Barrett s Esophagus in Women: Demographic Features and Progression to High-Grade Dysplasia and Cancer

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2005;3: Barrett s Esophagus in Women: Demographic Features and Progression to High-Grade Dysplasia and Cancer GARY W. FALK,* PRASHANTHI N. THOTA,* JOEL E. RICHTER,* JASON T. CONNOR, and DON M. WACHSBERGER* *Center for Swallowing and Esophageal Disorders, Department of Gastroenterology and Hepatology, and Department of Biostatistics, The Cleveland Clinic Foundation, Cleveland, Ohio Background & Aims: Barrett s esophagus is traditionally considered a disease of older white men. The aims of this study were to compare the demographic features of Barrett s esophagus in men and women and to determine the prevalence and incidence of high-grade dysplasia and cancer in these patients. Methods: All patients enrolled in the Cleveland Clinic Barrett s Esophagus Registry from were studied. Age, ethnicity, number of endoscopies, hiatal hernia size, length of Barrett s segment, and prevalence and incidence of high-grade dysplasia and cancer were compared between men and women. Results: There were 839 patients in the registry (628 men and 211 women). Barrett s segment length was greater in men than in women (mean, vs cm, respectively; P.003). There were no significant differences for other parameters. There were 114 prevalence cases of high-grade dysplasia or cancer (96 men, 18 women). Women were less likely to have prevalent high-grade dysplasia or cancer than men (odds ratio, 0.52; 95% confidence interval, ; P.015). There were 13 incidence cases of high-grade dysplasia or cancer (11 men, 2 women) during a mean follow-up of 4.72 years, which was similar in both genders with an incidence rate of 1 in 179 patient-years of follow-up for women and 1 in 91 patient-years of follow-up in men. Conclusions: Twenty-five percent of patients in our registry are women. The length of Barrett s esophagus is greater in men than in women, but other features are similar. The prevalence of high-grade dysplasia/cancer in women is approximately half that of men. Incidence rates for high-grade dysplasia/cancer are similar in men and women, although the number of cases is small. Barrett s esophagus is a complication that develops in a minority of patients with gastroesophageal reflux disease. It is associated with an increased risk of adenocarcinoma of the esophagus, a cancer whose incidence has increased rapidly since the 1970s and remains on the rise. Barrett s esophagus is typically considered to be a disease of older white men. 1 Although Barrett s esophagus has been described in women, little is known about the gender differences in this disease. 2,3 Several studies report a higher prevalence of Barrett s esophagus in men than in women. 4 8 There are even more striking gender differences with respect to the prevalence of high-grade dysplasia and adenocarcinoma in patients with Barrett s esophagus However, there have been few studies focusing on Barrett s esophagus in women. 22,23 The aims of our study were 2-fold: (1) to compare the demographic features of Barrett s esophagus in men and women and (2) to determine the prevalence and incidence of highgrade dysplasia and cancer in women with Barrett s esophagus compared with that in men. Methods Patient Population The study population consisted of all patients enrolled in the Cleveland Clinic Foundation Barrett s Esophagus Registry from 1979 November 2002 (Figure 1). The registry is a prospectively collected database of patients seen in the Department of Gastroenterology and Hepatology since 1979 with a diagnosis of Barrett s esophagus. The registry includes the following data on each patient: age, gender, race, number of endoscopies, length of Barrett s segment, size of hiatal hernia, and histologic findings. For the purposes of this study, 2 groups of patients were evaluated. The registry group was defined as all patients enrolled in the Barrett s esophagus registry with biopsy proven intestinal metaplasia in the esophagus. The surveillance group was defined as patients who had at least 2 surveillance endoscopies with at least 1 year of follow-up. Follow-up duration was defined as the time from the index endoscopy to the last surveillance endoscopy. The length of Barrett s esophagus and size of hiatal hernia were determined from measurements taken at the time of the index endoscopy. Abbreviations used in this paper: CI, confidence interval; OR, odds ratio by the American Gastroenterological Association /05/$30.00 PII: /S (05)

