CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

CANCER CHEMOPREVENTION

0889-8588/98 $8.00

+

.OO

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS David G. Beer, PhD, and Gary D. Stoner, PhD

The two principal types of esophageal cancer are (1) adenocarcinoma, derived from Barrett’s metaplasia, and (2) squamous cell carcinoma, which arises from epithelial dysplasia. Adenocarcinoma occurs frequently in the United States and rarely in the Middle East and far eastern countries, whereas the opposite is true for esophageal squamous cell carcinoma. Both cancers have a poor prognosis. Chemoprevention is likely to be a viable approach to reduce the incidence of and mortality from these diseases. Approaches for the chemoprevention of esophageal adenocarcinoma and squamous cell carcinoma are presented in this article, and they address events associated with both the initiation and promotion/progression stages of these diseases. The results of completed trials for the chemoprevention of these cancers are summarized, and suggestions are made for future trials involving additional strategies and endpoints. COLUMNAR-LINED ESOPHAGUS (BARRETT’S METAPLASIA)

In patients who have chronic gastroesophageal reflux disease (GERD), the normal squamous epithelium of the distal esophagus can From the Department of Surgery, Thoracic Surgery Research Laboratory, University of Michigan, Ann Arbor, Michigan (DGB); and Division of Environmental Health Sciences, Ohio State University School of Public Health, Columbus, Ohio (GDS)

HEMATOLOGY/ONCOLOGY CLINICS OF NORTH AMERICA VOLUME 12 NUMBER 5 * OCTOBER 1998

1055

1056

BEER & STONER

be replaced with a columnar-lined mucosa. This metaplastic mucosa, known as Barrett’s esophagus, is considered a premalignant condition and is associated with a 30 to 125 times higher incidence of adenocarci27, 46, lo4, Io5 Because the presence noma than in the general p~pulation.’~, of Barrett’s metaplasia is the most important risk factor for adenocarcinoma, patients with Barrett’s metaplasia are often enrolled in endoscopic surveillance programs involving routine biopsies to aid in cancer detection and to allow early surgical intervention. Although the need for endoscopic surveillance in all patients with Barrett’s mucosa remains controversial, a significantly increased 5-year survival among patients undergoing esophagectomy for adenocarcinoma was reported in those having routine endoscopic s~rveillance.~~ This was attributed to a higher percentage of early-stage cancers in the surveillance group. At the present time surveillance is the primary option offered to patients with Barrett’s metaplasia; however, chemoprevention may represent a potentially effective means to slow or prevent cancer development in these individuals. To help identify chemoprevention modalities that may be effective in reducing cancer development in patients with Barrett’s metaplasia, it is first necessary to identify those factors that put this epithelium at increased cancer risk.

Risk Factors and Chemoprevention Strategies in Patients with Barrett’s Esophagus Development of Barrett’s Metaplasia

It is unclear why only 3% to 12% of patients with chronic GERD develop Barrett’s metapla~ia.~~, *05 Identification of those factors that might predispose certain individuals to develop Barrett’s metaplasia, as well as those that might be protective for the patients who do not, may offer new chemopreventative approaches in addition to the gastric acid blockade medications and antireflux operations currently used. Genderor race-related differences may provide an important clue, because Barrett’s metaplasia is much more common in white men than in women, and it is infrequent in blacks and 26, 82 A familial association of Barrett’s metaplasia has also been 71 indicating that a genetic component either to the development of reflux or to Barrett’s metaplasia may be involved. lntestinalization of the Esophagus

The destruction of the distal normal squamous mucosa from chronic gastroesophageal reflux can result in its replacement by Barrett’s meta-

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1057

plasia, primarily because this metaplastic mucosa produces a protective mucus. The type of Barrett’s metaplasia that results, however, can represent a risk factor for further neoplastic development. There are three histologic types of Barrett’s esophagus,62 but it is the intestinal-type (specialized)epithelium that is most common and most often associated with high-grade dysplasia and adenocarcinoma.26The intestinal-type Barrett’s mucosa shares certain histologic features with normal small intestinal mucosa including goblet cells and the expression of brush border enzymes such as sucrase isomaltase,lo5aminopeptidase N;l and the calcium-regulated, actin-binding protein illi in.^^ Importantly, most dysplastic Barrett’s mucosa and the resulting adenocarcinomas also exlo6indicating their origin from the intestipress these intestinal nal-type Barrett’s metaplasia. None of these intestinal markers are expressed in the normal esophageal squamous mucosa or gastric mucosa, and thus they are specific to the intestinal-type Barrett’s mucosa. Cell Proliferation as a Target for Chemoprevention Studies The specific properties of intestinal-type Barrett’s metaplasia that underlie its increased cancer risk are of considerable interest, because paradoxically cancers of the small intestine are relatively infrequent. A comparison of the normal small intestinal mucosa with the intestinaltype Barrett’s metaplasia may suggest specific cellular properties affecting cancer risk as well as identify potential endpoints for chemoprevention studies. Cell proliferation can influence the potential cancer risk of a tissue, because rapidly cycling cells may be more likely to sustain mutations that can accumulate and lead to cancer development. In both the normal small intestine and esophageal mucosa, continuous cell proliferation is observed but is strictly confined to cells within the proliferative zones in these tissues, that is, within the crypts of the small intestine or in the basal zone of the esophageal squamous mucosa. Analyses of cell proliferation have been performed in Barrett’s metaplasia, using antibodies to specific nuclear antigens such as PCNA (proliferating cell nuclear antigen) and Ki-67.25,32 Intestinal-type Barrett’s mucosa was found to have the highest percentage of proliferating cells in both the crypt and glandular regions of all three types of Barrett’s metapla~ i a Furthermore, . ~ ~ the proliferative zone is more expanded and the percentage of proliferating cells was greater in intestinal-type Barrett’s mucosa. An expansion of the cell proliferation compartment was observed in Barrett’s mucosa with low-grade dysplasia, and in high-grade dysplasia cell proliferation was significantly increased in the upper crypt and surface cells.32 Because of the high level of cell proliferation in intestinal-we Barrett’s mucosa and the further increases observed in dysplastic mu-

1058

BEER 81 STONER

cosa, inhibitors of cell proliferation or inducers of cell differentiation may be useful as chemopreventative agents. In two small-scale studies, the effects of 13-cis-retinoic acid (RA) and the ornithine decarboxylase (ODC) inhibitor difluoromethylornithine (DFMO) were examined.23,72 ODC is a rate-limiting enzyme in polyamine synthesis and is necessary for cell proliferation. Substances that block ODC, such as retinoids, inhibitors of arachidonic acid, or the irreversible inhibitor DFMO, may be effective chemopreventatives. In 16 patients with Barrett’s metaplasia, Sampliner and Garewa17*gave RA at 1mg/kg/day for 6 weeks. Unfortunately, unacceptable toxicity and the production of esophageal ulcers were observed in some patients, and essentially no effect on the Barrett’s mucosa was observed. Based upon earlier studies showing elevated ODC in Barrett’s mucosa relative to both normal intestine or gastric mucosa, Gerner et alZ3treated eight Barrett’s patients with 1.5 gm/m2/ day of DFMO for 12 weeks and examined polyamine levels in these patients. Spermidine, but not spermine, content in both the Barrett’s and normal esophageal mucosa was reduced with DFMO and returned to control values upon discontinuation. The specific effects upon the histologic features of the Barrett’s mucosa were not discussed. A multiinstitutional phase I1 clinical trial is currently examining the effects of DFMO on intestinal-type and low-grade dysplastic Barrett’s metaplasia. Inhibitors of Metabolic Activation of Procarcinogens in Barrett’s Metaplasia The specific endogenous or environmental carcinogens that may play a definitive role in inducing genetic alterations in Barrett’s metaplasia are unknown, but they may arise from dietary, tobacco, or other sources. Because most environmental procarcinogens require metabolic activation, agents that act to inhibit this step might be chemopreventative in this mucosa. Polycyclic aromatic hydrocarbons are a class of chemicals that include such compounds as benzo(a)pyrene, which are present in tobacco smoke, charcoal-broiled foods, and in high levels in certain occupations. Recent studies from this laboratory indicate that intestinal-type Barrett’s mucosa expresses a number of different cytochrome P450 enzymes and can efficiently metabolically activate benzo(a)pyrene to forms that bind to DNA. Importantly, expression of cytochrome P450 enzymes was detected in cells within the proliferative zone of Barrett’s mucosa, whereas the expression of P450 in both the normal esophagus and small intestinal mucosa was distinctly separate from areas showing cell proliferation (Hughes et al, manuscript submitted). The combined capabilities of metabolic potential and cell proliferation in Barrett’s metaplasia may increase the chances of cells in this mucosa

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1059

sustaining genetic alterations. To date, compounds that are inhibitors of cytochrome P450 have not been examined in patients with Barrett’s metaplasia. Certain plant phenols that may be present in the human diet, such as ellagic acid, may be useful compounds for this purpose (reviewed in ref. 83).

