mouse intestinal fatty acid binding protein gene: nucleotide sequence. pattern of developmental and regional expression. and proposed structure of its pro- tein product. DNA Cell Biol 199211 1:31—41. (28) Shastri N. Gonzalez F. Endogenous generation and presentation of the ovalbumin peptide/Kb complex to T cells. J Immunol 1993;150:2724—36. (29) Palmiter RD. Sandgren EP. Avarbock MR. Allen DD. Brinster RL. Heter- ologous introns can enhance expression of transgenes in mice. Proc Natl Acad Sci U S A 1991;88:478—242. (30) Chirgwin JM. Pryybyla AE. MacDonald RJ. Rutter WJ. Isolation of bio- logically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 1979;18:5294—9. (31) Kim SK. Reed DS. Olson S. Schnell MJ, Rose JK. Morton PA. et al. Generation of mucosal cytotoxic T cells against soluble protein by tissue- specific environmental and costitnulatory signals. Proc Natl Acad Sci U S A 1998;95:10814—9. (32) Van Bleek GM. Nathenson SG. Isolation of an endogenously processed immunodominant viral peptide from the class I H-2Kh molecule. Nature 1990;348:213—6. (33) Carbone FR. Bevan MJ. Induction of ovalbumin-specific cytotoxic T cells by in vivo peptide immunization. J Exp Med 1989;169:60342. (34) Altman JD. Moss PA. Goulder PJ. Barouch DH. McHeyzer»Williams MG. Bell J1. et a1. Phenotypic analysis of antigen-specific T lymphocytes. Science 1996;274:94—6. (35) Murali-Krishna K. Altman JD. Suresh M. Sourdive DJ. Zajac AJ. Miller JD. et al. Counting antigen—specific CD8 T cells: a reevaluation of by» stander activation during viral infection. Immunity 1998;82177—87. (36) Vezys V. Olson S. Lefrancois L. Expression of intestine~specific antigen reveals novel pathways ofCD8 T cell tolerance induction. Immunity 2000; 122505—14. (37) Huang FP. Platt N. Wykes M. Major JR. Powell TJ. Jenkins CD. et al. A discrete subpopulation of dcndritic cells transports apoptotic intestinal ep- ithelial cells to T cell areas of mesenteric lymph nodes. J Exp Med 2000; 191:435—43. NOTES Supported by Public Health Service (PHS) grants DK57932 (National Institute of Diabetes and Digestive and Kidney Diseases) and Al41576 (National Institute of Allergy and Infectious Diseases [N1AlD]) from the National Institutes of Health. Department of Health and Human Services. V. Vezys was supported by PHS training grant T32-A107080 frotn the NIAID. We thank Dr. Jeffrey Gordon. Washington University. St. Louis. M0. for generously providing the intestinal fatty acid binding protein promoter. Inflammatory Cytokines and Mucosal Injury David A. Williams The cause of mucosal injury in inflammatory bowel disease (IBD) is not clear but likely involves infectious agents or other toxins followed by an abnormal immune response in genetically susceptible individuals. The inflammatory cyto- kines appear to play a key role in both the susceptibility of some individuals and the tissue damage that accompanies IBD. The generation of transgenic and gene-targeted (knock- out) animals has provided invaluable information regarding the cytokines and cellular immune effectors that are impor- tant in IBD. Information from these and other preclinical animal models, such as those involving interleukin 1], has led to human trials testing novel therapies for IBD and other diseases in which inflammation of the gut mucosa is an im- portant component. Thus, expression of inflammatory cyto- kines appears to be an important target for the development of novel therapies for IBD and other diseases in which in- testinal mucosal damage occurs, such as mucositis and graft- versus-host disease. [J Natl Cancer Inst Monogr 2001;29: 26-30] The etiology of inflammatory bowel disease (IBD), including Crohn's disease and ulcerative colitis. remains unknown and is likely multifactorial. General theories of contributing factors in— clude persistent or inciting infectious agents. a defective muco— sal barrier, and abnormal host immune responses to infection or environmental antigens (1). Genetic epidemiologic studies and more recent genome mapping studies (2,3) indicate that inher- ited factors may contribute to individual susceptibility to lBD. While little direct evidence exists supporting the role of abnor- mal host immune response in IBD in humans, a large body of data in experimental animal models and phenotypes of animals deficient in specific genes generated by homologous recombi- nation methods suggests that the immune system plays a key role in either initiating or maintaining the disease state (4,5). These observations have led to clinical studies targeting modi- fication of specific immune regulators (Table 1). In the case of interleukin (lL)-l l, a pleiotropic cytokine of mesenchymal ori- gin, observations in animal studies of mucosal damage by che- motherapy agents led to subsequent human trials in IBD (6). ABNORMAL HOST IMMUNE RESPONSES IN MUCOSAL DISEASE A variety of studies in animals have led to the hypothesis that an abnormal host immune response is an essential feature of mucosal disease [reviewed in (7)]. These studies suggest that various initiating events in a genetically susceptible individual lead to an imbalance between proinflammatory and anti— inflammatory cytokines (Table 2). T-helper (CD4) cells under the stimulation of specific cytokines differentiate into two sub— sets of cells called TH1 or TH: [reviewed in (8)]. In many studies, T cells of the THI subset appear to be an important mediator of mucosal inflammation. TH, cells are induced by lL—12 and in— terferon gamma (IFN y), whereas TH: cells are induced by IL— 26 IO and lL—4. T”, cells mediate various cellular immune re- sponses, including macrophage activation, leading to the production of proinflammatory cytokines. including lFN y, IL— 2, lL-12, tumor necrosis factor (TNF), nitrous oxide (NO), and lL—l. TH2 cells mediate hypersensitivity responses, reduce mac- rophage activation, and stimulate antibody responses. The re— sulting effects of stimulation by T lymphocytes on a variety of other inflammatory cells (mast cells, neutrophils, and natural killer cells) lead to the production of a large number of soluble inflammatory mediators that are increased in IBD, including arachidonic acid metabolites, toxic phagocytic products (oxygen metabolites, nitric oxide, collagenases, etc.), toxic lymphocyte products. neuropeptides, and various components of the plasma proteolytic cascades (1,4,5, 7. 9). The generation of various gene knockout and transgenic mouse strains has contributed substantial new understanding to the role of T cells in the development Of IBD [Table 3 and reviewed in (9)]. In particular, IBD develops in mice with al- terations in T-cell subpopulations and T-cell selection. including T-cell receptor-deficient ([0,1]), major histocotnpatibility com- plex class ll-deficient (II), and severe combined immunodefi- cient mice restored with CD45RBHl helper T cells (12,13); hu- man leukocyte antigen (HLA)-B27 rats (14); mice with targeted disruption ofcytokine genes, including lL-IO (l5), lL-2 (16,17), and transforming growth factor-B (18,19); and mice lacking sig— naling proteins important in T cells, including 01; subunit of G protein (20) or SMAD3 (2]). Although these animals have a variety of specific defects in immune function, as noted by Pow- rie (4), all share a common feature in that, in each, the T—cell— dependent regulatory system that normally protects the gut is disrupted. One implication Of these studies is that T lymphocytes play a critical role in the development and maintenance of oral tolerance (5). While direct evidence of the role of these genes in the development of [BD in humans is lacking, there are immu— noregulatory features in Crohn’s disease and ulcerative colitis, including a decreased ratio of lL—lra to IL-1 in mucosal biopsy specimens from ulcerative colitis and Crohn‘s patients (22). IL— lra is a circulating lL-l receptor, which, if present at high levels. negatively regulates the effects of lL-l. In addition, some ex— perimental evidence (23) suggests that monocytes from IBD patients may produce less lL-4, a key anti—inflammatory cyto— kine. IBD IN IL-lO-DEEICIENT MICE Mice deficient in lL—l0 have been especially useful in under- standing the role of inflammatory cytokines and other factors in Afli/ialiunx (giant/ml: Howard Hughes Medical Institute, Indiana University School of Medicine. Section of Pediatric Hematology/Oncology, Herman B Wells Center for Pediatric Research. and Department of Pediatrics, Indiana University School of Medicine, Indianapolis. Crn‘rm‘pmulcm'c In: David A. Williams, M.D.. Herman B Wells Center for Pediatric Research. 