2 1090 FALK ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 11 Patients were offered esophagectomy if high-grade dysplasia was confirmed by 2 expert gastrointestinal pathologists on 2 separate endoscopies. If patients were judged to be poor surgical candidates or if they declined surgery, options included surveillance every 3 months, photodynamic therapy, or endoscopic mucosal resection. Figure 1. Flow diagram of study patients in Barrett s registry. The surveillance group was defined as patients who had at least 2 surveillance endoscopies with at least 1 year of follow-up. Prevalence cases were patients who had high-grade dysplasia or cancer at the index endoscopy or within 1 year of index endoscopy. Endoscopy The diagnosis of Barrett s esophagus was established if the squamocolumnar junction was displaced proximal to the gastroesophageal junction, and intestinal metaplasia was detected by biopsy. 24 Patients with only cardiac or fundic type metaplasia were excluded from this study. The length of Barrett s epithelium was defined as the distance from the incisors to the distal end of the esophagus minus the distance from the incisors to the proximal margin of the Barrett sappearing epithelium. The distal end of the esophagus was defined by the most proximal margin of the gastric folds. A length 3 cm was considered to be short segment Barrett s esophagus, whereas a segment length 3 cm was considered to be long segment Barrett s esophagus. All patients with endoscopic findings suggestive of Barrett s esophagus had biopsies performed. If specialized intestinal epithelium was found, subsequent surveillance endoscopy was performed with a therapeutic endoscope. Four quadrant biopsies were taken every 2 cm with a jumbo biopsy forceps in patients without dysplasia and every 1 2 cm in patients with dysplasia, beginning at the gastroesophageal junction and extending cephalad. Any areas of mucosal irregularity or ulceration were also biopsied. Intervals of surveillance endoscopies varied over time in accordance with the evolving data on Barrett s esophagus. Endoscopic surveillance was typically done every 2 years in patients without dysplasia and more frequently in those with dysplasia. Histology Dysplasia in Barrett s esophagus was graded by using an established classification scheme 25,26 : (1) negative for dysplasia, (2) low-grade or indefinite for dysplasia, (3) high-grade dysplasia, and (4) intramucosal or submucosal adenocarcinoma. Low-grade dysplasia was reviewed by a single pathologist. If high-grade dysplasia was diagnosed, it was confirmed by a second pathologist. Incidence and Prevalence Cases Patients who had high-grade dysplasia or cancer at the index endoscopy or who developed high-grade dysplasia or cancer within 1 year of index endoscopy were considered to be prevalence cases. Incidence cases were defined as those who developed high-grade dysplasia or cancer at least 1 year after the index endoscopy. Statistical Analysis Baseline characteristics were compared between genders by using Wilcoxon rank sum tests for continuous variables and 2 tests for categorical variables. Results are expressed as mean 1 standard deviation unless otherwise stated. Logistic regression was used to identify risk factors for prevalent high-grade dysplasia or cancer. A multivariable logistic regression model was constructed that includes gender and all other predictors of prevalent high-grade dysplasia/ cancer at the P.05 level. Results are shown as odds ratios (ORs) with 95% confidence intervals (CIs). A Kaplan-Meier curve is provided for time to incident high-grade dysplasia/ cancer for men versus women. No exploratory Cox proportional hazards modeling could be performed because of the small number of incidence cases. Results Registry Patients The registry group consisted of 839 patients: 628 (74.9%) men and 211 (25.1%) women (Figure 1). The majority of these patients were from northeastern Ohio. Patients came from the following states: Ohio 82%, Pennsylvania 5%, West Virginia 2%, Michigan 2.4%, Indiana 1%, Florida 2%, and other states 5.6%. The demographics of the study population are shown in Table 1. The length of the Barrett s esophagus segment in men was greater than that in women ( cm versus cm, respectively; P.003). Forty-two percent of women had short segment Barrett s esophagus compared with 31% of men (P.0003). Of the men, 9.6% were younger than 40 years of age compared with 6.2% of the women with Barrett s esophagus (P.13). Otherwise, there were no significant demographic differences between men and women in terms of age, ethnicity, hiatal hernia size, and number of endoscopies (Table 1). Among the 839 patients in the registry group, a smaller proportion of women than men had dysplasia of any degree (women: 83.3% without dysplasia, 9% with