Inducers of Protective Mechanisms in Barrett’s Metaplasia Inhibition of covalent binding of carcinogens to DNA can be observed using compounds that increase inducible cellular detoxification reactions. One of the most important systems involves the glutathioneS-transferases (a,F, T,O), which catalyze the conjugation of a variety of structurally diverse endogenous and exogenous compounds with the nonprotein thiol g l ~ t a t h i o n e .Elevation ~~ of glutathione-S-transferase provides protection to exposure to carcinogens. Peters et aP4 have analyzed both glutathione-S-transferase enzyme activity and glutathione content in Barrett’s metaplasia, and they found them to be significantly less than in the gastric and duodenal mucosa from the same patients. Although specific forms, such as the glutathione-S-transferase-a, were relatively higher in the metaplastic mucosa, the glutathione-s-transferase-T form was significant lower, as was overall glutathione-s-transferase activity. This could imply a lower level of protection against cellular and genetic damage in patients with Barrett’s metaplasia. Compounds such as oltipraz (5[2-pyrazinyl]-4-methyl-2,3-dithiol-3-thione) can induce these protective detoxification enzymes and appear to show chemopreventative activity in humans (reviewed in ref. 102). Oltipraz can also increase glutathione content; however, the effects of this compound on prevention of cancer development in Barrett’s metaplasia have not yet been examined. Other dietary compounds such as isothiocyanates found in cruciferous vegetables, and polyphenols such as ellagic acid, also can increase glutathione-S-transferase as well as other detoxification enzymes. Dietary supplementation with the antioxidant butylated hydroxyanisole (BHA) is associated with inhibition of forestomach tumors in rodents.49Using short-term organ culture of the Barrett’s mucosa and normal esophagus from the same patient as well as Barrett’s adenocarcinoma cell lines, we have found induction of the glutathioneS-transferase-.rr form mRNA using both the glucocorticoid, dexamethasone, and BHA (Compton et al, manuscript submitted). These studies indicate that glutathione-S-transferase-n,the form that is low in Barrett’s metaplasia, is indeed inducible. Future studies will examine the effects of induction of glutathione-S-transferaseon the prevention of carcinogen-DNA binding in Barrett’s mucosa.

1060

BEER & STONER

Role of Diet and Obesity in Chemoprevention of Esophageal Adenocarcinoma

A recent study reported the analysis of a population-based, casecontrolled interview of 174 white men with esophageal adenocarcinoma in which the role of obesity and diet in the etiology of adenocarcinoma of the esophagus was in~estigated.'~ The analysis revealed that the risk of adenocarcinoma is greatest among those subjects with the greatest body mass and least among those with the lowest body mass. No significant associations were observed for total calories from food, number of meals per day, fat intake, or coffee and tea consumption. Interestingly, risk was highest among those individuals consuming the least amount of vegetables and raw fruit, with some indication of a dose response with the amount of consumed cruciferous vegetables. Therefore, diet appears to be an important source of compounds present in vegetables and fruit that may protect against development of adenocarcinoma. As indicated, induction of glutathione-S-transferasesmay be one factor involved in the effects of cruciferous vegetables. Raw vegetables and fruit also contain many substances such as carotenoids; antioxidants such as vitamins C, A, and E; dietary fiber; indoles; isothiocyanates; and dithiolthiones, which show cancer-protective effects in animal studies.66 Modulation of these components in the diet of those at highest risk of developing adenocarcinoma-that is, patients with intestinal-type Barrett's metaplasia-may be an effective chemopreventative strategy. Furthermore, increased body mass in obese individuals can predispose them to GERD, often as a consequence of hiatal hernia, and lead to reflux esophagitis, which is one risk factor for the development of Barrett's metap1a~ia.l~ It is unknown at what stage diet and obesity might influence esophageal cancer risk. It could either influence the initial development of Barrett's metaplasia or, once developed, its further progression towards adenocarcinoma. Chemoprevention of Esophageal Cancer Using Anti-inflammatory Agents Previous studies have indicated a protective role of regular use of aspirin and nonsteroidal anti-inflammatory drugs (NSAIDs) for cancers at a number of different tissue sites.21, 57, 74, 90, 91 In a recent large, population-based, case-control study, an association between current aspirin and NSAID use and reduced risk of esophageal adenocarcinoma, squamous cell carcinoma, and noncardia gastric adenocarcinoma was observed.ls An approximately 50% reduced risk of these cancers was observed in individuals reporting use of aspirin or NSAIDs once a week for 6 months or more. These results support previous studies of aspirin

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1061

use and reduced esophageal cancer risk.21,90 The mechanism underlying this reduced cancer risk may be due to the role of aspirin and NSAIDs as inhibitors of cyclo-oxygenase (reviewed in ref. SS), and thus in reducing prostaglandin synthesis. Prostaglandins have been shown to affect tumor cell growth, metastasis, and host immune function.88In colon adenocarcinoma cell lines, induction of apoptosis (programmed cell death) was also observed using treatment with NSAIDS.~~, 91,92 Aspirin and NSAIDs could be potent inhibitory agents that act at either one or more different stages of esophageal cancer development. For example, chronic gastroesophageal reflux is associated with esophagitis of varying severity as well as the development of Barrett's metaplasia.80-82, Io5 Chronic inflammation is often continued in patients with Barrett's metaplasia. Reactive oxygen species (ROS) are increased in certain inflammatory disorders such as ulcerative colitis and Crohn's disease, and recent studies indicate that ROS are elevated as well as in esophagitis and Barrett's metaplasia.60ROS can be produced by phagocytic cells such as neutrophils, macrophages, and epithelial cells, and they can result in damage to cellular DNA.Iol Activated macrophages also produce nitric oxide (NO) from arginine by the inducible NOsynthase pathway.8,40,47,66 Large amounts of NO may be produced during chronic inflammation and react with dissolved oxygen to produce N203 and N204, which can nitrosate amines. As mentioned, vegetables and fruit are good sources of vitamins and polyphenols that can inhibit nitrosation. The use of anti-inflammatory agents such as aspirin and NSAIDs therefore could potentially act at relatively early stages, reducing genetic damage either during esophagitis or from the inflammation associated with Barrett's metaplasia. Aspirin and NSAIDs may also be inhibitory to esophageal adenocarcinoma development by decreasing cell proliferation within Barrett's mucosa (or adenocarcinoma) via cyclo-oxygenase inhibition and decreased prostaglandin synthesis. Increased amounts of prostaglandins may affect cell mit~genesis,'~ although exceptions to this have been Increased prostaglandin synthesis occurs by cyclo-oxygenase-catalyzed metabolism of arachidonic acid, and there are two isoforms of cyclo-oxygenase: Coxl and Cox2 (reviewed in ref. 84). Coxl is constitutive, and Cox2 is not detectable in most tissues but is inducible by specific cytokines, serum, and certain tumor promoter^.'^, 5y Inflammation increases prostaglandin synthesis, which may be due to upregulation of Cox2, and inflammation is a recognized risk factor for epithelial carcinogenesis.'O' Interestingly, Cox2 is upregulated in many human cancers, including the related gastric carcinomas,7° and it is reported to be increased as well in Barrett's adenocarcinomas.20Selective Cox2 inhibitors may be attractive for chemoprevention because of fewer gastrointestinal side effects often associated with aspirin and NSAIDs. Inter-

