The incidence rates of adenocarcinoma of the esophagus. The Risk of Esophageal Adenocarcinoma After Antireflux Surgery. Methods Study Design

Transcription

1 GASTROENTEROLOGY 2010;138: The Risk of Esophageal Adenocarcinoma After Antireflux Surgery JESPER LAGERGREN,* WEIMIN YE,*, PERNILLA LAGERGREN,* and YUNXIA LU*, *Upper Gastrointestinal Research, Department of Molecular Medicine and Surgery, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden BACKGROUND & AIMS: The question of a possible preventive effect of antireflux surgery on the development of esophageal or cardia remains unsettled. We aimed to clarify whether antireflux surgery prevents later development of esophageal. METHODS: We performed a Swedish population-based cohort study of antireflux surgery from 1965 to Follow-up evaluation for cancer and censoring for death and emigration were achieved up to December 31, 2006, through linkages to nationwide registers of cancer, population, and emigration. The cancer incidence in the antireflux surgery cohort was compared with that in the corresponding Swedish population. Relative risks were presented as standardized incidence ratios (SIRs), that is, the observed number of cancer cases in the antireflux surgery cohort divided by the expected number. RESULTS: The antireflux surgery cohort comprised 14,102 persons, contributing 120,514 person-years at risk. Overall risk of esophageal (n 39) was increased 12-fold (SIR, 12.3; 95% confidence interval [CI], ). No risk decrease with time after antireflux surgery was found (P.86). After a postsurgical follow-up evaluation of 15 years or more, the SIR was 14.6 (95% CI, ). For the corresponding overall risk of cardia (n 21) the SIR was 4.4 (95% CI, ), without any major decrease in risk with time (P.20); the SIR was 3.1 (95% CI, ) after at least 15 years of follow-up evaluation. No association between antireflux surgery and gastric or esophageal squamous cell carcinoma was identified. CONCLU- SIONS: Antireflux surgery cannot be considered to prevent the development of esophageal or cardia among persons with reflux. Keywords: Neoplasm; Prevention; Fundoplication; Gastroesophageal Reflux; Esophagus; Cardia. The incidence rates of of the esophagus and the gastroesophageal junction (cardia) have been increasing rapidly in Western populations during the past few decades. 1,2 Despite attempts to improve the detection and treatment of these highly lethal cancers, the overall chance of 5-year survival remains only 10% 15%. 3,4 Preventive strategies are therefore highly warranted. Gastroesophageal reflux disease is the dominating risk factor for these tumors 5,6 and it repeatedly has been argued that antireflux surgery should have a cancer-preventive effect among persons suffering from such reflux. This is, however, a strong overinterpretation because the available literature has failed to provide any evidence supporting this hypothesis. 7,8 Although antireflux surgery could hinder gastroesophageal reflux mechanically, including any duodenopancreatic reflux, 9 the published studies addressing the effects of antireflux surgery have had too small sample sizes, too short follow-up evaluation, and too many problems with selection bias to allow conclusive results. 7,8 If the hypothesis that antireflux surgery will prevent the development of esophageal is true, a gradually or more abruptly decreasing risk of esophageal and cardia with increasing follow-up time after such surgery will be expected. To address this hypothesis and avoid the problems of power, length of follow-up evaluation, and selection that have been encountered in the previous studies of this topic, we conducted a complete nationwide Swedish cohort study with several decades of follow-up evaluation. Because s of the esophagus and cardia are related closely in terms of anatomy and etiology, with reflux being the main risk factor for both of these sites, we decided to address the risk of both these tumors in relation to antireflux surgery. Methods Study Design This was a Swedish population-based cohort study, using the entire Swedish population as a database, undertaken to address the risk of developing esophageal and cardia with increasing latency time after antireflux surgery. It was conducted during the period from January 1, 1965, through December 31, For comparison, the risks of developing gastric distal to the cardia and esophageal squamous cell carcinoma were analyzed. Abbreviations used in this paper: CI, confidence interval; SIR, standardized incidence ratio by the AGA Institute /10/$36.00 doi: /j.gastro

