sign in sign up
<<
Home , Esophageal cancer

Progress Review Groups

Report of the /Esophageal Progress Review Group

U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES National Institutes of Health National Institute December 2002

Stomach/Esophageal Cancers Progress Review Group Priority Recommendations

Population Studies Establish collaborations for conducting interdisciplinary, population-based, endoscopic, multi-institutional studies to identify populations at greatest risk for gastric cancer, esophageal , and esophageal squamous cancer, and to determine the prevalence and natural history of premalignant lesions. Prevention Develop prevention strategies based on the mechanisms of host/environment interactions that lead to metaplasia and neoplasia of the stomach and . Evaluate their effectiveness in at-risk populations. Patient/Provider Educate patients and their families, health care professionals, and the public Education regarding risk factors, risk reduction, and treatment options and outcomes for gastroesophageal cancers and their precursor states. Therapy Develop and test novel therapeutics, and optimize existing treatments for gastroesophageal cancers and their precursors, based on the identification and understanding of molecular pathways involved in oncogenesis, tumor response and resistance. Therapeutic Targets Define host and molecular/biologic tumor characteristics that will help customize treatment and best predict recurrence and/or survival. Markers and Molecular Profile the molecular, cellular, and epidemiological features of Profiling gastroesophageal tumors and their precursor lesions to identify diagnostic, prognostic, predictive, preventive, and therapeutic targets. Outcomes Develop and refine disease-specific, patient-oriented methods to assess quality of life, quality of care, and cost effectiveness of treatment in patients with gastroesophageal cancers and their precursors through all stages of disease and treatment, and include these instruments in clinical trials and observational studies. Host/Environmental Identify, develop, and validate genetic, biochemical, and biological markers Interactions that will help uncover host-environment interactions in esophageal and gastric carcinogenesis. Technologies for Develop noninvasive and minimally invasive technologies (e.g. serum Screening/Surveillance markers and imaging techniques) for screening and surveillance of premalignant and malignant gastroesophageal lesions. Preclinical Models Establish models to understand the biology of gastroesophageal cancers and their precursor lesions, and to stimulate prevention, diagnostic and treatment strategies.

REPORT OF THE STOMACH/ESOPHAGEAL CANCERS PROGRESS REVIEW GROUP

National Cancer Institute

Co-Chairs: Timothy J. Eberlein, M.D., Washington University School of Medicine Brian J. Reid, M.D., Ph.D., Fred Hutchinson Cancer Research Center

Executive Director: Ernest T. Hawk, M.D., M.P.H., National Cancer Institute

December 2002

From the Leadership

It is a great pleasure to submit this Report of the Stomach/Esophageal Cancers Progress Review Group (S/E PRG) to the Director and Advisory Committee to the Director of the National Cancer Institute (NCI). The S/E PRG accepted the charge of former NCI Director Dr. Richard Klausner to develop a national plan for stomach and esophageal cancer research over the next 5 years. The charge was advanced with the support of the current NCI Director, Dr. Andrew von Eschenbach. This report represents the collaborative efforts of the scientists, clinicians, industry representatives, and patient advocates who participated in the S/E PRG Roundtable Meeting. The priorities outlined in this report are a blueprint for progress toward preventing, diagnosing, and treating stomach and esophageal cancers. We look forward to discussing these priorities and the plan for their implementation with the leadership of the NCI.

Timothy J. Eberlein, M.D. Brian J. Reid, M.D., Ph.D. Ernest T. Hawk, M.D., M.P.H. Washington University School Fred Hutchinson National Cancer Institute of Medicine Cancer Research Center PRG Executive Director Cancer Research Center PRG Co-Chair PRG Co-Chair

We the undersigned members of the Stomach/Esophageal Cancers Progress Review Group concur with the enclosed report.

Timothy J. Eberlein, M.D. Brian J. Reid, M.D., Ph.D. PRG Co-Chair PRG Co-Chair Alvin J. Siteman Cancer Center Public Health Sciences Division Washington University School of Medicine Fred Hutchinson Cancer Research Center

Ernest T. Hawk, M.D., M.P.H. PRG Executive Director Division of Cancer Prevention National Cancer Institute

Jaffer Ajani, M.D. Martin Blaser, M.D. University of Texas New York University M.D. Anderson Cancer Center School of Medicine

Jo Ann Brooks, Ph.D. Donald Castell, Ph.D. Indiana University Medical Center Medical University of South Carolina

Daniel Coit, M.D., F.A.C.S. Pelayo Correa, M.D. Memorial Sloan-Kettering Cancer Center Louisiana State University Health Sciences Center

Thomas R. DeMeester, M.D. M. Brian Fennerty, M.D. University of Southern California Oregon Health and Sciences University Keck School of Medicine Division of

We the undersigned members of the Stomach/Esophageal Cancers Progress Review Group concur with the enclosed report.

Arlene A. Forastiere, M.D., Ph.D. Karen J. Goodman, Ph.D. Johns Hopkins University University of Texas Health Science Center Sidney Kimmel Comprehensive Cancer Center at Houston School of Public Health

Stanley R. Hamilton, M.D. Jimmie Holland, M.D. University of Texas Memorial Sloan-Kettering Cancer Center M.D. Anderson Cancer Center

Theodore Lawrence, M.D., Ph.D. Douglas S. Levine, M.D. University of Michigan AstraZeneca Department of Radiation

Stephen Meltzer, M.D. Cherie Nichols, M.B.A. University of Maryland Hospital National Cancer Institute

Roy C. Orlando, M.D. Tulane University Dawn Provenzale, M.D. Health Sciences Center Duke University Medical Center

Richard Sampliner, M.D. Anil K. Rustgi, M.D. University of Arizona University of Pennsylvania Arizona Health Sciences Center

We the undersigned members of the Stomach/Esophageal Cancers Progress Review Group concur with the enclosed report.

Robert Tell, M.S.H.A. Joel E. Tepper, M.D. Patient Advocate University of North Carolina at Chapel Hill School of Medicine

Jacques Van Dam Thomas Vaughan, M.D. Stanford University Medical Center Fred Hutchinson Cancer Research Center

Kenneth K. Wang, M.D. Michael Welch, Ph.D. Mayo Clinic Washington University Medical School

Acknowledgments

This report is the product of months of intense work that drew on the combined expertise and efforts of many individuals. The Stomach/Esophageal Cancers Progress Review Group (S/E PRG) particularly acknowledges the contributions of:

• The many scientists, clinicians, and advocates from across the country who generously gave of their time and knowledge and without whose participation this report would not have been possible.

• The staff of the NCI Office of Science Planning and Assessment (OSPA), under the leadership of Cherie Nichols, who provided ongoing guidance and technical support to the S/E PRG throughout the process of preparing this report. In particular, Deborah Duran, who coordinated the S/E PRG, as well as James Corrigan and Annabelle Uy.

• A group of experienced science writers, who provided excellent writing support during the Roundtable Meeting as well as during the subsequent development of this report: Laura Janusik, Lisa Chiu, Deborah Barnes, Jackie Beals, Deborah Berlyne, Laura Drake, Randi Henderson, Kit Johnston, Ramie Liebnitz, Cheryl Pellerin and Nancy Volkers.

• The staff of Palladian Partners, Inc., who provided excellent logistical support to the S/E PRG: Syreeta Tate, Ellie Dorsey, Jean Kazares, Bridgette Saunders, and Asia Walton.

TABLE OF CONTENTS

Overview ...... 1 Introduction ...... 4

SCOPE OF THE PROBLEM ...... 4 State of the Science ...... 6

OPPORTUNITIES FOR SCIENTIFIC ADVANCEMENT ...... 6 CHALLENGES TO BE ADDRESSED ...... 9 Recommendations ...... 9 Infrastructure: Resources and Partnerships...... 15

STOMACH/ESOPHAGEAL NEOPLASIA TRANSLATIONAL RESEARCH NETWORK ...... 15 Overview...... 15 SENTRNet’s Conceptual Foundation...... 15 Key Elements of SENTRNet’s Infrastructure/Components...... 18 Unique Management Strategy Concepts ...... 20 How SENTRNet Will Facilitate Progress Toward Achieving the PRG Recommendations ...... 22 Conclusion...... 27 References ...... 28 Appendix A: About the National Cancer Institute=s Progress Review Groups...... 29

CHARGE TO THE PRGS...... 29 THE PRG PROCESS ...... 29 THE PRG REPORT ...... 30 Appendix B: Breakout Reports...... 31

GUIDING PRINCIPLES ...... 31 Biology ...... 31 Etiology ...... 35 Genetics...... 39 Imaging/Technologies ...... 43 Outcomes/Education/Communication/Quality of Life...... 47 Predictive and Prognostic Markers...... 50 Prevention ...... 52 Surveillance/Databases ...... 56 Therapeutics ...... 59 Tumor Models...... 62 DISEASE SITES ...... 64 ...... 64 Gastric...... 67 Squamous...... 70 POPULATION MANAGEMENT...... 73 At Risk ...... 73 Premalignant...... 76 Localized Malignant...... 79 Late Malignant ...... 82 Appendix C: Stomach/Esophageal Cancers PRG Members...... 85 Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants ...... 87

Priorities of the Stomach/Esophageal Cancers Progress Review Group

Overview

Gastroesophageal cancers are an enormous basis for these cancers and the cause of morbidity and mortality worldwide. host/environment interactions underlying In the year 2000, it was estimated that more them are needed. Practical challenges than 1,288,000 new cases of include gaining access to adequate numbers gastroesophageal cancers were identified, of at-risk persons or cancer patients, and more than 984,000 people died from recruiting physicians with the expertise to them, making this combination of cancers manage the many at-risk patients, and the most common form of incident cancer dealing with a disparate collection of and the second most common cause of cancers (i.e., gastric adenocarcinoma, cancer death in the world.1 esophageal squamous cell , and esophageal adenocarcinoma). In the United States, gastroesophageal cancers are relatively uncommon; however, To capitalize on the current state of the esophageal cancer appears to be on the rise. science and to overcome these challenges, This suggests that there are many more collaboration across borders is needed. Only individuals at risk for the disease, although through collaborative efforts can enough their risk status may be unrecognized. Three patient data be collected to enhance the issues particularly relevant to gastro- understanding of the biology, etiology, esophageal cancers and their impact on the pathology, and treatment of the diseases. To U.S. population include: 1) the significant gain access to large numbers of at-risk morbidities associated with the diseases and patients for these cancers, there is a critical their treatments, 2) their almost uniformly need to involve gastroenterologists and other poor prognoses, and 3) their burden among specialists who perform esophagogastro- minorities. duodenoscopy in NCI-sponsored research efforts. Pathologists, molecular biologists, Several aspects of gastroesophageal cancers epidemiologists, and clinical trial provide opportunities for rapid scientific and researchers can then use these data to clinical advancements. The stomach and identify important new markers and targets. esophagus are relatively easy and safe to access, which can provide ample specimens There is a great opportunity-to-investment for research. Additionally, the technologic ratio related to research in these cancers. For advances in molecular profiling, imaging, instance, risk markers may be applied in and molecular targeting of preventive/ screening, diagnosis, and prognostication; therapeutic agents provide a foundation for interventive response markers may identify further developments to reduce the burden new approaches to modulate risk. Thera- of these cancers and perhaps others. At a peutic targets can inform agent identification minimum, advancing these important and development. Additionally, many research opportunities will improve the aspects of the molecular etiology of identification, care, and management of gastroesophageal carcinogenesis may be persons at risk for and living with shared with other cancers that are less gastroesophageal cancers. accessible or amenable to serial investi- gations; therefore, research findings may There are several challenges to advancing have broader implications. these research opportunities. At a molecular level, the development of tumor models and To develop these opportunities, the a further understanding of the molecular Stomach/Esophageal Cancers Progress

Priorities of the Stomach/Esophageal Cancers Progress Review Group 1 Review Group identified and prioritized carcinoma. A secondary challenge was to research recommendations that will advance make certain that the full spectrum of the the understanding of these cancers over the disease and the foundational scientific next 5 years. principles that underlie scientific advances in disease management were addressed. A Stomach and Esophageal Progress three-dimensional, overlapping approach Review Group Process was developed to ensure that ideas had multiple audiences. The attendees would Premise for Planning. The consider the cancers through the lenses of Stomach/Esophageal Cancers Progress Scientific Guiding Principles, Population Review Group (S/E PRG) held a planning Management, and Disease Sites. Figure 1 meeting January 24 and 25, 2002, to represents the conceptual foundation that organize the Roundtable Meeting that would organized the roundtable subgroup classify progress, identify gaps, and discussion format. This organization ensured highlight research opportunities across the that each topic would be discussed in continuum of stomach and esophageal adequate detail, with attention to the full cancers research. A primary challenge was spectrum of important issues. To guarantee identifying an approach that would include expert facilitation of each discussion topic, the various stomach and esophageal cancers. S/E PRG members were assigned to serve as It was agreed to focus on gastric co-chairs, and roundtable participants were adenocarcinoma, esophageal adeno- identified for each of the breakout carcinoma, and esophageal squamous cell subgroups.

Figure 1. PRG Organization

Scientific Guiding Principles

Roundtable Meeting. The S/E PRG session created three priorities and rationales Roundtable Meeting, which included addressing specific gastroesophageal approximately 112 participants, convened cancers. Afternoon sessions incorporated the May 5–7, 2002, at Westfield’s Conference morning sessions’ priorities and rationales Center in Chantilly, Virginia. Participants into their discussions to ensure that attended one session each on Scientific comprehensive views were represented in Guiding Principles, Population each session. Each subgroup’s report and Management, and Disease Sites. Each priorities can be found in Appendix B. The

2 Report of the Stomach/Esophageal Cancers Progress Review Group

Population Management and Disease Sites consensus session. The roundtable subgroups developed 21 priorities; each participants reached consensus on 10 high- priority had its own rationale, partnership priority research recommendations (Table 1) platforms, and resources needed to capitalize and a single partnership platform aimed at on existing opportunities and to overcome improving prevention, diagnosis, and current challenges. On the final day, the S/E treatment of stomach and esophageal PRG Leadership clustered the 21 priorities cancers. Recommendations and to elucidate similar recommendations. These corresponding rationales are detailed in the 21 priorities were presented to the Recommendations section that follows. roundtable participants at the final

Stomach/Esophageal Cancers Progress Review Group Priority Recommendations Table 1. Population Studies Establish collaborations for conducting interdisciplinary, population- based, endoscopic, multi-institutional studies to identify populations at greatest risk for gastric cancer, esophageal adenocarcinoma, and esophageal squamous cancer, and to determine the prevalence and natural history of premalignant lesions. Prevention Develop prevention strategies based on the mechanisms of host/environment interactions that lead to metaplasia and neoplasia of the stomach and esophagus. Evaluate their effectiveness in at-risk populations. Patient/Provider Educate patients and their families, health care professionals, and the Education public regarding risk factors, risk reduction, and treatment options and outcomes for gastroesophageal cancers and their precursor states. Therapy Develop and test novel therapeutics, and optimize existing treatments for gastroesophageal cancers and their precursors, based on the identification and understanding of molecular pathways involved in oncogenesis, tumor response and resistance. Therapeutic Targets Define host and molecular/biologic tumor characteristics that will help customize treatment and best predict recurrence and/or survival. Markers and Profile the molecular, cellular, and epidemiological features of Molecular Profiling gastroesophageal tumors and their precursor lesions to identify diagnostic, prognostic, predictive, preventive, and therapeutic targets. Outcomes Develop and refine disease-specific, patient-oriented methods to assess quality of life, quality of care, and cost effectiveness of treatment in patients with gastroesophageal cancers and their precursors through all stages of disease and treatment, and include these instruments in clinical trials and observational studies. Host/Environmental Identify, develop, and validate genetic, biochemical, and biological Interactions markers that will help uncover host-environment interactions in esophageal and gastric carcinogenesis. Technologies for Develop noninvasive and minimally invasive technologies (e.g., serum Screening/Surveillance markers and imaging techniques) for screening and surveillance of premalignant and malignant gastroesophageal lesions. Preclinical Models Establish models to understand the biology of gastroesophageal cancers and their precursor lesions, and to stimulate prevention, diagnostic, and treatment strategies.

Priorities of the Stomach/Esophageal Cancers Progress Review Group 3 The top research priority is the creation of a Gastroesophageal cancers are heterogeneous multi-institutional, multidisciplinary partner- with regard to their molecular and cellular ship of researchers focused on rapid genesis, specific risk factors, and histo- translational biomedical advances in these pathologic character. For this report, the cancers (detailed in the last section of the term “gastroesophageal cancers” encom- main report). This solution best addresses passes three distinct cancers: esophageal the S/E PRG recommendations within the , esophageal (or context of the current state of the science Barrett’s-related) adenocarcinoma, and and the incidence in the United States. The gastric adenocarcinoma. Although they are infrastructure is similar to cancer coopera- distinct entities that originate in the same tive groups and other clinically oriented general anatomic region of the digestive research consortia; however, two critical system, these three cancers share some features distinguish this initiative from characteristics. Most importantly, they tend others. First, innovative and progressive to remain clinically silent until late in the management strategies will facilitate disease process; thus, they are often effective and efficient components that will associated with later diagnoses, poorer foster group-wide priorities. These will prognoses, significant morbidities, and high include multi-institutional, inter-disciplinary mortality rates. collaborations and shared resources to enhance knowledge and reduce the burden . Worldwide, the incidence of stomach and esophageal cancers. Second, of stomach cancer is declining, pointing to a involvement of gastro-enterologists and critical environmental component in its other funding partners will provide access to etiology. Although the specific reason for patients at risk for gastroesophageal cancers this decline is unknown, the increased (a group not commonly cared for by consumption of fruits and vegetables and the specialists within traditional cancer decreased intake of salty foods, both at least cooperative groups) and diversify funding partially resulting from improved methods sources. Gastroesophageal tissues in food preservation and storage, are often representative of the full spectrum of credited. Despite this encouraging trend, pathogenesis will be secured. The shared stomach cancer is the fourth most common resources and the multidisciplinary experts new cancer diagnosis and the second leading will facilitate a true translational focus. cause of cancer mortality in the world, accounting for an estimated 876,341 new This kind of comprehensive partnership, cancer cases and 646,567 deaths worldwide which will help to overcome the scattering in 2000.1 The highest incidence of stomach of patients across the country and the limited cancer occurs in Japan and Eastern Asia; by resources of any one institution, is crucial to contrast, its incidence is relatively low in the NCI’s ability to make advances in Western Europe. combating stomach and esophageal cancers and addressing the S/E PRG recommen- Stomach cancer was the most common dations. This model is explained in detail in cancer in the United States during much of the Infrastructure section that follows. the early 20th century; however, its incidence has declined significantly since the 1950s. Introduction U.S. incidence rates for stomach cancer are higher for Asian/Pacific Islanders, blacks, Scope of the Problem and Hispanics than for whites or American Indians/Alaska Natives. Men are 1.5 to 2 Each year, gastroesophageal cancers account times as likely to develop stomach cancer as for an estimated 34,700 new cancer cases women are. and 25,000 deaths in the United States.2

4 Report of the Stomach/Esophageal Cancers Progress Review Group

Table 2. SEER Incidence Age-Adjusted Rates, 11 Registries, 1992–19993

ESOPHAGUS STOMACH Race/Ethnicity All Males Females All Males Females All 4.5 7.5 2.1 9.3 13.56.3 White 4.2 7.1 1.9 7.9 11.75.2 Black 8.0 12.9 4.4 13.9 19.69.9 American Indian/ Alaska Native 1.5 2.8 0.5 7.6 9.8 5.9 Asian or Pacific Islander 3.0 5.6 1.0 18.5 24.9 13.6 Hispanic 2.9 5.4 1.0 12.5 17.1 9.2

Age-adjusted rates, 2000. Rates are expressed as cases per 100,000.

Esophageal Cancer. Esophageal cancer has Epidemiologists have identified some been relatively uncommon in the United unexplained and remarkable differences in States, but recent trends are of concern. its distribution across the world. For Incidence rates for adenocarcinomas example, esophageal cancer is more involving the gastric cardia and lower common in developing countries than in the esophagus (Barrett’s esophageal United States. In addition, histopathologic adenocarcinoma) have increased markedly types vary between these regions, with since the mid-1970s. Among white males, squamous cell carcinoma dominating in the rate of esophageal adenocarcinoma has developing countries and adenocarcinomas increased more than 350 percent between becoming an increasing problem in the 1974 and 1994, making this one of the most United States. rapidly rising cancers in the U.S. population and suggesting that environmental factors Mortality. Stomach and esophageal cancers, play important roles in its etiology. In the while relatively uncommon in the United United States, Blacks have the highest rate States, are highly lethal. The estimated of esophageal cancer, primarily squamous overall 5-year survival rate is 22 percent for cell, almost double the incidence of all other stomach cancer and 14 percent for groups. Incidence is higher among men than esophageal cancer. In fact, mortality rates among women: men are 3 to 5 times as for these cancers approach their incidence likely to develop esophageal cancer as rates, suggesting that current treatment women are. options for these patients are limited and often ineffective. Notably, minorities tend to Esophageal cancer is the eighth most have disproportionately high mortality rates common new cancer diagnosis and the sixth from these cancers. For stomach cancer, the leading cause of cancer death in the world, rates are more than twice as high in blacks accounting for an estimated 412,327 new and Asian/Pacific Islanders, compared with cancer cases and 337,501 deaths in 2000.1 whites.

Priorities of the Stomach/Esophageal Cancers Progress Review Group 5 Table 3. SEER Mortality Age-Adjusted Rates, Total U.S., 1990–19993

ESOPHAGUS STOMACH Race/Ethnicity All Males Females All Males Females All 4.3 7.51.8 5.4 7.73.7 White 3.9 6.91.6 4.8 6.93.3 Black 8.1 14.33.8 10.3 15.27.1 American Indian/ Alaska Native 2.4 4.0 1.0 5.3 7.1 3.9 Asian or Pacific Islander 2.4 4.1 .96 10.4 13.6 8.0 Hispanic 2.4 4.4.91 7.2 9.75.3

Age-adjusted rates, 2000. Rates are expressed as cases per 100,000.

Three factors are associated with the poor stomach is removed, a patient should still be survival rates associated with able to eat fairly normally. However, if the gastroesophageal cancers. First, symptoms entire stomach is removed, a new eating are rare in the early stages of cancer pattern must be adopted, including frequent, development, often only occurring with small meals low in sugar and high in fat and advanced cancer; even then, they can be protein. For many patients, nonspecific. Second, although screening and and radiation also are necessary, so diagnostic techniques exist for these cancers, additional side effects are common. the risks, benefits, and feasibility of screening have not been adequately tested The incidence, morbidity, and mortality of and consequently are not routinely stomach and esophageal cancers make them recommended by physicians. Finally, an important health concern, particularly patients often fail to seek appropriate within minority sectors of the population. medical attention due to lack of knowledge The rapid rise of adenocarcinomas of the of important risk factors and symptoms distal esophagus and proximal stomach and/or reluctance to undergo invasive and among whites, as well as the high rates of relatively costly endoscopic procedures. esophageal cancer in blacks, suggests that Therefore, many patients present at late these cancers are an important health issue stages. today. With the increasing rates of Asian/Pacific Islander and Hispanic Morbidity. Patients diagnosed with stomach immigration, these cancers will be an and esophageal cancers often have important medical and public health issue in significant morbidities (e.g., difficulty the future as well. swallowing, painful swallowing, ) as a result of their cancer or the State of the Science treatments intended to help them. Necessary treatments often involve the removal of Opportunities for Scientific portions of the esophagus and/or stomach or the placement of a to maintain the Advancement patency of the GI tract. After , some esophageal cancer patients may need to Gastroesophageal cancers are significant receive nutrients directly into a vein or causes of morbidity and mortality in the through a feeding tube. If only a part of the United States; most individuals with these

6 Report of the Stomach/Esophageal Cancers Progress Review Group cancers are not identified until the cancer is studies suggest that symptoms of GERD advanced and symptomatic. In these cases, occur weekly in approximately 20 percent of even the best therapies are associated with Americans, implying that the number of significant morbidities and poor outcomes. persons at risk for these cancers may be Only a small fraction of patients is found to substantial. Of course, both of these diseases harbor preinvasive neoplastic lesions (e.g., are much more common in other parts of the Barrett’s ) while undergoing an world, so attention to the problem within endoscopic evaluation for nonspecific immigrants to the United States from Asian symptoms. Typically, such patients would or Central/South American countries is have periodic endoscopic surveillance, and a important. Indeed, the marked heterogeneity small number of them would develop early in the worldwide distribution of these stage cancer. Identified in this way, a patient cancers, as well as their rapidly changing would undergo an with the incidence patterns within the United States, potential for significant post-operative suggest that there are important environ- morbidities, and in most instances, would mental risk or preventive factors acting at have an excellent prognosis for long-term the molecular level, which, once identified, survival. may be employed to reduce the burden of these diseases. There are tremendous opportunities to improve the management and care of people Disease Pathogenesis. A detailed who have gastroesophageal cancers, as well understanding of a disease’s pathogenesis as those at risk. Three factors specific to and natural history is necessary to advance gastroesophageal cancers make scientific the care of affected individuals. and clinical progress imminently attainable. Gastroesophageal cancers are believed to First, the marked distributional develop over decades with few or no heterogeneity of gastroesophageal cancers presenting symptoms during most years. within the population suggests the presence Fortunately, the availability of of effective risk and preventive factors. esophagogastroduodenoscopy (EGD) allows Second, relatively easy and safe access to gastroenterologists to directly visualize the the stomach and esophagus is available lining of these organs and easily obtain through established technologies for serial for molecular, cellular, and endoscopic assessments with mucosal histopathologic assessments over time. biopsies. Finally, technologic advances in Thus, gastroenterologists can use EGD to genomics, proteomics, invasive and gain access to tissues at risk for these noninvasive imaging, as well as in the cancers, thus allowing improvements in the molecular targeting of preventive/ understanding of the disease process. therapeutic agents, can be applied for rapid advances in a highly translational Technologies. Recent technologic advances environment in which patients harboring are providing opportunities to reduce the preinvasive neoplasia undergo serial burden of gastroesophageal cancers. As surveillance of their gastroesophageal evidenced from the study of other organs, mucosa as a matter of standard care. knowledge of the molecular basis may lead to further understanding and identification Several common conditions—for example, of targets. Research may elucidate the gastroesophageal reflux disease (GERD) and identification of reliable indicators of cancer infection—place risk and patient response to interventions, as affected individuals at increased risk for one well as targets for preventive or therapeutic or more of these cancers. Recent cohort interventions. For example, with the advent

Priorities of the Stomach/Esophageal Cancers Progress Review Group 7 of genomic and proteomic technologies, it is and another to gastric adenocarcinoma. now possible to examine the genesis of Likewise, the molecular basis for the gastroesophageal neoplasia from the earliest relationship between other agents, such as molecular alterations through precursor bile or , and gastroesophageal lesions (cellular and tissue abnormalities) to cancers also may become evident. invasive and metastatic cancers. Charting Investigators will be able to develop the molecular changes underlying the prevention strategies based on the natural history of gastroesophageal cancers mechanism of protective host/environment during every step of this process may allow interactions. the identification of environmental risk factors, genetic risk factors, and/or Genes. Genomics-based methodologies may biomarkers that can be used for non-invasive facilitate drug discovery and the translation screening and diagnostic tests. of drugs from “bench to bedside”; they also may provide a means to monitor patient Bioinformatics. The advent of novel responses or resistance to new interventions analytic approaches has facilitated deriving applied with preventive or therapeutic intent. meaning from novel genomics and In addition, the examination of molecular proteomics data. These approaches include, genetic and host factors during clinical trials but are not limited to, hierarchical involving an intervention will provide agglomerative clustering, significance another level of understanding of patient analysis of microarrays, GeneFinder response and resistance. comparisons, artificial neural networks, and principal components analysis. Widespread Models. Scientific advances require the use use of such strategies will identify global of appropriate in vitro and in vivo models, patterns of gene and protein expression few of which exist for stomach and potentially useful a priori as early detection, esophageal carcinogenesis. The develop- diagnostic, prognostic, or risk assessment ment of improved tumor models—including tools. Moreover, these techniques will cell lines, xenografts, and animal models— identify individual genes or groups of genes therefore represents an opportunity for worthy of further study and useful in scientific advancement. Once standardized hypothesis generation, such as in the and validated, tumor models specific to implication of novel molecular pathways, gastroesophageal cancers will aid environmental and socioeconomic factors, investigators in the identification of dietary influences, genetic makeup, and the molecular and cellular changes that mark host response in the genesis or progression disease progression and improve of these important cancers. understanding of host/environment interactions. These models will provide a Molecular Profiling. Understanding the means for investigators to develop new molecular basis of these cancers also affords molecularly targeted approaches to the opportunity for more careful study of prevention and therapies. host-environment interactions. For example, epidemiological evidence indicates a Patient Care and Outcomes. In addition to relationship between infection by the molecular and technologic efforts, the bacterium H. pylori and the development of scientific community has the opportunity to gastric ulcers and adenocarcinoma. Genomic make significant advances in patient care and proteomic technologies will facilitate and morbidity reduction. Clinical the identification of genetic differences that researchers can develop disease-specific, may predispose one person to a gastric ulcer patient-oriented methods to assess the