1062

BEER & STONER

estingly, the selective Cox2 antagonist (SC58125) has been shown to inhibit the growth of ras-transformed rat intestinal cells.75 Induction of Apoptosis by ChemopreventiveAgents

A potential mechanism for the inhibitory effects of aspirin and NSAIDs on the incidence of esophageal adenocarcinoma,ls in addition to those discussed earlier, could be by induction of apoptosis via cyclooxygenase inhibition. Cells overexpressing Cox2 show increased expression of Bcl-2 and resistance to butyrate-induced a p o p t o ~ i s Treatment .~~ with sulindac sulfide (an NSAID) reversed the resistance to apoptosis caused by Cox2 overexpression. Although induction of apoptosis by these agents in esophageal adenocarcinomas has not yet been shown, recent evidence indicates there is reduced apoptosis in Barrett’s adenocarcinoma^.^^ In both normal esophagus and nondysplastic Barrett’s metaplasia, apoptotic cells are observed near the proliferative zones of these tissues, whereas in adenocarcinomas apoptosis is very infrequently detected. A decrease in the cell surface expression of the apoptosisinducing Fas receptor was observed in the Barrett’s adenocarcinomas but not in nondysplastic Barrett’s metaplasia. Loss of cell surface Fas expression in esophageal adenocarcinoma may represent one mechanism by which these cells escape immune cell-mediated apoptosis. In dysplastic Barrett’s metaplasia some loss of cell surface Fas expression was also observed, indicating that this change occurs relatively early in neoplastic progression of this tissue. Whether aspirin and NSAIDs might influence cell surface Fas expression in these tumors or dysplastic Barrett’s metaplasia will be of interest to determine. Dysplasia and Genetic Alterations as Biomarkers for Chemoprevention Studies In patients with Barrett’s metaplasia who are undergoing endoscopic surveillance, approximately 12% to 18% will develop dysplasia after 3 to 5.2 years.27,69 Dysplastic features include the acquisition of histologic changes such as cellular atypia, an increased nuclear to cytoplasmic ratio, loss of normal cellular architecture, and increased nuclear stratification. Because 10% to 25% of patients with low-grade dysplasia may progress to high-grade dysplasia, and nearly 68% to 100% of adeno69* dYScarcinomas are associated with dysplastic Barrett’s muc0sa,2~, plasia is an important indication of neoplastic progressional events in this metaplastic epithelium. As such, dysplasia is an excellent biomarker for potential chemoprevention studies in this mucosa. The molecular changes that are associated with or causally involved in the progression of Barrett’s metaplasia to low-grade dysplasia, high-grade dysplasia,

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1063

and finally adenocarcinoma is an area of active investigation. An extensive number of specific genetic changes have been detected in the neoplastic progression of Barrett’s metaplasia and in esophageal adenocarcinoma, and their review is beyond the scope of the present discussion. In general, however, these changes include the presence of aneuploidy, specific chromosomal abnormalities, alterations in tumor suppressor genes and oncogenes, changes in growth factor expression and their receptors, cell adhesion molecules, apoptosis, and cellular markers of differentiation. Many of the genetic alterations described could be suitable biomarkers for chemoprevention studies; however, the tumor suppressor gene p53 may be an excellent candidate because it is frequently altered by mutation, loss of heterozygosity, or both in esophageal adenocarcinomas, and alteration in this gene is a relatively early event occurring in dysplastic Barrett’s mucosa.10-12 Little information is available, however, regarding the modulation of any of these biomarkers by compounds that may be useful chemopreventative agents, and this represents an important area in which further studies are needed. Models of Barrett’s Adenocarcinoma

Chemoprevention studies of Barrett’s metaplasia and esophageal adenocarcinoma have been hampered by the lack of suitable animal models. In a canine model in which chronic gastroesophageal reflux is induced by surgery, a columnar mucosa is produced in the distal esophagus41;however, this mucosa is not of the intestinal-type and the animals do not develop adenocarcinoma. A rodent model has also been developed in which the esophagus and duodenum are surgically connected to induce gastroduodenal-esophageal re flu^.*^ With administration of carcinogens, tumors are produced but are primarily squamous cell carcinomas. Only a very small percentage of adenocarcinomas are observed, and those occurring often show a mixed squamous-adenocarcinoma histology. Recently, however, this rat model was modified so that following esophageal-duodenum anastomosis to induce gastroduodenalesophageal reflux, only iron dextran (50 mg / kg, intraperitoneally) was administered to the animals without any carcinogen treatment. The resulting iron overload results in apparently high levels of oxidative damage. Interestingly, putative intestinal-type Barrett’s esophagus was observed at the esophageal-duodenal anastomosis, which was followed by the development of dysplasia and adenocarcinoma. What must be clarified in this model, however, is whether this is actually intestinaltype Barrett’s metaplasia as the authors conclude24or, rather, cancer development in remodeled duodenum at the site of esophageal-anastomosis. At 31 weeks the putative Barrett’s metaplasia was an average of 0.22 mm, which is only tens of cells wide. Furthermore, the types of

1064

BEER & STONER

mucin detected and the presence of goblet cells may also be present in the duodenal mucosa and may not be proof of intestinal metaplasia. Nonetheless, development of adenocarcinomas in this model, using iron dextran treatment and no other carcinogen, represents an important finding and a potentially useful model for chemoprevention studies. In addition to the preceding animal models, information regarding the activities of potential chemopreventative agents can be also be obtained using organ-culture of Barrett’s metaplasia as described earlier, as well as cell lines of Barrett’s adenocarcinoma. A number of laboratories have reported the short-term culture of Barrett’s m~cosa,’~, and we have utilized organ culture to examine the metabolic and protective mechanisms present in this metaplastic tissue. We have also utilized three Barrett’s adenocarcinoma cell which will be useful for examining the direct effect of potential chemopreventative agents on fully transformed cells of this specific origin. 36799

Summary Chemoprevention is very likely to represent a useful approach to help reduce the risk of Barrett’s adenocarcinoma development. Reduction of cell proliferation in the intestinal-type Barrett’s metaplasia, induction of protective enzyme systems, or inhibition of metabolic capabilities all need to examined in this important and increasingly frequent disease. Chemoprevention strategies need to be designed for both the early and later stages of the disease process. For example, by understanding the risk factors in patients with GERD who go on to develop Barrett’s metaplasia, chemopreventative approaches might be developed to reduce the development of the most important risk factor for adenocarcinoma: intestinal-type Barrett’s metaplasia. In those individuals who have already developed intestinal-type Barrett’s metaplasia, chemoprevention may be used to reduce the further progression to dysplasia, a significant risk factor for subsequent adenocarcinoma development. Finally, in patients who develop dysplasia, new chemopreventative approaches could attempt to induce apoptosis in these neoplastic cells. New model systems should help in the identification of novel approaches for chemoprevention in patients at high risk for developing Barrett’s adenocarcinoma.

SQUAMOUS-LINED ESOPHAGUS (EPITHELIAL DYSPLASIA)

More than 90% of esophageal cancers worldwide are squamous cell carcinomas (SCCS).*~These tumors arise from epithelial dysplastic le-

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1065

sions and exhibit three types of growth patterns: (1)fungating, in which the tumor appears as an exophytic mass with intraluminal growth; (2) ulcerating, characterized by a large ulcer that may penetrate into the mediastinum; and (3) infiltrating, in which the tumor exhibits noncohesive growth in both horizontal and vertical planes of the esophagus. Microscopically, esophageal SCCs range from well-differentiated tumors that exhibit keratinization, intracellular bridges, moderate nuclear atypia, and minimal necrosis, to poorly differentiated lesions that can be difficult to categorize as squamous in origin, are mitotically active, and contain substantial necrosis. The prognosis for patients with advanced SCC of the esophagus is poor; the overall 5-year survival rate of patients with metastatic disease is less than 10%. The principal precursor lesion for esophageal SCC is epithelial dysp l a ~ i aMicroscopically, .~~ epithelial dysplasia is characterized by an accumulation of atypical cells with nuclear hyperchromasia, abnormally clumped chromatin, and loss of polarity and frequent mitoses that are often abnormal in shape. Several prospective studies suggest that esophageal SCC probably develops through a progressive sequence from lowgrade dysplasia to high-grade dysplasia and, eventually, carcinoma.2,39, 78 Thus, as discussed previously for Barrett’s esophagus and adenocarcinoma, a major approach for the chemoprevention of esophageal SCC is to prevent the progression of epithelial dysplasia to SCC.