3 November 2005 BARRETT S ESOPHAGUS IN WOMEN 1091 Table 1. Characteristics of Barrett s Esophagus Registry Patients Women Men P value Number 211 (25.1%) 628 (74.9%) Age (y) a Race.42 White 96.7% 96.1% Black 1.6% 2.87% Others 0.95% 1.6% Hiatal hernia size (cm) a Barrett s length (cm) a Long segment 97 (58%) 368 (69%).003 Short segment 69 (42%) 163 (31%) Number of endoscopies a a Results expressed as mean 1 standard deviation. low-grade dysplasia, 3.4% with high-grade dysplasia, and 4.3% with cancer as compared with men: 78.5% without dysplasia, 7.5% with low-grade dysplasia, 6% with high-grade dysplasia, and 8% with cancer). However, these differences were not statistically significant (P.12). Prevalence Cases of High-Grade Dysplasia or Adenocarcinoma There were 114 prevalent cases of high-grade dysplasia or adenocarcinoma in the study population (13.6%; 95% CI, 11.4% 16.2%). Among the 628 men, there were 96 prevalence cases (15.3%; 95% CI, 12.6% 18.4%), and among the 211 women, there were 18 prevalence cases (8.5%; 95% CI, 5.3% 13.4%). Among the prevalence cases, women were older than men (70.6 vs 64.1 years, respectively, P.016). Women were less likely to have prevalent high-grade dysplasia or cancer than men on univariate analysis (OR, 0.52; 95% CI, ; P.015). This finding remained significant on multivariate analysis (OR, 0.53; 95% CI, ; P.035), even when controlling for age (OR, 1.43; 95% CI, for an additional 10 years in age) and Barrett s segment length (OR, 3.4; 95% CI, for Barrett s segment length 3 cm). Otherwise, there Table 2. Prevalent Cases of High-Grade Dysplasia/Cancer: Gender Differences Women Men P value Number 18 (16%) 96 (84%).04 Age (y) a Race.53 White 100% 98% Black 0% 2% Others 0% 0% Hiatal hernia size (cm) a Barrett s length (cm) a Number of endoscopies a a Results expressed as mean 1 standard deviation. were no differences in the ethnicity, hiatal hernia size, Barrett s segment length, and number of endoscopies between men and women with prevalent high-grade dysplasia or cancer (Table 2). Surveillance Patients At study entry, there were 725 patients without high-grade dysplasia or cancer (532 men and 193 women). Of these patients, 289 (39.9%) had 2 or more endoscopies and 1 year or more of follow-up for a total of patient-years of follow up. These patients constituted the surveillance group. The mean age of this group was 57.5 years, and the mean duration of follow-up was 4.72 years (range, years; median, 4.1 years). A mean of 3.9 endoscopies were performed on these patients. There were no differences in the age, ethnicity, hiatal hernia size, Barrett s segment length, number of endoscopies, and follow-up between men and women in this group (Table 3). Incidence Cases of High-Grade Dysplasia or Adenocarcinoma Of the 289 surveillance group patients, 13 developed high-grade dysplasia or cancer (11 men and 2 Table 3. Barrett s Esophagus Surveillance Patients: Gender Differences Women Men P value Number 79 (27.3%) 210 (72.7%) Age (y) Race.16 White 96.2% 98.57% Black 2.53% 1.43% Others 1.26% 0% Hiatal hernia size (cm) a Barrett s length (cm) a Number of endoscopies a Follow-up (y) a Results expressed as mean 1 standard deviation.

4 1092 FALK ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 11 Figure 2. Kaplan-Meier curve demonstrating the time from entry into Barrett s registry to the development of incident high-grade dysplasia (HGD)/cancer. There was no significant difference in the risk of incident HGD/cancer between men and women in the surveillance group (P.45). women) during follow-up. No women developed cancer. Among men undergoing endoscopic surveillance, 8 developed high-grade dysplasia, and 3 developed cancer. Figure 2 shows that progression is similar between men and women with incident high-grade dysplasia/cancer. The Kaplan-Meier estimates of progression to high-grade dysplasia or cancer at 5 and 10 years after study entry are 5.5% (95% CI, 1.5% 9.2%) and 9% (95% CI, 0.9% 16.3%), respectively. The incidence rate was 1 per 179 patient-years of follow-up in women and 1 in 91 patientyears of follow-up for men. Because of the small number of patients with incident high-grade dysplasia or cancer, no formal statistical comparisons were