1044 West Walnut St.. Rm. 402. Indianapolis. IN 46202 (e-mail: dwilliam(atiupuicdu). See "Notes“ following “References.“ © Oxford University Press Journal of the National Cancer Institute Monographs No. 29, 200] Table I. Clinical trials for inflammatory bowel disease involving cytokines O Anti-interleukin I2 0 Interleukin 10 0 Interleukin 11 0 Antitumor necrosis factor-or Table 2. Inflammatory mediators postulated to be involved in mucosal pathology* O Pro-inflammatory IL-I, lL—6, TNF—(x. IFN—‘y, IL—2 0 Anti—inflammatory II.-Ira. TGF—B. IL-4. IL-IO, IL»II *IL : interleukin: TNF = tumor necrosis factor: IFN = interferon: lL—lra = circulating IL-I receptor: TGF : transforming growth factor. Table 3. Transgenic animals with inflammatory bowel disease* O Alterations in T-cell subpopulations or selection TC R (or/B) MHC class II SCID mouse restored with CD45RB'” helper T cells HLA-BZ7 rat 0 Cytokine knockout mice IL-Z lL-IO TGF—B O Signaling proteins G protein subunit (Gui2) SMAD3 *TCR = T-eeII receptor: MHC = major histocompatibility complex: SCID : severe combined immunodeficient: HLA = human leukocyte antigen; IL = interleukin: TGF : transforming growth factor. the development of mucosal inflammation. lL-IO was initially identified as an activity produced by TH: cells that inhibited the production of cytokines by THI cells (24). However. IL-lO has effects on a broad range of immune functions and is a potent suppressor of macrophage activation, inhibiting production of inflammatory cytokines. such as IL- I. IL-6. and TNF-0t. Kuhn et al. (15) described the development of mice deficient in IL-IO generated by homologous recombination targeting the IL—IO gene in embryonic stem cells. Surprisingly, these mice devel— oped normal numbers of B and T lymphocytes and demonstrated a normal immune response to T-ceIl—dependent immunizations. However. these mice developed an abnormal THI response to nematode infection with increased production of INF—y and IL— 5. The animals were also growth retarded and developed a mi— crocytic anemia that was likely caused by iron deficiency. The principal histopathologic finding in anemic and low-weight mice was a chronic enterocolitis involving the entire intestinal tract. The pathologic lesions in the intestine included mucosal inflam— mation, degeneration ofthe intestinal mucosa, and marked thick— ening of the mucosal wall associated with excessive regenerative hyperplasia. Mucosal surfaces demonstrated desquamation of the apical epithelia with superficial erosions and inflammatory exudates. Extensive histiocytic, lymphoid, macrophage, neutro- phil, eosinophil, and plasma cell infiltration was seen in the lamina propria and submucosa regions. The duodenum showed Journal of the National Cancer Institute Monographs No. 29. 200i the most extensive abnormalities, but changes were seen throughout the gastrointestinal tract. These pathologic changes were attenuated in mice bred in specific-pathogen-free (SPF) conditions, suggesting a role for microbial antigens in the se- verity or progression of the IBD. In a subsequent study. Berg et al. (25) demonstrated disease progression that occurred in mice kept in SPF conditions. These studies documented the fact that inflammatory changes first oc- curred in the cecum and ascending and transverse colon and ultimately involved the rectum and small intestine. Prolonged disease was associated with transmural lesions and increased incidence of adenocarcinomas. Mechanistically, Berg et a1. (25) demonstrated increased amounts of inflammatory cytokines (IL- la, TNF—oz, IL-6, IFN 'y, and NO) produced in colonic cultures of IL-lO-deficient mice. Purified CD4+ T cells derived from the colons of these mice also produced substantially more IFN y. Treatment of mice with anti-IFN y antibodies prevented colitis from developing, and administration of IL-IO substantially re— duced colitis, duodenitis, and the incidence of colorectal adeno— carcinoma. These studies also demonstrated genetic differences in disease susceptibility. since a marked difference in the devel— opment of intestinal lesions was seen in multiple inbred mouse strains. Kullberg et a1. (26) more recently demonstrated that IL—lO—deficient mice reared in SPF conditions that were experi- mentally infected with Helicobacter Impatiens developed chronic colitis associated with a THI cytokine response (TNF—or, IFN 'y. and NO). In this experimental infection with one specific microbe, neutralization of IFN y or IL—I2 in viva with antibodies resulted in a substantial reduction in intestinal inflammation. In summary, the data generated from lL-lO-deficient mice suggest a key role for proinflammatory cytokines. inheritable factors, and microbial antigens in the development and progression of IBD. IL-11 IN THE TREATMENT OF INFLAMMATORY DAMAGE OF THE INTESTINE As noted above, IL-1 1 is a pleiotropic cytokine of mesenchy— mal origin. The cloning of the complementary DNA (cDNA) responsible for lL-l 1 activity followed the development of more than 60 immortalized stromal cell lines from primate bone mar- row by using a recombinant retrovirus vector expressing simian virus 40 large T antigen (27). Of these cell lines. one designated PU-34 demonstrated the capacity to generate megakaryocytic cells from human CD34+ cells when used in long-term cultures. Conditioned medium from this cell line was shown to support proliferation of an IL-6-dependent plastnacytoma cell line, Tl I65, in the presence of excess neutralizing antibodies to lL-6. Because of data suggesting a role for lL-o in megakaryocyte colony growth in some systems, this activity was further studied. With the use of expression cloning methods and the lL-6- dependent plasmacytoma cell line, an lL-I I cDNA was cloned (along with multiple IL-6 cDNA clones) and was shown to have megakaryocyte colony-forming activity in vitro (28). Subse- quently, IL-I I was shown to stimulate the recovery of platelets in viva after cytoablative therapy (29) and in normal mice (30). Effects on platelet production were seen in a variety of preclini- cal models and in early—phase trials in humans [for a review. see (31)]. In 1998, the US Food and Drug Administration approved IL—1 1 (Neumega) as the first pharmacologic agent for the treat— ment of chemotherapy-induced thrombocytopenia. Human IL—I I is a l99—amino—acid protein with a molecular weight of approximately I9 kd [reviewed in (32)]. The gene 27 maps to 19ql3.3—ql3.4 in a region that contains several zinc finger genes and spans 7 kilobases (33). The protein, unlike many cytokines, is not glycosylated and has no cysteine residues or potential N-linked glycosylation sites. The cytokine probably has a structure with a four-helix bundle topology with two re— ceptor-binding sites located in the carboxyl terminus (34). Many mesenchymal cell lines have been shown to express IL-1 1, whereas IL—1 1 messenger RNA is abundant in the murine testis, hippocatnpal neuronal cells, and motor