Epidemiology and Etiology of Esophageal Squamous Cell Carcinoma

Esophageal SCC in humans occurs worldwide with a variable geographic distribution. It has been estimated to rank seventh in order of cancer occurrence, both sexes combined.61The highest incidence is in China, parts of Central Asia, India, the Transkei region of South Africa, and northern Iran.37There is a higher incidence of the disease in men, with a male to female ratio of 2:l or higher. The disease occurs consistently among the poor in most areas of the world, where the diet is often restricted and nutritional imbalance is common. The marked variation in the incidence of esophageal SCC throughout the world points to a cause for this disease in which nutritional and environmental factors play a significant role. Indeed, an increased risk for esophageal SCC has been correlated with tobacco use, consumption of alcoholic beverages, and the ingestion of salt-cured, salt-pickled, and moldy foodsa4Some of these products are frequently contaminated with N-nitrosamine carcinogens and/or fungal toxins. Extensive research in China and in the Transkei region of South Africa has suggested that Nnitrosamines in the diet are probable etiologic factors in these high-

1066

BEER & STONER

incidence areas.42Nitrites, nitrates, and secondary and tertiary amines, which can be converted to nitrosamines in the stomach, are also widely distributed in foodstuffs. Other factors associated with risk for development of esophageal SCC include deficiencies in various vitamins and minerals.@In addition, the drinking of hot beverages such as tea is thought to contribute to the occurrence of the disease. Finally, fungal dection and invasion in esophageal tissues leads to localized inflammation and irritation that may contribute to the development of esophageal cancer.

Approaches to the Prevention of Esophageal Squamous Cell Carcinoma In view of the exposures described, one approach to the prevention of esophageal SCC is through changes in lifestyle, that is, avoidance of tobacco and alcohol, and of high-salt foods that are contaminated with microbial toxins, nitrosamines, and nitrosamine precursors. This approach mandates educational efforts to inform the public of the major role of these lifestyle factors in the development of the disease. The inclusion in the diet of adequate quantities of vegetables and fruits is another approach, because these foodstuffs would provide necessary micronutrients, including vitamins and minerals, for the prevention of esophageal cancer. Indeed, epidemiologic studies have shown the protective effect of fruits, in particular, on the development of esophageal SCC.9Chemoprevention, to address factors associated with the etiology and progression of the disease, is another approach that has particular relevance in those high-incidence areas of the world where exposure levels to carcinogens are high and dietary deficiencies are common. The remainder of this article discusses chemoprevention approaches that either have been used or could be attempted for the prevention of esophageal SCC. Chemoprevention studies in a rat model of esophageal SCC and their relevance to human esophageal SCC are also described. Inhibitors of Metabolic Activation of Procarcinogens in Esophageal Mucosa The specific endogenous and environmental chemicals that play a role in the development of esophageal SCC are unknown, but undoubtedly they include compounds in tobacco smoke and in the diet. These include the tobacco-specific nitrosamines, of which N-nitrosonornicotine (NNN) is probably the most important for the esophagus, other nitrosamines such as N-nitrosodimethylamine and N-nitrosodiethylamine that are present both in tobacco smoke and in the diet, and any nitrosamines produced in the acidic environment of the stomach through the interac-

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1067

tion of nitrites with secondary and tertiary a m i n e ~The . ~ ~ detection of 06-methylguanine in the DNA of normal esophageal tissue taken from esophageal cancer patients in China suggests a role for methylating nitrosamines in esophageal cancer d e v e l ~ p m e n t .Other ~ ~ carcinogens of probable importance include the polycyclic aromatic hydrocarbons (PAHs)-for example, benzo(a)pyrene-that are found in tobacco smoke and in charcoal-broiled and smoked foods. In addition, fungal toxins produced by the Fusarium species, such as the trichothecenes and fumonisins present in food grains in some high-incidence areas for esophageal SCC, are likely to be causative agents of esophageal SCC.5,43 T-2 toxin, a trichothecene compound, has been shown to induce focal basal cell hyperplasia and dysplasia in cultured explants of human fetal esophaUS.^^ Toxins produced by other fungal species such as Alternaria and Geotvichium are also likely to contribute to the development of esophageal SCC. Metabolic studies conducted in our laboratory and in the laboratories of our collaborators have demonstrated the ability of human esophagus to metabolize several procarcinogens in tobacco smoke and in the diet including various PAHs and nitro~amines.~, 15, 2y Considerable variation was observed in the ability of esophageal tissues from different individuals to metabolize these procarcinogens into DNA-damaging species, which is undoubtedly associated with one’s risk of developing esophageal SCC through exposure to these chemicals. Several chemoprevention approaches could be used in attempts to reduce the ability of endogenous and environmental procarcinogens to induce genetic damage in esophageal mucosa. Some chemopreventives of potential use in this regard are suggested by investigations into the rat model of esophageal SCC. One approach would be to reduce the formation of nitrosamines from nitrites and secondary amines in the stomach by supplementing the diet with compounds such as ascorbic acid, various phenols, and/or sulfhydryl corn pound^.^^ These agents either decrease the production of nitrosamines or scavenge nitrite by reacting with it to form non-mutagenic nitrosamines. Of relevance to this approach is the observation that vitamin C supplementation was found to reduce the risk of esophageal cancer development in current tobacco smoker^.^ The dietary addition of lnhibitors of the metabolic activation of nitrosamines and PAHs might also be protective. This approach is suggested by the observation that human esophageal mucosa expresses cytochrome P450’s 1A2/1, 2B6, 2C, 3A, and 3A4/358and is capable of metabolizing nitrosamines and PAHs.~,15, 2y*50 Chemopreventive agents of potential use in inhibiting the metabolic activation of nitrosamines include phenethyl isothiocyanate and certain polyphenolic compounds such as ellagic acid or the green tea polyphenol, epigallocatechin-3-gallate (EGCG). These compounds have been shown to inhibit nitrosamine-induced DNA damage and tumorigenesis in the rodent

1068

BEER & STONER

esophagus.M,45, 85, 97 Epidemiologic studies indicate that green tea consumption protects against the development of esophageal SCC in humans, although this conclusion is controversial.22Potential inhibitors of the metabolic activation of PAHs in the esophagus include benzyl isothiocyanate and the polyphenols ellagic acid and curcumin, because these compounds are among the many chemopreventives known to exhibit protective effects against PAH-induced carcinogenesis in various animal model systems.85,loo The development of inhibitors of the metabolic activation of fungal toxins implicated in the etiology of human esophageal cancer is an important objective that has not been attempted to date. Inducers of Protective Mechanisms in Esophageal Mucosa

As stated earlier, inhibition of the covalent binding of carcinogens to DNA can be observed using compounds that increase inducible cellular detoxification reactions. One of the most important systems involves the glutathione-S-transferases (01, p, T, 0), which catalyze the conjugation of endogenous and exogenous compounds with the nonprotein thiol glutathione. Normal human esophageal epithelium has been shown to contain the a, p, and T classes of glutathione transferases, and the levels of these enzymes in the esophagus are significantly higher than in normal gastric ~ a r d i aIn . ~view ~ of this, the dietary addition of chemopreventives such as diallyl sulfide, sulforaphane, or oltipraz, which promote carcinogen detoxification through stimulation of glutathione-S-transferases and other phase I1 enzymes, could be '02, "' Diallyl sulfide is a potent inhibitor of nitrosamine carcinogenesis in the rat but no experimental studies in the esophagus have been undertaken with either sulforaphane or oltipraz. Other chemopreventives of potential use include ellagic acid and phenethyl isothiocyanate, which stimulate glutathione-S-transferase activities in the rat esophagus.', 94 Another parameter of importance in inhibiting carcinogenesis is the prevention of oxidative damage by the action of glutathione peroxidases. Glutathione peroxidases are enzymes that catalyze the reduction of organic hydroperoxides and hydrogen peroxide produced by cellular m e t a b ~ l i s mChemopreventives .~~ that stimulate the activities of glutathione peroxidases are likely to reduce oxidative damage. Although no data are available for the human esophagus, oltipraz and phenethyl isothiocyanate have been shown to induce glutathione peroxidases in the rat esophagus.94 Cellular Proliferation as a Target for Chemoprevention Studies An important strategy for inhibiting the development of esophageal SCC is to prevent the progression of premalignant lesions such as epithe-