performed. Discussion This study has focused on the epidemiologic features of Barrett s esophagus in women. In our study, we found that the male to female ratio of patients with Barrett s esophagus was 3:1. In addition, the length of Barrett s esophagus was shorter in women than in men. The prevalence of high-grade dysplasia and cancer was 2-fold higher in men than in women, whereas there were no gender differences in the incidence of high-grade dysplasia and cancer, although there were few such cases in either gender. Interestingly, women with prevalent high-grade dysplasia or cancer tended to be older than men with high-grade dysplasia or cancer, in agreement with other recent findings. 27,28 Few studies to date have examined the gender differences in Barrett s esophagus and risk of progression to cancer. The male to female ratio in previous studies of Barrett s esophagus varied from 1.3:1 4.6:1 4 20,23 (Table 4). Male sex itself appears to be a predictor of Barrett s esophagus. In one study of 502 patients with gastroesophageal reflux disease, male sex was associated with an OR of 2.6 for development of Barrett s esophagus. 7 There is a much greater disparity between sexes when considering adenocarcinoma. Review of various prevalence studies showed male to female ratios varying from 2:1 12:1 for high-grade dysplasia/cancer. 2,4,5,10,14,17,21 Data from the Surveillance, Epidemiology, and End Results (SEER) program revealed that the age-adjusted incidence rates for esophageal cancer are 6- to 8-fold higher in men than in women. 29 In incidence case series, the rates of cancer varied between 1 in 167 patient-years to 1 in 468 patient-years in women. 4,9 20 This suggests one of two possibilities: women with Barrett s esophagus might progress more slowly than men to cancer, or cancer, if it develops, occurs later in life. Interestingly, one possible explanation for this comes from the recent work of Van Blankenstein et al, 23 who found that there was a 20-year age shift in the prevalence of Barrett s esophagus in women compared with men. However, it remains unclear whether, in fact, Barrett s esophagus develops later in life in women, or whether women come to endoscopy later in life than men. Various factors are reported to be associated with a higher risk of esophageal adenocarcinoma. These include male sex, 30,31 Barrett s esophagus length, 15,16,31 33 smoking, 31,34 and high body mass index. 35,36 Other suggested risk factors include tall stature (more than 1.7 m) 30 and hypertension. 29,37 Because all these risk factors are more prevalent in men than in women, they likely place men at higher risk of developing adenocarcinoma than women. The role of estrogen and progesterone in the pathogenesis and epidemiologic risk for Table 4. High-Grade Dysplasia/Cancer in Barrett s Esophagus: Prevalence Series Reported in the Literature First author, year (reference number) Country Number of patients (% women) High-grade dysplasia/ cancer (% women) Cameron, 1985 (10) USA 122 (27.8) 18 (0) Williamson, 1991 (5) USA 236 (27.5) 65 (7) Miros, 1991 (14) Australia 133 (24) 9 (1) Wright, 1996 (17) United Kingdom 348 (34.7) 47 (2) Cox, 1997 (21) United Kingdom 183 (32.2) 20 (35) Caygill, 1999 (2) United Kingdom 2102 (40) 59 (25.4) Bani-hani, 2000 (4) United Kingdom 597 (44.2) 31 (25.8)

5 November 2005 BARRETT S ESOPHAGUS IN WOMEN 1093 Barrett s esophagus and adenocarcinoma is unknown. However, breast-feeding by women is associated with a reduced risk of developing esophageal adenocarcinoma. 38 Once in our endoscopic surveillance program, we did not observe any significant difference in the risk of developing high-grade dysplasia or cancer between men and women. This is in contrast to previous studies reporting lower incidence rates of cancer in women compared with men with Barrett s esophagus. 4,9,11 18 However, our small number of incident cases of both highgrade dysplasia and cancer raises the possibility that our observation might be a result of a type II error. Our study does have a number of limitations. This study is subject to selection bias, because this is a single center study from a tertiary referral center. As such, our findings might not accurately reflect those encountered in either community-based or Veterans Administration settings. Because it is a retrospective review of a database, there is a chance of underreporting of cases. The data collected for the registry are limited to age, gender, race, number of endoscopies, length of Barrett s segment, size of hiatal hernia, and histologic findings. Because several endoscopists and pathologists were involved in the collection of data and pathology interpretation, there is clearly a chance of interobserver variability in the measurement of length of the Barrett s segment and interpretation of dysplasia. 