neurons and sympathetic neurons of the spinal cord (35). lL-l 1 is a member of the IL—6 cytokine superfamily. The receptor for IL—1 1 contains a common GP130—signa1ing subunit and a specific or chain (36). Binding of 1L-11 stimulates receptor dimerization, activation, and phos— phorylation of Jak/Stat proteins. In an effort to develop tnore severe thrombocytopenia in a mouse model in which to test the stimulatory activity of IL—1 1, Du et al. (6) treated mice with the combination of total—body irradiation and 5-fluorouracil (S—FU). Mice treated with this combination therapy and IL—1 1 were noted to survive at signifi- cantly (P = .01) higher frequency than control mice. Neither an increase in leukocyte counts nor significant changes in bleeding explained the survival differences. Because deaths occurred rap— idly after cytoablative therapy, examination of the gut was un- dertaken; this analysis demonstrated marked changes in the mu- cosal architecture. Control mice demonstrated significant (P = .01) shortening of villus length and areas of ulcerations accom— panied by enteric bacterial foci in the liver. IL-ll treatment was associated with increased villus length, preserved villus/crypt ratios, and reduced incidence of hepatic bacterial foci. Subse- quent studies by many laboratories have demonstrated similar results in multiple models ofgut injury. including ischemia (37), burn (38). short gut (39), trinitrobenzene sulfonic acid (40), HLA-B27 rat (4]), 5-FU-induced mucositis (42), radiation therapy alone (43), and graft-versus-host disease (GVHD) (44). One study suggests a direct effect on mucosa] cells in vim (45). Studies on GVHD provide evidence of the mechanisms by which IL—1 1 may be acting in these varied models. GVHD is a multisystetn disease in which donor T cells directed against host antigen(s) mediate inflammation and tissue damage. Common end organs affected in GVHD include skin, gastrointestinal tract, liver, and blood cells (46). Clinical symptoms include exfolia- tive skin rashes, hepatitis, diarrhea, weight loss, immune— mediated blood cell destruction, and increased incidence of op- portunistic infections. Although the specific host antigens that are mediators of GVHD have not been identified, modulation of the incidence and severity of the disease occurs with T-cell depletion of the donor stem cell preparation (46). Hill et a1. (44) demonstrated that administration of IL-1 1 in an animal model of GVHD significantly (P = .05) reduces the incidence and sever- ity of intestinal complications and leads to increased survival of treated mice. These changes in the intestinal mucosa were as— sociated with a substantial reduction in IFN 'y and IL—2 secretion and an increase in IL-4 secretion. The TH2 polarization of the T—cell response also led to decreased lL-l2 production in mixed lymphocyte cultures in virru (44). It is interesting that systemic levels of TNF—(x, a potent inflammatory cytokine, were signifi- cantly (P = .01) reduced by IL-11 treatment. The authors hy— pothesized that IL-11 treatment reduced GVHD morbidity and mortality by polarization of donor T cells (to TH2 response), protection of the small intestine, and suppression of inflamma— tory cytokines. 28 Additional studies have also provided evidence that IL—11 has potent anti—inflammatory effects and may be acting in the gut by modulating macrophage cytokine production. Trepicchio et a1. (47) have demonstrated that administration of IL-1 1 to lipopoly- saccharide—treated mice reduced TNF-a, IL-1, and IFN y levels in viva. Treatment of isolated macrophages in vim) with IL-1 1 led to reduced production of these same cytokines and IL-12 p40 and NO. Subsequent studies demonstrated a dose—dependent de— crease in IL-12 production by macrophages after combined IFN y/Staplzy/ococcus uureus stimulation in viva. This decrease in expression of IL-12 was caused by transcriptional regulation (48). This transcriptional effect could be due to increased ex— pression of IkB-B and IKB-OL with subsequent decreased nuclear translocation of NF—kB in macrophages (49). Thus, studies sup- port the hypothesis that IL—ll may be beneficial in IBD by a combination of direct effects on enterocyte production/survival and modulation of immune responses, including systemic reduc- tion in inflammatory cytokines. Early studies in humans have supported the preclinical data presented above. A multicenter, double-masked, placebo— controlled study in 76 patients with IBD demonstrated the ex- pected increase in platelet counts and showed that 42% of IL— 11-treated patients had a positive clinical response in terms of IBD symptoms, versus 7% in the placebo group (50). This dose- escalation study showed minimal toxic effects with IL-1 1 given subcutaneously two times per week. The response was seen at a dose of 16 ug/kg per week. A multicenter, double-masked, pla— cebo—controlled phase II study has been recently reported in abstract form (5]). In this study, 148 patients with active disease (Crohn’s disease activity index [CDAI] >220) were given pla— cebo versus lL-l l in two schedules (15 (Lg/kg once a week or 7.5 (Lg/kg twice a week). Results in the once—a-week group showed a trend toward decreased mean percentage CDAI (32% versus 18% in placebo) and a substantial increase in remission rate (37% versus 16% in placebo, P<.05). Treatment using the twice-a-week schedule was effective but was associated with increased rates of side effects, including headache, edema, and increased platelet counts. These side effects were not noted at any increased frequency in the once-a—week treatment group compared with the placebo group. Thus, IL-ll appears to be both safe and effective in inducing remissions in a subset of patients with active Crohn’s disease. On the basis of these early clinical studies, a multi-institutional phase III trial is now under way (Schwertschlag U: personal communication). The expression of inflammatory cytokines appears to play a critical role in the development and progression of IBD. On the basis of a number of preclinical animal models, in which the expression of these cytokines is modulated, and early human clinical trials, expression of inflammatory cytokines appears to be an important target for the development of novel therapies for IBD and other diseases in which intestinal mucosal damage occurs, such as mucositis and GVHD. RELATIONSHIP or INFLAMMATORY CYTOKINES AND MUCOSAL INJURY IN CANCER The generation of mucositis and bowel injury accompanying chemotherapy and radiation therapy used in cancer treatment continues to be a major source of morbidity for many patients. In addition, these side effects of cancer therapy can frequently have adverse consequences on dose intensity and, therefore, compro- mise tnultimodality approaches to the cure of cancer. Mucosal Journal of the National Cancer Institute Monographs No. 29. 2001 injury is also a striking component of GVHD seen in the postal- logeneic stem cell transplant setting, a process that substantially increases morbidity and mortality in those using this therapeutic approach for the treatment of cancer. Although direct cytotox— icity of many chemotherapy agents and radiation to intestinal mucosa] cells undoubtedly plays a major role in the development of mucositis and other intestinal complications of these thera- pies, inflammatory cytokines, as discussed in this article, prob— ably also contribute substantially to the severity and mainte— nance of injury. FUTURE RESEARCH DIRECTIONS The use of transgenic and gene—targeted mice will continue to elucidate important mechanisms of mucosal injury. These stud- ies. therefore, provide logical and relevant targets for future pharmacologic intervention. Future studies are needed to con- tinue to define the basic mechanisms of mucosal cell growth and differentiation but also should aim to translate these findings into patient-focused research in the treatment of IBD, mucositis, and other gastrointestinal complications of cancer therapies REFERENCES (I) Sartor RB. Pathogenesis and immune mechanisms ofchronic inflammatory bowel diseases. Am J Gastroenterol 1997;92:554 IS. (2) Hugot JP. Laurent—Puig P, Gower-Rousseau C. Olson JM. Lee JC. Beaugerie L, et a]. Mapping of a susceptibility locus for Crohn‘s disease on chromosome 16. Nature 1996;379:821—3. Satsangi J. Parkes M. Louis E. Hashimoto L. Kato N. Welsh K. et a]. Two stage genome-wide search in inflammatory bowel disease provides evi— dence for susceptibility loci on chromosomes 3. 