8 Report of the Stomach/Esophageal Cancers Progress Review Group quality of life, quality of care, and cost Incomplete Network Systems. There effectiveness of preventive and therapeutic is a clear need for a coordinated, multi- approaches. In addition, education and institutional partnership that involves a wide outreach measures addressing risk factors range of research professionals, some of and screening that target the public and whom—particularly gastroenterologists— community physicians are likely to reduce are not adequately represented in existing morbidity and mortality from these diseases. NCI consortia. In addition to providing a means for rigorous study of current and new Integration/Translation. Preliminary data treatments, a network will provide from interdisciplinary translational research investigators with access to patients at in stomach and esophageal cancer are different stages of carcinogenesis, and to the promising. When interdisciplinary teams critical tissue samples needed for studies of involve combinations of molecular the biology and etiology of these cancers. biologists, gastroenterologists, pathologists, epidemiologists, medical/surgical oncolo- Lack of Awareness. Another challenge is gists, and other members, knowledge grows the underestimation and limited awareness quickly and exponentially. For example, of these cancers by the public and some a recent study of familial gastric cancer physicians. Unlike many other cancer identified a tightly linked mutation in cohorts, people with gastroesophageal E-cadherin that may provide insights into cancers have not had a specific advocacy the pathogenesis of the syndrome. In group or an important public figure to draw addition, two studies of patients with attention to the disease and advocate for Barrett’s metaplasia recently noted a resources. Public education in gastro- significant correlation between cytometric esophageal cancers might encourage at-risk aneuploidy and the future development of persons to seek earlier screening and dysplasia or cancer. diagnosis, thereby potentially reducing morbidity and mortality. Challenges To Be Addressed To capitalize on the scientific opportunities Low Incidence. The relative rarity of these and to overcome the challenges, the S/E cancers in the United States presents a PRG was convened by NCI and charged to significant research challenge. Individual provide recommendations to reduce the centers do not treat enough patients to burden of these cancers. conduct adequately powered clinical studies intended to evaluate the natural history of Recommendations these diseases, develop new methods for screening and surveillance, or test new The S/E PRG recognizes that these high- preventive and therapeutic interventions. As mortality cancers present a unique set of a result, few adequately powered studies challenges and opportunities. Gastro- addressing these issues have been esophageal cancers represent diverse completed, and most current care is based on that have rapidly changing observational data and expert opinion. To incidences, and their ethnic and gender transcend this challenge, NCI must evaluate disparities are not well understood. These the current research infrastructure and cancers also have long premalignant phases develop a means to coordinate research and that are uniquely accessible for endoscopic management of these relatively rare cancers visualization and ; this, combined in order to secure the critical mass of cases with the rapidly advancing field of cancer needed for scientific advancement. genetics, offers unparalleled opportunities

Priorities of the Stomach/Esophageal Cancers Progress Review Group 9 for mechanism-based approaches to esophageal adenocarcinoma, and gastro- screening, surveillance, prevention, early esophageal reflux and are risk detection, and treatment. factors for esophageal adenocarcinoma. Although these associations are clear, It is difficult for any one center in the United critical gaps exist in our knowledge of the States to generate enough cases to make an underlying mechanisms. Multidisciplinary, impact on gastroesophageal cancer tissue-based studies across the spectrum of morbidity and mortality. However, U.S. progression are needed to identify the cases could, if combined, provide data for linkages between molecular, cellular, and appropriately powered clinical and clinical pathogenesis. This knowledge then epidemiological studies. Through can be applied to better define the risk status collaboration of clinical, population, of individuals and groups. laboratory, and computational scientists, private industry, and the NCI, the burden Although gastroesophageal cancers are of these malignancies on society can be relatively uncommon in the United States, reduced. their predisposing conditions are relatively common: Barrett’s esophagus may be seen Therefore, the S/E PRG offers the following in as many as 10 percent of asymptomatic 10 prioritized research recommendations adults, and H. pylori infection rates may be and a high-priority infrastructure resource: as high as 40 percent. Despite the frequency of these conditions, the true population Population Studies: Establish prevalence and natural history of collaborations for conducting preinvasive neoplastic lesions in the interdisciplinary, population-based, stomach and esophagus are poorly endoscopic, multi-institutional studies to established because studies have come identify populations at greatest risk for mainly from single institutions. Endoscopy gastric cancer, esophageal adenocar- can safely and systematically visualize and cinoma, and esophageal squamous cancer, biopsy stomach and esophageal prema- and to determine the prevalence and natural lignant conditions prospectively, providing history of preneoplastic lesions. an unprecedented opportunity to define the prevalence and natural history of these Rationale conditions, establish risk stratification, and characterize the genetic and biological Patients with gastroesophageal cancers mechanisms of carcinogenic progression. usually present at late stages and have poor Data from these multi-institutional, multi- prognoses. To reverse this trend, patients at disciplinary studies must be aggregated to highest risk must be identified earlier. define statistically significant at-risk Recent epidemiologic studies suggest populations, risk and protective factors, and protective and risk factors for each of the the natural history of gastroesophageal gastroesophageal cancers. Non-steroidal neoplasia. anti-inflammatory drugs (NSAIDs) appear to be protective against esophageal Prevention: Develop prevention strategies squamous cell carcinoma, esophageal based on the mechanisms of adenocarcinoma, and gastric adenocar- host/environment interaction that lead to cinoma. A diet high in fruits and vegetables metaplasia and neoplasia of the stomach may protect against all gastroesophageal and esophagus. Evaluate their effectiveness cancers. H. pylori is a risk factor for gastric in at-risk populations. adenocarcinoma but is protective against

10 Report of the Stomach/Esophageal Cancers Progress Review Group

Rationale ethnicity, gender, and . For example, esophageal adenocarcinoma, Neoplastic progression in the stomach and which was rare two decades ago, now esophagus is a multi-decade process accounts for 60 percent of all esophageal characterized by genomic instability and the cancers. As many as 25 percent of gastric evolution of neoplastic clones, providing cancer patients receive no surgical time and targets for intervention long before treatment, even though they present at a the development of cancer. Premalignant treatable stage. These issues suggest that a conditions can be prospectively monitored critical lack of knowledge about risk factors, by endoscopic biopsy surveillance, risk reduction, treatment options and providing an unparalleled opportunity to outcomes among health care providers, understand the evolution and impact of risk patients, and the public. There is a lack of and protective factors in humans. public awareness of the scope, magnitude, Preliminary studies implicate , and premalignant stages of gastro- bile, H. pylori, diet, tobacco, NSAIDs, esophageal cancers. Specifically, there has obesity, and other exposures as risk and/or been a lack of focus on educating high-risk protective factors for these cancers, but little groups, including people with Barrett’s knowledge exists as to the molecular and esophagus, GERD, and H. pylori infection. cellular mechanisms involved. Knowledge In addition, the possible roles of alcohol, of the mechanisms that predispose people to tobacco, and diet in the etiology of these gastroesophageal cancers, especially at the cancers have not been emphasized. genomic, transcription (expression) and Furthermore, public education on risk proteomic levels, could identify novel factors, common presenting symptoms, and interventions to prevent these cancers. interventions has been inadequate to Additionally, insight could be gained that motivate the public to seek early diagnosis. may be useful in preventing or controlling Presumably, more at-risk patients would the evolution of intervention-resistant self-identify and seek treatment earlier if clones. Finally, after promising inter- risk profiles were more widely understood ventions have been identified, they must by the public, advocacy groups, and health be tested in adequately powered, well- care professionals. In addition, morbidity controlled clinical prevention trials that and mortality could decrease with well- allow prospective, tissue-based molecular developed tools, such as videos available in and cellular characterizations of response. physicians’ offices and user-friendly Web sites, to assist the educational process. Patient/Provider Education: Educate patients and their families, health care From primary prevention to survivorship professionals, and the public regarding risk and end-of-life issues, communication factors, risk reduction, and treatment empowers people to make informed cancer- options and outcomes for gastroesophageal related decisions and to engage in behaviors cancers and their precursor states. that will improve their health. To build on our progress in refining health Rationale communication theories and interventions, we must close major gaps in our Gastroesophageal cancers represent a understanding of how people access and use diverse group of malignancies, and each health information, as well as the subtype has a different risk profile. Some discrepancies between what is known and subtypes show recent rapid changes in what is practiced. incidence as well as striking variations by

Priorities of the Stomach/Esophageal Cancers Progress Review Group 11 The quality of cancer communication can be assessments with biopsies. Indeed, as enhanced by gaining a better understanding oncologic drug development orients itself of the information needs of patients, toward more molecularly targeted, families, and other decision makers involved “cytostatic” agents, serial assessment of in the choice of stomach and esophageal tissue-based response markers becomes cancer interventions. These findings can critical, particularly in early-phase clinical lead to accurate and balanced information trials intended to prioritize agents before about areas such as cancer prevention, entry into more costly phase III trials. diagnosis, treatment, and care. In addition to Knowledge from molecular inquiries during these findings, current dissemination therapy also may be useful to improve the systems for quality, evidence-based cancer accuracy and reliability of predictions education materials and the use of multi- regarding response, survival, and long-term media technologies must be evaluated and outcomes. improved. This information eventually will be applied to help people understand Therapeutic Targets: Define host and important health risks and to assist them in molecular/biologic tumor characteristics making informed choices despite exposure that will help customize treatment and best to contradictory or inaccurate health predict recurrence and/or survival. messages. Rationale Therapy: Develop and test novel therapeutics, and optimize existing Therapies for gastroesophageal cancers are treatments for gastroesophageal cancers largely empirical. Despite surgery and and their precursors, based on the adjuvant therapy, the majority of patients identification and understanding of with gastroesophageal cancers run a molecular pathways involved in substantial risk for both local and distant oncogenesis, tumor response and resistance. recurrences. There is an urgent need to identify molecular markers and pathways Rationale that confer sensitivity and resistance to existing and novel therapies. Advances in Gastroesophageal malignancies continue to genomic, transcription (expression) and be highly lethal, despite therapeutic proteomic technologies, combined with advances over the past 30 years. Although endoscopic access for biopsy of the stomach treatment has become more complex—often and esophagus, make it possible to safely involving combinations of surgery, evaluate neoplastic tissue before and after chemotherapy, and radiation—physicians therapy in a unique and unparalleled still do not know which patients benefit fashion. Insights gained from pre- and post- most from which approaches. The overall treatment tissue evaluation by a committed similarity of outcomes belies the biologic multidisciplinary team would guide and genetic heterogeneity of these development of customized therapy by malignancies and the potential for identifying markers of sensitivity or differential sensitivities to existing and resistance to existing and novel therapies. novel therapeutics. Identifying and under- Using these markers, clinical researchers standing molecular pathways involved in could better determine which patients were oncogenesis and tumor sensitivity, as well as most suitable for which therapies, as well as the evolution of therapeutic resistance, are develop therapies targeted specifically to facilitated by repeated endoscopic gastroesophageal cancers.

12 Report of the Stomach/Esophageal Cancers Progress Review Group

Markers and Molecular Profiling: Profile treatment, and include these instruments in the molecular, cellular, and epidemiologic clinical trials and observational studies. features of gastroesophageal tumors and their precursor lesions to identify Rationale diagnostic, prognostic, predictive, preventive, and therapeutic targets. Solid data exist that describe patterns of recurrences and survival in patients treated Rationale for gastroesophageal cancers, but information is limited on patients’ long-term Rapid technologic advances in genomics, functional outcome and quality of life. expression arrays, proteomics, and Patients with gastroesophageal cancers have bioinformatics, combined with the unique unique functional problems relating to both access of gastroesophageal cancers and their disease and treatment morbidity. Specific precursors to endoscopic biopsy, offer quality-of-life scales exist for stomach and innovative and exciting opportunities to esophageal cancers, but without longitudinal understand, prevent, and treat these studies across large populations, generali- malignancies. Endoscopic biopsies can be zations cannot be made. Challenges to the processed within seconds of being obtained, study of patient-centered outcomes include providing optimal material for genomic the small number of cases and high mor- expression and proteomic analyses. In tality rates. In addition, patient-centered addition, premalignant gastroesophageal issues vary across the premalignant to tissues are not typically removed according late malignant disease spectrum. Thus, to current standards of care, so molecular additional organ-specific instruments need findings may be evaluated with regard to to be created and utilized in large-scale prospective outcomes. These compre- clinical trials. Findings will inform hensive methodologies can give insight into assessments of quality-of-life and quality-of- the molecular genesis and signatures of care issues, cost of care, patient preferences, gastroesophageal cancers and their and quality of symptom management. precursors. This knowledge provides unprecedented opportunities to develop Host/Environment Interactions: Identify, molecular diagnostics for risk stratification develop, and validate genetic, biochemical, and to predict therapeutic response. and biological markers that will help uncover host/environment interactions in Such comprehensive methodologies also can esophageal and gastric carcinogenesis. identify molecular targets to prevent or treat cancer, as well as molecular endpoints to Rationale evaluate the success of interventions. New molecular markers and targets can aid in Gastroesophageal cancers encompass a developing novel, customized treatment diverse group of malignancies, each with strategies that are more effective and less identified protective and risk factors. Some toxic than existing regimens. environmental exposures appear to be risk (e.g., tobacco) or protective (e.g., a diet high Outcomes: Develop and refine disease- in fruits and vegetables, NSAIDs) factors for specific, patient-oriented methods to assess most gastroesophageal cancers, whereas quality of life, quality of care, and cost others appear to be specific for certain sites effectiveness of treatment in patients with or histologic subtypes. For example, gastroesophageal cancers and their gastroesophageal reflux and obesity precursors through all stages of disease and correlate with esophageal adenocarcinoma,

Priorities of the Stomach/Esophageal Cancers Progress Review Group 13 and H. pylori infection may even be a risk center clinical trials and comparisons among factor for gastric adenocarcinoma but a technologies are lacking. Such novel protective factor against esophageal noninvasive and minimally invasive adenocarcinoma. The mechanisms by which techniques could facilitate screening and these factors modulate cancer progression surveillance by making the procedures more are not well understood. The stomach and cost effective, acceptable, and available. esophagus are accessible to endoscopic biopsy and prospective evaluation and thus Preclinical Models: Establish models to are uniquely suited to research on understand the biology of gastroesophageal environmental exposures and host-response cancers and their precursor lesions and to relationships, which can be assessed stimulate novel prevention, diagnostic, and comprehensively using genomic expression treatment strategies. and proteomic technologies. Additionally, a new generation of molecular markers to Rationale identify at-risk patients for premalignant and malignant lesions will assist in both Few clinically relevant tumor models of developing new prevention strategies and stomach and esophageal cancers exist. stratifying patients for surveillance. Preclinical studies need well-characterized epithelial cell cultures, cell lines, xenografts, Technologies for Screening/Surveillance: and animal models that accurately represent Develop noninvasive and minimally invasive physiologically and genetically defined technologies (e.g., serum markers and stages in human gastroesophageal imaging techniques) for screening and carcinogenesis, including gastroesophageal surveillance of premalignant and malignant reflux, , H. pylori- gastroesophageal lesions. mediated progression, and gastroesophageal Rationale clonal evolution. Such preclinical tumor models could provide crucial mechanistic Gastroesophageal carcinogenesis occurs evidence for translation of advances in over decades, and the long premalignant laboratory research and host-environmental phases for these cancers provide great interactions into clinical prevention and opportunities for screening, surveillance, therapeutic trials. Preclinical tumor models early detection, and prevention. Unlike also provide an efficient use of resources to people with many other premalignant investigate the efficacy and safety of conditions, high-risk populations are therapeutic agents prior to human trials. identifiable, but many people do not seek Finally, gastroesophageal tumor models evaluation because of the invasive would be valuable assets for genomic procedures involved. Serologic markers to expression and proteomic exploration to identify persons at increased risk would identify new targets for intervention. permit cost-effective screening of the population. Novel imaging technologies for These recommendations constitute a screening (e.g., ultrathin endoscopes, scientific framework to translate bench, colorimetric devices, capsule endoscopy) bedside, and population advances into and surveillance (e.g., -induced improved care for patients with and at risk fluorescence spectroscopy, reflectance for gastroesophageal cancers. The S/E PRG spectroscopy, light-scattering spectroscopy, offers a high-priority infrastructure resource, trimodal spectroscopy, Raman spectroscopy, the Stomach/Esophageal Neoplasia optical coherence tomography, confocal Translational Research Network microendoscopy) show promise, but multi- (SENTRNet) to overcome challenges to

14 Report of the Stomach/Esophageal Cancers Progress Review Group efficient translation of these research opportunities with regard to their efficient recommendations, as described below. and effective contributions to key foundational elements of trial design. These Infrastructure: Resources include 1) better risk characterization for and Partnerships cohort identification/stratification; 2) agents/interventions with greater efficacy The primary challenge in achieving the S/E and/or safety; and 3) markers/endpoints with PRG’s recommendations is that no single greater accuracy and reliability. Once institution or organizational structure priorities are agreed upon, SENTRNet will currently possesses the resources necessary conduct laboratory, clinical, and population- to address the opportunities and challenges based studies to achieve the goals. specific to gastroesophageal cancers Population feedback will be used to refine outlined above. The rarity of these cancers interventions and improve patient care. means that no one center sees enough patients to conduct necessary clinical trials. The components of the model will include Additionally, a single center may not have experts from various disciplines and access to the tools, such as bioinformatics, institutions, a tissue repository, that may be critical to the success of trials. epidemiological data, surveillance, and the Likewise, the dearth of patients at any single technologies necessary to develop basic center prevents the development of quality- science discoveries that will translate into of-life interventions and assessment of high-quality prevention and treatment prevention strategies. Thus, it is imperative practices. SENTRNet will share scientific that NCI supports the multidisciplinary, leadership with a coordinating center that multi-institutional efforts of gastro- will utilize a business model approach and esophageal cancer research. foster linkages with academia, industry (both pharmaceutical and device-oriented), Accomplishing the priorities set forth by the consumers, and federal agencies. A mutual S/E PRG will require resources specifically dependence prototype for funding and designed to capitalize on translational collaborations will be employed to provide research opportunities and meet the incentives for productivity and rewards for challenges uniquely presented by gastro- effective partnerships and resource sharing. esophageal cancers. The progression of the disease, the late stage at diagnosis, the rarity The conceptual foundation, infrastructure, of each of these cancers, and the coordina- management strategies, and research tion of translationally oriented biomedical priorities are further described below. investigators capable of thorough evaluation of these diseases are best addressed through SENTRNet’s Conceptual Foundation the creation of a Stomach/Esophageal SENTRNet will have multi-agency, Neoplasia Translational Research Network multiinstitutional, and multidisciplinary (SENTRNet). collaborations with shared leadership. The The Stomach/Esophageal Neoplasia innovative infrastructure and unique management style are likely to enhance Translational Research Network collaborations as well as to facilitate achievement of the S/E PRG Overview recommendations.

SENTRNet (Figure 2) will be charged with In developing the idea for SENTRNet, prioritizing translational research existing institutions, consortia, and networks

Priorities of the Stomach/Esophageal Cancers Progress Review Group 15 devoted to clinical research (such as Cancer resources, and providing coordination and Cooperative Groups, Cancer Centers, the information dissemination across relevant Cancer Genetics Network [CGN], the Early research and practice enterprises. NCI is not Detection Research Network [EDRN], the in a position to bring the full scope of the Specialized Programs of Research S/E PRG recommendations to fruition alone. Excellence [SPORES], the Nonalcoholic Other potential support agencies include the Steatohepatitis consortium, and others) were National Institute of Allergy and Infectious considered and reviewed. Although each of Diseases (NIAID), the National Institute of these entities has strengths, none was ideally Diabetes and Digestive and Kidney Diseases suited to meet the unique translational (NIDDK), the National Institute of Nursing research challenges and opportunities Research (NINR), the National Institute on available in stomach and esophageal Aging (NIA), the Department of Veterans cancers. Success is dependent upon the Affairs (VA), the Department of Defense participation of gastroenterologists, as well (DOD), the Centers for Disease Control and as translationally focused interactions Prevention (CDC), the Centers for Medicare between diverse scientists working at the and Medicaid Services (CMS), and the laboratory, clinical, and population levels. In Agency for Healthcare Research and Quality addition, the proposed management system (AHRQ). The John E. Fogarty International for SENTRNet is unique. For all of these Center can be utilized to develop a better reasons, we felt that the research agenda understanding of the international impli- regarding stomach and esophageal cancers cations of these cancers and seek inter- would be best advanced by the creation of national research partners wherever SENTRNet. possible. In addition, SENTRNet will seek biotechnology, pharmaceutical, and medical Multi-Agency Partnerships and Linkages. device industry partners as well as solicit the A wide range of federal research agencies advice of the Food and Drug Administration should be invited to partner scientifically (FDA) to steer new devices and therapies and financially with NCI in this initiative as through the clinical trials process more a means of drawing on needed expertise and effectively and efficiently.

16 Report of the Stomach/Esophageal Cancers Progress Review Group

Figure 2. Schematic Diagram of the Structure and Function of SENTRNet.

Stomach/Esophageal Neoplasia Translational Research Network (SENTRNet) Foundation & Management Principles • Multi-institutional/agency/disciplinary • Synergy with existing institutions Clinical Research • Shared leadership Centers • Translational-focused • Business model • Mutual dependence • Managed progressive growth Administrative • Knowledge management Center Informatics Center Research Priorities Pathology Center • Population studies Research Products •Prevention To Improve Patient Translational Analytic • Patient/provider Care & Public Labs Center education • Therapy Health • Therapeutic targets • Mechanistic insights • Markers & molecular • Molecular targets Model- profiling • Risk measures oriented •Outcomes •New devices Human- • Host/environment • New interventions oriented interactions • New outcome measures • Technologies for • Patient & provider screening/surveillance education Patients with cancer • Preclinical models Persons at risk General population

SENTRNet will seek scientific Oncology, and others) and private cancer collaborations with other established NCI research foundations (e.g., the American programs, cooperative groups, and Cancer Society, Cancer Research consortia. For example, the SENTRNet Foundation of America) to promote Virtual Tissue Resource might partner with adequate representation of all interested already established programs, such as the parties. In addition, these affiliations will EDRN, Cooperative Human Tissue ensure that as many health care Network, Tissue Array Research Program, professionals and patients as possible will Tissue Expeditor, Tissue Locator, and have access to the new information and others. advances this partnership will provide.

Finally, SENTRNet also will seek to include Multidisciplinary and Multi-institutional. community physicians, consumer advocacy Representatives from many disciplines and groups, relevant professional associations institutions will be necessary to implement (e.g., the American Gastroenterology this comprehensive gastroesophageal Association, American College of cancers research plan, which will focus on Gastroenterology, American Society for the discovery and application of effective Gastrointestinal Endoscopy, American educational, prevention, and treatment College of Surgeons, Oncology Nursing strategies to improve patient care and the Society, American Society of Clinical public’s health. Disciplines and

Priorities of the Stomach/Esophageal Cancers Progress Review Group 17 representatives will include basic scientists, to evaluate devices and interventions more especially geneticists and cell biologists, efficiently and effectively. Finally, improved biostatisticians, epidemiologists, patient and provider education will apply pathologists, gastroenterologists, informatics these advances in a more equitable and specialists, nurses, psychologists, prevention productive manner. strategists, surgeons, radiation oncologists, medical oncologists, and consumers. These Key Elements of SENTRNet’s cross-disciplinary, multi-institutional Infrastructure/Components collaborations will be essential to facilitate timely discoveries that can improve Administrative Center. The administrative prevention, early detection, treatment, and center, under the scientific leadership of the comprehensive cancer care. The outcome of steering committee, will attend to all this collaboration could potentially reduce administrative details and coordinating the burden of disease and increase the functions. These functions will include the quality of life for those with or at risk for development and adoption of standard gastroesophageal cancers. common data elements, protocols, and informed consents; development and Shared Scientific Leadership. A steering monitoring of budgetary expenditures and committee coordinated through the product timelines; development and administrative center will provide scientific promulgation of incentives; facilitation of direction to SENTRNet to assure its efficient communication within and across the and productive translational focus. The network; promotion of industry relations; SENTRNet steering committee will consist and dissemination of information beyond of one representative from each member SENTRNet. A business administrator will institution and/or discipline, chosen to lead the center, which will follow a business reflect the diversity of research professionals model of management (described later). and advocates. Scientific priorities will be generated by the SENTRNet Steering Committee and Research Products. Clinical care is implemented by the other components. improved by the generation of high-quality reproducible data from unambiguously Informatics Center. Knowledge defined risks and benefits of tested devices, management is key to the usefulness of the drugs, and behaviors. Once these data are data generated. The network will assess the generated, they may be assimilated into market pulls and demands to determine what evidence-based recommendations and data is needed and to assess how, when, and clinical/public health practices. why this information is best translated into SENTRNet’s first priority is the generation improved practice. Internally, the data will of products that will expedite clinical trials be used to identify needs, track progress, to improve the clinical care of patients with, motivate future directions, and improve or at risk of, gastroesophageal cancers, delivery capacity. In addition, thereby improving the public’s health. bioinformatics tools will be developed for Specific research products will exploit new promoting, accessing, and adopting mechanistic insights gained in the course of evidence-based interventions. this research that will iteratively improve recognition of molecular targets and Pathology Center. The histopathologic development of new interventions. It is designation of premalignant conditions of expected that this research will suggest new the stomach and esophagus is hampered by risk and outcome measures that can be used tremendous inter- and intra-individual

18 Report of the Stomach/Esophageal Cancers Progress Review Group heterogeneity. Unfortunately, current through human-oriented and model-oriented standards of care depend upon pathologic translational laboratories. interpretations (e.g., low- versus high-grade dysplasia). For this reason, a pathology Human-oriented Translational Labs. These center devoted to histopathologic laboratories will focus on high-priority standardization is essential. Tissues research involving human specimens, subjected to molecular and cellular inquiries including both resected and pinch biopsy will first be characterized with regard to specimens. These units will develop so that accurate and reliable standard protocols for the handling and interpretations of molecular and cellular data transportation of human materials from the can be offered. moment they are obtained until they are received by the laboratories or virtual tissue Clinical Research Centers. These centers, repositories. Specimens delivered to these positioned throughout the United States, will laboratories will be subjected to a variety of serve as the essential links to persons at risk technologic inquiries, including expression for and living with gastroesophageal arrays, genetic and epigenetic assays, cancers, as well as their families. A team of assessments of cellular apoptosis and clinicians (e.g., gastroenterologists, GI proliferation, and proteomics. There are two surgeons, medical oncologists, psycholo- goals: first, to identify key aspects of human gists, nurses or clinical research associates, neoplastic progression for development as data or protocol managers) will staff the markers of risk, response, and therapeutic centers. These sites will develop clinical targets; second, to evaluate these markers as protocols; recruit, enroll, and manage the surrogate endpoints in epidemiological cohorts involved; assure the development of studies of risk and protective factors as well institutional tissue repositories; and monitor as in clinical treatment-response investi- the epidemiologic data collection to identify gations. The human-oriented translational risk and protective factors for gastro- labs will interact closely with established esophageal cancers and define the stages at NCI programs, such as the EDRN, to which they act during carcinogenesis. identify synergies and avoid overlap.

Analytic Center. The analytic center will Model-oriented Translational Labs. These assume responsibility for the development labs will improve gastroesophageal stem cell of management tools, such as electronic models; identify stem cell-specific databases and on-line study forms. promoters for use in animal models; and Additionally, they will interact with the develop, appropriate, and maintain cell lines. clinical research centers to develop clinical In addition, they will develop and maintain and population studies/protocols, empha- optimal animal models of these diverse sizing standardization of methods and tools. diseases, use available models to evaluate In addition, the analytic center will serve as preventive and therapeutic agents, and SENTRNet’s biostatistical unit, closely standardize criteria in mouse histopathology linked to bioinformatics, providing support related to premalignant and malignant in the development of studies and analyzing lesions. Previously established effective and data as they emerge. ineffective agents in humans will be back- validated by the models to definitively Translational Laboratories. A primary define them as reasonable positive or focus of SENTRNet is to move science into negative predictors of human efficacy. practice, and this can best be achieved Additionally, molecular and cellular

Priorities of the Stomach/Esophageal Cancers Progress Review Group 19 investigations will be carried out in these gastroesophageal cancers and contemporary models to identify risk markers, therapeutic models of translational research. Research targets, and response markers that may be conducted will be along the continuum from rapidly translated into similar aspects with basic to practice to population. humans. The model-oriented translational labs will interact closely with established Unique Management Strategy NCI programs, such as the Mouse Models of Concepts Human Cancers, to identify synergies and avoid overlap. SENTRNet’s concept is different from current cooperative efforts in a number of Virtual Tissue Repository. A number of ways. SENTRNet’s unique management the S/E PRG’s priorities rely on the strategy will help achieve the recommen- collection of tissues and/or blood of serially 0tions of the S/E PRG and overcome the acquired samples that can be linked to challenges of current models. SENTRNet patient demographic and medical infor- will be managed with a business model mation. The data can be used to exploit approach and a spirit of mutual dependence. genomics, proteomics, and translation of A strategy for progressive growth and the basic science findings into everyday patient utilization of technology will be integrated care. Therefore, SENTRNet will employ a throughout the components. virtual tissue bank with centralized reading as opposed to a centralized storage facility. Business Model Approach Strategy. Since 1955, NCI has established a number of The bank will be an electronic virtual tissue multi-institutional collaborative cancer repository where specimens are handled in a groups to conduct coordinated therapeutic standardized fashion and cataloged in a trials. However, gaining institutional centralized location, but specimens will be cooperation was challenging, even when housed at the individual institutions under governing boards had been established. In uniform storage conditions. Researchers addition, although cooperative groups participating in SENTRNet will have require detailed attention to complex incentives to share these tissue resources. budgets, complicated issues in personnel Tissue samples will be prioritized for use management, and multifarious structures of centrally for collaborative studies of highest tasks and authority, the groups’ scientists, impact. This component will establish clinicians, and government project officers linkages with other established tissue typically were not trained in business resources. The multi-institutional collection management strategies. A well-managed and access to tissues will facilitate new matrix organization simplifies and organizes discovery and rapid translation into daily the uses of information processing, highly clinical care of these rare cancers. specialized resources, staff, and equipment.