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1069

lial dysplasia into malignant SCC. The feasibility of this approach is underscored by the use of various endoscopic and cytologic methods to identify dysplastic lesions in the esophagus and to follow their progression. Importantly, the combined use of "balloon cytology" coupled with endoscopy in high-incidence areas of China has been useful in identifying individuals with premalignant lesions and improving their survival by clinical intervention. As might be expected, studies in highincidence areas for esophageal cancer in China have demonstrated a relationship between the pattern and rate of esophageal cell proliferation, as measured by the incorporation of tritiated thymidine into DNA, and the risk for the disease. Mufioz et a15'j reported that the rate of esophageal cell proliferation in the intermediate and superficial layers of the epithelium was more rapid in a high-risk group than in a lowrisk group. Another study indicated that the rate of cellular proliferation was lowest in normal esophageal epithelium and increased progressively in subjects having hyperplasia, mild dysplasia, and moderate d y s p l a ~ i a . ~ ~ Chemoprevention trials in China have been conducted in attempts to identify individual agents or combinations of agents that inhibit cell proliferation rates in the esophagus, but these have met with limited or no success. One study involving 3318 adult residents of Linxian, China evaluated the effect of daily supplementation of multiple vitamins and minerals for 30 months on proliferation rates in tissues from the following histologic categories: normal; acanthosis; esophagitis; squamous dysplasia; and SCC.'j*Analysis showed no treatment effect on the overall amount of squamous epithelial proliferation in any of the histologic categories. Similarly, supplementation of the diets of 200 subjects in a high-incidence area of China with calcium (1200 mg/day) for 11 months did not result in reduced rates of esophageal cell proliferation in either hyperplastic or dysplastic This result was disappointing in that high levels of calcium have been shown in vitro to inhibit the proliferation of, and stimulate the differentiation of, squamous epithelial cells of the e s ~ p h a g u sAlthough .~ cell proliferation rates were not measured, a study of the effects of antitumor-B (ATB; a mixture of Chinese herbs), retinamide (4-ethoxycarbophenylretinamide),and riboflavin supplementation in the diet of subjects diagnosed with mild or marked esophageal dysplasia in Hunan, China revealed reduced rates of cancer development from marked d y ~ p l a s i aATB . ~ ~ treatment for 3 and 5 years reduced the cancer development rate from marked esophageal dysplasia by 52% and 47%, respectively. Retinamide lowered the cancer development rate by 37% to 43%, and riboflavin from 22% to 35%; the riboflavin results were not significant. These observations suggest that compounds in antitumor-B and retinamide probably influence cell proliferation rates in the esophagus and deserve additional experimentation. In addition, the level of expression of ornithine decarboxylase (ODC) mRNA in esopha-

1070

BEER & STONER

geal SCC has been found to exceed that in normal esophageal mucosa by a factor of 15."O Thus, substances that block ODC such as difluoromethylomithine (DFMO), retinoids, certain polyphenols, and inhibitors of arachidonic acid may be effective chemopreventives. Chemoprevention Using Anti-inflammatory Agents and Inducers of Apoptosis The development of esophageal SCC in humans is associated with a significant inflammatory process.37Thus, it is not totally unexpected that an association was made recently between current aspirin and nonsteroidal anti-inflammatory drug (NSAID) use and a reduced risk of esophageal SCC.lS To date, there are no reports of the expression of the cyclooxygenase enzymes (Cox1 and Cox2) in normal or tumorous squamous epithelium of the esophagus or of prostaglandin levels in these tissues. Thus, this is an area for active investigation. Similarly, although the retinoids are known to stimulate apoptosis, there have been no studies of the effects of chemopreventives on the apoptotic rate in squamous epithelium of the esophagus. This is also a subject for active investigation. Rat Model for Chemoprevention of Esophageal Cancer The rat has been used almost exclusively as an animal model for SCC of the esophagus. The model has been described in detail elsewhere31and will be discussed only briefly here. Several nitrosamines act as fairly specific inducers of tumorigenesis in the rat esophagus, including NNN and N-nitrosomethylbenzylamine (NMBA). NMBA is by far the most potent inducer of esophageal tumors in rats; tumors can be induced in 15 weeks or less. A number of readily discernible preneoplastic lesions are produced as well, including simple hyperplasia, leukoplakia, and epithelial dysplasia. The majority of tumors induced are squamous papillomas, although a small incidence of basal or squamous cell carcinomas may also be detected. Owing to occlusion of the esophagus and/or induction of respiratory distress, a large esophageal papilloma can be life-threatening. Thus, many NMBA-treated animals are unable to survive long enough to develop a large number of carcinomas. A number of chemopreventive agents that influence the metabolism of carcinogens exhibit inhibitory activity against NMBA-induced esophageal tumorigenesis in rats. Diallyl sulfide (200 mg/kg) administered orally 3 hours prior to NMBA administration completely inhibited NMBA-induced preneoplastic lesions and tumors.9s As stated earlier, diallyl sulfide appears to act principally by stimulation of phase I1 enzymes involved in the metabolic detoxification of carcinogens. Ellagic

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1071

acid, when given in the diet for the duration of the experiment at concentrations of 0.4 and 4.0 g/kg, significantly inhibits NMBA-induced esophageal This effect appears to be antagonized by co-administration of 13-cis-retinoic acid.16 At least some of the effect of ellagic acid appears to be due to its inhibitory effect on the metabolic activation of NMBA. Dietary ellagic acid inhibits NMBA activation in the esophagus of rats: whereas ellagic acid added to in vitro cultures of rat esophagus inhibits both NMBA activation and NMBA-induced DNA adduct formation.45Another mechanism of ellagic acid inhibition appears to be due to its ability to stimulate hepatic glutathione transferase and other enzymes involved in the metabolic detoxification of carcinogens.I One of the most interesting groups of compounds evaluated in the rat esophagus thus far are the arylalkylisothiocyanates. Dietary phenethyl isothiocyanate (PEITC) completely inhibits NMBA-induced esophageal tumorigenesis when administered at concentrations of 3.0 mmol/ kg diet or g~-eater?~,*~ This effect seems to be related to inhibition of NMBA activation, because PEITC inhibits NMBA-induced DNA adduct formation in vivo and has no effects on NMBA tumorigenesis if administered after i n i t i a t i ~ n Phenylpropyl .~~ isothiocyanate is a considerably more potent inhibitor than PEITC, whereas benzyl isothiocyanate and 4-phenylbutyl isothiocyanate are decidedly less potent than PEITC.lo3 Interestingly, phenylhexyl isothiocyanate (PHITC) actually enhances NMBA-tumorigenesis,86but the mechanism of this enhancement does not appear to be due to either a stimulatory effect of PHITC on NMBA activation or an inhibitory effect of PHITC on DNA repair.53 Ellagic acid fed after initiation at a dietary concentration of 4 g/kg has a modest inhibitory effect on NMBA esophageal turn~rigenesis.~~ Green tea inhibits NMBA tumorigenesis when either intact NMBA or its lo8 precursors, sodium nitrite and methylbenzylamine, is administered.lo7, Tea and tea components are known to possess both anti-initiating and anti-progression activities109;thus, it is not surprising that decaffeinated green tea and decaffeinated black tea lnhibit NMBA-induced esophageal tumors when administered during or after NMBA admini~tration.~~ Both dietary sulindac (125 ppm) and supplemental dietary calcium (2%)failed to inhibit NMBA-induced tumorigenesis when administered during or after NMBA admini~tration.~~ Dietary selenium has no inhibitory effects on NMBA-induced esophageal tumorigenicity, DNA adduct formation, or oncogene expression.MThese results suggest that only a few chemopreventives have been shown to inhibit the progression stages of esophageal carcinogenesis in the rat; thus, the search for additional inhibitors of this stage of esophageal carcinogenesis is an important goal.