7 and 12. Nat Genet 1996114: 199—202. Powrie F. T cells in inflammatory bowel disease: protective and pathogenic roles. Immunity 1995;32171—4. Strober W. Ehrhardt RO. Chronic intestinal inflammation: an unexpected outcome in cytokine or T cell receptor mutant mice. Cell 1993;75:203—5. (6) Du XX. Doerschuk CM. Orazi A. Williams DA. A bone marrow stromal— derived growth factor. interleukin—1 l, stimulates recovery of small intesti— nal mucosa] cells after cytoablative therapy. Blood 1994;83:33—7. Strober W. Kelsal] B. Fuss I. Marth T. Ludviksson B, Ehrhardt R, et a]. Reciprocal IFN-gamma and TGF-beta responses regulate the occurrence of tnucosa] inflammation. Iinmunol Today 1997;18:614. Mosmann TR. Coffman RL. TH] and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol 1989;7z 145—73. Elson CO. Sartor RB. Tennyson GS. Riddel] RH. Experimental models of inflammatory bowel disease. Gastroenterology 1995;109:1344—67. (I0) Mixoguchi A, Mi/oguchi E, Chiba C. Spiekermann GM. Tonegawa S, Nagler-Anderson C, et a]. Cytokine imbalance and autoantibody production in T cell receptor—alpha mutant mice with inflammatory bowel disease. J Exp Med 1996;183:847-56. (II) Mombaerts P, Mi/oguchi E, Grusby MJ. Glimcher LH. Bhan AK. To- negawa S. Spontaneous development of inflammatory bowel disease in T cell receptor mutant mice. Cell 1993;75:274—82. (12) Morrissey PJ. Charrier K, Braddy S. Liggitt D. Watson JD. CD4+ T cells that express high levels of CD45RB induce wasting disease when trans~ ferred into congenic severe combined immunodeficient mice. Disease de- velopment is prevented by cotransfer of purified CD4+ T cells. J Exp Med 19932178223744. (/3) Powrie F. Leach MW. Mauze S. Caddle LB. Coffman RL. Phenotypically distinct subsets of CD4+ T cells induce or protect from chronic intestinal inflammation in C. B—17 scid mice. Int Immunol 199352146177]. (/4) Hammer RE. Maika SD. Richardson JA. Tang JP. Taurog JD. Spontaneous inflammatory disease in transgenic rats expressing HLA-BZ7 and human beta 2m: an animal model of HLA-B27-associated human disorders. Cell 1990;63:1099—112. (3 \_ (4 (5 (7 (x \_ (9 Journal of the National Cancer Institute Monographs No. 29. 200] (I5) Kuhn R. Lohler J. Rennick D. Rajewsky K, Muller W. Interleukin-10- deficient mice develop chronic cnterocolitis. Cell 1993;75:263—74. (/6) Sadlack B. Merl. H. Schorle H. Schimpl A, Feller AC. Horak I. Ulcerative colitis—like disease in mice with a disrupted interleukin-2 gene. Cell 1993: 75:253—61. (/7) Ehrhardt R0. Ludviksson BR. Gray B. Neurath M. Strober W. Induction and prevention of colonic inflammation in IL-2»deficicnt mice. J Immunol 1997;158:5664}. (/8) Shull MM. ()rmsby I. Kier AB. Pawlowski S, Diebold RJ, Yin M. et a]. Targeted disruption of the mouse transforming growth factor—beta I gene results in multifocal inflammatory disease. Nature 1992;359:693—9. (l9) Kulkarni AB. Huh CG. Becker D. Geiser A. Lyght M. Flanders KC. el al. Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death. Proc Natl Acad Sci U S A 1993; 90:770—4. (20) Rudolph U. Finegold MJ, Rich SS. Harriman GR. Srinivasan Y, Brabet P. et a]. Ulcerative colitis and adenocarcinoma of the colon in G alpha i2- deficient mice. Nat Genet 1995;10:143450. (2/) Yang X. Letterio JJ. Lechleider RJ. Chen L. Hayman R. Gu H. et a]. Targeted disruption of SMAD3 results in impaired mucosa] immunity and diminished T cell responsiveness to TGF—beta. EMBO J 1999;18:12807 91. Casini»Raggi V. Kam L. Chong YJ. Fiocchi C. Pizarro TI". Cominelli F. Mucosa] imbalance of ]L—1 and IL—l receptor antagonist in inflammatory bowel disease. A novel mechanism of chronic intestinal inflammation. J Immunol 19951154343440. Schreiber S, Heinig T, Panzer U, Reinking R, Bouchard A. Stahl PD. et a]. Impaired response of activated mononuclear phagocytes to interleukin 4 in inflammatory bowel disease. Gastroenterology 1995;108:21733. (24) Fiorentino DF. Bond MW. Mosmann TR. Two types of mouse T helper cell. IV. Th2 clones secrete a factor that inhibits cytokine production by Th1 clones. J Exp Med 1989;170:2081—95. Berg DJ. Davidson N. Kuhn R. Muller W, Menon S. Holland G, et a]. Enterocolitis and colon cancer in interleukin-10—deficient mice are associ« ated with aberrant cytokine production and CD4(+) THl—like responses. J Clin Invest 1996;98:1010—20. (26) Kullberg MC. Ward JM, Gorelick PL. Caspar P. Hieny S. Cheever A. et a]. Helicnbut'lw' lie/mucus triggers colitis in specific-pathogen—free interleu- kin»10 (IL-1())-deficient mice through an IL—12- and gamma interferon- dependent mechanism. Infect Immun 1998;66:5157—66. (27) Paul SR, Yang YC. Donahue RE. Goldring S, Williams DA. Stromal cell-associated hematopoiesis: immortalization and characterization of a primate bone marrow—derived stromal cell line. Blood 1991;77:1723—33. (28) Paul SR. Bennett F. Calvetti JA. Kelleher K. Wood CR. O'Hara RM Jr. et a]. Molecular cloning of a cDNA encoding interleukin 1]. a novel stromal cell-derived lymphopoietic and hematopoietic cytokine. Proc Natl Acad Sci U S A 1990;87:7512—16. (29) Du XX. Neben T. Goldman S. Williams DA. Effects of recombinant human interleukin-l ] on hematopoietic reconstitution in transplant mice: accelera- tion of recovery of peripheral blood neutrophils and platelets. Blood 1993: 81:27—34. (30) Neben TY. Loebelenz J. Hayes L. McCarthy K. Stoudemire J. Schaub R. et a1. Recombinant human interleukin—I 1 stimulates megakaryocytopoesis and increases peripheral platelets in normal and splencctomized mice. Blood 1993;81:901—8. (3/) Du X. Williams DA. Interleukin-l 1: review of molecular. cell biology and clinical use. Blood 1997;89:38977908. (32) Du XX, Williams DA. Interleukin-l 1. a multifunctional growth factor de- rived from bone marrow stromal cells. In: Remick DG. Friedland JS. edi- tors. Cytokines in health and disease. 2"d ed. New York (NY): Dekker: 1997. p. 153—65. (33) McKinley D. Wu Q. Yang-Feng T. Yang YC. Genomic sequence and chromosomal location of human interleukin—11 gene (IL 11). Genomics 1992;13:814719. (34) Czupryn MJ, McCoy JM. Scoble HA. Structure—function relationships in human interleukin—ll. Identification of regions involved in activity by chemical modification and site-directed mutagenesis. J Biol Chem 1995: 2702978445. (35) Du X. Everett ET, Wang G. Lee WH. Yang 2. Williams DA. Murine interleukin-1 1 (IL-1 1) is expressed at high levels in the hippocampus and (22 \— (23 \_ (25 \_ 29 (36) (37) (38) (39) (40) (4/) (42) (43) (44) 30 expression is developmentally regulated in the testis. J Cell Physiol 1996: 1682362—72. Hilton DJ. Hilton AA. Raicevic A. Rakar S. Harrison—Smith M. Gough NM. et al. Cloning ofa murine IL—1 1 receptor alpha—chain; requirement for gpl30 for high affinity binding and signal transduction. EMBOJ 1994:13: 4765—75. Du X. Liu Q. Yang Z. Oraxi A. Rescorla FJ. Grosfeld JL. et al. Protective effects of interleukin-11 (IL—11) in a murine model of isehemic bowel necrosis. Am J Physiol 1997;272:0545—52. Schindel D. Maze R, Liu Q. Williams DA. Grosfeld J. Interleukin-ll improves survival and reduces bacterial