The management of SENTRNet is as The predicted future of limited budgets and important as its multidisciplinary, multi- high performance demands sets the stage for institutional structure in assuring its shared resources and better management of translational function. SENTRNet will participating individuals, organizations, and utilize unique management strategies that assets. Cooperation can enhance the cost to build upon the current research consortia benefit ratio; thus, partners will be required and cooperative groups, and SENTRNet will to cooperate with SENTRNet’s concept of adopt advances from other disciplines to the matrix organizational governing resolve many issues specific to structure for shared resources and better

20 Report of the Stomach/Esophageal Cancers Progress Review Group business management. Those electing to a framework, SENTRNet might reward participate in SENTRNet will receive productive collaborative research with training in general managerial skills, data additional funding and easier access to management, systems evaluation, resource needed resources. allocation and utilization, strategic planning, assessment, and monitoring. Those trained Mutual dependence strategy applied to the will evidence an overall awareness of the virtual tissue resource will prevent any interrelationship and interdependency of single institution or individual from various financial, economic, and wielding undue influence over the use of administrative considerations within a tissue. By applying the mutual dependence business environment. This training will be strategy, the Steering Committee or its required of participating investigators. designee (i.e., the Repository Subcom- mittee) will be able to prioritize utilization Mutual Dependence Strategy. SENTRNet of the tissue and types of studies, while the will address operational challenges by investigators develop and maintain the tissue encouraging a spirit of mutual dependence resource. The group will place the highest between partners. In the spirit of mutual priority on collaborative studies of the dependence, collaborations and partnerships greatest potential impact; however, an will be stressed, but each of the components individual also may utilize the tissue for will be funded and peer-reviewed separately. scientific discoveries based on his/her own The concept of mutual dependence initiative. capitalizes on the strengths and needs of SENTRNet’s partners and creates incentives Managed Progressive Growth Strategy. for collaboration. For example, by creating The intended scope of SENTRNet’s an interface through which laboratory operation is quite large; however, the group specialists gain access to patients and human will start on a smaller scale and move to a tissues, clinical researchers have direct larger effort to ensure that all participants access to laboratory expertise. Mutual are trained and ready to participate in the dependence, as opposed to individual group’s organizational structure and accomplishments, rewards those who share knowledge management systems. Training resources, encourages productivity, will be continuous as management, promotes resourcefulness, strengthens technology, and scientific advances collaborations, and provides latitude for progress. Standardized protocols and creative endeavors, especially for procedures will be established across translational progress. partnering sites to protect data quality; to produce quality analyses; and to facilitate Those who participate in SENTRNet’s the development, translation, application, matrix organizational structure must and dissemination of scientific advances into demonstrate effective business management quality care. Standardized protocols will aid of their site, sharing of resources, in ensuring aggregation of data and collaborative discoveries (according to maintaining a focus on SENTRNet’s foundational review criteria that will be collaborative intent during the managed developed and employed), and translational growth process. productivity. Decisions related to continued support will be based on these accomplish- Knowledge Management and Technology ments and not mere membership. Advances Strategy. The goals of Performance matters and is more important knowledge management are twofold: it than participation. Using this philosophy as provides a means to share information and it

Priorities of the Stomach/Esophageal Cancers Progress Review Group 21 establishes user-friendly resources to obtain typically been accumulated at single centers essential information to enhance individual or in large studies evaluating only patients and collective productivity. Knowledge who have progressed to cancer. Although management entails the storing of electronic these studies have made significant files that have been created, edited, and advances, existing approaches contain tailored to a particular need. Information critical gaps in the accumulation of technology infrastructures facilitate the knowledge necessary to translate the distribution of knowledge specific to the research advances into improved patient organizational needs for which it was care and public health. developed. The electronic management of information facilitates knowledge-based SENTRNet provides a vehicle to translate categorization of new and archived data. research results more rapidly than existing mechanisms. The patient/research SENTRNet will develop and capitalize on participant is the foundation of SENTRNet’s the daily use of a tailored knowledge research approach, and all clinical, management system that will enhance epidemiological, and laboratory data are communication capabilities of partners; directly linked to a specific endoscopy for a facilitate information sharing; provide a specific patient. These results can be more mechanism for information transfer; and rapidly generalized than those of single foster standardized data collection within the centers because of the multi-institutional partnership. In addition, the SENTRNet nature of SENTRNet. For instance: system will develop and establish technology capable of linking databases • The clinical research centers will include from various sources in an effort to speed gastroenterologists, who are presently research and to aid in accomplishing the S/E not well represented in the NCI research PRG’s priorities. and prevention community. These specialists are critical because they can Therefore, an integrated organizational provide access to tissue from their matrix infrastructure, combined with sound patients at risk for and with stomach and management, will support multi-institutional esophageal cancers. This access is and multidisciplinary research. SENTRNet essential to delineate the natural history will facilitate progress toward achieving all of neoplastic lesions, validate biomarkers 10 of the S/E PRG recommendations. for risk stratification, and understand the effects of environmental risk and How SENTRNet Will Facilitate protective factors on the premalignant Progress Toward Achieving the PRG . Recommendations • SENTRNet’s epidemiology section

within the analytic center will develop Population Studies. The unique standardized procedures and accessibility of stomach and esophageal questionnaires that can be applied premalignant conditions to endoscopic uniformly to patients in the clinical biopsy is a driving force for translational research centers at the time of research on of these organs. endoscopy. Current knowledge concerning the natural history of stomach/esophageal premalignant • Tissue obtained by endoscopic biopsy conditions, risk and protective factors that can be evaluated by the Translational modulate progression to cancer, and Laboratory to validate markers for biomarkers for risk stratification have

22 Report of the Stomach/Esophageal Cancers Progress Review Group

clinical risk stratification and provide Prevention. SENTRNet’s research platform surrogate endpoints to determine the provides a vehicle for rapidly translating mechanisms by which risk and protective hypotheses on risk and preventive factors factors modulate progression to cancer. generated from preclinical and epidemiologic studies (such as those that • These data can be related to the will be done within the model-oriented pathology center, minimizing the translational labs and the analytic center) confusion caused by significant intra- into efficient prevention trials in the clinical and interobserver variations in dysplasia research centers. The development of those diagnosis and in single-center trials will be facilitated by other components studies. of SENTRNet in several ways. Advances in the understanding of the prevalence and • Database components of SENTRNet can natural history of neoplasia (derived from be linked with existing databases, such as epidemiologic research within SENTRNet) SEER, the gastrointestinal databases of can provide solid data for power/sample size existing cooperative cancer groups, the calculations of clinical trials, thereby National Coalition of Cancer Survivor- improving their efficiency. For example: ship, and the NIDDK database. The aggregated information may fill gaps in • Validated risk markers can improve risk the information that any one database can estimates, thereby allowing intervention provide. The technology developed for trials targeted more specifically toward linking these databases also can be used persons most likely to benefit from them. for the network’s communication and bioinformatics efforts. • Tissue-based surrogate endpoints can be derived from an understanding of the Thus, SENTRNet can provide a clinical, mechanisms by which the risk and pro- epidemiologic, laboratory, pathology, and tective factors modulate carcinogenesis. analytic research infrastructure. Those • Translational Laboratories that have working within the infrastructure will be gained experience in population studies better able to estimate the prevalence of can provide high-throughput data for significant risk factors, to determine tissue-based inclusion/exclusion criteria efficiently the natural history of and surrogate endpoints in prevention premalignant disease states, and to trials. determine the mechanisms by which risk and protective factors modulate progression • The pathology center can provide as a prelude to prevention studies. In standardized interpretations to minimize addition, they will validate markers that the confounding influence of observer identify patients at high risk for progression variation in dysplasia diagnosis. to cancer, and establish or merge databases for hypothesis generation and disease • The analytic center can provide data modeling. SENTRNet’s contributions to analysis. population studies will facilitate the accomplishment of other S/E PRG priorities, As new information is gained, prevention including prevention, patient/provider strategies can be refined to improve cost education, markers and molecular profiling, effectiveness. Thus, data obtained by outcomes, and host/environmental SENTRNet population studies (such as interactions. information on the mechanisms by which NSAIDs and a diet high in fruits and

Priorities of the Stomach/Esophageal Cancers Progress Review Group 23 vegetables are protective against insights. That resource then can serve as a gastroesophageal cancers) can fuel common information platform available for randomized clinical trials to develop broad dissemination to health care providers, effective prevention strategies for these advocates, and patients via informatics links cancers. Similarly, SENTRNet population to each partner’s existing Web sites, as well studies that lead to a better understanding of as by other methods of communication. the mechanisms by which risk factors (such as gastroesophageal reflux, obesity, and H. Therapy and Therapeutic Targets. pylori) promote esophageal and gastric SENTRNet can provide an integrated, adenocarcinomas can be translated comprehensive tissue-based approach to efficiently into prevention trials. clinical trials that is beyond the capacity of any single cancer center. SENTRNet’s SENTRNet’s research platform provides a gastroenterologists can obtain endoscopic vehicle for prevention studies to synergize biopsies before and after treatment. This with other S/E PRG priorities, including collection provides a valuable tissue patient provider education, therapy and repository to be used by SENTRNet therapeutic targets, markers and molecular molecular biologists and epidemiologists in profiling, outcomes, host/environmental an effort to investigate host and tumor interactions, and preclinical models. characteristics predictive of positive and negative responses to existing and novel Patient/Provider Education. Previously, therapies. Furthermore, comprehensive critical knowledge concerning at-risk genomic expression array and proteomic symptoms, risk stratification, clinical trials, analyses of these tissues can identify novel and outcomes has been scattered among a therapeutic targets that can be investigated number of centers across the United States, in SENTRNet tumor models and used to with inadequate compilation and identify promising new interventions. The standardization of data and knowledge in a interventions then can be tested in clinical central resource. trials in an effort to improve the efficacy and safety of available therapeutic options. SENTRNet, through its centralized review, Ultimately, SENTRNet will evaluate compilation, and updating of relevant contributions of different treatment information on at-risk subsets, risk modalities for esophageal and gastric stratification, clinical trials, and patient- cancers as well as the proper integration and oriented measures, can provide a unique sequencing for optimal treatment outcomes. resource for research into the informational needs of patients with and at risk for cancer. Some therapeutic targets and treatment Once those needs are better defined, modalities are on the verge of break- SENTRNet can serve as a platform for the throughs, and they can be accelerated in dissemination to the public of knowledge clinical trials organized by SENTRNet. For concerning gastroesophageal cancers. More example, exciting new agents that specifically, SENTRNet can partner with selectively modulate the behavior of cancer NCI’s Physician Data Query, AHRQ, and cells (i.e., cytostatic anti-cancer drugs) the Centers for Disease Control and promise to be less toxic and more effective Prevention—in collaboration with than current drugs. The effectiveness of representative professional organizations, these agents, particularly in early phases of advocacy groups, and private cancer drug development, can best be demonstrated foundations—to develop a central, standard by evaluating key biologic parameters resource of disease-specific knowledge and before and after treatment in gastro-

24 Report of the Stomach/Esophageal Cancers Progress Review Group esophageal tissues. In addition, the providing characterization of tissues at effectiveness of traditional interventions is multiple stages of carcinogenesis, including less than optimal, but additional research on its treatment, SENTRNet may create patients and their tumors before and after synergy with the Cancer Genome Anatomy intervention may better explain the Project. mechanisms underlying response and resistance, thereby providing insights to Outcomes. SENTRNet’s multi-institutional, guide the development of new, more multidisciplinary team can facilitate a rich effective combinations. understanding of how people use health information and access communication Thus, SENTRNet provides a translational technologies of all kinds. SENTRNet can research model from the laboratory to evaluate clinical and outcomes data related standard practice. First, SENTRNet will to symptoms, quality of life (QoL), and determine which patients and tumors are quality of care (QOC) in people with gastric most likely to respond to existing therapies. and esophageal cancers and in those at risk Second, SENTRNet will develop novel for these conditions. Cross-sectional studies therapeutic approaches based on the using existing databases can then provide a molecular pathways of gastroesophageal more formal assessment. SENTRNet can carcinogenesis (interacting with established utilize and adapt current esophageal and NCI programs such as the Chemoprevention gastric-specific QoL and QOC instruments Agent Development Program and the Drug to measure disease stage and treatment- Therapeutics Program). Finally, SENTRNet related outcomes. SENTRNet’s will evaluate the most effective therapies in infrastructure may best address the clinical trials and move them efficiently into following: standard practice. • Many patients with stomach and Markers and Molecular Profiling. esophageal cancers present at a late stage SENTRNet will have an unparalleled that reduces their chance of survival, repository of normal, premalignant, leaves little opportunity for non-surgical pretreatment malignant, and posttreatment interventions, fosters surgical treatment malignant tissues that can be used for that renders potential disabilities, and molecular marker development ranging from increases their likelihood of co- exploratory preclinical investigations to morbidity. SENTRNet’s research marker validation to risk stratification and initiatives can address important treatment response studies. Many of these questions related to communication of activities may create synergy with existing health messages, treatment decisions, NCI initiatives and services, including the QOC, long-term follow-up, and QoL. Specimen Resource Locator, Tissue Expediter, and the EDRN. This partnership • The number of cancer survivors is platform will allow laboratory experts on expected to increase as more people stomach and esophageal carcinogenesis in undergo cancer screening, as screening one institution to collaborate with other technologies improve, and as new clinical and population-based research at therapies are introduced. While some of other institutions. These collaborations can these cancer survivors are cured of their support other S/E PRG priorities, including original malignancies, they may have population studies, prevention, therapy and health-limiting impacts and related side therapeutic targets, host-environmental effects that remain poorly documented or interactions, and preclinical models. By understood. In addition, many of these

Priorities of the Stomach/Esophageal Cancers Progress Review Group 25 individuals may be at risk for the are involved. For example, endoscopic development of new tumors. The length biopsies with standardized pathology of time projected for survival and the risk interpretations from patients with known of co-morbid conditions often exert an environmental exposures can be evaluated impact on both cancer treatment and by comprehensive genomic, expression posttreatment follow-up care. Thus, there array, and proteomic methods. These is a need for more research on the complex tissue-based data then can be identification, prevention, treatment, analyzed in the analytic center, and the post-therapeutic surveillance, and care of resulting hypotheses can be clinically tested. the broad spectrum of conditions These studies might define strategies either experienced by survivors of stomach and to avoid or reduce adverse exposures, to esophageal cancers. identify genetic susceptibility far in advance of clinical disease, to identify appropriate • Evidence suggests that some patients treatment regimes, and to take special with cancer do not receive newer, more precautions for people at high risk. effective treatments. Moreover, in some SENTRNet’s research infrastructure cases, there remains substantial provides a means to create synergy between disagreement or uncertainty about what host/environmental interaction research constitutes optimal care, especially from advances and other S/E PRG priorities, the patient’s perspective. Clearly, too including population studies, prevention, many patients face significant financial patient/provider education, therapy and difficulties and other barriers to therapeutic targets, markers and molecular obtaining appropriate and timely care. As profiling, and preclinical models. society wrestles with how to make health care more accessible to more people, it is Technologies for Screening/Surveillance. critically important to advance a Although endoscopic visualization and comprehensive research agenda that biopsy offer the potential for early detection, includes finding ways to improve existing methods are not cost effective. outcomes and the quality of the cancer Advances in imaging technology and care. Additionally, a greater under- molecular biopsy characterization have the standing of the factors that impede potential to improve risk stratification, and access, regardless of race/ethnicity, the development of serum assays can make income and geographic location, is population-based screening possible. needed. Stomach and esophageal cancers However, such advances have typically been provide an opportunity for such studies. suggested in single-center studies without rapid translation into more definitive, multi- Host/Environment Interactions. Early center investigations that would be efforts to discover how genes and necessary to define the risks and benefits of environmental factors interact to cause these technologies. Current funding cancer are showing promise, but they also mechanisms through NCI or NIDDK do not highlight the complexity of the puzzle. match the developmental potential of these SENTRNet’s research in this area can new technologies, nor do they provide for uncover elements of gene-environment adequate comparisons between novel and interactions that lead to improvements in existing approaches that have a direct impact preventing and controlling stomach and on patient care and the public’s health. esophageal cancers. SENTRNet’s research platform will accelerate discovery and SENTRNet will dramatically improve the translation because all necessary elements rapid translation of research observations

26 Report of the Stomach/Esophageal Cancers Progress Review Group into clinical and public health interventions. can facilitate the achievement of many S/E Collaborations among SENTRNet’s clinical PRG priorities. However, esophageal and research centers, translation laboratories, gastric tumor models face many challenges, pathology center, and analytic center can which have been discussed previously. In collectively offer what is not possible now. spite of these challenges, there have been The collaborations can develop a menu of limited successes in animal models. These markers for stomach and esophageal cancers include transgenic and knockout mice with and their precursor states. They can conduct alterations in APC, SMAD-4, TFF-1, TGF- multicenter, multidisciplinary studies for beta 1, RUNX3, CDX 2 cyclin D1 and validation of existing and novel biomarkers. EGFR, as well as animal models of H. pylori Finally, they can provide specified standards infection, Barrett’s esophagus, and of performance for tissue and blood esophageal adenocarcinoma, and primary collections, quality control, database cell cultures of Barrett’s esophagus. tracking, prioritization, and specimen distribution. SENTRNet can work with NCI’s Mouse Models of Human Cancer Consortium to SENTRNet clinical research centers also can create robust mouse models, using hybrid develop, compare, and validate novel techniques (genetic approaches, surgery, H. imaging techniques, including ultrathin pylori infection) to define underlying endoscopes, colorimetric devices, optical molecular mechanisms of esophageal or detection of dysplasia, autofluorescence, gastric carcinogenesis, as well as to allow Raman spectroscopy, light-scattering for preclinical evaluation of novel spectroscopy, and synchronous chemoprevention and therapeutic agents. luminescence. These devices can have a Additionally, SENTRNet can develop and profound impact on identifying the true characterize primary cells, immortalized incidence of both esophageal and gastric cells, transformed cells, organ cultures, and cancers by identifying patients at risk for the organotypic cultures for studying stem-cell disease and interrupting such conditions as biology, intestinal metaplasia, and cancer in the “dysplasia-carcinoma” sequence far in esophagus and stomach, as well as develop advance of cancer. As serum markers immunocompetent rodent models of become available, SENTRNet’s virtual advanced disease. SENTRNet also can tissue and serum repository will provide the define the genetic factors that regulate best means available to rapidly validate and epithelial cell responses to injury that lead to compare them to imaging and tissue-based esophageal and gastric cancers, but which markers in phase III studies. SENTRNet’s are difficult to study directly in humans. components will allow for the rapid comparison of endoscopic, imaging, and Conclusion molecular risk stratification for translation to improved patient care in phase IV studies. The S/E PRG has identified 10 recommen- SENTRNet is the only potentially available dations essential to making progress in vehicle by which imaging and molecular understanding and treating stomach and screening and surveillance can be evaluated esophageal cancers. The best approach to in definitive biomarker studies to demon- achieve those recommendations is through a strate reduction in mortality in patients who multi-institutional and multidisciplinary have stomach and esophageal cancers. group employing unique strategies to Preclinical Models. Valid preclinical enhance effectiveness, collaboration, and models of gastroesophageal carcinogenesis quality in a translational research enterprise

Priorities of the Stomach/Esophageal Cancers Progress Review Group 27 that fluidly moves advances between the lab, 2Jemal A, Thomas A, Murray T, Thun M. the clinic, and the population. SENTRNet is Cancer statistics, 2002. CA Cancer J Clin necessary to advance the scientific priorities 2002; 52:23–47. of the S/E PRG. 3Ries LAG, Eisner MP, Kosary CL, Hankey References BF, Miller BA, Clegg L, Edwards BK. SEER Cancer Statistics Review, 1973–1999. 1Ferlay J, Bray F, Pisani P, Parkin DM. Bethesda, MD: National Cancer Institute, GLOBOCAN 2000: Cancer incidence, 2002. mortality, and prevalence worldwide; ed Version 1.0, International Agency for Research on Cancer Press, 2001, vol 2002.

28 Report of the Stomach/Esophageal Cancers Progress Review Group

Appendix A

About the National Cancer Institute’s Progress Review Groups

Appendix A: About the National Cancer Institute=s Progress Review Groups

The National Cancer Institute (NCI) THE PRG PROCESS supports basic, clinical, and population-based research to elucidate the PRG members are selected from among biology, etiology, early detection, prominent members of the scientific, prevention, and treatment of cancers of medical, and advocacy communities and various organ sites. These research efforts from industry to represent the full spectrum have produced a substantial base of of scientific expertise required to make knowledge that, while providing a wealth of comprehensive recommendations for the new scientific opportunities that can further NCI’s cancer research agenda. The advance our knowledge and progress against membership is also selected for its ability to these diseases, also requires that the Institute take a broad view in identifying and determine the best uses for its resources. prioritizing scientific needs and oppor- tunities that are critical to advancing the To help ensure the wise use of resources, field of cancer research. NCI has established Progress Review Groups (PRGs) to assist in assessing the The leadership of each PRG finalizes an state of knowledge, reviewing the Institute=s agenda and process for a PRG Planning research portfolio, and identifying scientific Meeting. At the Planning Meeting, priorities and needs for its large, participants are identified to take part in a site-specific research programs. subsequent Roundtable meeting. Topics are identified for Roundtable breakout sessions CHARGE TO THE PRGS to which participants will be assigned and for which the PRG members will serve as Each PRG is charged to: co-chairs.

• Identify and prioritize scientific research A PRG Roundtable brings together in an opportunities and needs to advance open forum approximately 100B180 leading medical progress against the cancer(s) members of the relevant cancer research, under review. medical, industry, and advocacy communities to formulate key scientific • Define the scientific resources needed to questions and priorities for the next 5B10 address these opportunities and needs. years of research on specific cancers. As part of the process, the NCI provides the • Compare and contrast these priorities PRG Roundtable with an analysis of its with the current NCI research portfolio. portfolio of cancer research in the relevant organ site. This analysis is intended to • Prepare a written report that describes enable the Roundtable to compare and findings and recommendations. contrast identified scientific priorities with • Discuss a plan of action with NCI the research currently being done under the leaders to ensure that the priority areas Institute=s auspices. Input from the Round- are addressed. table is used by the PRG in delineating and prioritizing recommendations for research, The following section details the process related scientific questions, and resource and used to execute these charges. infrastructure needs. At its discretion, the

Appendix A: About the National Cancer Institute’s Progress Review Groups 29 PRG may solicit additional input from the is widely disseminated and integrated into research and advocacy communities through the Institute=s planning activities. At this workshops, ad hoc groups, or by other meeting, the PRG and NCI identify the means. The PRG also may consider the research priorities that ongoing NCI deliberations of previously convened expert initiatives and projects do not address. Then groups that have provided relevant cancer the PRG and NCI discuss a plan for research information. implementing the highest research priorities of the PRG. This plan becomes a blueprint THE PRG REPORT for tracking and hastening progress against the relevant cancer. After the Roundtable, the PRG=s recommendations are documented in a draft PRG reports on breast cancer; prostate report, multiple iterations of which are cancer; ; ; reviewed by the PRG leadership and PRG cancer; brain tumors; leukemia, members. The final draft report is then , and myeloma; gynecologic submitted for deliberation and acceptance by cancers; and kidney/bladder cancers, in the NCI Advisory Committee to the addition to this PRG report on stomach/ Director. After the report is accepted, the esophageal cancers are available online at PRG meets with the NCI Director to discuss http://planning.cancer.gov. the Institute=s response to the report, which

30 Report of the Stomach/Esophageal Cancers Progress Review Group

Appendix B

Breakout Reports

Guiding Principles Biology

Co-Chairs: Correa, Pelayo and Orlando, Roy C.; Peek, Richard

Participants:

Coit, Daniel Rothman, William Stoner, Gary Dawsey, Sandy Souza, Rhonda Wojcik, Brian Karpeh, Martin

Overview/Background Described below are some of the molecular Information/Barriers and cellular changes triggered by infection with the bacterium H. pylori. These will be The biology of squamous and columnar used as a model to explore other pathways epithelia is of critical importance in of carcinogenesis in the esophagus and understanding their transition to stomach stomach that occur within the context of and esophageal cancers. The carcinogenic inflammatory states. The example of H. process is driven by three main influences: pylori infection provides a paradigm of a (1) an injurious chemical or infectious agent; known initiator that, through induction of an (2) host factors that modulate the response inflammatory response, leads to neoplastic to the injurious agent(s), such as immune transformation. responses (IL-1 beta), protective molecules (MUC-1), and adhesion molecules (E- Gastric adenocarcinoma is the second cadherin); and (3) environmental factors that leading cause of cancer mortality determine the host response to the worldwide. Infection with H. pylori carcinogenic agents. The carcinogenic significantly increases the risk of the two process appears to be closely related to predominant histologic subtypes of gastric chronic active , which may adenocarcinoma. The more common type is lead to a neoplastic or a nonneoplastic intestinal-type gastric cancer. The other outcome. form is diffuse gastric cancer, which originates within a background of superficial Gastroesophageal carcinogenesis consists of gastritis and may occur spontaneously a series of events that can be viewed as without well-defined glandular structures. those that initiate and subsequently mediate Intestinal-type gastric cancer progresses genetic changes within epithelial cell DNA. through well-defined histologic steps, These genetic changes then provide a including atrophic gastritis, intestinal survival advantage for affected cells, metaplasia and dysplasia, which are resulting in . An underlying followed by frank adenocarcinoma. The first theme that unifies stomach and esophageal stage is the transition from normal mucosa cancers is that both develop in response to to chronic superficial gastritis. chronic inflammation or injury. Another common theme, exclusive of initiating The relationship between H. pylori infection factors that may not always be identifiable, and gastric cancers stems from interactions is that the mediation of disease progression between the bacterium and the host. Some of may be via inflammation-induced DNA the genetic components of H. pylori that damage, due to the production of superoxide increase the risk of cancer include the cag radicals and other products of oxidation. pathogenicity island and the vacuolating

Appendix B: Breakout Reports 31 cytotoxinVacA, the effects of which are less inflammation, and production of oxygen- well understood. Components of the cag free radicals that induce DNA damage over pathogenicity island induce (1) increased time. Animal models may mimic aspects of release of the proinflammatory cytokine IL- the process that occur in humans. For 8, and (2) profound morphologic example, after H. pylori-induced derangements of gastric epithelial cells inflammation, intestinal metaplasia may following bacterial attachment. The result from overexpression of Cox-2 (in molecular pathways that underlie these mice), which hastens the progression from events are (1) activation of NF kappa B and atrophic gastritis to intestinal metaplasia. MAP kinase signaling; and (2) translocation Other mediators of intestinal-type gastric of the CagA protein into the host cell, which cancer include exogenous causes of induces the morphologic derangements of hypochlorhydria and hypergastrinemia, such the cells themselves. The latter phenotype as vagotomy and acid-suppression therapy. mirrors mitogenic stimulation with growth factors. More virulent strains of H. pylori The paradigm for H. pylori-induced (e.g., cag-positive strains) may increase the inflammation in the stomach can also be risk of cancer by inducing a more intense applied, in principle, to esophageal inflammatory response and by mimicking squamous cancer and adenocarcinoma, the effects of growth factor stimulation. although the initiating factors are distinctly different. The initiators of squamous cancer Host polymorphisms within immune- include achalasia, which results in stasis of response genes, including IL-1-beta and ingested food; infection with human TNF-alpha, also influence disease risk. papillomavirus (HPV); and environmental Specifically, the IL-1 gene cluster contains factors such as high alcohol intake, smoking, several informative polymorphisms that can and lye ingestion. Genetic risk factors also be correlated with either increased or play a role in squamous carcinoma, decreased IL-1-beta production. Studies including the inherited disorder, tylosis, have shown that persons who possess which results in squamous cell polymorphisms associated with high levels hyperproliferation. The initiation of of IL-1-beta production have an increased esophageal adenocarcinoma is primarily risk of atrophic gastritis and gastric cancer, through reflux esophagitis-induced Barrett’s but these relationships only hold sway in H. metaplasia. pylori-infected persons. Therefore, a synergism exists between H. pylori infection Strategies for cancer protection include and host genotype. A consequence of long- removal of the initiating events (where term H. pylori infection is the development identified) and/or inhibition of one or more of hypochlorhydria, which allows the of the inflammatory mediators, such as Cox- overgrowth of non-H. pylori pH-sensitive 2, IL-1, and/or NF kappa B activation. bacteria, conversion of ingested nitrites to However, these relationships are complex, N-nitrosamines, and an increased risk of because while infection with H. pylori gastric cancer. increases the risk of distal gastric cancer, it may protect against the development of Chronic H. pylori infection also leads to esophageal adenocarcinoma. This hypergastrinemia, which can stimulate complexity underscores the importance of epithelial cell growth. Thus, multiple factors precisely identifying the mechanisms contribute to the survival of the mutagenic through which inflammation can induce the cell in an environment of genetic instability: carcinogenic cascade. hyperproliferation in the presence of

32 Report of the Stomach/Esophageal Cancers Progress Review Group

Future Research Directions Three Scientific Priorities and Rationales Future research in stomach and esophageal cancers should include the following: Priority 1: Define the genetic bacterial and host factors that regulate epithelial cell • Identification and appropriate eradication responses to injury that lead to esophageal (where possible) of initiating factors. and gastric cancer.