1072

BEER & STONER

Summary

Esophageal SCC is a complex disease involving multiple etiologic factors. A number of preventive approaches could be taken to reduce the occurrence of the disease including changes in lifestyle and improved nutrition, for example, the inclusion of higher quantities of fruits and vegetables in the diet. Unfortunately, these primary prevention approaches are not easily implemented and often fall short in achieving marked reductions in disease occurrence. Chemoprevention offers another approach to reducing the risk of esophageal SCC that is likely to be useful, even though the clinical trials to date have not resulted in the identification of agents that produce marked inhibitory effects on the development of the disease. Given esophageal SCC’s complex etiology, it would appear that the most effective chemoprevention strategy would be to employ agents that reduce mutational events associated with exposure to esophageal carcinogens in combination with agents that inhibit the progression of epithelial dysplasia to esophageal SCC. The feasibility of addressing carcinogen-induced mutational events is underscored by the fact that many of the suspected esophageal carcinogens are known, and inhibitors of these carcinogens have been identified in animal model systems. In addition, biomarkers to assess the efficacy of anti-initiation agents, such as levels of phase I and I1 enzyme activities and of carcinogen:DNA adducts, can be measured. The identification of agents that inhibit the progression of dysplastic lesions to esophageal SCC has proven difficult; however, the results of the trial with ATB and retinamide are encouraging. Clearly, it seems important to identify the active chemopreventives in the antitumor-B herbal mixture. Further studies to identify strong inhibitors of tumor progression in the rat model for esophageal SCC are also needed. Biomarkers of cell proliferation (eg, PCNA, Ki67), cell differentiation (keratins), apoptosis, gene expression (EGFR, cyclin D1, p53), and nuclear/nucleolar morphometry can be used in studies to assess the efficacy of chemopreventives to either reverse esophageal dysplastic lesions or slow their rate of progression. The development of viable approaches toward the chemoprevention of esophageal SCC is truly an important goal in view of the poor prognosis of this disease.

References 1. Ahn D, Putt D, Kresty L, et al: The effects of dietary ellagic acid on rat hepatic and esophageal mucosal cytochromes P450 and phase I1 enzymes. Carcinogenesis 17821428, 1996 2. Anani PA, Gardiol D, Savary M, Monnier F:An extensive morphological and compar-

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1073

ative study of clinically early and obvious squamous cell carcinoma of the esophagus. Pathol Res Pract 187214-219, 1991 3. Autrup H, Stoner GD: Metabolism of N-nitrosamines by cultured human and rat esophagus. Cancer Res 42:1301-1311, 1982 4. Babcock MS, Marino MR, et al: Clonal growth and serial propagation of rat esophageal epithelial cells. In Vitro 19:403415, 1983 5. Bacon CW, Norred WP, Marijanovic DR, et al: The natural occurrence, mutagenic and other toxicological implications of fusarin C from Fusavium moniliforme. In Llewellyn GC, ORear CE (eds): Biodeterioration Research Three. New York, Plenum Press, 1990, pp 103-112 6. Barch DH, Fox CC: Selective inhibition of methylbenzylnitrosamine-inducedformation of esophageal Oh-methylguanine by dietary ellagic acid in rats. Cancer Res 48:7088-7092, 1988 7. Barone J, Taioli E, et a1 Vitamin supplement use and risk for oral and esophageal cancer. Nutr Cancer 18:3141, 1992 8. Bartsch H, Ohshima H, Pignatelli B, et al: Human exposure to endogenous N-nitroso compounds: Quantitative estimates in subjects at high risk for cancer of the oral cavity, oesophagus, stomach and urinary bladder. Cancer Surv 8:335, 1989 9. Block G, Patterson B, Subar A: Fruit, vegetables, and cancer prevention: A review of the epidemiological evidence. Nutr Cancer 18:l-29, 1992 10. Blount PL, Galipeau PC, Sanchez CA, et al: 17p Allelic losses in diploid cells of patients with Barrett’s esophagus who develop aneuploidy. Cancer Res 542292, 1994 11. Blount PL, Meltzer SJ, Yin J, et al: Clonal ordering of 17p and 5q allelic losses in Barrett’s dysplasia and adenocarcinoma. Proc Natl Acad Sci USA 90:3221, 1993 12. Blount PL, Ramal S, Raskind WH, et al: 17p Allelic deletions and p53 protein overexpression in Barrett’s adenocarcinoma. Cancer Res 51:5482, 1991 13. Brown LM, Swanson CA, Gridley, et al: Adenocarcinoma of the esophagus: Role of obesity and diet. J Natl Cancer Inst 87104, 1995 14. Cameron AJ, Ott BJ, Payne WS: The incidence of adenocarcinoma in columnar-lined (Barrett’s) esophagus. N Engl J Med 313:857, 1985 15. Castonguay A, Stoner GD, et al: Metabolism of tobacco-specific N-nitrosamines by cultured human tissues. Proc Natl Acad Sci USA 80669445697, 1983 16. Daniel EM, Stoner GD: The effects of ellagic acid and 13-cis-retinoic acid on Nnitrosobenzylmethylamine-induced esophageal tumorigenesis in rats. Cancer Lett 56~117-124,1991 17. DuBois RN, Awad J, Morrow J, et al: Regulation of eicosanoid production and mitogenesis in rat intestinal epithelial cells by transforming growth factor and phorbol ester. J Clin Invest 93:493, 1994 18. Farrow DC, Vaughan TL, Hansten PD, et al: Use or aspirin and other nonsteroidal anti-inflammatory drugs and risk of esophageal and gastric cancer. Cancer Epidemiol Biomarkers Prev 797, 1998 19. Fitzgerald RC, Omary MB, Triadafilopoulos G: Dynamic effects of acid on Barrett’s esophagus. J Clin Invest 98:2120, 1996 20. Fu S, Ramanujam KS, Meltzer SJ, et al: Expression and localization of inducible nitric oxide synthase and cyclooxygenase-2 in Barrett’s esophagus. Gastroenterology 13:A516, 1997 21. Funkhouser EM, Sharp GB: Aspirin and reduced risk of esophageal carcinoma. Cancer (Phila) 76:1116, 1995 22. Gao YT, McLaughlin JK, et al: Reduced risk of esophageal cancer associated with green tea consumption. J Natl Cancer Inst 86355, 1994 23. Gemer EW, Garewal HS, Emerson SS, et al: Gastrointestinal tissue polyamine contents of patients with Barrett’s esophagus treated with a+difluoromethylornithine. Cancer Epidemiol Biomarkers Prev 4:325, 1994 24. Goldstein SR, Yang G-Y, Curtis SK, et al: Development of esophageal metaplasia and adenocarcinoma in a rat surgical model without the use of carcinogen. Carcinogenesis 18:2265, 1997 25. Gray MR, Hall PA, Nash J, et al: Epithelial proliferation in Barrett’s esophagus by proliferating cell nuclear antigen immunolocalization. Gastroenterology 103:1769,1992