26,39 However, to deal with the former issue, the length of Barrett s segment was determined from measurements taken at the time of the index endoscopy only. Furthermore, most of the biopsies were read by a gastrointestinal pathologist. It is important to recognize that the definition of Barrett s esophagus has evolved over time from the columnar-lined esophagus to at least 3 cm of endoscopically apparent columnar epithelium in the distal esophagus to the requirement of intestinal metaplasia within the 3-cm segment to intestinal metaplasia anywhere in the tubular esophagus without a minimum length requirement. 24,40 Furthermore, the classic landmark definition for the proximal margin of gastric folds was first described in This brings up the possibility of misclassification bias, although this limitation should apply equally to male and female patients in our study population and thus should not detract from the overall findings of our study. The definitions of incidence and prevalence are a subject of debate. Because the cancer or high-grade dysplasia cases developing within the first year are most likely due to sampling error rather than true incidence cases, we included them under the prevalent cases category. Several other studies have used a similar approach in defining incidence and the prevalence cases. 4,17 20 Another limitation is that surveillance data are available on only 405 of our patients, and few cases of high-grade dysplasia or esophageal adenocarcinoma developed in either gender. Despite the above limitations, we believe that this study offers a realistic snapshot of features of Barrett s esophagus in women encountered in a tertiary care center. Nevertheless, we do not know whether these findings reflect differences in health-seeking behavior of women compared with men. In conclusion, men outnumber women with Barrett s esophagus by a ratio of 3:1 and have longer segments of Barrett s esophagus. This might partly explain the higher prevalence rates of high-grade dysplasia or cancer in men than in women. However, once in a surveillance program, the risk of developing high-grade dysplasia or cancer is low in both men and women. References 1. Cameron AJ, Lomboy CT. Barrett s esophagus: age, prevalence, and extent of columnar epithelium. Gastroenterology 1992;103: Caygill CP, Reed PI, Hill MJ, et al. An initial comparison of nine centers registering patients with the UK National Barrett s Oesophagus Registry (UKBOR). Eur J Cancer Prev 1999;8: Conio M, Cameron AJ, Romero Y, et al. Secular trends in the epidemiology and outcome of Barrett s oesophagus in Olmsted County, Minnesota. Gut 2001;48: Bani-hani K, Sue-Ling H, Johnston D, et al. Barrett s oesophagus: results from a 13-year surveillance programme. Eur J Gastroenterol Hepatol 2000;12: Williamson WA, Ellis FH, Gibb SP, et al. Barrett s esophagus: prevalence and incidence of adenocarcinoma. Arch Intern Med 1991;151: Conio M, Filiberti R, Blanchi S, et al. Risk factors for Barrett s esophagus: a case-control study. Int J Cancer 2002;97: Campos GMR, DeMeester SR, Peters JH, et al. Predictive factors of Barrett s esophagus: multivariate analysis of 502 patients with gastroesophageal reflux disease. Arch Surg 2001;136: Caygill CP, Reed PI, Johnston BJ, et al. A single centre s 20 years experience of columnar-lined (Barrett s) oesophagus diagnosis. Eur J Gastroenterol Hepatol 1999;11: Spechler SJ, Robbins AH, Rubins HB, et al. Adenocarcinoma and Barrett s esophagus: an overrated risk? Gastroenterology 1984; 87: Cameron AJ, Ott BJ, Payne WS. The incidence of adenocarcinoma in columnar-lined (Barrett s) esophagus. N Engl J Med 1985;313: Robertson CS, Mayberry JF, Nicholson DA, et al. Value of endoscopic surveillance in the detection of neoplastic change in Barrett s oesophagus. Br J Surg 1988;75: Ovaska J, Miettinen M, Kivilaakso E. Adenocarcinoma in Barrett s esophagus. Dig Dis Sci 1989;34: Hameeteman W, Tytgat GNJ, Houthoff HJ, et al. Barrett s esophagus: development of dysplasia and adenocarcinoma. Gastroenterology 1989;96: Miros M, Kerlin P, Walker N. Only patients with dysplasia