• Identification of novel means to suppress Rationale: It is essential to understand the the inflammatory response. For example, underlying genetic factors and molecular inhibition of NF kappa B activation may mechanisms that regulate progression to be an important strategy for cancer cancer, because only a minority of patients prevention, either by disrupting the with identifiable risk factors progress to inflammatory process or normalizing develop esophageal or gastric cancers. For imbalances in cell-cycle dynamics. gastric carcinoma, H. pylori genotypes that may augment the risk of neoplastic • Suppression of the hyperproliferative transformation need to be identified. This response, irrespective of its initiating and will help identify specific patient cohorts for mediating events. screening and treatment in the future.

• Identification of additional strain-specific Priority 2: Define the environmental and H. pylori virulence determinants, which host factors that regulate inflammatory may in turn identify persons at increased responses to epithelial cell injury in risk for gastric cancer. esophageal and gastric mucosa that lead to cancer. • Identification of specific bacterial strains in conjunction with a particular Rationale: There is increasing recognition host background to allow focused that inflammatory responses play a central therapeutic interventions, rather than role in progression to cancer. In addition, indiscriminately treating all persons who targeting inflammation may be an effective are colonized with H. pylori, because in method of preventing the development of some persons the bacterium may protect malignancy. against the development of esophageal adenocarcinoma. Priority 3: Develop more relevant animal model systems for understanding upper GI • Identification of biomarkers in Barrett’s carcinogenesis, including squamous and metaplasia to serve as predictors of adenocarcinomas of the esophagus and cancer risk. stomach.

• Characterization of the biology of Rationale: Animal models are essential to Barrett’s metaplasia as it relates to its understanding the development of gastric origination and protection against reflux- and esophageal cancers and identifying host- induced esophageal injury. factor interactions that increase disease risk. Studies of these models are a key step in the • Establishment of animal models to study process of translating basic research into the pathogenetic pathways for both more effective methods for screening and squamous and adenocarcinoma of the treatment. esophagus.

Appendix B: Breakout Reports 33 Infrastructure Needed To • Development of SPORES—the Accomplish Priorities establishment of tissue banks and cell lines that are available to all Partnership Platforms investigators.

Gastric and esophageal cancers are a global Expected Resources To Overcome problem and account for a high proportion Limitations of Previous Research and To of cancer deaths worldwide. International Capitalize on Existing Opportunities collaborations in high-prevalence populations will facilitate a more efficient • Development of gerbil and other animal and more complete understanding of the models disease and the development of more effective interventions and treatments. – Establish cell lines that represent normal cells as well as all stages of Partnership platforms for these studies cancer progression include investigations of all age groups, including children (NICHD), immunologic – Develop specific reagents, such as studies (NIAID), and collaborative studies antibodies, to be employed in gerbil of basic biology, specimen collection, and studies development of screening and prevention (NIDDK). Specific recommended initiatives – Establish H. pylori strain repositories include: – Establish tissue-specimen banks, • Multi-institutional funding initiatives including blood banks, to identify genetic factors, having as a goal the • Dual-investigator RO1’s development of a nationwide tissue bank.

34 Report of the Stomach/Esophageal Cancers Progress Review Group

Etiology

Co-Chairs: Blaser, Martin S. and Vaughan, Thomas; Nyren, Olof

Participants:

Bernstein, Leslie Fox, James Tell, Robert Chow, Wong-Ho Gammon, Marilie D. Ward, Mary H. Daschner, Phillip Mayne, Susan Weston, Allan

Overview/Background comparison and has changed little in recent Information/Barriers years. Many studies, but not all, have shown a correlation between the prevalence of Non-cardia gastric adenocarcinoma, esophagitis and esophageal squamous cell esophageal adenocarcinoma, and esophageal , but the importance of squamous cell carcinoma are each a distinct esophagitis remains poorly defined. entity with its own uniquely identified causes. Although they share a superficial The increased incidence of adenocarcinoma anatomical relationship, each is distributed has been recognized since the mid-1980s. in different populations in differing Gastroesophageal reflux is identified as a amounts. key risk factor: people with weekly reflux symptoms demonstrate a 5- to 8-fold higher Gastric cancer remains the second most risk of developing this cancer. Even more common cancer worldwide, although a striking is that persons with Barrett’s decline over the past 30–40 years has esophagus, estimated to develop in 10–15% decreased the age-specific incidence of this of people with chronic reflux, are recog- cancer in most western countries by 50%. nized as having at least 30–40 times the Due to the aging of the world’s population, incidence as in the general population. and a steep gradient in occurrence among the elderly, the actual numbers of victims In considering major secular factors in are increasing. Efforts to improve treatment cancer etiology and distribution, the have had only limited success, and the age- following relationships are of interest: specific decrease is likely attributable to decreased exposure to causal factors. Most • Fruits and vegetables (anti-oxidants) are prominent among these is infection by H. protective in all three types of cancers. In pylori, now identified as the strongest and the stomach, nearly 85% of studies have most important risk factor. revealed a decreased risk of gastric cancer with a high intake of raw Squamous cell carcinomas comprise the vegetables, and the evidence is similarly majority of esophageal cancers in the world, strong for the beneficial effects of citrus with the majority of cases occurring in fruit. The preventive effect of fruit and developing countries. Areas of China, vegetable intake is also strong in central Asia, and southern Africa have esophageal cancer. A strongly reduced extremely high mortality, but with large risk of squamous cell carcinoma has been variations in occurrence over short reported along with intake of vegetables, distances. In the U.S., the occurrence of fresh fruits, and vitamin C. Although the squamous cell carcinoma is low by data are less plentiful, studies have also

Appendix B: Breakout Reports 35 identified an inverse relationship that remains at its increased level until between fruit and vegetable consumption more than 20 years after smoking and esophageal adenocarcinoma. cessation.

• A review of literature through the early • All three cancers are more common in 1990s revealed no substantial effect of men than in women. alcohol consumption on the risk of stomach cancers. By contrast, alcohol • Low socioeconomic status correlates consumption has been identified as a with increased frequency of all three major cause of squamous cell carcinoma. cancers. The strength of this effect is That cancer is more frequent in people strongest for squamous and weakest for who drink alcoholic beverages, with the adenocarcinoma. risk depending more on mean daily intake than on the length of the habit. • All three cancers are rare before the age The risk of esophageal cancer returns to of 50. Their incidence then rises with age the level for abstainers only after 10 but never reaches a plateau. years without drinking alcohol. Alcohol consumption has no significant effect on • In the U.S., gastric cancer is more the risk of adenocarcinoma. common in non-whites than in whites. Squamous cell carcinoma occurs 6 times • Smoking tobacco contributes to the risk more frequently in blacks than in whites; of these cancers, with a 1.5–2.5-fold adenocarcinoma occurs 5 times more increase in the risk of stomach cancer frequently in whites than in blacks. among current smokers. Several studies have shown increasing risk with longer • The incidence of gastric and squamous and heavier tobacco use. For squamous cell carcinoma co-vary, while adeno- cell carcinoma, tobacco use is one of the carcinoma is often reciprocal in its major causative factors. The life-time occurrence. duration of cigarette smoking is a significant variable. For adeno- A summary of many of these secular factors carcinoma, smoking increases the risk in cancer etiology appears in Table 1.

36 Report of the Stomach/Esophageal Cancers Progress Review Group

Table: Etiology-1

Direction and Magnitude of the Association Risk/protective factor Gastric (a) Squamous Adeno (a) Fruits/vegetables ↓↓ ↓↓ ↓↓ Alcohol (b) -- ↑↑↑ -- Smoking ↑ ↑↑↑ ↑↑ Male sex (c) ↑ ↑↑ ↑↑↑ Low Socioeconomic Status ↑↑↑ ↑↑↑ ↑ Age (d) ↑↑↑ ↑↑↑ ↑↑↑ Ethnicity (e) Black ↑↑ ↑↑↑ ↓↓↓ Hispanic ↑↑ -- -- Asian ↑↑↑ ↑↑ ↓↓↓ Native American ↑↑ Unknown Unknown

(a) Excludes cardia. (b) There may be an inverse association with wine. (c) Squamous and adeno may be closer. (d) Over age 50, no plateau with age, typical of epithelial cancer. (e) Compared with whites.

Three Scientific Priorities and Conversely, there is preliminary evidence Rationales that the presence of H. pylori is associated with protection against adenocarcinoma of Priority 1: Identify and explore the the esophagus and precursor lesions. This relationship of H. pylori to gastric cancer point needs clarification because the and esophageal adenocarcinoma in terms implications are opposite to those for gastric of physiology, premalignant lesions, and cancer. interaction with other factors (e.g., age of acquisition, strain differences, ethnicity, Priority 2: Identify the causation, normal host susceptibility, and exogenous variation, and pathophysiologic conse- exposures). quences of reflux, and its interrelationship with BMI, fat distribution, and other Rationale: There is a wide body of evidence factors in the development of esophageal that H. pylori is the single most important and gastric cardia adenocarcinomas and risk factor identified for gastric cancer; their precursors. however, because of its high prevalence, it is clearly not sufficient to explain these Rationale: There is a wide body of evidence cancers. Therefore, a deeper understanding that reflux and anthropometric measures are of the steps from H. pylori acquisition to very important risk factors for adenocar- development of these cancers must be cinoma of the esophagus. Despite this determined, including the relationship with importance, our knowledge base is modifying factors. insufficient. There are important

Appendix B: Breakout Reports 37 definitional questions and questions about Infrastructure Needed To natural history and mechanisms. Accomplish Priorities

Priority 3: While independent etiologic Partnership Platforms factors have been identified in the development of esophageal and gastric Large, collaborative cohort studies could be cancers and their precursors, the priority is coordinated through such agencies as NCI to understand how these factors interact in and NIDDK. affecting the disease continuum and explaining patterns of incidence in the Expected Resources To Overcome population. Limitations of Previous Research and To Capitalize on Existing Opportunities Rationale: Existing evidence suggests that these cancers are multifactorial diseases, and • Investment in/improvement of databases reliance on approaches directed toward with emphasis on enrollment of minority single risk factors are unlikely to be populations sufficient to provide a complete explanation of etiology. • Archival specimen and tissue banking

38 Report of the Stomach/Esophageal Cancers Progress Review Group

Genetics

Co-Chairs: Meltzer, Stephen and Fennerty, Brian M.; Powell, Steven

Participants:

Chak, Amitabh Mori, Yuriko Taylor, Philip Christie, Adrian Moss, Steven Triadafilopoulos, George Henley, Donald Romero, Yvonne

Overview/Background fication. Examples of genes activated in Information/Barriers gastric and/or esophageal cancers include c- myc, c-erbB-2, and those encoding Many molecular alterations have been and EGF-R. DNA amplification occurs at described in gastric and esophageal the chromosomal loci 7q, 17q, and 20q. carcinogenesis. The genetic features of these Inactivation, another category of genetic cancers can potentially be used to develop alteration, is exemplified by E-cadherin, ways of assessing risk, improving detection, p53, , APC, hMLH1, and by genetic loci identifying prognostic markers, and strati- showing allelic loss, such as 4q, 5q, 8p, 9p, fying therapies. However, such advances 17p, 18p, and18q. Mechanisms of gene will require careful prioritization of inactivation include point mutation, allelic strategies in order to discover and validate loss, and hyper-methylation. For example, both inherited and acquired molecular hyper-methylation of the hMLH1, E- alterations in these cancers and their cadherin, and APC genes has been reported precancerous states. Three research areas in gastric epithelia and tumors (Tamura et adequately describe approaches to al., JNCI, 2000; Fleisher, A.S. et al., Cancer advancement in this field: laboratory Research, 1999, Oncogene, 2000; Tamura et research, technology, and goal setting. Each al., Oncogene, 2000). Finally, further area is described in detail below. research is needed on altered gene expression in order to establish the clinical Laboratory Research or biological significance of global gene expression patterns as well as to advance The first area—laboratory research into the understanding of the role of aberrant genetics of gastric and esophageal cancers— expression of individual genes. Genes involves investigation into several categories already known to be important in these of genetic alterations. For example, genetic cancers include those encoding Cox-2, instability is a hallmark of cancer. This iNOS, growth factors and their receptors, category of abnormality includes two and the proline-rich differentiation gene subcategories: chromosomal instability esophagin. (aneuploidy) and deficient DNA mismatch repair (microsatellite instability). This latter Another important area of laboratory mechanism involves targets that occur research is studying the molecular genetics downstream in the pathway of disordered of gastric and esophageal cancers to mismatch repair, such as TGF-beta1 RII, distinguish hereditary from somatic gene MSH-3, MSH–6, BAX, and ActRII. Genetic alterations. Although most abnormalities activation also involves several mechanisms, that have been described are somatic, some such as point mutation and DNA ampli- are altered in the germ line. One prime

Appendix B: Breakout Reports 39 example of germ line alteration is E- suggest that esophageal cancers and their cadherin in familial gastric cancer. Similar premalignant precursor lesions can be alterations in germ line abnormalities of accurately diagnosed based on molecular esophageal cancer still need to be phenotyping (Selaru et al., 2002; Xu et al., investigated. 2002).

Finally, laboratory research is also needed to A second technologic advance, assess the role of infection on the advent of instabilotyping, has been used to show that gastric, esophageal adenocarcinoma, and gastric and other cancers have a unique esophageal squamous cancers. For example, profile of mutations. This technique has also H. pylori infection is very common, yet less resulted in the discovery of several novel than 2% of those infected ultimately develop candidate tumor suppressor genes (Mori et gastric cancer. El-Omar, et al. (Nature, al., 2001; 2002). 2000) determined that common polymorphisms that exist in the population One technology in particular, SNPs, can be for the Interleukin I (IL-1) gene strongly used to assess risk of upper gastrointestinal influence, either positively or negatively, the cancers. Although studies of risk related to gastric response to H. pylori. Other studies SNPs have been limited, SNPs hold promise in Portugal and UK/Poland have confirmed as potential risk factors themselves and as and extended this concept for TNF-α, IL-10, factors that influence environmental and the IL-1 receptor. Thus, the pro/anti- exposures. SNPs under study include those cancer risk is related not only to the type of for activation (P450s) and bacterium, but to the genetically determined metabolism (glutathione s-transferase), response to the bacterium. nutrient metabolism (folate, selenium, and proteins), and alcohol metabolism.

Technology Goal Setting

Technological advances, the second research Goal setting defines the third research area area to advance gastric and esophageal and qualifies as an approach to advance the cancers, have changed both basic research understanding of gastric and esophageal and clinical investigation. Some techniques cancers by focusing investigation and have been perfected, while others have only generating collaborations. In particular, the recently been developed, but all deserve discovery of new genes and biomarkers consideration as emerging approaches to the should be emphasized and priorities set for genetic understanding of gastric and their further study. The function of novel esophageal cancers. These technologies genes should be determined by using in vitro include genomics, proteomics, bioinfor- and in vivo models. Noninvasive or matics, flow cytometry, immunohisto- minimally invasive “bedside” assays need to chemistry, comparative genomic be developed in order to translate bench hybridization (CGH), single nucleotide discoveries to the clinic. Importantly, polymorphisms (SNPs), fluorescent in situ putative or potential genetic markers must hybridization (FISH), microarray, DNA be validated, because the clinical utility and mutational screening, imaging technologies “generalizability” of these markers have not (e.g., endoscopic/whole body), methylation been definitively established. These markers discovery platforms, and systems biology need to be assessed for clinical utility, approaches (i.e., multidisciplinary technical reliability, translation into high integration). For example, bioinformatics throughput assays, and recognition of their studies based on cDNA microarray data general importance in other cancers.

40 Report of the Stomach/Esophageal Cancers Progress Review Group

Future Research Directions epigenetic alterations in GAS malignant and premalignant lesions can be determined. Gastric, esophageal Adenocarcinoma, and esophageal Squamous (GAS) cancers are Priority 2: Identify and discover novel relatively underresearched, are much more genes and biomarkers in GAS neoplasia. lethal than other cancers, have had relatively fewer candidate genes identified, and have Rationale: In GAS cancers and their had fewer opportunities for intellectual precursor conditions, little is known of the sharing than have other types of cancer. genome for genetic and epigenetic Only a small portion of the genome has been abnormalities. The majority of alterations investigated, and the identified genes have remains to be discovered. This situation not been translated to the clinic. Current stands in stark contrast to that in other funding for these cancers has not provided cancers, where many more viable candidate sufficient incentives for sharing resources genes have already been identified and across disciplines and institutions. characterized. Moreover, many of the Traditional funding mechanisms have candidate genes identified in other cancer encouraged single-center academic types have been found to be uninvolved or institutional studies, thereby limiting clinically insignificant in GAS lesions. A community patient enrollment and broader palette of genetic alterations and participation. Therefore, a novel guiding candidate genes in GAS neoplasia would not principle to address these needs is proposed: only increase the basic understanding of these diseases but also yield potential VIDA: Validate, Identify/Discover, and biomarkers for validation and adaptation Adapt. (priorities 1 and 3).

To form a large, multi-institutional, trans- Priority 3: Adapt current and future disciplinary, patient-centered, academic and technologies and biomarkers to the clinical community-based consortium of basic, arena. translational, and clinical investigators to achieve the following priorities in gastric, Rationale: In GAS cancers and their esophageal adenocarcinoma, and squamous precursor lesions, both current and future esophageal (GAS) carcinogenesis. markers must be clinically measurable. One unique advantage of GAS premalignant Three Scientific Priorities and tissues is that they remain in situ, in contrast Rationale for Each to preneoplastic lesions arising in other organ systems. Thus, proposed technologies Priority 1: Validate genetic alterations in and biomarkers should utilize this unique GAS neoplasia. advantage.

Rationale: In GAS and their precursor Infrastructure Needed To conditions, no previously identified genetic Accomplish Priorities alterations of early detection, prognostic, or diagnostic biomarkers have been reliably Partnership Platforms validated. Currently, validation strategies have been predominantly applied to small, A SWOG-like structure needs to be formed geographically localized patient populations. as an incentive to cross-institutional and Current technology permits following cross-disciplinary fertilization and subjects endoscopically and longitudinally, collaboration in GAS studies. This structure, so the clinical significance of genetic and known as VIDA (Validation, Isolation,

Appendix B: Breakout Reports 41 Discovery, and Adaptation), would be hMLH1 gene promoter in human gastric composed of a large network of cancers with microsatellite instability. gastroenterologists, epidemiologists, Cancer Res. 1999; 59(5):1090–1095. pathologists, bioinformaticists, and experts 3 in other appropriate disciplines. VIDA’s Tsuchiya T, Tamura G, Sato K, Endoh Y, purpose would be to promote large-scale Sakata K, Jin Z, Motoyama T, Usuba O, cooperation, patient enrollment, and Kimura W, Nishizuka S, Wilson KT, James biomarker validation. SP, Yin J, Fleisher AS, Zou T, Silverberg SG, Kong D, Meltzer SJ. Distinct Expected Resources To Overcome methylation patterns of two APC gene Limitations of Previous Research and To promoters in normal and cancerous gastric Capitalize on Existing Opportunities epithelia. Oncogene 2000; 19(32):3642- 3646. • A specified line item, Southwest Oncology Group-like, cooperative group 4Selaru FM, Zou T, Xu Y, Shustova V, Yin mechanism for this consortium of J, Mori Y, Sato F, Wang S, Olaru A, Shibata gastroenterologists and other specialists D, Greenwald BD, Krasna MJ, Abraham studying GAS neoplasia. JM, Meltzer SJ. Global gene expression profiling in Barrett’s esophagus and • Cooperative mechanisms that produce esophageal cancer: a comparative analysis incentives for sharing of resources using cDNA microarrays. Oncogene 2002; among community-based or academic 21(3):475–478. investigators. GAS carcinogenesis is uniquely suited to make use of this 5Xu Y, Selaru FM, Yin J, Zou TT, Shustova mechanism because of the ability to V, Mori Y, Sato F, Liu TC, Olaru A, Wang serially access these lesions longi- S, Kimos MC, Perry K, Desai K, Greenwald tudinally, as well as with detailed spatial BD, Krasna MJ, Shibata D, Abraham JM, mapping, to improve the under-standing Meltzer SJ. Artificial neural networks and of neoplastic progression in these gene filtering distinguish between global diseases. gene expression profiles of Barrett’s esophagus and esophageal cancer. Cancer • Research strategies devolving from this Res. 2002; 62(12):3493–3497. GAS human model may be applicable to cancers arising in other organ sites. 6Mori Y, Yin J, Rashid A, Leggett BA, Young J, Sims L, Kuehl PM, Langenberg P, References Meltzer SJ, Stine OC. Instabilotyping: 1Tamura G, Yin J, Wang S, Fleisher AS, comprehensive identification of novel Zou T, Abraham JM, Kong D, Smolinski cancer-related genes by large-scale probing KN, Wilson KT, James SP, Silverberg SG, for mutations in coding region micro- Nishizuka S, Terashima M, Motoyama T, satellites. Cancer Res. 2001; 61(16):6046– Meltzer SJ. E-Cadherin gene promoter 6049. hypermethylation in primary human gastric 7 carcinomas. J Natl Cancer Inst. 2000; Mori Y, Sato F, Selaru FM, Olaru A, Perry 92(7):569–573. K, Kimos MC, Tamura G, Matsubara N, Wang S, Xu Y, Yin J, Zou T-T, Leggett B, 2Fleisher AS, Esteller M, Wang S, Tamura Young J, Nukiwa T, Stine OC, Abraham G, Suzuki H, Yin J, Zou TT, Abraham JM, JM, Shibata D, Meltzer SJ. Instabilotyping Kong D, Smolinski KN, Shi YQ, Rhyu MG, reveals novel unique mutational spectra in Powell SM, James SP, Wilson KT, Herman microsatellite-unstable gastric cancers. JG, Meltzer SJ. Hypermethylation of the Cancer Res. 2002; 62(13):3641–3645.

42 Report of the Stomach/Esophageal Cancers Progress Review Group

Imaging/Technologies

Co-Chairs: Welch, Michael and Van Dam, Jacques; Fischman, Alan

Participants:

Balakrishnan, Krishna Georgakoudi, Irene Michaels, Margo Burakoff, Peter Knisley, Eric Sivak, Jr., Michael V. Dehdashti, Farrokh MacAulay, Calum

Overview/Background tissue cytopathology include immuno- Information/Barriers cytochemistry and PCR amplification.

There is a dramatic, ongoing increase in A number of studies have demonstrated the the incidence of adenocarcinoma of the feasibility and safety of endoscopy, without esophagus and gastric cardia. For both sedation, using a variety of instruments tumors, precursor lesions have been identi- including narrow diameter endoscopes. fied and are identifiable endoscopically. In Early instruments were relatively unsatis- the past 20 years, remarkable advances in factory in terms of the resolution and imaging and technologies in esophageal and illumination. Newer technology has made gastric cancers have occurred. Advances in these small diameter instruments equivalent endoscopic technology, non-endoscopic or nearly equivalent to standard endoscopes. optical imaging techniques, radiological However, the sensitivity and specificity of techniques, and other new technologies these instruments for the detection of provide opportunities to affect the course Barrett’s esophagus and cardia cancer is not of esophageal and gastric cancers. fully established. At the same time, the technology has advanced more quickly than Advances in endoscopy include new data can be accumulated from clinical trials, ultrathin endoscopes capable of being used primarily because most published data from clinically without sedation. New high- trials have been from single centers, both magnification endoscopes with or without within and outside of the United States. the use of exogenous dyes (chromo- endoscopy) provide high-resolution images Another difficulty with using endoscopy of gastrointestinal mucosa, which correlate, without sedation is that expert endoscopists in many cases, with histopathological have not acquired all of the available diagnosis. New endoscopic staging screening data. It is not known whether this technologies have advanced in just the past imaging method can be used by primary 10 years to become the state-of-the-art for care physicians as an office-based esophageal and gastric cancers. Endoscopic procedure. This would require the ultra-sonography (EUS), which combines development of additional cost-effective the benefits of medical ultrasound with the technology largely related to the processing access of endoscopy, provides unparalleled and disinfecting of the instruments. imaging and staging accuracy. The advent of EUS-guided fine needle aspiration provides A number of high-resolution adjuncts to a tissue diagnosis in the case of metastatic endoscopy are currently under development regional lymph nodes in patients with and are the focus of intense research. Referred to collectively as “optical biopsy,” esophageal and gastric cancer. Adjuncts to

Appendix B: Breakout Reports 43 these techniques include laser-induced precancerous and cancerous cells. Capsule fluorescence spectroscopy, reflectance “endoscopy” is the newest non-endoscopic, spectroscopy, light-scattering spectroscopy, non-radiological imaging technique capable trimodal spectroscopy, Raman spectroscopy, of imaging the stomach. Sometimes referred optical coherence tomography, and confocal to as a “remote” or “wireless” endoscopy, microendoscopy. The precursor for the system is a pill-sized unit containing a esophageal and cardia cancers is one camera, battery, and telemetry unit. Images proposed target for imaging by these highly from within the GI tract are transmitted to a advanced techniques. Preliminary reports receiver worn as a harness by the patient. suggest that one or more of these imaging Currently in development, advanced capsule techniques may be capable of detecting devices are capable of real-time imaging and mucosal dysplasia. This could have a propulsion within the stomach so that a profound impact on the incidence of both complete gastric examination may be esophageal and gastric cancers by performed. identifying patients at risk for the disease and interrupting the “dysplasia-carcinoma” Three Scientific Priorities and sequence. Rationales

High-resolution spiral computed Priority 1: Develop and implement tomography scans, magnetic resonance improved screening modalities that do not imaging, and FDG-Positron Emission require sedation for esophageal and gastric Tomography (PET) scanning represent cancers. nonendoscopic advances in imaging. The latter technology may rival the sensitivity Rationale: Available data from studies and specificity of EUS as a staging tool for indicate that a screening endoscopy without patients with advanced disease. With the use sedation is well-tolerated and acceptable to of neoadjuvant therapy, assessment of patients. However, these studies also response has been very important. However, indicate that as many as 40% of patients the conventional imaging modalities are decline to undergo unsedated screening quite limited in prediction of response to endoscopy. This problem needs to be therapy or in monitoring therapy. PET with addressed through the development of Flourine-18-fluoro-2 deoxy-D-Glucose improved screening modalities and (FDG), which localizes rapidly to many educational programs for physicians and human cancers following intravenous patients. Ultimately, screening of large injection, has been shown to have potential patient populations for esophageal and for use in this setting. gastric cancers will enable improved

Contrast agents have the potential to be a identification of the risk factors associated powerful adjunct to current radiological with progression to cancer, and a better imaging techniques. Contrast agents under understanding of the molecular, bio- development for use in conjunction with chemical, and morphological changes nuclear and magnetic resonance imaging are associated with the progression or regression capable of quantifying tissue metabolic of esophageal and gastric cancers. processes, angiogenesis, apoptosis, hypoxia, receptors, enzymes, and the degree of Priority 2: Develop improved imaging cellular proliferation. Stomach and techniques and contrast agents specific to esophageal cancers offer a unique stomach/esophageal cancers. opportunity, because oral delivery of contrast agents is possible to target Rationale: Using imaging, it is possible to molecular markers expressed on define a signature of cancerous and

44 Report of the Stomach/Esophageal Cancers Progress Review Group precancerous cells. Various optical imaging may have application in a specific group techniques are in development for imaging of patients in a similar way that PET is stomach and esophageal cancers; these currently being used. include trimodal spectroscopy, optical coherence tomography, and confocal • Imaging can be used to define doses of microendoscopy. In the future, clinical therapeutic agents. For example, optical comparisons and evaluations will need to be techniques can be used to define thera- conducted. Due to cost and the high level of peutic doses of agents for photodynamic technology required, it will be necessary to therapy based upon the activity of the partner with industry to continue the optical agent. Therapeutic drugs can be development of suitable instruments and labeled with optical or nuclear probes, equipment. and the pharmacokinetics of uptake and biological effect can be quantified prior Priority 3: Develop and evaluate imaging to determining the dosing of the techniques to define or predict responses to therapeutic agent. new therapies. Infrastructure Needed To Rationale: Using imaging, it is possible to Accomplish Priorities predict therapeutic response soon after initiating therapy or to tailor dosimetry to Partnership Platforms the physiology of the individual patient to optimize its effectiveness. Early diagnosis • Partnerships must be developed between will avoid the morbidity and expense the various groups developing imaging associated with ineffective treatments. As techniques and applying them. new therapies evolve, innovative probes and techniques would be adopted for monitoring • Individuals developing optical probes therapy. and agents (in gastroenterologists and engineers) need to partner with As examples: radiologists and imaging scientists in developing CT, MRI, and PET • In the case of esophageal cancer, techniques and probes. persistent FDG uptake in a lesion after treatment with radiation could represent Expected Resources To Overcome either residual tumor or inflammation Limitations of Previous Research and To induced by radiation. Development of an Capitalize on Existing Opportunities agent to distinguish between these states would be a major advance. • Current advances in imaging have been largely due to ad hoc collaborations • Local and systemic disease evaluation. between physician scientists and Specifically, the use of PET with FDG physicist/engineers. Such interdisci- (or new radiotracers) as a tool to evaluate plinary collaborations should be the effectiveness of new therapeutic encouraged. strategies. • Establish and fund imaging centers of

excellence for laboratory and clinical • Local disease evaluation. Optical research. techniques such as Optical Coherence Tomography (OCT), which provides • Foster interdisciplinary research by cross-sectional images of tissues in situ, creating RFAs that support research

Appendix B: Breakout Reports 45 conducted by collaborations between • Encourage increased collaboration radiologist/molecular biologists, between scientists in academia and gastrointestinal endoscopists/engineers, industry in the development of imaging and/or applied physicists and other instrumentation and probes. One method scientists. to encourage such collaboration is through the SBIR/STTR grant • Support the training and education of mechanisms. patients and physicians regarding the benefits of screening for precursors of esophageal and gastric cancers in patients at risk for the disease.