translocation and bone tnarrow suppression in burned mice. J Pediatr Surg 199713231275. Liu Q. Du XX. Schindel DT. Yang ZX. Rescorla FJ. Williams DA. et al. Trophic effects of interleukin-1] in rats with experimental short bowel syndrome. J Pediatr Surg 1996;31:1047—51. Qiu BS. Pfeiffer CJ. Keith JC. Protection by recotnbinant human interleu- kin—ll against experimental TNB-induced colitis in rats. Digest Diseases Sci 1996;41:1625—30. Keith JC Jr. Albert L. Sonis ST. Pfeiffer CJ. Schaub RG. IL-l I. a pleio- tropic cytokine: exciting new effects of IL—1 l on gastrointestional tnucosal biology. Stem Cells 1994;12:79790. Sonis ST. Van Vugt AG. McDonald J. Dotoli E. Schwertschlag U. Szklut P. et al. Mitigating effects of interleukin II on consecutive courses of 5—fluorouraci|»induced ulcerative mucositis in hamsters. Cytokine 199719: 605—12. Potten CS. Interleukin—11 protects the clonogenic stem cells in marine small—intestinal crypts from impairment of their reproductive capacity by radiation. Int J Cancer 1995;62:356—61. Hill GR. Cooke KR, Teshima T. Crawford JM. Keith JC Jr. Brinson YS. et al. Interleukin-ll promotes T cell polarization and prevents acute graft— versus-host disease after allogeneic bone marrow transplantation. J Clin Invest 1998;102:1154}. (45) Booth C. Potten CS. Effects of lL-l l on the growth of intestinal epithelial cells in l’ffl‘l). Cell Prolif 1995;28:581—94. (46) Ferrara J. Lipton J. Hellman S. Burakoff S. Mauch P. Engraftment follow- ing T—ccll-depleted marrow transplantation. I. The role of major and minor histocompatibility barriers. Transplantation 1987;43:461—7. (47) Trepicchio WL. Bozza M. Pedneault G. Dorner AJ. Recombinant human lL-ll attenuates the inflammatory response through down-regulation of proinflammatory cytokine release and nitric oxide production. J Immunol 1996;157:3627—34. (48) Leng SX. Elias JA. Interleukin-11 inhibits macrophage interleukin—12 pro— duction.J Immunol 1997;159:2161—8. (49) Trepicchio WL. Wang L. Bozza M. Dorner AJ. IL—ll regulates macro- phage effector function through the inhibition of nuclear factor-kappaB. J Immunol 1997;159:5661—70. (50) Sands BE. Bank S. Sninsky CA. Robinson M. Kati. S. Singleton JW. et al. (5/ Preliminary evaluation of safety and activity of recombinant human inter— leukin 1 l in patients with active Crohn's disease. Gastroenterology 1999'. l 17:58—64. Sands B. Winston B. Sal/berg B. Barish C. Safdi M. Wruble L. et al. A randomized. double-masked. placebo-controlled study of recombinant hu- man interleukin eleven (RHlL-I I) in Crohn‘s disease subjects not receiving prednisone [abstract]. Gastroenterology 199911 162A811. \— NOTES D. A. Williams is a patent holder on the use of interleukin ll (IL-1 1) in the treatment of gut damage. He receives payments from Children's Hospital. Bos— ton . MA. based on milestones set forth in an IL—1 1 agreement between Genetics Institute (Cambridge. MA) and Children's Hospital. I thank Eva Meunier and Sharon Smoot for excellent administrative assistance and Drs. James Keith. Ullrich Sehwertschlag. Andrew Dorner. Sandy Goldman. Stephen Sonis. and Steven Clark for many helpful discussions. Journal of the National Cancer Institute Monographs No. 29. 2001 Infection and Mucosal Injury in Cancer Treatment Shahab A. Khan, John R. Wingurd The oral and gastrointestinal mucosa acts as an important mechanical barrier that prevents local or systemic invasion by microorganisms. Cytotoxic chemotherapy-induced muco- sal injury (MI) of oral cavity and intestinal epithelium oc- curs in many patients treated for malignancy. Compromise of the mucosal barrier can contribute to local invasion by colonizing microorganisms and, subsequently, to systemic infection. Historically, gram-negative bacteremia has been the most problematic bacterial infection in neutropenic pa- tients, but its incidence has reduced over time because of the use of prophylactic antibiotics. There has been a shift in the type of infecting organisms responsible for bacteremia in these patients, from predominantly gram-negative organ- isms to gram-positive cocci. The viridans group of strepto- cocci is composed of the most frequent bacterial pathogens associated with M1. When speciated, oral colonizers such as Streptococcus mitis, Streptococcus oralis, and Streptococcus sangulis II are the most frequently identified pathogens. Other systemic infections caused by vancomycin-resistant enterococci, Stenotrophomonas maltophilia, and Candida spe- cies have also been associated with MI after cancer treat- ment. Infection can also exacerbate MI after cancer treat- ment. The best recognized example is herpes simplex virus type 1 (HSV-l). Latent virus is frequently reactivated in HSV-seropositive patients; this reactivation leads to stoma- titis, which can be indistinguishable from MI caused by cy- toreductive therapies. Antiviral prophylaxis or treatment can control the virus-induced MI and bring about overall amelioration of MI. Recognition of this infectious cause of MI is important in order for clinicians to anticipate and minimize oral toxicity and to facilitate optimal delivery of the antineoplastic regimen. [J Natl Cancer Inst Monogr 2001;29:31—6] Mucosal injury (MI) can lead to a variety of systemic con- sequences. These include impaired oral intake of fluid and nu- trients. leading to dehydration and malnutrition; pain; nausea; vomiting; abdominal cramping; and diarrhea. The mucosa of the oral cavity and gastrointestinal (GI) tract also serves as an im- portant mechanical barrier that helps to prevent a local or sys- tetnic invasion of various microbes and the absorption of mi— crobial products that are normally present in the oral cavity and the lumen of the gut (I). Derangement in the barrier function of the GI tract plays a central role in the pathophysiology of sys— temic infection. shock. and sepsis syndrome. In this article. we will examine two propositions. The first is that Ml is a major contributor to the development of systemic infection by com- mensal colonizing organisms and that this presents a serious challenge to optimal management of the cancer patient. The second proposition exatnines the notion that certain microorgan- isms exacerbate MI. which. in turn. can increase the suscepti— bility for systemic infection from other commensal organisms. Infectious causes of MI are indistinguishable from cytotoxic drug-induced MI and can be confused with Ml from the anti— Journal of the National Cancer Institute Monographs No. 29. 2001 neoplastic regimen. Infection—induced MI may necessitate dose reduction or modification of the antineoplastic regimen. which may compromise the ultimate benefits of the treatment regimen. EFFECTS or CANCER TREATMENTS ON THE ORAL AND GI MUCOSA Cytotoxic chemotherapy is known to cause Ml both in the oral cavity (2w6) and to mitotically active intestinal crypt cells (7). The manifestations of oral mucositis include erythema, ulcer formation. bleeding, and exudates. Methotrexate (7). 