6 1094 FALK ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 3, No. 11 progress to adenocarcinoma in Barrett=s esophagus. Gut 1991;32: Iftikhar SY, James PD, Steele RJ, et al. Length of Barrett s oesophagus: an important factor in the development of dysplasia and adenocarcinoma. Gut 1992;33: Van der Burgh A, Dees J, Hop WCJ. Oesophageal cancer is an uncommon cause of death in patients with Barrett s esophagus. Gut 1996;39: Wright TA, Gray MR, Morris AI, et al. Cost-effectiveness of detecting Barrett s cancer. Gut 1996;39: Ferraris R, Bonelli L, Conio M, et al. Incidence of Barrett s adenocarcinoma in an Italian population: an endoscopic surveillance programme. Eur J Gastroenterol Hepatol 1997;9: Drewitz DJ, Sampliner RE, Garewal HS. The incidence of adenocarcinoma in Barrett=s esophagus: a prospective study of 170 patients followed 4.8 years. Am J Gastroenterol 1997;92: Weston AP, Badr AS, Hassanein RS. Prospective multivariate analysis of clinical, endoscopic, and histological factors predictive of the development of Barrett s multifocal high-grade dysplasia or adenocarcinoma. Am J Gastroenterol 1999:94: Cox MA, Nwokolo CU, Loft DE. Screening for Barrett s cancer is worthwhile: a prospective single center experience. Endoscopy 1997;29:E Banki F, DeMeester SR, Mason RJ, et al. Barrett s esophagus in females: a comparative analysis of risk factors in females and males. Am J Gastroenterol 2005;100: Van Blankenstein M, Looman CW, Johnston BJ, et al. Age and sex distribution of the prevalence of Barrett s esophagus found in a primary referral endoscopy center. Am J Gastroenterol 2005; 100: Sampliner RE. The practice parameters committee of the American College of Gastroenterology: updated guidelines for the diagnosis, surveillance, and therapy of Barrett s esophagus. Am J Gastroenterol 2002;97: Reid BJ, Haggitt RC, Rubin CE. Observer variation in the diagnosis of dysplasia in Barrett s esophagus. Hum Pathol 1988;19: Montgomery E, Bronner MP, Goldblum JR, et al. Reproducibility of the diagnosis of dysplasia in Barrett s esophagus: a reaffirmation. Hum Pathol 2001;32: Solaymani Dodaran M, Logan RFA, West J, et al. Risk of oesophageal cancer in Barrett s esophagus and gastro-oesophageal reflux. Gut 2004;53: Wong A, Fitzgerald RC. Epidemiological risk factors for Barrett s esophagus and associated adenocarcinoma. Clin Gastroenterol Hepatol 2005;3: El-Serag HB, Mason AC, Petersen N, et al. Epidemiological differences between adenocarcinoma of the esophagus and adenocarcinoma of the gastric cardia in the USA. Gut 2002;50: Zhang Z, Kurtz RC, Sun M, et al. Adenocarcinoma of the esophagus and gastric cardia: medical conditions, tobacco, alcohol and socioeconomic factors. Cancer Epidemiol Biomarkers Prev 1996;5: Menke-Pluymers MBE, Hop WC, Dees J, et al. Risk factors for the development of an adenocarcinoma in columnar-lined (Barrett) esophagus. Cancer 1993;72: Rudolph R, Vaughan TL, Storer BE, et al. Effect of segment length on risk for neoplastic progression in patients with Barrett s esophagus. Ann Intern Med 2000;132: O Connor JB, Falk GW, Richter JE. The incidence of adenocarcinoma and dysplasia in Barrett s esophagus: report on Cleveland Clinic Barrett s esophagus registry. Am J Gastroenterol 1999;94: Gammon MD, Schoenberg JB, Ahsan H, et al. Tobacco, alcohol and socioeconomic status and adenocarcinoma of the esophagus and gastric cardia. J Natl Cancer Inst 1997;89: Lagergren J, Bergstrom R, Nyren O. Association between body mass and adenocarcinoma of the esophagus and gastric cardia. Ann Intern Med 1999;130: Vaughan TL, Kristal AR, Blount PL, et al. Nonsteroidal anti-inflammatory drug use, body mass index, and anthropometry in relation to genetic and flow cytometric abnormalities in Barrett s esophagus. Cancer Epidemiol Biomarkers Prev 2002;11: Lagergren J, Bergstrom R, Adami HO, et al. Association between medications that relax the lower esophageal sphincter and risk for esophageal adenocarcinoma. Ann Intern Med 2000;133: Cheng KK, Sharp L, McKinney PA, et al. A case-control study of oesophageal adenocarcinoma in women: a preventable disease. Br J Cancer 2000;83: Kim SL, Waring JP, Spechler SJ, et al. Diagnostic inconsistencies in Barrett s esophagus. Gastroenterology 1994;107: Attwood SE, Morris CD. Who defines Barrett s oesophagus: endoscopist or pathologist? Eur J Gastroenterol Hepatol 2001;13: McClave SA, Boyce HW, Gottfried MR. Early diagnosis of columnar-lined esophagus: a new endoscopic criterion. Gastrointest Endosc 1987;33: Address requests for reprints to: Gary W. Falk, MD, Department of Gastroenterology and Hepatology, Desk A-30, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio falkg@ ccf.org; fax: (216)