46 Report of the Stomach/Esophageal Cancers Progress Review Group

Outcomes/Education/Communication/Quality of Life

Co-Chairs: Brooks, Jo Ann; Holland, Jimmie and Provenzale, Dawn; Lipscomb, Joseph

Participants:

Bloom, Bernard Helft, Paul Tepper, Joel Donaldson, Molla Hornbook, Mark Valentine, Janet Frazzitta, Bart Rabeneck, Linda Hayman, James Rowland, Julia

Overview/Background Cancer (EORTC QLQ-C30) consists of a Information/Barriers core module covering physical, emotional, and social aspects, accompanied by a Each year in the United States, there are disease-specific set of questions for 21,600 new cases of gastric cancer and esophageal (OES24 module) and gastric 12,400 deaths. Survival rates for esophageal (STO 22 module) cancers. The Functional cancer are worse, with 13,100 new cases Assessment of Cancer Therapy – General annually and approximately 12,600 deaths. (FACT-G), the U.S. counterpart to the While studies of diagnosis and treatment EORTC, addresses both generic and disease- have focused on the traditional outcomes of specific issues. The disease-specific issues disease-free survival and tumor progression, are investigated through the esophageal and there have been few studies focusing on gastric subscales. For both instruments, the patient-centered outcomes directed toward general components are relevant to a broad health-related quality of life (HRQoL), range of patients with cancer. Both patient care experiences, symptom instruments have provided rich information management, and total economic burden. in the general core areas through a large Education for the public and professional number of studies. However, few studies sectors regarding both cancers in regard to have used these instruments in patients with their presenting symptoms, interventions, and stomach and esophageal cancers, so disease- treatment options has been inadequate. specific information on HRQoL is severely Furthermore, since many patients are lacking. Additionally, HRQoL issues, diagnosed in late stages of stomach and including the ability to perform activities of esophageal cancers, there are virtually no daily living, to work, and to attend school, as outcome studies in patients identified in well as common symptoms resulting from premalignant phases. disease and treatment have not been among outcome variables in large-scale clinical There is an increased symptom burden trials and observational studies. resulting from both disease and treatment in advanced stages of stomach and esophageal Currently, a range of validated symptom cancers. Currently, only two well-validated identification and management instruments disease-specific instruments exist for exists for measuring pain, , fatigue, measuring HRQoL in stomach and anxiety, depression, and , but none esophageal cancers: the EORTC QLQ-C30 of these instruments has been studied instrument and the FACT-Gastric and FACT- specifically for stomach and esophageal Esophageal instruments. The European cancers. Nor do these instruments address Organization for Research and Treatment of differences in terms of ethnicity, culture,

Appendix B: Breakout Reports 47 race, gender, or health literacy, which would those with Barrett’s esophagus, GERD, and potentially influence responses to HRQoL H. pylori infection. For others, the possible questions. These instruments have generally role of alcohol, tobacco, and diet in these been used in studies for as cancers has not been emphasized. opposed to earlier stages of care, and they Furthermore, public education on risk have been used for multiple cancer sites but factors, common presenting symptoms, and not specifically for gastric and esophageal interventions has been inadequate to motivate cancers. Additionally, these instruments have the public to seek early diagnosis. Finally, not been integrated into clinical trials and the public has not received information about observational studies on a large-scale basis. the possible treatment options available for Thus, little is known about symptom premalignant disorders. management for stomach and esophageal cancers. Barriers to the study of patient-centered outcomes in stomach and esophageal cancers A number of evidence-based interventions relate to the small number of cases, the high have been tested for the management of mortality rate, and the consequent need for symptoms (e.g., pain, dysphagia) associated multidisciplinary, multicenter studies to with disease and treatment of gastric and generate an adequate sample size. Studies are esophageal cancers. However, few clinical also constrained by changes in the incidence trials have employed these interventions in and demographic characteristics of patients patients with these specific cancers. who develop these cancers. In addition, Additionally, a body of science building on patient-centered issues are very different HRQoL issues is nonexistent. Thus, across the spectrum of disease stages, from information on best practice interventions premalignant to late malignant. Finally, there and management of symptoms is lacking. is a lack of evidence-based information on the efficacy of early diagnosis and treatment In terms of education, the evidence and of these cancers. These factors all serve as consensus-based treatment guidelines for barriers to public education. esophageal and gastric cancers have not been uniformly disseminated and implemented. Thus, the quality of care (QOC) varies TWO SCIENTIFIC PRIORITIES AND widely by geographic region, socioeconomic RATIONALES level, and provider preferences. In addition, there has been a lack of an interdisciplinary Priority 1: Evaluate clinical and patient/ approach that includes not only chemo- family data available for patient-centered therapy, radiation, and surgery, but also outcomes related to symptoms, HRQoL, and management, pain control, counsel- QOC in gastric and esophageal cancer. This ing, and concern for psychological, social, should include assessment of the aggregate existential, and spiritual issues. For those economic burden of these cancers patients who survive for longer periods, there considering years of productive life lost, has not been adequate attention to physical negative impacts on survival, cost of care and psychosocial rehabilitation measures. and caregiving, and other factors. Then conduct a cross sectional study using exist- For the general public, there is a lack of ing databases to provide a more formal public awareness of the scope and magnitude assessment. Importantly, all clinical trials of stomach and esophageal cancers, and their and observational studies should include premalignant and preventive aspects. patient-centered outcomes for HRQoL and Specifically, there has been a lack of focus QOC. on educating high-risk groups including

48 Report of the Stomach/Esophageal Cancers Progress Review Group

Rationale: There is a lack of data on patient- • ASCO symptom management to add centered outcomes related to symptoms, modules HRQoL, and QOC. • Partnering among agencies Priority 2: Utilize and adapt current esophageal and gastric-specific HRQoL and • American Cancer Society QOC instruments to measure disease stage and treatment-related outcomes. • Quality of Cancer Care Committee Tools for developing and testing care experiences—including pain control, • Professional organizations–thoracic treatment options, effects of therapy, and surgeons, general others—are critical for measuring patient- centered outcomes. Ensure that patient • Recommend that patient advocates be advocates are involved in all aspects of involved across the spectrum evaluation and development of measures of patient-focused issues and QOC. • Foundations (e.g., The Robert Wood Johnson Foundation, Pew Memorial Rationale: Current instruments have not Trust, Kaiser Family Foundation, and been widely used and tested in observational others) studies and clinical trials. Patients bring a unique perspective regarding their • Veteran’s Health Administration experiences and responses concerning these diseases. Expected Resources To Overcome

Limitations of Previous Research and To Infrastructure Needed To Capitalize on Existing Opportunities Accomplish Priorities • Multimodality and multicenter clinical Partnership Platforms trials and observational studies

• New health Web sites • Well-defined cohorts

• Pharmaceutical firms • Availability of databases (e.g., SEER,

Medicare, etc.) • Academic and high-volume institutions

Appendix B: Breakout Reports 49 Predictive and Prognostic Markers

Co-Chairs: Ajani, Jaffer A. and Hamilton, Stanley R.

Participants:

Conner, Jerry Lugo, Tracy Shibata, David Kumar, Rakesh Okunieff, Paul Srivastava, Sudhir Lawrence, Theodore S. Selaru, Florin M.

Overview/Background normal tissue. Molecular markers may help Information/Barriers alleviate the problem of widespread tissue damage by allowing physicians to test A predictive marker is an indicator of tumors for susceptibility to these toxins. In response to therapy, preferably defining a addition, biological modulators of toxicity patient’s survival after treatment. A are becoming clinically available by prognostic marker is an indicator of the identifying prognostic markers and natural history of the disease, and it is used integrating them into clinical staging for to help define patients with high and low better prediction of the natural history of risks of death that result from the inherent carcinomas and premalignant conditions. heterogeneity of the disease process. There are currently no well-defined These markers are applicable to screening pathways for validating predictive and patient populations with normal risk, prognostic markers and incorporating them surveillance of patient populations at into routine clinical practice. Gene arrays of increased risk, diagnosing symptomatic Barrett’s esophagus tissue indicate that the patients, determining disease stage, and number of molecular pathways and targets prevention strategies. These principles are are limited. Since esophageal tumors are applicable to adenocarcinomas of the similar to other cancer tumors, esophagus esophagus, the esophagogastric junction, cancers could finally provide a window on and the stomach as well as to esophageal these markers. Finally, a clear infrastructure squamous cell carcinoma. These principles does not exist that permits multicenter also are applicable to premalignant routine molecular correlative studies of conditions, such as Barrett’s metaplasia. predictive and prognostic markers in clinical trials of esophageal and gastric cancers. The era of molecular biology has yielded a Gastric and esophageal cancers affect many; plethora of potential predictive and however, there is little information from prognostic markers, but none has achieved multicenter trials. routine clinical use because methods for assessing response or prognosis are neither Three Scientific Priorities and standardized nor correlative. Applying Rationale for Each individual predictive markers is complicated by the use of multimodality therapies, which Priority 1: Develop rapid, flexible, and employ chemotherapeutic agents with adequate funding mechanisms for differing mechanisms of action, biological cooperative groups and institutions effects, and radiotherapy. In addition, these conducting clinical trials. They will focus therapies often cause substantial damage to on engaging in real-time collaborative

50 Report of the Stomach/Esophageal Cancers Progress Review Group studies of existing markers for validation of Priority 3: Identify specific and laboratory methodologies, tissue and blood quantitatively valid molecular pathways collection, quality control, database involved in oncogenesis, tumor response, tracking, prioritization, and specimen tumor progression, and normal tissue distribution, with minimum standards of tolerance. performance. Rationale: Premalignant and malignant Rationale: The NCI research portfolio progression can be identified for squamous heavily emphasizes clinical trials and and adenocarcinoma of the esophagus and of biologic studies. Large numbers of patients the stomach, and these organs allow for with esophageal or stomach cancers are access to tissue both preoperatively and after enrolled in therapeutic trials, but therapy. Insights would be valid for much of translational studies of correlative markers adult cancer, amplifying the impact for are not constant features of the protocols. understanding and treating cancer. In Current funding mechanisms for such malignant tumors, targets for therapy (such translational studies are out of phase with as enzymes, receptors, genes, and proteins) clinical trial development and have also been described; however, their implementation. There is an urgent need to clinical implementation and validation of support cooperative groups and institutions methodology is lagging. Additionally, conducting clinical trials for marker studies. reducing toxicity should be a major goal, given the limited success of existing Priority 2: Development and validation of therapies. novel methods for both clinical and operative molecular staging, particularly by means of molecular markers and imaging Infrastructure Needed To techniques (molecular and/or functional). Accomplish Priorities

Rationale: Despite curative surgery and Partnership Platforms adjuvant therapy, the majority of patients with gastric and esophageal cancers suffer • Multicenter trials from both local and distant recurrence. Conventional staging techniques, • Partner with insurance companies to particularly T and N, do not adequately explore endoscopic screening as a predict heterogeneity of patient outcomes. preventative measure, similar to Japan An integrated molecular staging might predict patient outcomes more accurately. Expected Resources To Overcome Novel molecular staging techniques may Limitations of Previous Research and To assist in guiding operative treatment, (e.g., Capitalize on Existing Opportunities extent of surgery/lymphadenectomy, use of intraperitoneal chemotherapy and/or • Coordinator for multicenter trials biologic therapy) as well as the use of appropriate adjuvant chemotherapy, • Tissue banks need to be reorganized to , and biologic therapy. This include untreated tissue advance will establish a novel panel of prognostic markers that will supersede the • Expedited grant process current parameters.

Appendix B: Breakout Reports 51 Prevention

Co-Chairs: DeMeester, Tom R. and Goodman, Karen J; Falk, Gary

Participants:

Fontham, Elizabeth O’Toole, Liam Tell, Robert Forastiere, Arlene Richmond, Ellen Wu, Anna Lines, Stephen

Overview/Background is difficult to confirm when studies do not Information/Barriers report results separately for these two types. Difficulties in determining whether a cancer The decline in gastric cancer has been in this region originated in the stomach or viewed as a public health triumph in the esophagus may contribute to this problem. United States, yet it is still an important Little research in the United States has cause of death for high-risk groups. focused on the two major subtypes of gastric Although the incidence of esophageal cancer carcinoma, intestinal and diffuse, perhaps in the United States is even lower than because the distinction may not be recorded stomach cancer, it accounts for slightly more routinely and is therefore frequently deaths each year due to dismal survival unavailable. There has been minimal rates. Thus, consideration of prospects for investigation of rare subtypes such as gastric prevention is warranted. Specifically, have lymphoma. the high-risk groups been appropriately and fully identified, and what are the most The more common stomach and esophageal effective means for reducing their risk? cancer subtypes (non-cardia gastric carcinoma and esophageal squamous cell Prevention research on stomach and carcinoma) occur most frequently in esophageal cancers in the last decade has populations of low socioeconomic status. focused on four disease subtypes: non-cardia Esophageal, and perhaps cardia, and cardia adenocarcinoma of the stomach; adenocarcinomas appear to be increasing in adenocarcinoma; and squamous cell more affluent populations. The major shared carcinoma of the esophagus. Nearly all risk factors for non-cardia gastric carcinoma stomach cancers are adenocarcinomas; and esophageal squamous cell carcinoma are however, recent research has revealed low intake of fruits and vegetables and potentially different etiologies according to (although the effect of whether the site is the cardia or more distal smoking appears stronger for esophageal (“non-cardia”). Adenocarcinomas of the cancer). esophagus and gastric cardia have been increasing in some population groups at Other major risk factors for esophageal alarming rates. Some studies group squamous cell carcinoma are alcohol esophageal adenocarcinomas with cardia consumption and nutrient deficiencies. adenocarcinomas, in part because the Suspected modifiable risk factors, for which available number of cases is often small. current evidence is less convincing, include Emerging epidemiologic evidence suggests hot food and drink, pickled vegetables, that esophageal and cardia adenocarcinomas moldy food (mycotoxins), nitrosamines, and may have a common etiology; however, this human papillomavirus. Other major risk

52 Report of the Stomach/Esophageal Cancers Progress Review Group factors for non-cardia gastric carcinoma are Gastric carcinoma of the intestinal type is H. pylori infection and a high intake of associated with identifiable premalignant preserved (salted, pickled) foods. lesions, such as atrophic gastritis, intestinal metaplasia, and dysplasia. Diffuse-type Risk factors for cancers of the gastric carcinoma has not been linked to gastroesophageal junction are beginning to identifiable premalignant lesions. Population emerge from recent research. These screening for detection of premalignant subtypes are associated with gastric lesions or early invasive cancers is gastroesophageal reflux. Suspected not considered cost-effective for low-risk modifiable risk factors for adenocarcinoma populations. This approach has not been of the esophagus include obesity, high advocated in the United States, where there intake of fat and vitamin A, low intake of has been little evaluation of its cost- fiber, smoking, and perhaps alcohol intake. effectiveness in high-risk groups. Population Suspected modifiable risk factors for cardia screening has been used successfully in adenocarcinoma are similar. The evidence Japan, where gastric cancer rates are high. for risk factors specific to cancers of the gastroesophageal junction comes from a Squamous cell carcinoma of the esophagus small body of studies; therefore, the risk is associated with identifiable premalignant patterns need to be confirmed in more lesions such as chronic esophagitis, atrophy, extensive research. and dysplasia, but the predictive value of cytology has not been considered adequate Squamous cell carcinoma of the esophagus for population screening. Screening trials in is considered preventable through reductions high-risk populations in China have had in smoking and alcohol consumption, and equivocal results. Esophageal improvements in basic nutrition. For non- adenocarcinomas are generally preceded by cardia gastric carcinoma, perhaps the reflux esophagitis and Barrett’s esophagus. greatest promise is in interventions aimed at eliminating H. pylori infection through Currently, more than 26 NCI projects are treatment or immunization. However, related to the prevention of stomach and research suggests that a vaccine will not be esophageal cancers. These include projects available in the near future. Meanwhile, on dietary interventions for general cancer studies have focused on evaluating the cost- prevention (2 projects); cancer awareness effectiveness of screening for H. pylori for minority populations (5 projects); infection and treating those who are screening for early detection of esophageal positive. Given concerns about potential cancer in China (1 project); dietary adverse consequences of H. pylori treatment intervention for preventing disease of asymptomatic individuals, there have progression in Barrett’s esophagus patients been calls for intervention trials to assess (1 project); laboratory research focused on benefits relative to adverse effects. mechanisms of carcinogenesis and potential Chemoprevention trials also need to be chemotherapeutic agents (4 projects); human evaluated for cost-effectiveness. Most are chemoprevention trials for the esophagus aimed at preventing progression of and stomach in China (1 project) and for the premalignant lesions through combinations stomach alone in Colombia and Mexico (2 of micronutrient supplementation and H. projects); improving cost-effectiveness of pylori eradication. Recent studies have strategies for early detection of esophageal shown that and garlic may have adenocarcinoma (1 project); identification of protective effects. biomarkers of risk of disease progression in Barrett’s esophagus (1 project); identifying

Appendix B: Breakout Reports 53 prognostic markers and improving treatment appropriateness of plans for screening and outcomes in stomach and esophageal cancer treatment (within the context of risk-profile patients (3 projects); and observational models), and (b) provide guidelines for studies to identify modifiable risk factors or assessing effectiveness of prevention potential preventive agents (5+ projects). measures in low-, middle-, and high-risk groups. There are three major barriers. First, there are small numbers of cases of these cancers, Priority 3: Develop cost-effective which limits subgroup analysis. Second, prevention strategies for reducing mortality there is a lack of uniformity in classifying of stomach and esophageal cancers. neoplasms by subsite/subtype, particularly regarding the location of adenocarcinomas Rationale: Prevention research needs to proximal to the gastroesophageal junction. weigh the costs and benefits of intervention Finally, sampling variability in ascertaining at three levels: preventing disease onset by intermediate endpoints leads to classification modifying risk factors, early detection of errors. asymptomatic disease, and minimizing potential adverse consequences of treatment. Three Scientific Priorities and Again, without increasing the power of Rationales studies through broad collaborative net- works, the cost-effectiveness of screening Priority 1: Develop risk profile(s) for and treatment plans cannot be assessed stomach and esophageal cancers that serve accurately. In particular, controversy over as a basis for subsequent intervention the need for and effectiveness of H. pylori (similar to the GAIL model for breast screening and treatment cannot be resolved cancer). until research examines the costs, including adverse effects, and benefits of such an Rationale: Some known risk factors can be approach. used to develop preliminary risk profiles. However, more etiological and Infrastructure Needed To epidemiological information is needed. The Accomplish Priorities goal is to define the populations at risk so prevention efforts can be targeted Partnership Platforms effectively. Due to the relatively low incidence of these cancers, accurate • Partnership with NIDDK (National estimates of risk require research efforts that Institute of Diabetes and Digestive, and involve a broad collaborative network across Kidney Diseases) institutions and geographic regions to develop and maintain a comprehensive • Pharmaceutical companies (these exposure database, to increase the statistical already share some structures for power of studies, and to develop uniform collaborative networks, but what is questionnaires and diagnostic classifications. missing is uniting under a common agenda) Priority 2: Develop a menu of biomarkers of risk for stomach and esophageal cancers • Academic centers and their precursor states. • Patient advocacy groups Rationale: Intermediate disease endpoints should be established that can (a) inform the

54 Report of the Stomach/Esophageal Cancers Progress Review Group

Expected Resources To Overcome • The validity of new prevention research Limitations of Previous Research and efforts will be maximized through To Capitalize on Existing Opportunities collaborations that allow the use of uniform methods in study design, • Infrastructure: All of the priorities and conduct, and analysis. collaborative efforts outlined above depend on establishing a broad • The development and maintenance of collaborative network for increasing the shared databases will maximize the power of research through large, efficiency of the research. multicenter studies. Given that numbers of cancer cases at any given institution • Patient advocacy group stimulation can are low, potentially useful information support research efforts. Such groups about risk factors, surveillance, and capitalize on patient involvement to outcomes is dispersed and isolated rather provide information for databases. Many than aggregated. Existing data about patients with Barrett’s esophagus are screening, early detection, intervention, motivated and compliant. Creative and social, behavioral, dietary, approaches are needed to stimulate microbiological, or genetic predictors advocacy for stomach and esophageal must be aggregated for increased cancer subtypes that occur primarily in statistical power. hard-to-reach populations.

Appendix B: Breakout Reports 55 Surveillance/Databases

Co-Chairs: Levine, Douglas and Sampliner, Richard, Sharma, Prateek

Participants:

Blount, Patricia Queirolo, Lewis Spechler, Stuart Cameron, Alan J. Sandler, Robert Lieberman, David Sonnenberg, Amnon

Overview/Background duration of follow-up necessary to document Information/Barriers effective screening and surveillance.

Intestinal metaplasia is the premalignant Compared to esophageal adenocarcinoma, lesion for adenocarcinoma of the esophagus the incidence of gastric adenocarcinoma is and stomach. In the United States, at least twofold higher. The incidence of esophageal adenocarcinoma has the most proximal gastric cancer was rising rapidly rapidly rising incidence of all cancers in before recently leveling off. Barrett’s White males. Current screening and esophagus is associated with and may be the surveillance for esophageal adenocarcinoma premalignant lesion for this cancer. The is based on endoscopic biopsies and the incidence of distal gastric cancer has histologic confirmation of Barrett’s decreased dramatically as has the prevalence metaplasia. If Barrett’s metaplasia is of H. pylori infection. H. pylori infection diagnosed, follow-up endoscopic biopsy results in a sequence of mucosal changes, surveillance for evidence of dysplasia and/or including intestinal metaplasia, which can early adenocarcinoma is warranted. No lead to gastric adenocarcinoma. prospective evidence exists that shows that screening and surveillance of Barrett’s The overall incidence of esophageal esophagus reduces the mortality from squamous carcinoma is approximately the adenocarcinoma of the esophagus. Most same as that of esophageal adenocarcinoma, patients who develop esophageal but it is at least twice as high in Black males adenocarcinoma are unaware that they have than in White males. A premalignant lesion Barrett’s esophagus, and they are not in a for esophageal squamous cell carcinoma is surveillance program. not recognized in the United States; how- ever, in China, dysplasia, diagnosed by non- Barriers to effective screening and endoscopic brush cytology, is commonly surveillance of Barrett’s esophagus include recognized prior to the development of the expense of screening the at-risk squamous cell cancer as part of mass population for the presence of Barrett’s population screening. esophagus; the expense of endoscopy with biopsy in low-risk Barrett’s patients; the The major barrier to screening and sur- technical difficulties of performing intensive veillance of these cancers is the lack of a systematic biopsy protocols and targeting well-defined, precancerous condition that is small areas of endoscopically invisible endoscopically visible. An important need is dysplasia or cancer; the inter-observer to identify patients at risk for gastric disagreement in the reading of dysplasia; adenocarcinoma and esophageal squamous and the large numbers of patients and long cell carcinoma who might benefit from

56 Report of the Stomach/Esophageal Cancers Progress Review Group screening and surveillance procedures. Priority 2: Conduct population-based There is the global issue of what magnitude screening to identify patients at high risk of increased cancer risk warrants screening for premalignant disease states. and surveillance. Rationale: The prevalence of premalignant NCI funding addresses two areas that may disease states in the general population is improve the clinical effectiveness and cost- currently undefined. Therefore, risk effectiveness of screening and surveillance stratification criteria to increase cost- strategies for Barrett’s esophagus. These effectiveness of screening have not been areas include the optical detection of developed. Endoscopy is currently the dysplasia by such methods as screening method of choice; however, newer autofluorescence, RAMAN spectroscopy, technologies that are more effective, more light-scattering spectroscopy, and cost-effective, better tolerated, and safer synchronous luminescence. Development of need to be developed and promoted. biomarkers in tissue samples—including array analysis, peptides, DNAploidy, p53 Priority 3: Establish or merge databases mutation, Cox-2 expression, angiogenesis for hypothesis generation and disease factors, retinoic acid receptor, iNOS, and modeling to understand the natural history telomerase—are being evaluated. NCI is of these diseases. also supporting an epidemiologic study of Barrett’s esophagus. Rationale: More detailed databases will improve the identification of at-risk Three Scientific Priorities and populations as well as assist in assessing Rationales costs of disease management and impacts on the quality of life of patients with Priority 1: Establish a clinical research premalignant conditions and cancer. Access infrastructure with multispecialty and to greater information can help guide the multi-institutional centers to perform development of methodologies for clinical surveillance to determine the natural trials and protocols. Finally, more history of the premalignant disease states. information from more sites will increase the generalizability of the findings. Rationale: Current studies do not allow risk stratification of patients with premalignant Infrastructure Needed To disease (epidemiologic and biomarkers), Accomplish Priorities given the lack of validated prognostic markers. They do not provide statistical Partnership Platforms power for hypothesis testing, given the lack of sufficiently large sample sizes with • Link independent research centers for adequate clinical outcomes. Finally, current expanding the databases tissue repositories are not linked to the clinical databases and actively managed • Take advantage of multidisciplinary patient populations in an effort to translate expertise for research on natural history basic research into clinically relevant of disease, new technologies for information. An infrastructure of this screening, and surveillance magnitude and detail would correct for all of these. • Validate prognostic markers

Appendix B: Breakout Reports 57 Expected Resources To Overcome • Centralized tissue banks, pathology Limitations of Previous Research and To readings, and biomarker assessment Capitalize on Existing Opportunities • Standardized disease classification, data, • Patient cohorts and tissue collection

• Multidisciplinary expertise (intellectual synergies)

58 Report of the Stomach/Esophageal Cancers Progress Review Group

Therapeutics

Co-Chairs: Castell, Donald O; Hundahl, Scott and Rothenberg, Mace

Participants:

Bowersox, Jon Jatoi, Aminah Patterson, Reese Govindan, Ramaswamy Leichman, Lawrence Willett, Christopher Haller, Daniel G.

Overview/Background Further, sub-optimal surgical treatment Information/Barriers appears common. Clearly, strategies are needed to enhance the safety and efficacy of Even though the esophagus and stomach surgical treatment. share close proximity, cancers of these organs are distinct diseases. They differ both Recently, the use of adjuvant postoperative in etiology and in their reaction to therapy. In chemoradiation resulted in about 40% addition, esophageal cancer comprises two survival, double the survival rate of surgery alone (Macdonald et al., 20014). A recent different types: squamous and 5 adenocarcinoma. Nevertheless, both surgical analysis (Hundahl et al., 2002 ) esophageal and gastric cancers require a revealed that most cases had a high likely- coordinated, interdisciplinary approach to hood of residual regional disease that could therapy. have been addressed by the surgeon; an index of residual regional nodal disease proved a Stomach cancer. Worldwide, stomach significant independent predictor cancer accounts for 9.9% of all reportable of survival. Patients with a low residual- cancers, ranks as the second most frequently disease index displayed a 60% survival rate reported , and is responsible for compared with 25% for the rest of the group. 12.1% of all cancer deaths (Parkin et al., Importantly, adjuvant chemoradiation 19991). In the United States, stomach cancer appeared to improve survival in all surgical- was the most common solid tumor in the pathologic subgroups. Enhanced local- early 1900s, but it is now relatively regional treatment appears to enhance uncommon, accounting for less than 2% of survival, and chemoradiation increases reported cancers (Greenlee et al., 20012). survival for all subgroups. Nonetheless, at least 40% of the cases with apparent local- The median age of afflicted patients in the regional disease still recur. Also, the optimal United States is 68 years, and the impact of sequence of systemic treatment (biological or co-morbid disease on treatment selection chemotherapeutic) and local-regional must be considered. As many as 25% of (surgical or radiotherapeutic) treatments gastric cancer patients receive no surgical remains undefined. treatment despite presenting in a treatable stage (Hundahl et al., 20003). Surgery, the For patients with metastatic disease being mainstay of current treatment, carries notable treated with multi-agent chemotherapy morbidity and mortality. For example, in- regimens, relative response rates of 50–60% hospital surgical mortality in New York State are common. However, the complete for gastrectomy for cancer is 6.2%, with a response rate remains less than 10%, and clear volume-to-mortality relationship. survival remains brief (8 to 10 months).