5-flouro— uracil. cisplatin (8), cytarabine (9). etoposide. and radiation therapy (XRT) (10) have been shown to have mucosal—damaging effects. Most of the patients treated for head and neck cancer and almost half of the patients receiving chemotherapy for non-head and neck cancer develop oral complications (II). The course of oral mucositis after standard— or high—dose chemotherapy parallels the neutropenia that occurs following such therapy. The onset of oral mucositis occurs near the nadir of neutrophil count. and its resolution parallels hematologic recovery (12). Slavin et al. (9) described the natural history of cytotoxic therapy-induced intestinal damage. Initial injury began during the first week of cytotoxic therapy and was characterized by replacement of normal crypts of mucous-secreting cells by atypical undifferentiated cells. During subsequent weeks, the injury progressed to a second stage. which consisted of cellular necrosis, a lack of mitotic activity. disappearance of villous sur— face. and complications by various infections. Finally. the re- covery phase followed. when mitotic activity returned and cells regenerated. differentiated, and covered the denuded surface. There are several studies (13) of D-xylose absorption tests that have been used as a measure of functional integrity of the in- testinal tnucosal barrier. Studies in patients with acute myeloid leukemia (AML) receiving remission induction therapy have shown malabsorption of D-xylose during weeks 2 and 3 after chemotherapy, secondary to gut epithelial damage (14,15). The magnitude of intestinal epithelial damage as measured by D— xylose malabsorption was strongly correlated with the induction regimen. The effect of radiation therapy on oral cavity primarily results from local tissue changes. These changes are initiated by a re- duction in the proliferation of basal epithelial cells. causing at- rophy (II). The damage of connective tissue may lead to an increase in vascular permeability and tissue edema (10). Oral complications of cancer chemotherapy may be direct somatotoxicity of chemotherapy against basal epithelium: indi- Aflilimion of ant/mm: University of Florida College of Medicine. Gaines— villc. (‘1)l'!"(’.\’[?()lld€ll('(‘ to: John R. Wingard. M.D.. Division of Hematology. Uni- versity of Florida College of Medicine. 1600 SW Archer Rd. PO. Box |00277. Gainesville. FL HMO—0277 (e—mail: wingajr®medicineuflcdu). Soc “Note" following “References." 0 Oxford University Press rect somatotoxicity through the patient’s inability to contain lo— cal, minor oral disease during myelosuppression: or a combina— tion of both (1/). Local infection can produce inflammatory changes that further exacerbate MI. The degree of MI is dependent on the dose intensity of the treatment regimen. Mucositis is particularly frequent in the bone marrow transplant (BMT) population because of the intensive conditioning regimen. Approximately 75% of patients develop some degree of mucositis after the conditioning regimen, which consists of high—dose chemotherapy or combined Chemoradia— tion. Over two thirds of patients with leukemia and one third of those with non—Hodgkin’s lymphoma develop MI (1]). Patients with solid tumors are at lower risk of developing MI (40%), except for patients with head and neck cancer who receive com— bined XRT and chemotherapy. Virtually all of these patients develop mucositis. The effects of XRT on the mouth primarily result from local changes. Consequently, the total dose of XRT to the oral cavity and dose rate are directly related to the extent of MI. The Ml is noted at a level of 20 Gy when XRT is administered at a rate of 200 cGy daily (1]). MI of the oral cavity is frequently accompanied by oral in— fections. Viral, bacterial, and fungal infections are all common. In some clinical settings, systemic bacterial or fungal infections may be more common in patients with mucositis. MI AS CONTRIBUTORY FACTOR TO SYSTEMIC INFECTION Bacterial Infections Bacteremia from gram-negative rods has been the most prob— lematic bacterial infection in chemotherapy—induced neutrope- nia. The GI tract is a major source of bacteria in patients who develop MI as a result of chemotherapy (16). Between 25% and 50% of cases of septicemia in neutropenic cancer patients appear to originate from oral colonizing bacteria (17). The incidence of gram—negative bacterial infections in neu- tropenic patients has decreased over time, perhaps because of both the prophylactic use of broad-spectrum antibiotics in neu— tropenic patients and the empiric use of systemic broad— spectrum antibiotics at first sign of fever in neutropenic patients (18) (Table 1). Nevertheless, studies (18—25) have shown that bacteremia caused by gram—positive organisms is becoming more common. At present, gram-positive bacteria represent the overwhelming majority of neutropenic systemic infections. Fur— thermore, substantial proportions of these gram-positive bacte- rial pathogens are viridans group streptococci. Viridans streptococci are now the second most common ge- Table l. Infectious pathogens encountered in cytotoxic- induced myelosuppression'l' Systemic pathogen arising from gut* Relative frequency Relative severity G NR + +++ 0 PC +++ ++ Candida ++ +++ As'pz’rgillus + +++ HS V +++ + C MV + +++ *GNR = grammegative rods; GPC : gram-positive cocci; HSV : herpes simplex virus; CMV : cytomegalovirus. j‘+, less frequent: ++, frequent: +++. more frequent. 32 nus of bacteria isolated from blood culture after coagulase- negative staphylococci. They can be responsible for up to 39% of bacteremia cases in neutropenic population (26). Several au— thors (27~29) have suggested that oropharyngeal lesions were the most probable portal of entry for viridans streptococci that caused bacteremia. Other investigators (30) have suggested that the rest of the digestive tract, particularly the stomach and lower respiratory tract, might also be portals of entry. In a recent review of literature on bacteremia caused by viridan streptococci in neutropenic patients (26), the most frequently isolated species in blood culture were Streptococcus min's, Strepmcoccus ora/is. and Streptm'm-c-us sanguis [1. Various risk factors have been identified, and the presence of oropharyngeal mucositis was a statistically significant independent factor in most of these stud— ies (263235). Other risk factors included severe neutropenia, prophylactic antibiotic treatment with co-trimaxazole or quino— lone, chemotherapy involving high doses of cytarabine, GI tox- icity requiring antacids or H2 blockers, and heavy colonization by viridans streptococci (27—34). Bochud et al. (32) reviewed 26 episodes of viridans streptococcal bacteremia that occurred in 25 neutropenic patients undergoing intensive chemotherapy for hematologic malignancies. Multivariate analysis of predisposing factors showed that the presence of mucositis was an important independent risk factor for the development of viridans strepto- coccal bacteremia. Pharyngeal lesions were statistically signifi- cantly more frequent in case patients (85%) than in the control patients (55%) (P = .0l). Multivariate analysis of risk factors showed that mucositis was among the three independent predic— tors for the development of viridans streptococcal bacteremia (P = .02). Ruescher et al. (35) reported on 24 patients who were treated with high-dose chemotherapy and an autologous BMT for he— matologic malignancies and who had developed bacteremia with a-hemolytic streptococci. Of these 24 patients with bacteremia, 14 (62%) had ulcerative mucositis, compared with 16 (36%) of 45 patients in the control population (P<.05). Patients with ul- cerative mucositis were found to be three times as likely to develop a-hemolytic streptococcal bacteremia as those without ulcerative mucositis (odds ratio = 3.02). Streptococcal organisms are the most frequent bacterial pathogens associated with M1. However, systemic infections by other bacteria and fungi have also been implicated as sequel- ae of MI. Vancomycin-resistant enterococci (VRE) are rapidly increas- ing causes of infection in hospitalized patients and are associated with considerable morbidity (36). Mucositis has been implicated as a possible contributory factor associated with invasive VRE infection, In one study reported by Kuehnert et al. (37). 