1074

BEER & STONER

26. Haggitt RC, Dean PJ: Adenocarcinoma in Barrett’s epithelium. In Spechler SJ, Goyal RK (eds): Barrett’s Esophagus: Pathophysiology, Diagnosis, and Management. New York, Elsevier Science Publishing, 1985, p 153 27. Hameetman W, Tytgat GNJ, Houthoff HJ, et al: Barrett’s esophagus: Development of dysplasia and adenocarcinoma. Gastroenterology 961249, 1989 28. Hamilton SR, Smith RR: The relationship between columnar epithelial dysplasia and invasive adenocarcinoma arising in Barrett’s esophagus. Am J Clin Pathol87301,1987 29. Harris CC, Autrup H, Stoner GD, et a1 Metabolism of benzo(a)pyrene, N-nitrosodimethylamine, and N-nitrosopyrrolidine and identification of the major carcinogenDNA adducts formed in cultured human esophagus. Cancer Res 39:44014406, 1979 30. Hayes JD, Pulford DJ: The glutathione S-transferase supergene family: Regulation of GST and the contribution of the isoenzymes to cancer chemoprotection and drug resistance. Crit Rev Biochem Mol Biol 30:445, 1995 31. Hecht SS, Stoner GD: Lung and esophageal carcinogenesis. In Aisner J, Arriagada R, Green MR, et a1 (eds): Comprehensive Textbook of Thoracic Oncology. Baltimore, Williams and Wilkins, 1996, pp 25-50 32. Hong MK, Laskin WB, Herman BE, et al: Expansion of the Ki-67 proliferative compartment correlates with degree of dysplasia in Barrett’s esophagus. Cancer 75:423, 1995 33 Hsia C-C, Tzian B-L, Harris CC: Proliferative and cytotoxic effects of Fusaviurn T: toxin on cultured human fetal esophagus. Carcinogenesis 4:1101-1107, 1983 34. Hu G, Han C, Wild CP, et a1 Lack of effects of selenium on N-nitrosomethylbenzylamine-induced tumorigenesis, DNA methylation, and oncogene expression in rats and mice. Nutr Cancer 18:287-295, 1992 35. Hughes SJ, Nambu Y, Soldes OS, et al: Fas/APO-1 (CD95) is not translocated to the cell membrane in esophageal adenocarcinoma. Cancer Res 575571, 1997 36. Khan SM, Pillay SP, Papadimos D, et al: Barrett’s oesophagus: A technique for the culture of Barrett’s oesophageal cells. J Gastroenterol Hepatol 12:606, 1997 37. Krasna MJ, Wolfer R S Esophageal carcinoma: Diagnosis, evaluation, and staging. In Aisner J, Arriagada R, Green MR, et a1 (eds): Comprehensive Textbook of Thoracic Oncology. Baltimore, Williams and Wilkins, 1996, pp 563-584 38. Kumble S, Omary MB, Fajardo LF, et a1 Multifocal heterogeneity in villin and EpCAM expression in Barrett’s esophagus. Int J Cancer 66:48, 1996 39. Kuwano H, Watanabe M, Sadanaga N, et al: Squamous epithelial dysplasia associated with squamous cell carcinoma of the esophagus. Cancer Lett 72141-147, 1993 40. Leaf CD, Wishnok JS, Tannenbaum SR Mechanisms of endogenous nitrosation [review]. Cancer Sum 8:323, 1989 41. Li H, Walsh TN, ODowd G, et a1 Mechanisms of columnar metaplasia and squamous regeneration in experimental Barrett’s esophagus. Surgery 115:176, 1994 42. Li MH, Ji C, Cheng SJ: Occurrence of nitroso compounds in fungi-contaminated foods: A review. Nutr Cancer 8:63-69, 1986 43. Lou Y, Yoshizawa T, Katayama T Comparative study on the natural occurrence of fusarium mycotoxins (trichothecenes and zearalenone) in com and wheat from hghand low-risk areas for human esophageal cancer in China. Appl Environ Microbiol 56~3723-3726, 1990 44. Mandal S, Stoner G D Inhibition of N-nitrosobenzylmethylamine-inducedesophageal tumorigenesis in rats by ellagic acid. Carcinogenesis 11:55-61, 1990 45. Mandal S, Shivapurkar NM, et a1 Inhibition of N-nitrosobenzylmethylamine metabolism and DNA binding in cultured rat esophagus by ellagic acid. Carcinogenesis 9~1313-1316,1988 46. Mann NS, Tsai MF, Nair PK Barrett’s esophagusin patients with symptomatic reflux esophagitis. Am J Gastroenterol 84:1494, 1989 47. Marletta MA: Mammalian synthesis of nitrite, nitrate, nitric oxide, and N-nitrosating agents. Chem Res Toxic01 1:249, 1988 48. Mamett LT: Aspirin and the potential role of prostaglandins in colon cancer. Cancer Res 52:5575, 19’92 49. McLellan LI, Harrison DJ, Hayes JD Modulation of glutathione-S-transferases and

-

CLINICAL MODELS OF CHEMOPREVENTION FOR THE ESOPHAGUS

1075

glutathione peroxidase by the anticarcinogen butylated hydroxyanisole in murine extrahepatic organs. Carcinogenesis 13:2255, 1992 50. Mirvish SS: Role of N-nitroso compounds (NOC) and N-nitrosation in etiology of gastric, esophageal, nasopharyngeal and bladder cancer and contribution to cancer of known exposures to NOC. Cancer Lett 93:17-48, 1995 51. Moore JH, Lesser EJ, Erdody DH, et a1 Intestinal differentiation and p53 gene alterations in Barrett’s esophagus and esophageal adenocarcinoma. Int J Cancer 56:487, 1994 52. Morse MA, Stoner G D Cancer chemoprevention: Principles and prospects. Carcinogenesis 141737-1746, 1993 53. Morse MA, Lu J, et al: Mechanism of enhancement of esophageal tumorigenesis by 6-phenylhexyl isothiocyanate. Cancer Lett 112119-125, 1997 54. Morse MA, Zu H, Galati AJ, et al: Dose-related inhibition by dietary phenethyl isothiocyanate of esophageal tumorigenesis and DNA methylation induced by Nnitrosomethylbenzylamine in rats. Cancer Lett 72:103-110, 1993 55. Mufti SI, Zirvi KA, Garewal HS: Precancerous lesions and biologic markers in esophageal cancer. Cancer Detect Prev 15:291, 1991 56. MuAoz N, Lipkin M, Crespi M, et al: Proliferative abnormalities of the esophageal epithelium in Chinese populations at high and low risk for esophageal cancer. Int J Cancer 36:187-189, 1985 57. Muscat JE, Stellman SD, Wynder EL: Nonsteroidal anti-inflammatory drugs and colorectal cancer. Cancer (Phila) 74:1847, 1994 58. Nakajima T, Wang R-S, et al: Expression of cytochrome P450s and glutathione Stransferases in human esophagus with squamous-cell carcinomas. Carcinogenesis 1 71477-1481, 1996 59. OBanion MK, Sadowsky HB, Winn V, et al: A serum and glucocorticoid-regulated 4kilobase mRNA encodes a cyclooxygenase-related protein. J Biol Chem 26623261, 1991 60. Olyaee M, Sontag S, Salman W, et a1 Mucosal reactive oxygen species production in oesophagitis and Barrett’s oesophagus. Gut 37168,1995 61. Parkin DM, Stjernsward J, Muir CS: Estimates of the worldwide frequency of twelve major cancers. Bull World Health Organ 62:163-182, 1984 62. Paul1 A, Trier JS, Dalton MD, et al: The histologic spectrum of Barrett’s esophagus. N Engl J Med 295:476, 1976 63. Peizhong L, Jinshen Z, Zhenpeng R, et al: Studies on medicamentous inhibitory therapy for esophageal precancerous lesions-3- and 5-year inhibitory effects of antitumor-B, retinamide and riboflavin. Proceedings of the Chinese Academy of Medical Sciences and Peking University Medical College 5:121-128, 1990 64. Peters WH, Roelofs HM, Hectors MP, et al: Glutathione and glutathione S-transferases in Barrett’s epithelium. Br J Cance 65. Peters WHM, Wobbes T, et al: Glu sferases in esophageal cancer. Carcinogenesis 14:1377-1380,1993 66. Potter JD, Steinmetz K Vegetables, fruit and phytoestrogens as preventative agents. In Stewart BW, McGregor D, Kleihues P (eds): Principles of Chemoprevention. IARC Sci Pub1 139. Lyon, France, IARC, 1996, p 61 67. Poynton AR, Walsh TN, OSullivan S, et a1 Carcinoma arising in familial Barrett’s esophagus. Am J Gastroenterol 91:1855, 1996 68. Rao M, Liu F-S, Dawsey SM, et al: Effects of vitamin/mineral supplementation on the proliferation of esophageal squamous epithelium in Linxian, China. Cancer Epidemiol Biomarkers Prev 3:277-279, 1994 69. Reid BJ, Blount PL, Rubin CE, et al: Flow cytometric and histological progression to malignancy of a cohort. Gastroenterology 1021212, 1992 70. Ristimaki A, Honkanen N, Jankala H, et al: Cycloxygenase-2 gene expression in human gastric carcinoma. Cancer Res 571276, 1997 71. Romero Y, Cameron AJ, Locke GR 111, et a1 Familial aggregation of gastroesophageal reflux in patients with Barrett’s esophagus and esophageal adenocarcinoma. Gastroenterology 1131449, 1997