Appendix B: Breakout Reports 59 There is no consensus regarding the optimal to include pre- and post-tissue collection and chemotherapeutic regimen for these patients. analysis. This could provide important predictive and prognostic information to Esophageal cancer. One of the remarkable guide future research directions. characteristics of esophageal cancer during the past 30 years has been the change in Three Scientific Priorities and predominant histologic subtype from Rationale for Each squamous cell carcinoma to adenocarcinoma. There has been a 10- to 20-fold increase in Priority 1: Develop meaningful anatomic the incidence of esophageal adenocarcinoma and biological subsets of esophageal and over the past 20 years, which represents the gastric cancers. largest increase of any solid tumor during this period. Currently, the incidence of Rationale: There are at least four major adenocarcinoma exceeds that of squamous anatomic subsets of gastroesophageal cell carcinoma. The site of the primary lesion neoplasms. Overall similarity of outcomes has migrated down from mid-esophagus to belies the biological heterogeneity of these the lower esophagus/GE junction. All of malignancies and the potential for differential these changes have definite therapeutic sensitivity to newer therapies. implications. Priority 2: Evaluate contributions of Historically, surgery has been the cornerstone different treatment modalities for of treatment for patients with local (stage I) esophageal and gastric cancers and their or locally advanced (stages IIA to III) cancer. proper integration and sequencing for However, the high rate of unresectability, optimal treatment outcomes. coupled with the high rate of extra-regional disease, has challenged the notion of single- Rationale: Existing predictive models for modality treatment of esophageal cancer. gastric cancers based on clinicopathologic During the past 15 years, studies of combined features may help in selection of optimal generated encouraging surgical treatment. In addition, molecularly results in phase II and phase III trials. based models could be developed to predict Combined chemoradiotherapy is a valid response, survival, and long-term outcomes alternative to surgery for patients with stage to specific therapeutic interventions. II or III (T3) disease. Whether long-term Different treatments have different outcomes survival is increased through the use of in mortality and morbidity. For example, trimodality therapy is an area of active surgery avoids adverse events associated investigation. In addition, the preferred with chemoradiotherapy, whereas chemo- timing of chemoradiotherapy in relation to radiation avoids surgical-associated surgery remains unknown. mortality.

Esophageal cancer is highly symptomatic: Priority 3: Expand a clinical trials network 90% of patients have dysphagia and weight for these diseases to include multispecialty loss at presentation. Fifty percent have representation. (pain on swallowing). Because many patients are symptomatic, symptom Rationale: Low accrual makes it difficult to palliation is an important goal of therapy. test new therapies. Trials should be expanded to include participation from gastro- Given the ease of access to the esophagus, enterologists, epidemiologists, surgeons, therapeutic trials should make every attempt pathologists, basic scientists, diagnostic

60 Report of the Stomach/Esophageal Cancers Progress Review Group radiologists, nutritionists, statisticians, 2Greenlee RT, Hill-Harmon MB, Murray T, specialists in outcome measures, and Thun M. Cancer statistics 2001. CA Cancer J representatives of industry. Clin 2001; 51(1):15–36.

Infrastructure Needed To 3Hundahl SA, Phillips JL, Menck HR. The Accomplish Priorities National Cancer Data Base Report on poor survival of U.S. gastric carcinoma patients Partnership Platforms treated with gastrectomy: Fifth edition, American Joint Committee on Cancer • Practitioner-focused tissue-acquisition staging, proximal disease, and the “different programs oriented toward outcome- disease” hypothesis. Cancer 2000; linked integrated research. 88(4):921–932.

• Clinical trial network explicitly to bring 4Macdonald JS, Smalley SR, Benedetti J, in all the different types of specialists. Hundahl SA, Estes NC, Stemmermann GN, Haller DG, Ajani JA, Gundersun LL, Jessup Expected Resources To Overcome JM, Martenson JA. Chemoradio-therapy after Limitations of Previous Research and To surgery compared with surgery alone for Capitalize on Existing Opportunities adenocarcinoma of the stomach or gastroesophageal junction. N Engl J Med Central office to coordinate and facilitate 2001; 345(10):725–730. attracting interested practitioners to contribute tissue and to collaborate in clinical 5Hundahl SA, Macdonald JS, Benedetti J, trials Fitzsimmons T, for the Southwest Oncology References Group and the Gastric Intergroup. Surgical treatment variation in a prospective, 1Parkin DM, Pisani P, Farlay J. Global cancer randomized trial of chemoradiotherapy in statistics. CA Cancer J Clin 1999; 49(1):33– gastric cancer: the effect of undertreatment. 64. Ann Surg Oncol 2002; 9(3):278–286.

Appendix B: Breakout Reports 61 Tumor Models

Co-Chairs: Wang, Kenneth and Rustgi, Anil; Wang, Timothy

Participants:

Beer, David MacDonald, John Silberg, Debra Burgart, Lawrence Navtej, Buttar Tobey, Nelia Ilson, David

Overview/Background The mouse offers multiple opportunities for Information/Barriers genetic approaches to understanding molecular mechanisms underlying Models of esophageal and gastric cancers esophageal and gastric cancers. Mutations in are necessary for dissecting biological, a number of genes have led to the biochemical, and genetic pathways and for development of premalignant or malignant applying innovative diagnostic and tumors of the stomach. These include therapeutic strategies. mutations in APC, SMAD-4, TFF-1, TGF- beta 1, and RUNX3. The Cdx2 gene can Historically, the vast majority of models induce intestinal metaplasia in the stomach. have been based on carcinogen application Perturbations in cyclin D1 or EGFR have in rodents, especially rats. A number of been demonstrated to induce esophageal carcinogen-based animal models for upper- squamous dysplasia in transgenic mice. GI cancers have been described in the past When cyclin D1 mice are bred into a p53- with respect to stomach cancer. The deficient background, there is development carcinogen MNNG has been used in the rat; of esophageal squamous cancer. However, and in the mouse, NMU has been used with genetic models that recapitulate Barrett’s variable success. In recent years, greater esophagus are lacking. Apart from these attention has been paid to H. pylori species considerations, in vitro or cell-culture based as the more relevant and physiologic models are in a nascent stage. carcinogen for inducing stomach cancer. H. pylori has been shown to induce gastric Currently, primary mouse and human cancer in mouse, ferret, and Mongolian esophageal squamous epithelial cells have gerbil models. been established in culture. Recently these cells were placed in organotypic culture to With respect to esophageal cancer, DMBA recapitulate the stratified squamous and NMBA have been used in the rat, and to epithelium. However, the role of oncogenes a lesser extent in the mouse, for inducing and tumor-suppressor genes in these cell- squamous papillomas and squamous cell culture models requires elucidation. There carcinoma. These lesions are accentuated in has been limited success in maintaining the setting of various mineral deficiencies. primary cultures of Barrett’s esophageal specimens. These particular cultures From a surgical viewpoint, rats have been maintain the genotypic profile of the original subjected to esophageal jejunostomy with tissues. There has been a wealth of resulting Barrett’s esophagus and utilization of transformed esophageal cancer esophageal adenocarcinoma. These lesions cell lines. are accelerated in a p53-deficient background in the mouse.

62 Report of the Stomach/Esophageal Cancers Progress Review Group

With respect to stomach cancer, there has cancer, and test chemoprevention and been considerable work with transformed therapeutic agents. Cooperative groups that and non-transformed cell lines in culture and share resources in these investigations in nude mice and some recent work with would decrease duplicate efforts. gastric cancer spheroids. There are no cell lines representative of intestinal metaplasia Priority 3: Develop an immunocompetent of the stomach. rodent model of advanced disease.

Barriers to esophageal and gastric tumor Rationale: There is a need to develop models include a lack of identification of biomarkers for diagnosis and treatment stem cells and markers for esophagus and stratification as well as to test targeted stomach, a lack of stem-cell-specific therapeutic agents. promoters for use in animal models, a lack of centralized core facilities for cell lines Infrastructure Needed To and animal models, a lack of uniform Accomplish Priorities criteria in mouse histopathology, and a lack of physiologic tools and approaches to Partnership Platforms animal models. Core facilities for the following: Three Scientific Priorities and Rationale for Each • Mouse models: Surgery, H. pylori infection, breeding Priority 1: Create robust mouse models using hybrid techniques (genetic • Morphology: interpretation, approaches, surgery, H. pylori infection) to microdissection define underlying molecular mechanisms of esophageal and gastric carcinogenesis. • Technology (devices) and imaging

Rationale: Defining underlying molecular • Genomics/proteomics mechanisms would assist in elucidating biological and genetic mechanisms as well • Drug prioritization for testing as in developing surrogate biomarkers. There is a strong need to test the efficacy of Expected Resources To Overcome chemoprevention and therapeutic agents. Limitations of Previous Research and To Capitalize on Existing Opportunities Priority 2: Develop and characterize primary cells, immortalized cells, • NCI (UO1, SPORE, PO1) transformed cells, organ cultures, and organotypic cultures for studying stem-cell • AACR, AGA, ASCO, ACS biology, intestinal metaplasia, and cancer in esophagus and stomach. • Industry (biotechnology, pharmaceutical, animal labs) Rationale: There is strong need to test biological mechanisms, investigate stem-cell • Cooperative oncology groups biology, investigate stepwise progression to

Appendix B: Breakout Reports 63 Disease Sites Adenocarcinomas

Co-Chairs: Tepper, Joel; Forastiere, Arlene and Blaser, Martin J; Spechler, Stuart

Participants:

Beer, David Hamilton, Frank Sampliner, Richard Bernstein, Leslie Hamilton, Stan Silberg, Debra Buttar, Navtej Ilson, David Sivak, Michael Cameron, Alan J. Kumar, Rakesh Souza, Rhonda Chak, Amitabh Levine, Douglas S. Srivastava, Sudhir Christie, Adrian J. Lieberman, David Stoner, Gary Conner, Jerry Lines, Stephen Tell, Robert Dehdashti, Farrokh MacAulay, Calum Tobey, Nelia DeMeester, Tom R. Mayne, Susan Triadafilopoulos, George Falk, Gary Nyren, Olof Vaughan, Thomas Frazzitta, Bart Powell, Steven Wang, Kenneth K. Gammon, Marilie D. Richmond, Ellen Weston, Allan Govindan, Ramaswamy Rodriguez, Luz M. Willett, Christopher Hayman, James Romero, Yvonne Wu, Anna Henley, Donald Rothenberg, Mace

Overview/Background With the increased occurrence of both Information/Barriers esophageal and gastric cardia adenocar- cinomas, it is important to identify these Barrett’s esophagus, a metaplastic change in tumors as either esophageal or gastric in the esophageal lining from normal squamous origin for both diagnostic and therapeutic epithelium to columnar intestinal-type reasons. The distinction between tumors of epithelium, is recognized as a common the proximal stomach or the distal esophagus sequela of gastroesophageal reflux disease is frequently a difficult one, especially when (GERD), developing in approximately 10– a tumor straddles the gastroesophageal (GE) 15% of persons with reflux disease. Because junction. For glandular cancers that cross the people with Barrett’s esophagus display 30– GE junction, this situation is even more 40 times the incidence of esophageal complex, because glandular elements may adenocarcinoma found in the general arise from either side of the GE junction. population, GERD and Barrett’s esophagus To date, no test is able to determine have been identified as major risk factors unequivocally where the tumor arose. A predictive of esophageal adenocarcinoma. major problem confounding investigations of Each year, approximately 0.5% of patients tumors of the GE junction is the lack of with Barrett’s esophagus develop esophageal standardized anatomic landmarks that could adenocarcinoma. The tumors occur clearly delimit the extent of the gastric predominantly among White males, and the cardia. The gastric cardia has been variously incidence has quadrupled since the mid- described as comprising a rim as wide as 1-2 1970s. During the same period, the incidence cm to as little as 1-4 mm adjacent to the GE of adenocarcinomas of the gastric cardia junction. Factors such as hiatal hernia or the (upper stomach) has also increased distortion accompanying a lesion make dramatically. anatomical localization even more difficult.

64 Report of the Stomach/Esophageal Cancers Progress Review Group

The common precursor lesion to these three Three Scientific Priorities and anatomic areas (distal esophagus, GE Rationale for Each junction, and gastric cardia) is intestinal metaplasia. However, the pathways to Priority 1: Elucidate the mechanisms by intestinal metaplasia differ depending on its which host and environmental factors site of origin. In the esophagus, GERD interact in the development of metaplasia in appears to be a key in the development of the stomach and esophagus and its intestinal metaplasia. In addition to GERD progression to cancer, and apply this and Barrett’s esophagus, factors that increase knowledge to develop prevention strategies, the risk of developing intestinal metaplasia improve therapeutics, and diagnostics. include obesity, diet, and perhaps smoking. The roles of other potential environmental Rationale: These cancers represent a exposures are yet to be evaluated. multidecade process, and they progress in an orderly fashion from intestinal metaplasia to Despite the strong evidence of a relationship dysplasia and invasive disease. Preliminary between GERD and adenocarcinomas, 40% studies implicate disorders involving gastric of patients diagnosed with these tumors give acid, bile, exposure to nitrosamines, and no history of GERD or Barrett’s esophagus. possibly a protective effect of H. pylori Large gaps in current epidemiological data colonization in the development of these about GERD and Barrett’s esophagus have conditions. Ascertainment of mechanisms, resulted from difficulties in defining, especially at the molecular level, may allow recognizing, and verifying both conditions. preventive steps. In addition, environmental risk factors remain ill defined. An effective screening Priority 2: Develop a molecular protocol has yet to be developed, because it characterization of adenocarcinoma of the is difficult to identify those actually at risk. esophagus, GE junction, and cardia for classification and staging, and compare Most clinical trials fail to distinguish among with other foregut malignancies to help adenocarcinomas from the esophagus, GE define causation and to develop and apply junction, cardia, or distal gastric cancers. In novel and specific therapies. fact, even the most recent staging system does not distinguish between these cancers. Rationale: The mortality rates for these As a result, very little data exist on either adenocarcinomas remain unacceptably high single or combined modality therapies compared to many other cancers for which focused on this entity. Incorporating major therapeutic advances have occurred. chemotherapy and radiation into primary Currently, the clinical management of treatments has made some progress, but the adenocarcinomas of the distal esophagus, benefits of platinum-based therapies have GE junction, and cardia has not been reached a plateau. Both screening and distinguished from that of squamous cancers therapies present significant quality of life and distal gastric cancers. These tumors are issues as they can cause patient morbidity. arbitrarily classified as either gastric or Further improvements in outcomes will esophageal cancers. This arbitrary require a major change in strategy, such as categorization impairs knowledge of incorporating molecular characterization of causation and true incidence, as well as the adenocarcinomas in order to optimize making it difficult to define differences that both diagnostic and therapeutic protocols. could be important in the development of

Appendix B: Breakout Reports 65 novel therapeutics. There is an urgent need to Infrastructure Needed To identify potential markers of response or Accomplish Priorities resistance to therapy and molecular pathways that could be targeted by specific therapies. Partnership Platforms

Priority 3: Target screening to populations • Develop a multidisciplinary consortium at greatest risk by first defining the to identify the epidemiological, prevalence of premalignant lesions and molecular, and clinical/pathologic associated risk factors in populations of parameters of esophageal, GE junction/ diverse ethnicity not seeking medical cardia adenocarcinoma development and attention, and then defining the natural translate them into clinical trials history of these lesions. • Partner with industry Rationale: The vast majority (95%) of patients presenting with adenocarcinoma of • Create interinstitutional cooperation with the esophagus were not known to have any NIAID, NIDDK, NIA, Department of premalignant lesions. Thus, there is a need to Veterans Affairs, DOD, and CDC detect more quickly individuals at increased risk. In particular, knowledge of the extent of Expected Resources To Overcome these lesions in members of minority groups Limitations of Previous Research and To is limited, and there is good reason to suspect Capitalize on Existing Opportunities that important differences exist. Under- standing the natural history of these lesions is • Support for tissue acquisition, important to develop strategies for classification, and storage appropriate interventions. Better knowledge • Bioinformatics and biostatistical core of risk factors should lead to improved diagnostics for identification of at-risk • Imaging facilities individuals.

66 Report of the Stomach/Esophageal Cancers Progress Review Group

Gastric

Co-Chairs: Ajani, Jaffer; Coit, Daniel and Correa, Pelayo; Macdonald, John

Participants:

Bloom, Bernard Mayne, Susan Rothman, William Burgart, Lawrence Meltzer, Stephen Selaru, Florin Chow, Wong-Ho Michaels, Margo Shibata, David Daschner, Phillip Mori, Yuriko Wang, Timothy Fontham, Elizabeth Moss, Steven Ward, Mary Fox, James Patterson, Reese Welch, Michael Goodman, Karen Peek, Richard, Weston, Allan Helft, Paul Rabeneck, Linda Willett, Christopher Hundahl, Scott Richmond, Ellen Lugo, Tracy Rodriguez, Luz

Overview/Background surgery alone demonstrated significant Information/Barriers improvement in disease-free and overall survival. In the United States, postoperative Gastric cancer will occur in 21,900 patients chemoradiation is now considered a standard this year and cause 13,500 deaths in the of care for patients at risk for recurrence United States. Worldwide, gastric cancer following resection. will occur in 798,000 patients. The most common known antecedent to the The use of preoperative chemotherapy or development of gastric cancer is prior chemoradiation therapy (neoadjuvant infection with H. pylori with subsequent therapy) produces objective responses in as development of chronic gastritis. Host and many as 50% of patients with primary environmental factors modulate the process gastric cancer. This therapeutic approach has of carcinogenesis. H. pylori infection is not not been evaluated in phase III trials and, inevitably associated with the development thus, must be considered investigational in of gastric cancer, and it is unknown whether the management of gastric cancer. the eradication of H. pylori will decrease the risk of gastric cancer. Conventional staging techniques (particularly T and N) do not adequately The conventional therapy of primary gastric predict the heterogeneity of patient cancer is based upon enbloc surgical outcomes. Novel molecular staging resection of the stomach tumor and the techniques may assist not only in more draining lymph nodes. In the United States, accurately predicting outcome but also in overall 5-year survival after gastric resection guiding treatment decisions, including extent is approximately 20%. Because of the high of surgery, use of adjuvant chemotherapy, relapse rate after gastric resection, extensive radiation therapy, and/or biologic therapy. studies of adjuvant chemotherapy have been performed. There is no solid evidence that Current measurements of outcome in patients benefit from this approach. patients treated for gastric cancer are However, a recent NCI intergroup phase III inadequate. Assessment of tumor- and trial of postresection chemoradiation versus treatment-related morbidity is critical, and

Appendix B: Breakout Reports 67 data measuring quality of life (QoL) are include the application and/or development scant. There is a need to disseminate of disease-specific QoL instruments. guidelines for optimal treatment in this disease. Rationale: While the clinical endpoints of relapse and death are often reported, there Three Scientific Priorities and are very few tools to measure the functional Rationale for Each outcome and QoL of patients treated for gastric cancer. Tumor and treatment-related Priority 1: Use multidisciplinary research morbidity is substantial and often affects not approaches to understand the interactions only relapse/survival rates, but also treatment of various strains of H. pylori, host factors, decisions. QoL tools and functional and other lifestyle/environmental factors in measurements become imperative as gastric carcinogenesis. increasing numbers of patients are cured of disease, as a result of early detection or Rationale: Chronic inflammatory states are multimodality therapy. Long-term functional commonly associated with carcinogenesis. sequelae in patients treated for gastric cancer H. pylori infection is recognized as a are undefined. common precursor to gastric cancer. The availability of human, animal, and H. pylori Infrastructure Needed To genomics offer a unique opportunity to study Accomplish Priorities the mechanisms of gastric carcinogenesis.

This type of approach may evolve into a Partnership Platforms template for the study of other cancers.

Priority 2: Develop novel methods, using • Establish close interactions with the NCI molecular profiling of gastric neoplasia, to Office of Communications to use the stratify patients into risk groups to help resources of the Federal Government to direct therapeutic decision-making. This distribute uniform information in a would include sequential and anatomic thoughtful and effective way. mapping of alterations in tumor compared to adjacent nonmalignant tissue. It would • Emphasize interactive relationships and also include genomic and bioinformatic incentives to encourage adherence to approaches to create comprehensive guidelines. profiles of these lesions. • Establish partnerships with other NIH Rationale: Conventional staging is institutes, Department of Defense, and inadequate for assessing prognosis and Veterans Administration, and industry. optimizing treatment decisions. The International collaborations in high- application of current therapies is largely prevalence populations will facilitate empiric. The opportunity to understand more efficient and complete under- molecular profiles may lead to the standing of disease and the development identification of new targets and new of more effective interventions. therapies. Expected Resources To Overcome Priority 3: Measure outcomes of diagnostic Limitations of Previous Research and To and treatment strategies, including early Capitalize on Existing Opportunities detection, response to treatment, survival, QoL, and quality and cost of care in • Explore the development of an patients with gastric cancer. This would international, interactive,

68 Report of the Stomach/Esophageal Cancers Progress Review Group

interdisciplinary, multi-institutional • Establish an international H. pylori consortium for gastric cancer, possibly species bank. joining with other gastrointestinal disease.

Appendix B: Breakout Reports 69 Squamous

Co-Chairs: Hamilton, Stanley R.; Castell, Donald O. and Orlando, Roy C; Leichman, Lawrence

Participants:

Brooks, Jo Ann Ilson, David Silberg, Debra Dawsey, Sandy Jatoi, Aminah Sivak, Michael DeMeester, Thomas Knisley, Eric Sonnenberg, Amnon Donaldson, Molla Okunieff, Paul Stoner, Gary Fischman, Alan Queirolo, Lewis Taylor, Philip Georgakoudi, Irene Rowland, Julia Tobey, Nelia Hamilton, Frank Sandler, Robert Wojcik, Brian Holland, Jimmie

Overview/Background evidence for the phase of initiation and Information/Barriers progression at which the effects may occur.

Esophageal squamous cancer is uncommon Also, the incidence rate is nearly equaled by in the U.S. population, accounting for 6,000 the mortality rate, and medical care for cases each year, with the incidence patients with advanced disease is complex decreasing. The histopathologic subtype of and expensive, indicating that improvements squamous cell carcinoma is now less in therapy and end-of-life care are needed. common than esophageal adenocarcinoma. The NCI funding portfolio for esophageal However, the occurrence of esophageal squamous cell carcinoma is small and heavily squamous cell carcinoma is known to relate weighted toward treatment research, to well-recognized socioeconomic, lifestyle, especially clinical trials that also include and demographic factors that identify high- patients with esophageal adenocarcinoma. In risk groups. The vast majority of cases occur terms of absolute dollars and as a proportion in males, and the incidence is about 15 per of research funding, there is a much lower 100,000 population in non-White males, as allocation of NCI funds dedicated to the contrasted with about 2 per 100,000 in White biology of squamous esophageal cancers than males. other tumor types (7% versus 24%). The characteristics of esophageal squamous cell Few cases occur in the absence of known carcinoma and the current research predisposing conditions that include tobacco environment suggest opportunities for use (in common with other upper- initiatives to attempt to improve population aerodigestive squamous cell carcinomas), incidence, morbidity, and mortality of alcohol consumption, a history of caustic individuals with the disease. injury to the esophagus, human papilloma virus infection, or tylosis and other rare Three Scientific Priorities and genetic syndromes. Nonsteroidal anti- Rationale for Each inflammatory drugs and a diet rich in fruits and vegetables are reported to have Priority 1: Further define the molecular protective benefit, giving additional clues to events involved in the multistage process of possible prevention strategies but without squamous cell carcinoma development in

70 Report of the Stomach/Esophageal Cancers Progress Review Group the human esophagus with a focus on agents that increase disease risk. Conse- different ethnic groups and geographic quently, study of animal models is key in the locations. Clarify the similarities and process of translating basic research into differences in development of esophageal effective clinical methods for prevention, adenocarcinoma. screening, and treatment. Moreover, in this low-prevalence disease, cell line xenografts, Rationale: To elucidate the process of and animal models provide an economic use oncogenesis in squamous cell esophageal of resources from which valuable material cancers, the molecular genetic events can be obtained to study diagnostic, involved in this process must be defined. therapeutic, and chemopreventive This opportunity is afforded by the approaches. esophageal mucosa because it can be targeted for serial biopsies over time. Consequently, Infrastructure Needed To this characteristic enables serial examination Accomplish Priorities of the molecular processes in oncogenesis, tumor prevention (environmental risks and Partnership Platforms chemoprevention), and tumor progression before and after treatment. Develop consortia of investigators who treat patients with squamous cell esophageal Priority 2: Characterize the molecular, carcinoma to collaborate in studies of its cellular, and epidemiological features of epidemiology and natural history. squamous cell carcinoma of the esophagus, Collaborative groups should collect cancer with the goal of using these findings to and premalignant tissue, and contribute to the identify diagnostic, prognostic, predictive, conduct of the genetic and biology studies and therapeutic targets. outlined in the above priorities.

Rationale: Intratumoral markers and targets • Establish a clinical research infrastructure are essential to developing molecular and with multispecialty and multiinstitu- physiological imaging and diagnostic tional centers to perform surveillance in strategies that are less invasive, and treatment an effort to determine the natural history strategies that are more efficacious and less of the premalignant state. toxic than existing modalities. Therapeutic programs should be developed to enhance • Minorities and high-risk populations quality of life (QoL) considerations for the should be targeted for prevention efforts. affected patient population. The QoL issues in cancer survivors and predictors of survivorship in patients who Priority 3: Develop clinically relevant have esophageal cancer should be human or genetically defined animal studied. Public education programs models of established squamous cell should be developed to publicize the links carcinoma and its premalignant phase. between esophageal cancer, smoking, and alcohol. Rationale: Few clinically relevant animal models of esophageal squamous cell • Continue the clinical trials of esophageal carcinomas exist. Animal models are squamous carcinoma as a feature of GI essential to understanding the development Committees of the existing Cooperative of esophageal cancers and to identifying host Oncology Groups. interactions with various environmental

Appendix B: Breakout Reports 71 Expected Resources To Overcome tissue and blood collection, quality Limitations of Previous Research and To control/quality assurance of research Capitalize on Existing Opportunities materials, database tracking, prioritization, and specimen distribution • Increased funding from a variety of with established minimum standard of sources to support human and animal performance in the clinical trials setting. tissue banks as well as databases. • Development of consortia of • Rapid, flexible, and adequate funding investigators, NCI, and local agencies mechanisms for cooperative groups and where these cancers are prevalent. institutions conducting clinical trials to enable them to engage in real time, • Sponsorship of nationwide workshops for collaborative studies to validate existing investigators and public health personnel markers based on levels of evidence. to highlight the relationship between These resources should be applied to smoking and squamous cell carcinomas validation of laboratory methodologies, of the esophagus and other organs.

72 Report of the Stomach/Esophageal Cancers Progress Review Group

Population Management At Risk

Co-Chairs: Tell, Robert; Meltzer, Stephen and Rustgi, Anil; Christie, Adrian

Participants:

Balakrishnan, Krishna Michaels, Margo Sivak, Jr., Michael V. Blaser, Martin J. Moss, Steven Souza, Rhonda Daschner, Phillip Nyren, Olof Srivastava, Sudhir Donaldson, Molla Orlando, Roy C. Tobey, Nelia A. Goodman, Karen J. Peek, Richard Vaughan, Thomas Hornbrook, Mark C. Queirolo, Lewis Willett, Christopher Kumar, Rakesh Rabeneck, Linda Wu, Anna Lieberman, David Romero, Yvonne

Overview/Background Little accurate information on gastric cancer, Information/Barriers esophageal adenocarcinoma, and esophageal squamous cancer is readily available to the Currently, there is no way to identify general public. With the advent of an everyone who is at risk for gastric cancer, enormous array of information on the esophageal adenocarcinoma, and esophageal Internet, many patients who are diagnosed squamous cancer. The mortality rate of with Barrett’s esophagus quickly obtain approximately 95% indicates that a majority erroneous or misleading information, or none of patients are not presenting with early- at all. They need tools to effectively evaluate stage cancer. Moreover, adequate markers this information, to learn about their options, are not available to predict which patients and to make informed decisions. In addition, will develop Barrett’s esophagus, gastric some general practitioners, other primary intestinal metaplasia, or squamous dysplasia. care physicians, and gastroenterologists have insufficient knowledge of gastric cancer, To the best of our knowledge, existing esophageal adenocarcinoma, and esophageal markers have only been tested in patients squamous cancer, or their precursors, due to with Barrett’s esophagus or gastric cancer, the rarity of these diseases. Providing esophageal adenocarcinoma, and esophageal information to practitioners as well as to squamous cancer, and have not been tested educational media that can best meet the in the general population, which includes the needs of those at risk is important. at-risk population. For example, inactivation of the tumor-suppressor gene p53 is known Our understanding of host/environment to occur early in esophageal interactions in gastric cancer, esophageal adenocarcinogenesis, but it has never been adenocarcinoma, and esophageal squamous studied in asymptomatic patients without cancer is limited. For example, we know that Barrett’s esophagus. Similarly, acid and H. Pylori induce cyclooxygenase hypermethylation of the familial polyposis (Cox)-2 expression in vitro. However, we gene APC occurs in the tissues of 92% of know little about other pathways or genes tumors and 25% of sera from patients with involved in the host’s response to esophageal adenocarcinoma. environmental risk factors. An improved understanding of these interactions would

Appendix B: Breakout Reports 73 generate new biomarkers to identify patients adenocarcinoma, and esophageal at risk for premalignant and malignant squamous cancer. lesions. Moreover, insights into the biology of these interactions could have far-reaching Rationale: A study of this magnitude would ramifications for other human premalignant assist in defining lifestyle, dietary, environ- syndromes, particularly those related to mental, and genetic factors affecting risk, chronic inflammatory states. which is important because cancers of the esophagus and stomach often present at a Many millions of dollars currently are spent late stage. With improved screening on treating both premalignant and malignant procedures to identify patients at risk, gastric and esophageal lesions. For example, more patients could be treated earlier. proton pump inhibitors are used widely (and Additionally, a large study could determine perhaps indiscriminately) to treat a broad whether endoscopy reduces mortality, which, array of symptoms. More precise diagnosis in turn, would help better define risk factors and disease classification could result in as a platform for stratifying who should and more discriminate use of these agents, should not be screened. yielding significant cost savings. Priority 2: Educate health care Several barriers to identifying the at-risk professionals and the general public population exist. First, no uniform regarding risk for gastric cancer, classification system is available for the esophageal adenocarcinoma, and different cancer subtypes. Moreover, esophageal squamous cancer and their interventions affect the natural history of precursor states. these diseases at both the testing (e.g., H. pylori) and treatment levels (e.g., proton Rationale: Presumably, more at-risk patients pump inhibitors, antibiotics, nonsteroidal would self-identify and seek treatment anti-inflammatory agents). In addition, with earlier if a risk profile were disseminated to the decreasing prevalence of H. pylori and the public and health care professionals. the relatively low prevalence of gastric Mortality and morbidity could decrease with cancers, esophageal adenocarcinomas, and well-developed tools to establish an esophageal squamous cancers, studies with educational infrastructure and disseminate large numbers of patients are difficult to information. conduct. International collaborations would be instrumental in addressing this need. Priority 3: Identify, define, and validate Finally, the genomic diversity of populations biomarkers (genetic, biochemical, within the United States and in other biological) that stem from interactions countries makes population studies between host and environmental factors challenging. specific to esophageal and gastric carcinogenesis (e.g., H. pylori, acid, bile, Three Scientific Priorities and nitrosamines), using appropriate in vivo Rationale for Each and in vitro models.