738 cancer patients admitted into the hospital had at least one stool specimen obtained for VRE. Nineteen cases of VRE bacteremia were identified. When case patients were compared with control patients, the presence of mucositis, among other factors, was statistically significantly associated with VRE bloodstream in— fection (P<,0l) in univariate analysis. When the independent importance of various risk factors identified in univariate analy— ses was tested in multivariate analysis using logistic regres— sion models, only mucositis remained statistically significantly associated with VRE bacteremia. Furthermore. when the sever— ity of mucositis was assessed quantitatively, the risk of VRE bacteremia increased with increasingly severe mucositis (P<.003); this finding remained valid after adjusting for severity Journal of the National Cancer Institute Monographs No. 29, 2001 of illness and degree of neutropenia. Kuehnert et a]. hypoth- esized that the association of mucositis with VRE bacteremia may be due to diffuse GI mucosa] breakdown, which promotes bloodstream entry by gut—colonizing VRE. In recent years, the emergence of increasing bloodstream an- aerobic infections in neutropenic patients, formerly rare, has been described. Most of these patients have oral mucositis or periodontal disease (38). Labarca et al. (39) reported Sien()rrophomanas malmphi/ia bacteremia in a cluster of eight allogeneic BMT patients. In addition to other associated factors identified when infected patients were compared with control patients, severe mucositis was identified as one of the risk factors (P = .028). Fungal Infection Invasive fungal infections are frequent in patients undergoing cancer chemotherapy that results in prolonged neutropenia and after a BMT (40). As many as 40% of patients undergoing BMT develop invasive fungal infection when neutropenia persists for more than 20 days (41). Candida and Aspergillus sp. are the most frequent causes of fungal infection in leukemia patients undergoing chemotherapy and in BMT patients. Invasive fungal disease in these patients is associated with a high mortality rate (approximately 50%—90%) (42,43). Candida species are commensal organisms that reside normally on the oral mucosa and in the lumen of the GI tract. They not only can cause local infection of the oral mucosa, which is painful, but also can result in esophageal candidiasis or in systemic dissemination. Systemic fungal infections are dif- ficult to recognize and respond poorly to treatment (44). The intact mucosa is an important host defense against systemic Candida infection in neutropenic patients (14). Wingard et a]. (44) reported on 89 consecutive patients treated intensively for leukemia or undergoing BMT for a 12- month period. They observed 18 episodes of Candida sepsis in 17 patients (19%). Three (5%) of 60 patients colonized by Can— dida albicans in their mucosa became infected, while 14 (56%) of 25 patients colonized by C. tropicalis became infected (P3001). These data suggest that C. tropical/is is a more viru— lent systemic pathogen than C. albicans in neutropenic cancer patients, despite being a less frequent colonizer of mucosal sur— face (45). When examined in animal models of Candida viru- lence, no difference in virulence was noted between C. albicans and C. tropicalis when the organisms were given intravenously. However, after the organisms were inoculated into the esopha— gus in mice given chemotherapy that induced damage of the gut mucosa (Fig. l) and neutropenia, C. fI‘0[)i(‘£l/i.8‘ isolates were substantially more virulent than C. albicans isolates. In dose— response assays, systemic invasive infection occurred at inocu- lation doses more than lOO—fold less with C. trapicalis isolates than with C. albicans isolates (46,47). Bow et a]. (14) studied the relationship of cytotoxic regimens with intestinal mucosa] damage and fungal colonization in the pathogenesis of invasive fungal disease in 138 patients under- going induction therapy for untreated AML. They used weekly D-xylose absorption tests ([3) for evaluation of the functional integrity of the upper GI tract and to measure small intestinal epithelial damage. Their results suggested that pathogenesis of invasive fungal disease is linked to cytotoxic therapy-related gut epithelial damage in the setting of fungal colonization of the gut. Patients in whom invasive fungal disease developed had Journal of the National Cancer Institute Monographs No. 29, 200] Fig. 1. A) Cross-sectional histology of normal gastrointestinal tract. Note the elongated villi. single-cell epithelial lining, lack of cellular infiltrate in the sub- mucosa, and scant submucosal blood vessels. B) Following cytarabine, the villi are markedly blunted, and there is denudation of the epithelial mucosa, marked inflammatory infiltration, and dilatation of submucosa] blood vessels, with scat— tered hemorrhages. lower serum D—xylose levels (indicative of greater intestinal ep— ithelial damage), with the maximal difference noted at weeks 2 (P = .0288) and 3 (P = .0019) of chemotherapy, than did uninfected patients. Bow et a]. speculated that damaged mucosal surface may facilitate infection by promoting adherence, local proliferation, and translocation of microorganisms colonizing these surfaces. In this study, gut epithelial damage was maximal during weeks 2 and 3 of induction therapy, which was coinci- dental with the neutrophil nadir. In another report (13), neutro— penic colitis and hepatosplenic fungal infection were also cor- related with the D-xylose malabsorption. The mean serum D-xylose levels during week 2 of chemotherapy in AML pa- tients were lower among subjects who developed neutro- penic enterocolitis (P = .002) and hepatosplenic candidiasis (P = .002) (15). Neutropenic enterocolitis was strongly corre— lated with the development of candidemia (P = .005). Various other reports have identified mucositis as a risk fac- tor for fungemia among the patients receiving antineoplastic therapy. In one report (48) of 41 episodes of breakthrough fungemia occurring in cancer patients receiving antifungal pro— phylaxis, mucositis was identified as one of the risk factors for breakthrough fungemia (34.2% versus 13.1%; P<.05). While each of these studies had methodological differences, each supports the concept that MI offers a portal of entry into the systemic circulation for commensal oral and GI bacteria. Thus, the ability to treat and prevent the severe MI would be an im- portant too] to decrease the rate of bacterial infections in this patient population. These studies in aggregate suggest that systemic infections resulting from MI occur in more intensively treated patients DJ 1») (acute leukemia, induction therapy. and BMT) or are more prevalent in regimens that cause greater degrees of MI and that children are as vulnerable as adults. The reasons are not clear: Oral commensal organisms appear to be more frequent systemic pathogens than GI-colonizing organisms. despite the fact that there is a substantially greater burden of gut—colonizing organ— isms compared with oral colonizers and the fact that there is a much greater surface area of gut mucosa compared with oral mucosa. INFECTION CONTRIBUTING TO MI Herpes simplex virus type 1 (HSV—l) causes the most com— mon symptomatic oral viral infection. HSV seropositivity is an indicator for latent or persistent infection. which may reactivate from a variety of stimuli such as chemotherapy or radiotherapy. This risk for reactivation correlates with the dose intensity of antineoplastic therapy. Reactivation occurs in up to 70%—80% of seropositive BMT and acute leukemic patients (49,50). Reacti— vation rates are lower in less intensively treated patient groups. HSV reactivation occurs in 38%—60% of non-Hodgkin’s lym- phoma patients under treatment and in 15%—20% of patients receiving chemotherapy or radiotherapy for head and neck can— cer (51—62). The frequency of HSV reactivation in various an- tineoplastic treatment settings, especially solid tumor treatment regimens. is not well established. In many patients who are seropositive for HSV, the virus is reactivated after the chemo— therapy (63). Resultant HSV—induced mucositis may be difficult to differentiate from Ml from direct damage caused by chemo- therapy. since the telltale labial blister. the pathogenomonic fea- ture of reactivation of HSV. may not be present. Deep and extensive oral ulcerations may occur because of HSV—l (Fig. 2). In patients treated with high-dose chemotherapy with BMT or after intensive chemotherapy for leukemia. HSV-l mucosal in- fection can also spread contiguously along the mucosal surface, resulting in esophagitis. tracheitis. 