1076

BEER & STONER

72. Sampliner RE, Garewal HS: Phase I1 trial of 13-cis-retinoic acid (isotretinoin) in Barrett's esophagus. Gastroenterology 94:A396, 1988 73. Schmidt HG, Riddell RH, Walther B, et al: Dysplasia in Barrett's oesophagus. J Cancer Res Clin Oncol 110:145, 1985 74. Schreinemachers DM, Everson RB: Aspirin use and lung, colon, and breast cancer incidence in a prospective study. Epidemiology 5:138, 1994 75. Sheng GG, Shao J, Sheng H, et al: A selective cyclooxygenase 2 inhibitor suppresses the growth of H-ms-transformed rat intestinal epithelial cells. Gastroenterology -, 113:1%83,1997 76. Shiff ST. Rieas " 8: Nonsteroidal anti-inflammatorv drues and colorectal cancer: Evolving concepts of their chemopreventive actions. Gastroenterology 113:1992, 1997 77. Shiff SJ, Koutsos MI, Qiao L, et al: Nonsteroidal anti-inflammatory drugs inhibit the proliferation of colon adenocarcinoma cells: Effects on cell cycle and apoptosis. Exp Cell Res 222:179, 1996 78. Shu YJ, Uan XQ, Jin SP: Further investigation of the relationship between dysplasia and cancer of the esophagus (Chinese). Chin Med J 6:39-41, 1981 79. S i g h JC, Barch DH, Stoner G D Effects of dietary phenethyl isothiocyanate, ellagic acid, sulindac and calcium on the induction and progression of N-nitrosomethylbenzylamine-induced esophageal carcinogenesis in rats. Carcinogenesis 16:llOl-1106, 1995 80. Spechler SJ: Endoscopic surveillance for patients with Barrett's esophagus: Does the cancer risk justify the practice? Ann Intern Med 106:902, 1987 81. Spechler SJ: The frequency of esophageal cancer in patients with Barrett's esophagus. Acta Endoscopica 22:541, 1992 82. Spechler SJ, Goyal R K Barrett's esophagus. N Engl J Med 315:362, 1986 83. Stoner GD, Mukhtar H: Polyphenols as cancer chemopreventive agents. J Cell Biochemistry 22:169-180, 1995 84. Stoner GD, Rustgi AK: Biology of esophageal squamous cell carcinoma. Gastrointestinal Cancers: Biology, Diagnosis, and Therapy 8:141-146, 1995 85. Stoner GD, Morrissey DT, et al: Inhibitory effects of phenethyl isothiocyanate on Nnitrosobenzylmethylamine carcinogenesis in the rat esophagus. Cancer Res 51:206> 2068, 1991 86. Stoner GD, Siglin JC, et al: Enhancement of esophageal carcinogenesis in male F344 rats by dietary phenylhexyl isothiocyanate. Carcinogenesis 162473-2476, 1995 87. Streitz JM, Andrews CW, Ellis FH: Endoscopic surviellance of Barrett's oesophagus. Does it help? J Thorac Cardiovasc Surg 105:383, 1993 88. Subbaramaiah K, Zakim D, Weksler B, et al: Inhibition of cyclooxygenase: A novel approach to cancer prevention. Proc SOCExp Biol Med 216:201, 1997 89. Suh 0, Mettlin C, Petrelli NJ: Aspirin use, cancer, and polyps of the large bowel. Cancer (Phila) 721171, 1993 90. Thun MJ, Namboodiri MM, Calle EE, et al: Aspirin use and risk of fatal cancer. Cancer Res 53:1322, 1993 91. Tsujii M, DuBois RN: Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase-2. Cell 83:493, 1995 92. Tsujii M, Kawano S, DuBois RN: Cyclooxygenase-2expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci USA 94:3336, 1997 93. Umbenhauer D, Wild CP, Montesano R, et al: O6Methyldeoxyguanosine in oesophageal DNA among individuals at high risk of oesophageal cancer. Int J Cancer 36:661665, 1985 94. van Lieshout EMM, Peters WHM, Jansen JBMJ: Effect of oltipraz, a-tocopherol, pcarotene and phenethylisothiocyanate on rat oesophageal, gastric, colonic and hepatic glutathione, glutathione S-transferase and peroxidase. Carcinogenesis 171439-1445, 1996 95. Wang LD, Lipkin M, Qui SL, et al: Labeling index and labeling distribution of cells in esophageal epithelium of individuals at increased risk for esophageal cancer in Huixian, China. Cancer Res 502651-2653, 1990 96. Wang L-D, Qui S-L, Yang G-R, et al: A randomized double-blind intervention study ,I

3

"

CLINICAL MODELS OF CHEMOI’REVENTION FOR THE ESOPHAGUS

1077

on the effect of calcium supplementation on esophageal precancerous lesions in a high-risk population in China. Cancer Epidemiol Biomarkers Prev 2:71-78, 1993 97. Wang ZY, Wang LD, Chen WF, et al: Inhibition of N-nitrosomethylbenzylamine (NMBzA)-induced esophageal tumorigenesis in rats by decaffeinated green tea and black tea. Proc Am Assoc Cancer Res 34:A746, 1993 98. Wargovich MJ, Woods C, et al: Chemoprevention of N-nitrosomethylbenzylamineinduced esophageal cancer in rats by the naturally-occurring thioether, diallyl sulfide. Cancer Res 48:6872-6875, 1988 99. Washington K, Gottfried MR, Telen MJ: Tissue culture of epithelium derived from Barrett’s esophagus. Gut 35:879, 1994 100. Wattenberg LW: Inhibitory effects of benzyl isothiocyanate administered shortly before diethylnitrosamine or benzo(dpyrene on pulmonary and forestomach neoplasia in A/J mice. Carcinogenesis 8:1971-1973, 1987 101. Weitzman SA, Gordon LI: Inflammation and cancer: Role of phagocyte-generated oxidants in carcinogenesis. Blood 76:655, 1990 102. Wilkinson J, Clapper ML: Detoxification enzymes and chemoprevention. Proc SOC Exp Biol Med 216:192, 1997 103. Wilkinson JT, Morse MA, et al: Effect of alkyl chain length on inhibition of Nnitrosomethylbenzylamine-induced esophageal tumorigenesis and DNA methylation by isothiocyanates. Carcinogenesis 16:lOll-1015, 1995 104. Williamson WA, Ellis FH Jr, Gibb SP, et al: Barrett’s esophagus: Prevalence and incidence of adenocarcinoma. Arch Intern Med 151:2212, 1991 105. Winters C Jr, Spurling TJ, Chobanian SJ, et al: Barrett’s esophagus. A prevalent, occult complication of gastroesophageal reflux disease. Gastroenterology 92118, 1987 106. Wu GD, Beer DG, Moore JH, et al: Sucrase-isomaltase gene expression in Barrett’s esophagus and adenocarcinoma. Gastroenterology 105:837,1994 107. Xu Y, Chi H: The effect of Chinese tea on occurrence of esophageal tumor induced by N-nitrosomethylbenzylamineformed in vivo. Biomed Environ Sci 3:406-412, 1990 108. Xu Y, Han C: The effect of Chinese tea on occurrence of esophageal tumor induced by N-nitrosomethylbenzylamine in rats. Biomed Environ Sci 335-42, 1990 109. Yang CS, Wang ZY Tea and cancer. J Natl Cancer Inst 85:1038-1049, 1993 110. Yoshida M, Hayashi H, et al: Elevated expression of the omithine decarboxylase gene in human esophageal cancer. Cancer Res 52:6671-6675, 1992 111. Zhang Y, Talalay P, Cho C-G, Posner GH: A major inducer of anticarcinogenic protective enzymes from broccoli: Isolation and elucidation of structure. Proc Natl Acad Sci USA 89:2399-2403, 1992

Address reprint requests to David G. Beer, PhD Department of Surgery Thoracic Surgery Research Laboratory University of Michigan B560, MSRB2 Box 0686 Ann Arbor, MI 48109