Priority 1: Implement broad-based, Rationale: Genomics and proteomics could integrated, population-based, endoscopic, assist in discovering novel genes and multi-institutional studies to define biomarkers, which, in turn, could assist in environmental, clinical, and laboratory developing strategies for more effective risk markers in an effort to identify groups at stratification and prevention of cancer in the risk for gastric cancer, esophageal stomach and esophagus.

74 Report of the Stomach/Esophageal Cancers Progress Review Group

Infrastructure Needed To adenocarcinoma, and esophageal Accomplish Priorities squamous cancer.

Partnership Platforms • Use information technology to increase awareness. • Form a national consortium of gastroenterologists, epidemiologists, • Develop tissue banks (esophagus, pathologists, bioinformatics specialists, stomach) for large, multicenter studies. and other related specialists to conduct multicenter, interdisciplinary studies. A • Develop blood banks (DNA, RNA, H. proposed name for this consortium is pylori genotyping) for large multicenter Validate, Identify, Discover, and Adapt studies, as well as host/environment (VIDA). interactions.

• HMO Research Network. • Develop questionnaires for demo- graphics, dietary factors, environmental • Tap into existing professional societies factors. and advocacy groups, and encourage the creation of new advocacy groups, where • Linkages to existing and new databases. a need exists. • Develop common data elements. • Industrial partnerships to move new technologies forward. • Provide incentives for collaboration.

• Enhance and expand large-scale studies • Bioinformatics and statistics core by incorporating community hospitals as (medical and genomic) to support well as multiple academic centers. priority 1.

Expected Resources To Overcome • Identify the minority of patients with Limitations of Previous Research and To known risk factors who develop Capitalize on Existing Opportunities esophageal and gastric cancers.

• Attract more researchers to the field of gastric cancer, esophageal

Appendix B: Breakout Reports 75 Premalignant

Co-Chairs: DeMeester, Tom R; Fennerty, Brian and Sampliner, Richard

Participants:

Beer, David Goldblum, John R. Mayne, Susan Bernstein, Leslie Govindan, Ramaswamy Provenzale, Dawn Blount, Patricia Hamilton, Frank A. Rodriguez, Luz Cameron, Alan J. Hamilton, Stanley R. Rothman, William Castell, Donald Hundahl, Scott Selaru, Florin M. Correa, Pelayo Knisley, Eric Silberg, Debra Dawsey, Sandy Leichman, Lawrence Spechler, Stuart Falk, Gary Levine, Douglas S. Triadafilopoulos, George Georgakoudi, Irene MacAulay, Calum Wang, Timothy C.

Overview/Background histology is required for diagnosis of these Information/Barriers lesions that cannot be detected by a routine history and physical examination. Another There are well-described and identifiable barrier to early identification of these premalignant lesions that precede esophageal diseases is that gastric and squamous squamous cell carcinoma, esophageal precursor lesions cannot be seen in a routine adenocarcinoma, and gastric endoscopic screening examination, and adenocarcinoma: squamous dysplasia, symptoms related to these lesions or their Barrett’s esophagus, and gastric intestinal associated etiologic conditions can overlap metaplasia. These premalignant lesions are broadly with other diseases. For example, well described in part because they remain in many patients visit their physicians for situ for long periods, which allows them to reflux, and approximately 12% of these be identified and studied. Some groups, such patients have Barrett’s esophagus; however, as African American males, are at higher risk cancer is rarely identified at this early stage, for developing esophageal squamous cell and many of those with adenocarcinoma do carcinoma, and other contributing exposures, not have reflux. Conversely, 95% of those such as alcohol consumption and smoking, with adenocarcinoma have not been have been linked to these diseases. Other diagnosed previously with Barrett’s groups, such as Caucasian males, are at esophagus. higher risk for developing esophageal adenocarcinoma, a cancer that has been Little is known regarding the natural history linked to gastro-esophageal reflux disease. of these lesions. As yet, it is unclear whether Of particular interest, esophageal current screening or surveillance strategies adenocarcinoma is the cancer with the most are effective in preventing malignant rapidly rising incidence in the United States. transformation or reducing mortality from Helicobacter infection is strongly associated these cancers. Many other factors regarding with gastric adenocarcinoma across ethnic these lesions also remain unknown, groups. including who is and is not at risk for acquiring them, their biology, including the Barriers exist to the identification of process of carcinogenesis, the optimal premalignant lesions. Endoscopic biopsy management of individuals at risk, and what

76 Report of the Stomach/Esophageal Cancers Progress Review Group can be done to reduce their risk. What we esophageal squamous cell, adenocarci- know is limited because precancer is not noma, and gastric cancer by performing a always a reportable disease; existing cancer population-based endoscopic screening registries rarely include these premalignant study. lesions; and burdening physicians with recording more information is often viewed Rationale: Impact of cancer outcomes as an additional barrier. requires a better characterization of these cancers in their premalignant disease states. Clinical issues include whom, when, and Additionally, there are large potential how often, or even whether, we should populations at risk for these cancers (GERD, screen and survey individuals at risk. H. pylori, , smoking, obesity), thus lending credibility to population-based • What technologies are applicable for prevalence studies. Within these at-risk improving screening and surveillance populations, high-risk and low-risk or no- strategies? risk subjects can be identified through endoscopic biopsy and stratified for longi- • What is the optimal treatment of early tudinal study, because the lesions are not neoplastic and nonneoplastic typically removed, unlike other premalignant premalignant upper gut lesions? processes (e.g., the adenomatous polyp). Further adenocarcinoma of the esophagus • What is the natural history of these remains the cancer with the most rapidly lesions? rising incidence in the United States, so an • What are the outcomes of untreated and accurate estimate of the prevalence of its treated lesions? precursor lesions and evaluation of their population distribution are essential. Such a • Are screening, surveillance, and research initiative would allow for the treatment of these lesions cost-effective? establishment of risk stratification (molecu- lar, environmental, and epidemiological) for Without answers to these questions, it is these lesions. An additional benefit of this difficult for us to focus on preventing the initiative is the potential to measure patient- disease. centered issues, such as the impact of identifying these premalignant lesions on Currently, NIH funding for studies of the functional status and quality of life. Finally, premalignant lesions that precede these establishment of risk stratification allows a cancers is very limited and includes one concentration of resources directed at early detection project for squamous individuals and populations at risk of disease dysplasia in China; one risk stratification progression. project for Barrett’s esophagus in the United States; and two funded gastric preneoplasia Priority 2: Establish risk stratification for studies in Mexico. these premalignant lesions by the formation of a multi-institutional cohort registry of Three Scientific Priorities and patients drawn from screening studies and Rationales current surveillance practices.

Priority 1: Establish the prevalence of Rationale: Establishment of a cohort of preneoplastic lesions (squamous dysplasia, subjects with upper gastrointestinal Barrett’s esophagus, and gastric intestinal premalignant lesions would provide metaplasia) in the U.S. population for opportunities for addressing the natural

Appendix B: Breakout Reports 77 history of these lesions, opportunities for Infrastructure Needed To further risk stratification of patients with Accomplish Priorities these lesions, and opportunities for investigating the genetic and biologically Partnership Platforms controlling events in the carcinogenic process. This, in turn, would provide oppor- • No single center possesses the resources tunities for intervention studies aimed at necessary for addressing any of these preventing disease progression and would three priorities. Moreover, single center serve as a resource for evaluating and studies, by nature, preferentially exclude developing novel invasive and noninvasive community-based individuals with these diagnostic technology applications. lesions. Interinstitutional cooperation is Importantly, the premalignant lesions of mandatory. these cancers remain in situ, which allow them to be followed and studied over time. • Development of imaging technologies This makes them a good model for studying capable of identifying premalignant carcinogenesis in these and other cancers. lesions and indicators of neoplastic progression are likely to require funding Priority 3: Establish noninvasive and collaboration from industry as well technologies, such as serum markers and as from NIH. imaging techniques, for screening and surveillance of these premalignant lesions. Resources Needed To Overcome Limitations of Previous Research and To Rationale: At present, there is little or no Capitalize on Existing Opportunities funding on a national level for development of new technologies that has enormous • All three priorities require coordination potential for identifying these early lesions. of cross-institutional, cross-disciplinary, Accurate identification of these lesions, multicenter studies designed by especially in asymptomatic people, would interested investigators in this field, allow real population screening, which including epidemiologists, would give us true prevalence figures and let gastrointestinal endoscopists, basic us see the geographic and population scientists, and pathologists. variability of these lesions. Unlike some other malignant processes, higher risk • Central to this process is a multi- populations for these diseases are identi- institutional registry that would include fiable, and the malignancies are associated epidemiological and biological infor- with larger organ field defects that can be mation, as well as a tissue repository of more reliably studied (e.g., long-segment well-characterized individuals with these Barrett’s esophagus). Additionally, the same premalignant lesions. Without such an new technologies can probably be used for infrastructure, many testable hypotheses earlier detection and treatment of these and recommended studies cannot be curable premalignant and early malignant performed. lesions, which should reduce the morbidity and mortality of these cancers.

78 Report of the Stomach/Esophageal Cancers Progress Review Group

Localized Malignant

Co-Chairs: Coit, Daniel and Lawrence, Theodore; Haller, Daniel G.

Participants:

Brooks, Jo Ann Forastiere, Arlene Rowland, Julia Burgart, Lawrence Fox, James Sandler, Robert Chak, Amitabh Helft, Paul Sharma, Prateek Chow, Wong-Ho Ilson, David Stoner, Gary Conley, Barbara Karpey, Martin Van Dam, Jacques Conner, Jerry Lines, Stephen Weston, Allan Dehdashti, Farrokh Powell, Steven Fontham, Elizabeth T.H. Rothenberg, Mace

Overview/Background downstage tumors and cure some patients, Information/Barriers and may be considered as sole treatment for patients when nonsurgical palliation alone is Patients with localized malignancies of the considered. Many patients treated with stomach and esophagus have a wide variety curative intent now receive chemotherapy, of biologic and anatomic features at radiation, and surgery, although the relative presentation. In addition to traditional contribution of each modality to ultimate staging for each tumor (based on outcome remains uncertain. characteristics of the , lymph nodes, and distant metastases), there are at For patients with gastric cancer, cure rates least three distinct anatomic sites vary widely, depending on the stage and site (esophagus, GE junction, stomach) and two of tumor, with distal lesions having higher (adenocarcinoma and squamous cure rates than proximal lesions with surgery cell). This heterogeneity makes clinical alone. Based on the recent U.S. intergroup research in patients with localized trial results, a standard of care for patients malignancies in these sites even more with gastric cancer who are at risk for challenging. recurrence following complete resection is postoperative adjuvant chemoradiation. Surgery is the mainstay of treatment for Future trials are being designed to assess tumors of the upper , the optimal sequence of treatments with a wide variety of outcomes depending (preoperative vs. postoperative) and the not only on tumor stage, surgical expertise, modifications of systemic treatment to and technique, but also on the increasingly reduce the risk of distant failure. common application of multimodality treatments. While the incidence of both squamous cell carcinoma of the esophagus and adeno- For esophageal cancer, surgery alone carcinoma of the distal stomach appears to remains a standard for most patients, but be decreasing, adenocarcinoma of the gastric with a low likelihood of cure, even in cardia and gastroesophageal junction appear apparently localized disease. to be increasing more rapidly than any other Chemoradiation may significantly human cancer. Some of the precursor lesions for this entity have been identified (gastric

Appendix B: Breakout Reports 79 intestinal metaplasia, esophageal Barrett’s predictors of partial or complete response epithelium), but the host factors governing to both conventional and novel therapies. progression to malignancy are not fully characterized. Furthermore, once the Rationale: Treatment approaches for carcinoma is diagnosed, long-term prognosis esophageal and stomach tumors have is poor, even after multimodality therapy for become more complex, with more patients apparent localized disease. The effects of receiving surgery, chemotherapy, and patient, tumor, and treatment-related factors radiation. However, it remains unclear in outcome are not well understood. whether all patients require such treatments. In particular, the role of nonsurgical local Patients with esophageal and gastric cancers modalities (such as endoscopic mucosal are older and often present with significant resection, stenting and photodynamic medical comorbidities that can limit therapy) in patients with esophageal cancer treatment options. These comorbidities often needs to be further evaluated. To understand define a patient’s tolerance of and recovery which patients would best benefit from what from intensive treatment programs. These are now considered standard treatments, the patients often have unique functional role of chemoradiation in gastric and problems arising from both disease and esophageal cancers needs to be explored. treatment-related morbidity. Long-term Advances in biologic markers and imaging survival data do not adequately describe should be exploited to better help in patient these outcomes of treatment in this group of selection and determining response to patients. therapy. Some intermediate measures of success (including serial biopsy before, Compared to clinical research in other during, and after treatment as well as tumors, there has been relatively little work imaging) would enable clinical researchers in predictive and prognostic markers, even to assess efficacy of traditional and targeted retrospectively, to select optimal therapy for biologic therapies outside of the framework individual patients or groups of patients. of large-scale randomized trials. Barriers to clinical research in these tumors include the lack of biologic markers, strong Priority 2: Apply and refine patient- biases on the part of both patients and centered methods to assess specific short physicians for or against certain therapies, and long-term tumor-and treatment-related and suboptimal mechanisms - beyond the quality of life issues in patients with national cooperative groups - for collection localized esophageal and gastric cancer. of data and for testing of hypotheses in this These would include assessment of pain, relatively uncommon malignancy. nutrition, swallowing, fatigue, and diarrhea as well as quality and cost of care Three Scientific Priorities and issues. Rationales Rationale: Information characterizing long- Priority 1: Optimize available treatment term functional outcome is scant even modalities and promote the development of though good data exist describing patterns of novel targeted therapeutics, e.g., recurrence and survival of patients treated radiosensitizing agents, systemic agents, for localized gastric and esophageal cancers. and minimally invasive resection and These patients have unique functional ablation techniques. Develop predictors of problems related to both disease and response that may affect treatment treatment-related morbidity. While some selection, including molecular and imaging data exist for quality of life outcomes,

80 Report of the Stomach/Esophageal Cancers Progress Review Group additional organ-specific instruments need predicting toxicity of therapy. Esophageal to be developed, validated, and applied. In and stomach cancers offer ease of access for addition, assessment of quality and cost of serial biopsies to assess the impact of care as well as patient preferences should be therapy, which makes them unique for these incorporated into clinical trial design. purposes.

Priority 3: Define host and tumor Infrastructure Needed To characteristics to best predict relapse and Accomplish Priorities survival for patients with localized cancer of the esophagus or stomach. These Partnership Platforms include genetic, molecular, biochemical, imaging, and other clinical factors, Cooperative groups as well as patient sociodemographic characteristics. Unique characteristics of • Collaboration with NIDDK esophageal and stomach cancers permit the serial sampling of tumor before, during, • Department of Defense; Department of and after treatment. Veterans Affairs, Research Wing

Rationale: Currently, the treatment patterns Expected Resources To Overcome for large groups of patients with esophageal Limitations of Previous Research and To or gastric tumors are largely based on Capitalize on Existing Opportunities empirical data. The problem is that the different primary sites should be considered • Cooperative national database of patients as distinct diseases. Molecular and genetic with gastric and esophageal cancers markers should be obtained to rationally select both the need for and type of therapy • National tissue bank for study of for the individual patient or subset of molecular profile of gastric and patients. With the introduction of new esophageal cancer chemotherapeutic agents and targeted biologic therapies, predictive markers for • National multi-institutional consortium response or resistance are important in for gastric and esophageal cancer designing treatment programs and in

Appendix B: Breakout Reports 81 Late Malignant

Co-Chairs: Holland, Jimmie and Wang, Kenneth; Jatoi, Aminah

Participants:

Ajani, Jaffer MacDonald, John Taylor, Philip Bloom, Bernard Mori, Yuriko Tepper, Joel Buttar, Navtej Okunieff, Paul Ward, Mary Fischman, Alan Patterson, Reese Welch, Michael Frazzitta, Bart Richmond, Ellen Wojcik, Brian Gammon, Marilie Sonnenberg, Amnon

Overview/Background dysphagia. Thermal ablative therapies, such Information/Barriers as Nd:YAG laser therapy or , have also been used to open the Among the approximately 30,000 U.S. esophageal lumen and may be tolerated patients diagnosed annually with gastric or better, but often they do not offer durable esophageal cancer, the vast majority will palliation. Gastric and esophageal cancers eventually face widespread metastatic also affect nutritional status, which can be disease, thereby confronting the prospect of enhanced by novel enteral access devices an incurable cancer and a limited life. such as percutaneous jejunostomy and Curative therapy for these cancers can be . achieved only rarely. The primary considerations for these cancers have been Chemotherapy. Chemotherapy has palliative therapies and treatments designed provided patients a survival advantage. In to extend survival. three previous trials, chemotherapy was compared to best supportive care, and Palliation. Palliative therapy for upper although benefits were modest, a statistically gastrointestinal tumors has focused on significant survival advantage was observed maintaining the patency of the lumen in among chemotherapy-treated patients in order to allow nutrition, medications, and each of these trials. salivary secretions to pass. The most commonly applied methods of achieving This survival advantage has spawned this goal have been the placement of renewed interest in testing other to alleviate severe dysphagia, as well as chemotherapeutic agents in this setting. chemotherapy and radiation. Gastric cancers However, recent studies suggest two rarely obstruct because of the larger recurrent and concerning themes. First, diameter of the lumen in the stomach. conventional chemotherapy appears to be Modern stents can be made from flexible reaching a plateau with respect to its plastic materials, metallic expandable mesh, efficacy. For example, a promising metallic stents that are coated with a plastic treatment regimen for stomach cancers is material, or metallic mesh stents that contain combined administration of , flaps to prevent reflux of ingested material. , and 5-. This regimen Metallic stents have become commonly used yielded a response rate as high as 70% in in the esophagus because of their ease of previous phase II trials. However, the placement and longer-term palliation of treatment regimen provides only a modest

82 Report of the Stomach/Esophageal Cancers Progress Review Group survival advantage over a previously used when these symptoms are the most common. regimen of 5-fluorouracil, adriamycin, and More importantly, it is critical that patients methotrexate. The median survival times for with esophageal and gastric tumors be these regimens were 8.9 versus 5.7 months, studied in late stages to determine the respectively. complex distressing symptoms and to conduct symptom-control trials using the Second, high toxicity rates remain a major modalities currently available, while also problem. The testing of newer drugs, such as exploring novel interventions. the taxanes and camptothecans, in combi- nation with other agents, has provided Three Scientific Priorities and response rates that approach 50%, but only Rationale for Each at the cost of severe toxicity that also occurs in approximately 50% of patients. Thus, Priority 1: Develop specific, molecularly although the modest benefits of chemo- targeted therapies for late-stage gastro- therapy protect from nihilism, there is a esophageal cancers based on knowledge of clear mandate to explore other strategies to molecular pathways important in tumor improve treatment efficacy and to reduce progression, response to therapy, and toxicity. normal tissue tolerance. Identify molecular markers that could be assessed by Patient Concerns. Patients with late stage nationally available bioinformatics esophageal or gastric cancer are resources to define patients who would experiencing the physical symptoms of respond to nonsurgical treatments. poorly controlled disease such as anorexia, pain, , fatigue, and obstruction, and Rationale: Current therapies for late-stage the recognition that therapies are not gastroesophageal cancer are unsatisfactory. curative. This combination of physical and It is important to develop new therapies that existential concerns (anxiety, depression, decrease mortality and minimize damage to seeking of the meaning of life and death) normal tissue. Molecular markers of lead to levels of distress that are substantial susceptibility to these toxicities are being and are experienced not only by the patient, discovered, and biological modulators are but by the family as well. While there has becoming clinically available. Because of been increasing attention to the control of their accessibility, gastroesophageal cancers symptoms near the end of life, few studies are ideal candidates for testing novel have addressed the problems of patients with therapies and for identifying surrogate tumors of these two sites. The interruption markers of response. The accessibility also of gastrointestinal function adversely affects means that tissue can be acquired serially, nearly every aspect of daily living and limits which would enhance the ability to establish meaningful social interactions with family tissue databanks to study these diseases. and others, which often occur around food. Priority 2: Design and conduct clinical There is a range of assessment tools trials by multidisciplinary investigators to available that validly measure subjective test the efficacy of new diagnostic and symptoms: pain, nausea and , treatment modalities for late-stage dysphagia, fatigue, anxiety, depression, and gastroesophageal cancers. These should delirium related to treatment toxicities. It is include measurements of QoL, cost- important to review this body of information effectiveness, best supportive care, and and its relevance to these tumors, patient education in the non-curative particularly in relation to late stage disease, management of late disease.

Appendix B: Breakout Reports 83 Rationale: Currently, there are very few advocacy groups—particularly the National clinical trials for advanced-stage gastro- Coalition of Cancer Survivorship. esophageal cancers, and virtually no symptom management studies related to Expected Resources To Overcome these cancers. Limitations of Previous Research and To Capitalize on Existing Opportunities Priority 3: Develop validated tumor models for late-stage gastroesophageal cancers to • Investigate and define optimal facilitate the development and testing of information networks, including the NCI new drugs that would allow effective Office of Communication, to: treatment of advanced tumors. – Inform patients Rationale: Tumor models of late-stage disease are necessary to elucidate biological – Educate physicians about the and genetic mechanisms of cancer standards of care for advanced progression. They are needed to test the disease (e.g., physicians need to efficacy of therapeutic agents and permit the anticipate B12 deficiency, the need application of genomics and proteomics. for bone-density scans, and the

possibility of H. pylori infection). It Infrastructure Needed To may be possible to approach this Accomplish Priorities objective by modifying widely disseminated Clinical Practice Partnership Platforms Guidelines.

Establish partnerships for research and • Develop an Internet-based information education with other governmental agencies, system for public distribution. It should cooperative groups, community oncologists, include information about post-operative private foundations, relevant professional complications, nutritional needs, etc. organizations, industry, and patient

84 Report of the Stomach/Esophageal Cancers Progress Review Group

Appendix C

Stomach/Esophageal Cancers PRG Members

Appendix C: Stomach/Esophageal Cancers PRG Members

Timothy J. Eberlein, M.D. M. Brian Fennerty, M.D. PRG Co-Chair Professor of Medicine Bixby Professor and Chair Division of Gastroenterology Washington University School of Medicine Oregon Health and Sciences University

Brian J. Reid, M.D., Ph.D. Arlene A. Forastiere, M.D., Ph.D. PRG Co-Chair Professor, Department of Oncology Head, Gastrointestinal Oncology Program Sidney Kimmel Comprehensive Cancer Center Fred Hutchinson Cancer Research Center Johns Hopkins University

Ernest T. Hawk, M.D., M.P.H. Karen J. Goodman, Ph.D. PRG Executive Director Assistant Professor Chief, Gastrointestinal and Other Department of Epidemiology Cancers Research Group University of Texas Health Science Center National Cancer Institute at Houston School of Public Health

Jaffer A. Ajani, M.D. Stanley R. Hamilton, M.D. Professor of Medicine Professor and Head, Department of Pathology Department of Gastrointestinal Division of Pathology and Laboratory Medicine Medical Oncology University of Texas M.D. Anderson University of Texas M.D. Anderson Cancer Center Cancer Center Jimmie Holland, M.D. Martin J. Blaser, M.D. Chairman, Department of Psychiatry and Chairman, Department of Medicine Behavioral Sciences New York University School of Medicine Memorial Sloan-Kettering Cancer Center

Jo Ann Brooks, Ph.D. Theodore S. Lawrence, M.D., Ph.D. Assistant Professor Professor and Chairman Department of Thoracic Surgery Department of Radiation Oncology Indiana University Medical Center University of Michigan

Donald O. Castell, M.D. Douglas S. Levine, M.D. Professor of Medicine Executive Director of Clinical Research Medical University of South Carolina AstraZeneca

Daniel Coit, M.D., F.A.C.S. David Lieberman, M.D. Chief, Gastric and Mixed Tumor Service Chief, Divison of Gastroenterology Memorial Sloan-Kettering Cancer Center Oregon Health and Sciences University

Pelayo Correa, M.D. Stephen Meltzer, M.D. Boyd Professor, Department of Pathology Professor of Medicine Louisiana State University Health University of Maryland Hospital Sciences Center Cherie Nichols, M.B.A. Thomas R. DeMeester, M.D. Director Professor and Chairman, Department of Surgery Office of Science Planning and Assessment Keck School of Medicine National Cancer Institute University of Southern California

Appendix C: Stomach/Esophageal Cancers PRG Members 85 Roy C. Orlando, M.D. Joel Tepper, M.D. Professor of Medicine and Physiology Professor and Chairman Tulane University Health Sciences Center Department of Radiation Oncology University of North Carolina at Chapel Hill Dawn Provenzale, M.D. School of Medicine Associate Professor and Director of Gastrointestinal Outcomes Research Jacques Van Dam, M.D., Ph.D. Duke University Medical Center Professor and Clinical Chair Division of Gastroenterology Anil K. Rustgi, M.D. Stanford University Medical Center T. Grier Miller Associate Professor of Medicine and Genetics Thomas Vaughan, M.D. Chief of Gastroenterology Head, Program in Epidemiology University of Pennsylvania Fred Hutchinson Cancer Research Center

Richard Sampliner, M.D. Kenneth K. Wang, M.D. Professor of Medicine Associate Professor of Medicine Arizona Health Sciences Center Mayo Clinic University of Arizona Michael Welch, Ph.D. Robert Tell, M.S.H.A Professor, Department of Radiology Patient Advocate Washington University Medical School

86 Report of the Stomach/Esophageal Cancers Progress Review Group

Appendix D

Stomach/Esophageal Cancers PRG Roundtable Participants

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants

Timothy J. Eberlein, M.D.* David Beer, Ph.D. PRG Co-Chair Professor, Department of Surgery Bixby Professor and Chair University of Michigan Medical School Department of Surgery MSRB2, B560, Box 0686 Director, Alvin J. Siteman Cancer Center 1150 West Medical Center Drive Washington University School of Medicine Ann Arbor, MI 48109-6860 660 South Euclid Avenue, Campus Box 8109 Telephone: (734) 763-0325 St. Louis, MO 63110 Fax: (734) 763-0323 Telephone: (314) 362-8086 E-mail: [email protected] Fax: (314) 454-1898 E-mail: [email protected] Leslie Bernstein, Ph.D. Chair in Cancer Research, AFLAC, Inc. Brian J. Reid, M.D., Ph.D.* Professor, Keck School of Medicine PRG Co-Chair University of Southern California Head, Gastrointestinal Oncology Program 1975 Zonal Avenue, KAM 506 Public Health Sciences Division Los Angeles, CA 90033 Fred Hutchinson Cancer Research Center Telephone: (323) 442-1619 1100 Fairview Avenue North, Mailstop C1-157 Fax: (323) 442-1992 Seattle, WA 98109 E-mail: [email protected] Telephone: (206) 667-6792 Fax: (206) 667-6132 Martin J. Blaser, M.D.* E-mail: [email protected] Chairman, Department of Medicine New York University School of Medicine Ernest T. Hawk, M.D., M.P.H.* 550 First Avenue, Room NBV-16N1 PRG Executive Director New York, NY 10016 Chief, Gastrointestinal and Other Telephone: (212) 263-6394 Cancers Research Group Fax: (212) 263-7700 Division of Cancer Prevention E-mail: [email protected] National Cancer Institute Executive Plaza North, Room 2141 Bernard Bloom, Ph.D. 6130 Executive Boulevard Research Professor, Department of Medicine Bethesda, MD 20892-7322 University of Pennsylvania Telephone: (301) 594-2684 3615 Chestnut Street Fax: (301) 435-6344 Philadelphia, PA 19104-2676 E-mail: [email protected] Telephone: (215) 898-7178 Fax: (215) 573-8684 Jaffer A. Ajani, M.D.* E-mail: [email protected] Professor of Medicine Department of Gastrointestinal Patricia Blount, M.D. Medical Oncology Affiliate Investigator University of Texas M.D. Anderson Fred Hutchinson Research Cancer Center Cancer Center 1100 Fairview Avenue North, Room C1-157 1515 Holcombe Boulevard, Box 426 Seattle, WA 98109 Houston, TX 77030-4009 Telephone: (425) 401-6689 Telephone: (713) 792-2828 Fax: (425) 401-6689 Fax: (713) 745-1163 E-mail: [email protected] E-mail: [email protected]