0r pneumonitis. It is frequently difficult to distinguish between infectious and noninfectious oral mucositis caused by chemotherapy or irradia— tion. For example, when phase I dose escalation studies were performed for etoposide in a stem cell rescue setting. severe mucositis was reported to be the dose-limiting toxic effect (64). When phase I dose—escalation studies ofetoposide were repeated with acyclovir prophylaxis to prevent HSV reactivation. the Fig. 2. Diffuse ulcerations of the oral mucosa after cytotoxic chemotherapy. 34 maximally tolerated dose of etoposide. evaluated at 50% higher doses (65,66). was not achieved. clearly indicating that much of the formerly described MI attributed to etoposide—direct cyto- toxicity was instead caused by reactivation of HSV. Ulcerative mucositis is still seen after administration of etoposide with acyclovir prophylaxis. especially at the high doses used in con- ditioning regimens in the BMT setting, but it is less severe. Acyclovir. a nucleoside analogue, which is selectively phos- phorylated by a virus—specified thymidine kinase targeting the viral DNA polymerase. is highly effective in preventing MI from HSV-1 and has been shown to be effective prophylactically in prospective randomized trials (51.54.66.67). Before acyclovir‘s prophylactic use. HSV—infected patients had mortality rates from HSV-l infection as high as 5%—IO% after BMT (68). Acyclovir prophylaxis can have secondary benefits in the reduction of the risk for systemic infection from streptococcal bacteria coloniz- ing the mucosa. This was amply illustrated in 60 consecutive BMT patients in which the risk for streptococcal bacteremia was 25% in 30 patients not treated with acyclovir prophylaxis but 0% in 30 consecutive patients in which acyclovir prophylaxis was given (69). Oral cytomegalovirus (CMV)—associated infection of the lip, labial mucosa. tongue. and pharynx has rarely been described in immunocompromised patients (70.71) and can be an infrequent infectious cause of MI. Some reports have described CMV in- fection of the tongue following BMT (72). CMV esophagitis and gastritis. while more common in acquired immune deficiency syndrome. are less frequently seen after BMT. Few case reports have described CMV esophagitis or gastritis either accompany- ing the more commonly seen CMV pneumonitis or coinfecting with HSV in patients receiving immunosuppressive therapy for treatment of graft-versus-host disease (73—75). CMV infection is an infrequent cause of colitis in BMT patients (76). CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH We conclude that the intact mucosa is an important host defense against systemic infection in neutropenic patients. MI is an important and identifiable risk factor for various bacterial and fungal infections, including viridans streptococcal, enterococcal. anaerobic bacteria. and certain gram-negative bacteria, in pa- tients receiving cytotoxic chemoradiotherapy for the treatment of various malignancies. Moreover. the risk of invasive fungal disease is linked to cytotoxic therapy-related oral and gut epi- thelial damage in the setting of fungal colonization of the oral cavity and the gut. The distinction between an infectious etiology of MI as op— posed to regimen-related tissue damage is crucial to the optimal delivery of the antineoplastic regimen. Direct cytotOXiCity dur— ing the course of repeated cycles of chemotherapy may neces- sitate dose reduction in subsequent courses of treatment. Such dose reductions may compromise the ultimate therapeutic con- trol of the underlying neoplasm. since dose intensity has been shown in a number of studies (77—80) of certain neoplasms to affect not only remission rates but also survival. If an infectious etiology for MI were the case. then treatment of the infection during that given course would be appropriate. and subsequent secondary prophylaxis during subsequent courses of treatment to suppress further reactivation and MI would be appropriate to facilitate the delivery of the entire treatment dose. Moreover. amelioration of MI would make the antineoplastic regimen more Journal of the National Cancer Institute Monographs N0, 29. 2001 palatable to the patient and facilitate better compliance. At pres— CHI. no specific therapies are proven to be effective to either treat or prevent M1 secondary to cytotoxic chemoradiotherapy. Thus. new modalities to treat and prevent severe MI and the ability to understand the early steps in pathogenesis of infections at the mucosal barrier would improve the outcome of these groups of patients. reducing the treatment-related morbidity and mortality. REFERENCES (I) Fink MP. Gastrointestinal mucosal injury in experimental models of shock. trauma, and sepsis. Crit Care Med 1991;19:627—411. (2) Kolbinson DA. Schubert MM. Flournoy N. Truelove EL. Early oral (3 (4 (5 (6 \_ (7 \_ (8 \_. (9 (/0) (ll) ([2) (/3) (/4) (/5) (I6) (I7) (/8) (/9) (20) changes following bone marrow transplantation. Oral Surg Oral Med Oral Pathol 1988;66:130—8. Chapko MK. Syrjala KL. Schilter L. Cummings C. Sullivan KM. Cherno- radiotherapy toxicity during bone marrow transplantation: time course and variation in pain and nausea. Bone Marrow Transplant 1989;42181—6. Woo SB. Sonis ST. Monopoli MM. Sonis AL. A longitudinal study of oral ulcerative mucositis in bone marrow transplant recipients. Cancer 1993:72: 1612-7. McGuire DB. Yeager KA. Dudley WN. Peterson DE. Owen DC. Lin LS. et al. Acute oral pain and mucositis in bone marrow transplant and leuke— mia patients: data from a pilot study. Cancer Nurs 1998;21:385—93. Pico JL. Avila»Garavito A, Naccachc P. Mucositis: its occurrence. conse- quences. and treatment in the oncology setting. Oncologist 19983446751. Baskerville A. Batter-Hatton D. Intestinal lesions induced experimentally by methotrexate. BrJ Exp Pathol 1977;58:663—9. Allan SG. Smyth JF. Small intestinal mucosal toxicity of cis—platinum— comparison of toxicity with platinum analogues and dexamethasone. Br J Cancer 1986;53:355—60. Slavin RE. Dias MA. Saral R. Cytosine arabinoside induced gastrointesti- nal toxic alterations in sequential chemotherapeutic protocols: a clinical— pathologic study of 33 patients. Cancer 1978;42:1747—59. Baker DG. The radiobiological basis for tissue reactions in the oral cavity following therapeutic x-irradiation. A review. Arch Otolaryngol 1982:1082 21—4. Sonis ST. Oral complications. In: Holland JF. Bast RC Jr. Moston DL. editors. Cancer medicine. Baltimore (MD): Lippincott; 1997. p. 3255764. Lockhart PB. Sonis ST. Relationship of oral complications to peripheral blood leukocyte and platelet counts in patients receiving cancer chemo- therapy. Oral Surg 1979;48:21—8. Craig RM. Atkinson AJ. D«Xylose testing. A review. Gastroenterology 1988295222341. Bow E. Loewen R. Cheang M. Schacter B. Invasive fungal disease in adults undergoing remission—induction therapy for acute myeloid leukemia: the pathogenetic role of the antileuketnic regimen. Clin Infect Dis 1995:le 361-9. Bow EJ. Loewen R. Cheang MS. Shore TB. Rubinger M. Schacter B. Cytotoxic therapy-induced D