*PRG Members

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants 87 Jon C. Bowersox, M.D., Ph.D. Alan J. Cameron, M.D. Director of Medical Affairs Professor, Department of Medicine Johnson and Johnson/Ethicon Mayo Clinic Endo-Surgery, Inc. 200 First Street, S.W. 4545 Creek Road, ML-21 Rochester, MN 55905 Cincinnati, OH 45242 Telephone: (507) 288-3198 Telephone: (513) 337-3108 Fax: (507) 284-0538 Fax: (513) 337-7328 E-mail: [email protected] E-mail: [email protected] Donald O. Castell, M.D.* Jo Ann Brooks, Ph.D.* Professor of Medicine Assistant Professor Medical University of South Carolina Department of Thoracic Surgery Clinical Science Building, Suite 202 Indiana University Medical Center 96 Jonathan Lucas Street 545 Barnhill Drive, Room EH 215 Charleston, SC 29425 Indianapolis, IN 46202 Telephone: (843) 792-7522 Telephone: (317) 274-3940 Fax: (843) 792-8395 Fax: (317) 278-3996 E-mail: [email protected] E-mail: [email protected] Amitabh Chak, M.D. Lawrence Burgart, M.D. Assistant Professor Pathologist, Department of Anatomic Pathology Case Western Reserve University Mayo Clinic University Hospitals of Cleveland 200 First Street, S.W. 11100 Euclid Avenue Rochester, MN 55905 Cleveland, OH 44106-8066 Telephone: (507) 284-3883 Telephone: (216) 844-5386 Fax: (507) 284-1599 Fax: (216) 983-0347 E-mail: [email protected] E-mail: [email protected]

Navtej Buttar, M.D. Wong-Ho Chow, Ph.D. Researcher Senior Investigator Mayo Clinic Occupational Epidemiology Branch, DCEG 200 First Street, S.W. National Cancer Institute Rochester, MN 55905 Executive Plaza South, Room 8100 Telephone: (507) 266-0132 6120 Executive Boulevard Fax: (507) 255-7612 Rockville, MD 20852 E-mail: [email protected] Telephone: (301) 435-4708 Fax: (301) 402-1819 Kevin M. Callahan, Ph.D. E-mail: [email protected] Deputy Director Office of Science Planning and Assessment Adrian J. Christie, M.D. National Cancer Institute Director, Department of Pathology Building 31, Room 11A03 St. John Macomb Hospital 31 Center Drive 11800 East 12 Mile Road Bethesda, MD 20892-2590 Warren, MI 48093-3494 Telephone: (301) 402-7519 Telephone: (810) 573-5026 Fax: (301) 435-3876 Fax: (810) 573-5007 E-mail: [email protected] E-mail: [email protected]

*PRG Members

88 Report of the Stomach/Esophageal Cancers Progress Review Group

Daniel Coit, M.D., F.A.C.S.* Phillip Daschner Chief, Gastric and Mixed Tumor Service Program Administrator Memorial Sloan-Kettering Cancer Center Division of Cancer Biology 1275 York Avenue National Cancer Institute New York, NY 10021-6007 Executive Plaza North, Room 5014 Telephone: (212) 639-6325 6130 Executive Boulevard Fax: (212) 717-3400 Rockville, MD 20852 E-mail: [email protected] Telephone: (301) 496-9740 Fax: (301) 496-2025 Barbara Conley, M.D. E-mail: [email protected] Chief, Diagnostics Research Branch, CDP National Cancer Institute Sandy Dawsey, M.D. Executive Plaza North, Room 6035A Senior Investigator 6130 Executive Boulevard Cancer Prevention Studies Branch Rockville, MD 20852 National Cancer Institute Telephone: (301) 496-1591 6116 Executive Boulevard, Suite 705 Fax: (301) 402-7819 Bethesda, MD 20892-8314 E-mail: [email protected] Telephone: (301) 594-2930 Fax: (301) 435-8644 Jerry Conner E-mail: [email protected] 2607 Rock Port Circle Garland, TX 75044 Farrokh Dehdashti, M.D. Telephone: (972) 530-2317 Associate Professor of Radiology E-mail: [email protected] Washington University School of Medicine 510 South Kingshighway Boulevard Pelayo Correa, M.D.* St. Louis, MO 63110 Boyd Professor, Department of Pathology Telephone: (314) 362-7418 Louisiana State University Health Fax: (314) 362-1032 Sciences Center E-mail: [email protected] 1901 Perdido Street New Orleans, LA 70112 Thomas R. DeMeester, M.D.* Telephone: (504) 568-6035 Professor and Chairman, Department of Surgery Fax: (504) 599-1278 Keck School of Medicine E-mail: [email protected] University of Southern California 1510 San Pablo Street, Suite 514 James Corrigan, Ph.D. Los Angeles, CA 90033-4612 Chief, Program Assessment Branch Telephone: (323) 442-5925 Office of Science Planning and Assessment Fax: (323) 442-5872 National Cancer Institute E-mail: [email protected] Building 31, Room 11A03 31 Center Drive, MSC 2590 Molla Donaldson, Ph.D. Bethesda, MD 20892-2590 Senior Scientist for Quality of Care Telephone: (301) 496-5515 Research and Policy Fax: (301) 435-3876 Outcomes Research Branch, AOP E-mail: [email protected] National Cancer Institute Executive Plaza North, Room 4028 6130 Executive Boulevard Rockville, MD 20852 Telephone: (301) 435-1638 Fax: (301) 435-3710 E-mail: [email protected]

*PRG Members

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants 89 Deborah Duran, Ph.D. Elizabeth T.H. Fontham, Ph.D. PRG Coordinator Professor and Chairman, Department of Office of Science Planning and Assessment Public Health and Preventive Medicine National Cancer Institute Louisiana State University Health Building 31, Room 11A03 Sciences Center 31 Center Drive, MSC 2590 1600 Canal Street, Suite 800 Bethesda, MD 20892-2590 New Orleans, LA 70112 Telephone: (301) 496-5515 Telephone: (504) 599-1396 Fax: (301) 435-3876 Fax: (504) 568-6905 E-mail: [email protected] E-mail: [email protected]

Gary Falk, M.D. Arlene A. Forastiere, M.D., Ph.D.* Staff Gastroenterologist, Professor, Department of Oncology Department of Gastroenterology Sidney Kimmel Comprehensive Cancer Center Center for Swallowing and Johns Hopkins University Esophageal Disorders 1650 Orleans Street, Room G-90 The Cleveland Clinic Foundation Baltimore, MD 21231 9500 Euclid Avenue, Desk A-30 Telephone: (410) 955-9818 Cleveland, OH 44195 Fax: (410) 614-9861 Telephone: (216) 444-1762 E-mail: [email protected] Fax: (216) 444-6302 E-mail: [email protected] James Fox, D.V.M. Professor and Director M. Brian Fennerty, M.D.* Division of Comparative Medicine Professor of Medicine Massachusetts Institute of Technology Division of Gastroenterology Building 16, Room 825 Oregon Health and Sciences University 77 Massachusetts Avenue 3181 S.W. Sam Jackson Park Road, MC PV-310 Cambridge, MA 02139 Portland, OR 97201-3098 Telephone: (617) 253-1735 Telephone: (503) 494-3787 Fax: (617) 252-1877 Fax: (503) 494-7556 E-mail: [email protected] E-mail: [email protected] Bart Frazzitta Alan J. Fischman, M.D., Ph.D. 11 Crane Court Director, Nuclear Medicine Division Manalapan, NJ 07726 Massachusetts General Hospital Telephone: (732) 446-9509 Professor Fax: (732) 446-0005 Harvard Medical School E-mail: [email protected] 55 Fruit Street, Tilton 201 Boston, MA 02114 Marilie D. Gammon, Ph.D. Telephone: (617) 726-8353 Associate Professor Fax: (617) 726-6165 Department of Epidemiology E-mail: [email protected] University of North Carolina at Chapel Hill 2102C McGavran-Greenberg Hall Pittsboro Road, Campus Box 7400 Chapel Hill, NC 27599-7435 Telephone: (919) 966-7421 Fax: (919) 966-2089 E-mail: [email protected]

*PRG Members

90 Report of the Stomach/Esophageal Cancers Progress Review Group

Irene Georgakoudi, Ph.D. Stanley R. Hamilton, M.D.* Research Scientist, Spectroscopy Laboratory Professor and Head, Department of Pathology Massachusetts Institute of Technology Division of Pathology and Laboratory Medicine 77 Massachusetts Avenue, Room 6-014 University of Texas M.D. Anderson Cambridge, MA 02139 Cancer Center Telephone: (617) 253-9487 1515 Holcombe Boulevard, Box 85 Fax: (617) 253-4513 Houston, TX 77030 E-mail: [email protected] Telephone: (713) 792-2040 Fax: (713) 792-4094 Karen J. Goodman, Ph.D.* E-mail: [email protected] Assistant Professor Department of Epidemiology James Hayman, M.D. University of Texas Health Science Center Associate Professor, Department of at Houston School of Public Health Radiation Oncology 1200 Herman Pressler, Room W906 University of Michigan Medical School Houston, TX 77030 Box UH-B2C490-0010 Telephone: (713) 500-9268 1500 East Medical Center Drive Fax: (713) 500-9329 Ann Arbor, MI 48109 E-mail: [email protected] Telephone: (734) 936-4288 Fax: (734) 763-7370 Ramaswamy Govindan, M.D. E-mail: [email protected] Assistant Professor, Department of Medicine Washington University School of Medicine Paul Helft, M.D. Wohl Hospital, Room 108 Assistant Professor, Department of Medicine 4960 Childrens’ Place Division of Hematology/Oncology St. Louis, MO 63110 Indiana University Cancer Pavilion Telephone: (314) 362-4819 535 Barnhill Drive, Room 414 Fax: (314) 362-7086 Indianapolis, IN 46202 E-mail: [email protected] Telephone: (317) 278-6942 Fax: (317) 278-4190 Daniel G. Haller, M.D. E-mail: [email protected] Professor, Department of Medicine University of Pennsylvania Donald Henley, M.A. 3400 Spruce Street, 16 Penn Tower 5944 McDonie Avenue Philadelphia, PA 19104-4202 Woodland Hills, CA 91367 Telephone: (215) 662-6318 Telephone: (818) 883-1249 Fax: (215) 349-5326 Fax: (818) 883-0833 E-mail: [email protected] E-mail: [email protected]

Frank A. Hamilton, M.D., M.P.H. Jimmie Holland, M.D.* Chief, Digestive Diseases Program Chairman, Department of Psychiatry National Institute of Diabetes and and Behavioral Sciences Digestive and Kidney Diseases Memorial Sloan-Kettering Cancer Center 2 Democracy Plaza, Room 669 1242 Second Avenue 6707 Democracy Boulevard New York, NY 10021 Bethesda, MD 20892-5450 Telephone: (212) 639-3004 Telephone: (301) 594-8877 Fax: (212) 717-3763 Fax: (301) 480-8300 E-mail: [email protected] E-mail: [email protected]

*PRG Members

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants 91 Mark C. Hornbrook, Ph.D. Eric Knisley Associate Director Vice President and General Manager Center for Health Research, Northwest Region Fujinon Incorporated Kaiser Permanente 10 High Point Drive 3800 North Interstate Avenue Wayne, NJ 07470 Portland, OR 97227-1110 Telephone: (973) 633-5600 Telephone: (503) 335-6746 Fax: (973) 633-8818 Fax: (503) 335-2428 E-mail: [email protected] E-mail: [email protected] Rakesh Kumar, Ph.D. Scott Hundahl, M.D., F.A.C.S. Professor Medical Director, The Queen’s Cancer Institute University of Texas M.D. Anderson The Queen’s Medical Center and Cancer Center University of Hawaii 1515 Holcombe Boulevard, Box 108 1301 Punchbowl Street, Tower 6 Houston, TX 77030-4009 Honolulu, HI 96813 Telephone: (713) 745-3558 Telephone: (808) 537-7353 Fax: (713) 745-2050 Fax: (808) 537-7080 E-mail: [email protected] E-mail: [email protected] Theodore S. Lawrence, M.D., Ph.D.* David Ilson, M.D., Ph.D. Professor and Chairman Associate Attending Physician Department of Radiation Oncology Memorial Sloan-Kettering Cancer Center University of Michigan 1275 York Avenue, Suite H912 B2C502 University Hospital New York, NY 10021 1500 East Medical Center Drive Telephone: (212) 639-8306 Ann Arbor, MI 48109 Fax: (212) 717-3320 Telephone: (734) 647-9955 E-mail: [email protected] Fax: (734) 763-7371 E-mail: [email protected] Aminah Jatoi, M.D. Assistant Professor Lawrence Leichman, M.D. Mayo Clinic Chief, Division of Medical Oncology 200 First Street, S.W. Albany Medical College Rochester, MN 55905 47 New Scotland Avenue Telephone: (570) 284-3902 Buffalo, NY 14263 Fax: (570) 538-1803 Telephone: (518) 262-8269 E-mail: [email protected] Fax: (518) 262-6556 E-mail: [email protected] Martin Karpeh, M.D. Associate Attending Surgeon Douglas S. Levine, M.D.* Memorial Sloan-Kettering Cancer Center Executive Director of Clinical Research 1275 York Avenue AstraZeneca New York, NY 10021 725 Chesterbrook Boulevard, Suite E-3C Telephone: (212) 639-8056 Wayne, PA 19087-5677 Fax: (212) 794-5847 Telephone: (610) 695-4306 E-mail: [email protected] Fax: (610) 695-1162 E-mail: [email protected]

*PRG Members

92 Report of the Stomach/Esophageal Cancers Progress Review Group

David Lieberman, M.D. Susan Mayne, Ph.D. Chief, Divison of Gastroenterology Associate Professor Oregon Health and Sciences University Yale University School of Medicine Portland VAMC P3-GI P.O. Box 208034 3710 S.W. Veterans Hospital Road 60 College Street Portland, OR 97207 New Haven, CT 06520-8034 Telephone: (503) 273-5318 Telephone: (203) 785-6274 Fax: (503) 220-3426 Fax: (203) 785-6980 E-mail: [email protected] E-mail: [email protected]

Stephen Lines Stephen Meltzer, M.D.* Program Manager Professor of Medicine AstraZeneca University of Maryland Hospital 725 Chesterbrook Boulevard, Suite E-2C 22 South Greene Street, Room N3W62 Wayne, PA 19078 Baltimore, MD 21201 Telephone: (610) 695-1253 Telephone: (410) 706-3375 Fax: (610) 695-1245 Fax: (410) 706-1325 E-mail: [email protected] E-mail: [email protected]

Joe Lipscomb, Ph.D. Margo Michaels, M.P.H. Chief, Outcomes Research Branch, DCCP Survivor and Public Education Section National Cancer Institute Office of Education and Special Initiatives Executive Plaza North, Room 4005 National Cancer Institute 6130 Executive Boulevard 6116 Executive Boulevard, Suite 202 Rcokville, MD 20852 Bethesda, MD 20852-8334 Telephone: (301) 402-3590 Telephone: (301) 594-8993 Fax: (301) 435-3710 Fax: (301) 594-7063 E-mail: [email protected] E-mail: [email protected]

Calum MacAulay, Ph.D. Yuriko Mori, M.D. Head, Cancer Imaging Department Postdoctoral Fellow British Columbia Cancer Agency University of Maryland, Baltimore 601 West 10th Avenue 655 West Baltimore Street Vancouver, British Columbia V5Z 1L3 Bressler Research Building, Room 8-012 Canada Baltimore, MD 21201 Telephone: (604) 877-6098, ext. 3109 Telephone: (410) 706-3375 Fax: (604) 877-6077 Fax: (410) 706-1325 E-mail: [email protected] E-mail: [email protected]

John MacDonald, M.D. Steven Moss, M.D. Medical Director and Chief Associate Professor of Medicine Gastrointestinal Oncology Services Rhode Island Hospital St. Vincent’s Comprehensive Cancer Center Brown University 325 West 15th Street 593 Eddy Street, Room APC 445 New York, NY 10011 Providence, RI 02903 Telephone: (212) 604-6011 Telephone: (401) 444-6713 Fax: (212) 604-6039 Fax: (401) 444-2939 E-mail: [email protected] E-mail: [email protected]

*PRG Members

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants 93 Cherie Nichols, M.B.A.* Richard Peek, M.D. Director Associate Professor of Medicine Office of Science Planning and Assessment Division of Gastroentology National Cancer Institute Vanderbilt University School of Medicine Building 31, Room 11A03 C-2104 Medical Center North 31 Center Drive 1161 21st Avenue South Bethesda, MD 20892-2590 Nashville, TN 37232 Telephone: (301) 496-5515 Telephone: (615) 322-5200 Fax: (301) 435-3876 Fax: (615) 343-6229 E-mail: [email protected] E-mail: [email protected]

Olof Nyren, M.D., Ph.D. Steven Powell, M.D. Professor of Clinical Epidemiology Associate Professor of Medicine Department of Medical Epidemiology Digestive Health Center of Excellence Karolinska Institute University of Virginia Health Systems Box 286, S-171 77 P.O. Box 800708 Berzelius vag 15C Charlottesville, VA 22908-0708 S-171 77 Stockholm Telephone: (434) 243-2718 Sweden Fax: (434) 924-0491 Telephone: 46-8-728-6195 E-mail: [email protected] Fax: 46-8-314-975 E-mail: [email protected] Dawn Provenzale, M.D.* Associate Professor and Director of Paul Okunieff, M.D. Gastrointestinal Outcomes Research Professor and Chair Duke University Medical Center Department of Radiation Oncology 508 Fulton Street, Building 16, Room 70 University of Rochester Durham, NC 27705 601 Elmwood Avenue, Box 647 Telephone: (919) 286-2287 Rochester, NY 14642 Fax: (919) 416-5839 Telephone: (585) 275-5575 E-mail: [email protected] Fax: (585) 275-1531 E-mail: [email protected] Lewis Queirolo, Ph.D. 440 Eagle Crest Road Roy C. Orlando, M.D.* Camano Island, WA 98282 Professor of Medicine and Physiology Telephone: (360) 387-4652 Tulane University Health Sciences Center Fax: (360) 285-6471 1430 Tulane Avenue, SL-35 E-mail: [email protected] New Orleans, LA 70112-2699 Telephone: (504) 588-5606 Linda Rabeneck, M.D. Fax: (504) 587-2188 Professor of Medicine and Director E-mail: [email protected] Division of Gasteroenterology University of Toronto Reese Patterson 238 Davenport Road, Suite 319 6005 Glenmary Road Toronto, Ontario M5R 1J6 Knoxville, TN 37919 Canada Telephone: (865) 588-8169 Telephone: (416) 471-3407 Fax: (416) 323-6129 E-mail: [email protected]

*PRG Members

94 Report of the Stomach/Esophageal Cancers Progress Review Group

Ellen S. Richmond, M.S., R.N., C.S., William Rothman, M.S.A.K. A.O.C.N. University of Michigan School of Public Health Program Director and Clinical Trials 2861 Stratford Street Nurse Specialist Oak Park, MI 48237 Gastrointestinal and Other Cancer Telephone: (248) 968-2437 Research Group E-mail: [email protected] National Cancer Institute Executive Plaza North, Room 2148 Julia Rowland, Ph.D. 6130 Executive Boulevard Director, Office of Cancer Survivorship, DCCPS Rockville, MD 20852 National Cancer Institute Telephone: (301) 435-2466 Executive Plaza North, Room 4086 Fax: (301) 435-6344 6130 Executive Boulevard E-mail: [email protected] Rockville, MD 20852 Telephone: (301) 402-2964 Luz Rodriguez, M.D. Fax: (301) 594-5070 Surgical Oncologist and Staff Physican E-mail: [email protected] Genetics Branch, CCR National Cancer Institute Anil K. Rustgi, M.D.* National Naval Hospital T. Grier Miller Associate Professor of Building 8, Room 5101 Medicine and Genetics 8901 Wisconsin Avenue Chief of Gastroenterology Bethesda, MD 20889 University of Pennsylvania Telephone: (301) 435-1460 415 Curie Boulevard Fax: (301) 496-0047 Philadelphia, PA 19104 E-mail: [email protected] Telephone: (215) 898-0154 Fax: (215) 573-5412 Yvonne Romero, M.D. E-mail: [email protected] Consultant, Division of Gastroenterology and Hepatology Richard Sampliner, M.D.* Mayo Clinic Professor of Medicine 200 First Street, S.W. Arizona Health Sciences Center Rochester, MN 55905 University of Arizona Telephone: (507) 284-8714 P.O. Box 245028 Fax: (507) 284-0538 1501 North Campbell Avenue, Room 6406 E-mail: [email protected] Tucson, AZ 85724-5018 Telephone: (520) 626-6119 Mace Rothenberg, M.D. Fax: (520) 629-4737 Ingram Associate Professor of Cancer Research E-mail: [email protected] Vanderbilt-Ingram Cancer Center Vanderbilt University Robert Sandler, M.D. 777 Preston Research Building Professor of Medicine and Epidemiology Nashville, TN 37232-6307 University of North Carolina at Chapel Hill Telephone: (615) 343-8422 719 Burnett Womack Building, CB 7080 Fax: (615) 343-7602 Chapel Hill, NC 27599-7080 E-mail: mace.rothenberg@ mcmail.vanderbilt.edu Telephone: (919) 966-0090 Fax: (919) 966-2478 E-mail: [email protected]

*PRG Members

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants 95 Florin M. Selaru, M.D. Amnon Sonnenberg, M.D., M.Sc. Faculty Research Associate Staff Physician University of Maryland, Baltimore University of New Mexico 655 West Baltimore Street, BRB 8-012 New Mexico VA Health Care System Baltimore, MD 21201 VA Medical Center, Room 111F Telephone: (410) 706-3375 1501 San Pedro Drive, S.E. Fax: (410) 706-1325 Albuquerque, NM 87108 E-mail: [email protected] Telephone: (505) 265-1711, ext. 4513 Fax: (505) 256-5751 Prateek Sharma, M.D. E-mail: [email protected] Assistant Professor, Gastrointestinal Section University of Kansas Medical School and Rhonda Souza, M.D. VA Medical Center Assistant Professor of Medicine 4801 East Linwood Boulevard Gastrointestinal Department Kansas City, MO 64128 Dallas VA Medical Center Telephone: (816) 861-4700, ext. 6737 University of Texas Southwestern Fax: (816) 922-4692 Medical School E-mail: [email protected] 4500 South Lancaster Road, MC 111B1 Dallas, TX 75287 David Shibata, M.D. Telephone: (214) 857-0301 Assistant Professor of Surgery Fax: (214) 857-0328 University of Maryland School of Medicine E-mail: [email protected] 22 South Greene Street, Room N4W58 Baltimore, MD 21201 Stuart Spechler, M.D. Telephone: (410) 328-7320 Chief, Division of Gastroenterology Fax: (410) 328-8118 Dallas VA Medical Center E-mail: [email protected] University of Texas Southwestern Medical School Debra Silberg, M.D., Ph.D. 4500 South Lancaster Road Assistant Professor Dallas, TX 75216 University of Pennsylvania Telephone: (214) 374-7799 415 Curie Boulevard, 650 CRB Fax: (214) 857-1571 Philadelphia, PA 19104 E-mail: [email protected] Telephone: (215) 898-0157 Fax: (215) 573-2024 Sudhir Srivastava, Ph.D., M.P.H. E-mail: [email protected] Chief, Cancer Biomarkers Research Group, DCP National Cancer Institute Michael V. Sivak, Jr., M.D. Executive Plaza North, Room 3142 Chief, Division of Gastroenterology 6130 Executive Boulevard Case Western Reserve University Rockville, MD 20852 School of Medicine Telephone: (301) 435-1594 University Hospitals of Cleveland Fax: (301) 402-8990 11100 Euclid Avenue, Wehrn 253 E-mail: [email protected] Cleveland, OH 44106 Telephone: (216) 844-7344 Fax: (216) 844-7371 E-mail: [email protected]

*PRG Members

96 Report of the Stomach/Esophageal Cancers Progress Review Group

Gary Stoner, Ph.D. George Triadafilopoulos, M.D. Professor and Chairman Professor of Medicine Division of Environmental Health Sciences Department of Gastroenterology 111-G1 Cancer Hospital and Research Institute Stanford University Ohio State University School of Public Health Palo Alto VA Health Care System 300 West 10th Avenue, Room 1148B 3801 Miranda Avenue Columbus, OH 43210-1240 Palo Alto, CA 94304-6187 Telephone: (614) 293-3268 Telephone: (650) 493-5000, ext. 64485 Fax: (614) 293-3333 Fax: (650) 856-8024 E-mail: [email protected] E-mail: [email protected]

Philip R. Taylor, M.D., Sc.D. Annabelle Uy, M.S. Chief, Cancer Prevention Studies Branch, CCR Program Analyst National Cancer Institute Office of Science Planning and Assessment 6116 Executive Boulevard, Room 705 National Cancer Institute Rockville, MD 20852 Building 31, Room 11A03 Telephone: (301) 594-2932 31 Center Drive, MSC 2590 Fax: (301) 435-8645 Bethesda, MD 20892-2590 E-mail: [email protected] Telephone: (301) 496-5515 Fax: (301) 435-3876 Robert Tell, M.S.H.A.* E-mail: [email protected] Patient Advocate 29401 Windmill Court Jacques Van Dam, M.D., Ph.D.* Farmington Hills, MI 48334 Professor and Clinical Chair Telephone: (248) 626-6410 Division of Gastroenterology E-mail: [email protected] Stanford University Medical Center 300 Pasteur Drive, Room H-1121, MC 5202 Joel Tepper, M.D.* Stanford, CA 94305-5202 Professor and Chairman Telephone: (650) 736-0431 Department of Radiation Oncology Fax: (650) 723-8305 University of North Carolina at Chapel Hill E-mail: [email protected] School of Medicine Clinical Cancer Center, Room 1043 Thomas Vaughan, M.D.* 101 Manning Drive, Campus Box 7512 Head, Program in Epidemiology Chapel Hill, NC 27599-7512 Fred Hutchinson Cancer Research Center Telephone: (919) 966-0400 P.O. Box 19024 Fax: (919) 966-7681 1100 Fairview Avenue North E-mail: [email protected] Seattle, WA 98109 Telephone: (206) 667-4738 Nelia A. Tobey, Ph.D. Fax: (206) 667-4787 Research Professor of Medicine E-mail: [email protected] SL 35-Gastroenterology Department Tulane University School of Medicine Andrew von Eschenbach, M.D. 1430 Tulane Avenue Director New Orleans, LA 70112 National Cancer Institute Telephone: (504) 568-0811, ext. 5847 Building 31, Room 11A48 Fax: (504) 587-2188 31 Center Drive E-mail: [email protected] Bethesda, MD 20892-2590 Telephone: (301) 496-5615 Fax: (301) 402-0338 E-mail: [email protected]

*PRG Members

Appendix D: Stomach/Esophageal Cancers PRG Roundtable Participants 97 Kenneth K. Wang, M.D.* Allan Weston, M.D. Associate Professor of Medicine Chief, Gastrointestinal-Hepatology Section Mayo Clinic Kansas City VA Medical Center Main Alfred Building, Room 430 4801 East Linwood Boulevard 200 First Street, S.W. Kansas City, MO 64128 Rochester, MN 55905 Telephone: (816) 861-4700 Telephone: (507) 284-2174 Fax: (816) 922-4692 Fax: (507) 255-7612 E-mail: [email protected] E-mail: [email protected] Christopher Willett, M.D. Timothy C. Wang, M.D. Professor, Department of Radiation Oncology Chief, Gastroenterology Division Massachusetts General Hospital University of Massachusetts Medical School Harvard Medical School Lazare Research Center, Suite 208 100 Blossom Street, Cox 3 364 Plantation Street Boston, MA 02114 Worcester, MA 01605 Telephone: (617) 724-1548 Telephone: (508) 856-4778 Fax: (617) 726-3603 Fax: (508) 856-4770 E-mail: [email protected] E-mail: [email protected] Brian Wojcik, Ph.D. Mary H. Ward, Ph.D. Scientific Review Administrator Tenure-track Investigator Division of Examural Activities Occupational Epidemiology Branch, DCEG National Cancer Institute National Cancer Institute 6116 Executive Boulevard, Room 8013 Executive Plaza South, Room 8104 Rockville, MD 20852 1620 Executive Boulevard Telephone: (301) 402-2785 Rockville, MD 20852 Fax: (301) 496-6497 Telephone: (301) 435-4713 E-mail: [email protected] Fax: (301) 402-1819 E-mail: [email protected] Anna Wu, Ph.D. Professor, Keck School of Medicine Michael Welch, Ph.D.* University of Southern California Professor, Department of Radiology 1441 Eastlake Avenue, MC9175 Washington University Medical School Los Angeles, CA 90089 Campus Box 8225 Telephone: (323) 865-0484 510 South Kingshighway Boulevard Fax: (323) 865-0139 St. Louis, MO 63110 E-mail: [email protected] Telephone: (314) 362-8436 Fax: (314) 362-8399 E-mail: [email protected]

*PRG Members

98 Report of the Stomach/Esophageal Cancers Progress Review Group

NIH Publication Number 03-5314 December 2002 T916