2 1298 LAGERGREN ET AL GASTROENTEROLOGY Vol. 138, No. 4 Study Cohort The study cohort consisted of all patients who had undergone antireflux surgery with total or partial fundoplication for a diagnosis representing gastroesophageal reflux from 1965 to 2006, as recorded in the Swedish Inpatient Register. The Inpatient Register, administered by the National Board of Health and Welfare, was initiated in and includes data on patient age, sex, national registration number (a unique identification number assigned to every resident in Sweden), up to 6 discharge diagnoses and 6 surgical procedures, and the dates of each hospitalization. Sixty percent of the Swedish population were covered by this register in 1969 and 85% were covered in 1983, and since 1987 the coverage has been 100%. Because there has been virtually no private in-hospital care in Sweden and patients have been obliged to use a hospital in their county of residence, a study based on the Inpatient Register can be considered population-based. From the originally extracted records of 16,821 patients, 2719 records were excluded for the following reasons: (1) antireflux surgery had been conducted for reasons other than reflux (ie, anomalies of the diaphragm, achalasia, cardiospasm, or paraesophageal hernia with obstruction; n 2011), (2) prevalent cancers (n 646), (3) cancers discovered for the first time at autopsy (n 37), or (4) incorrect national registration numbers (n 25). Thus, the final antireflux surgery cohort comprised a total of 14,102 patients. Follow-up Evaluation All cancers that were diagnosed during follow-up evaluation of the cohort ( ) were identified through linkage of the national registration number of each cohort member to the Swedish Cancer Register, a nationwide register started in 1958 and administered by the National Board of Health and Welfare. The completeness regarding esophageal cancer has been found to be 98%. 10 For correct exclusion of person-time no longer at risk of being identified in the Cancer Register, the dates of death and emigration were collected. This was accomplished through linkage of the national registration number of each cohort member to the Total Population Register and Swedish Emigration Register, respectively, nationwide registers with complete and updated information regarding exact dates of death and emigration. Comparison Cohort The incidence of cancer among the cohort members was compared with the incidence in the corresponding Swedish background population (ie, the population of corresponding age, sex, and calendar year). The cancers occurring in the comparison population were assessed through the data on the cancer incidence and prevalence in the Cancer Register. Any changes with time after the study exposure (ie, antireflux surgery) were thereby compared with a stable baseline risk of cancer in the population corresponding in age, sex, and calendar year of the study to the cohort members. Statistical Analyses Person-time at risk was accumulated from the date of surgery until the first occurrence of any cancer, death, emigration, or the end of observation (December 31, 2006), whichever came first. Cancers found incidentally for the first time at autopsy were excluded so as to avoid any detection bias owing to a higher autopsy rate among those undergoing antireflux surgery compared with the background population. The relative risk was estimated as the standardized incidence ratio (SIR), which was defined as the ratio of the observed to the expected number of cancers in the study cohort. The expected number of cancers was calculated by multiplying the observed person-time by cancer incidence rates specific for age, sex, and calendar year. The expected rates were derived from the Cancer Register data through the entire Swedish population and aggregated into 5-year intervals. The SIRs inherently were adjusted for some potential confounding factors (age, sex, and calendar year), because the incidence in the observed cohort was compared with the corresponding incidence in the age-, sex-, and calendar year matched general population. Confidence intervals (CIs) of SIRs were calculated on the assumption that the observed number of events followed a Poisson distribution. 11 Ethics The study was approved by the Regional Ethics Committee in Stockholm. Results Study Participants The 14,102 patients representing the antireflux surgery study cohort had 120,514 person-years of follow-up evaluation. Some characteristics of the cohort members and the observed cases of esophageal, cardia, and noncardia gastric are presented in Table 1. Male patients were in the majority and the mean ages were higher in all case groups than in the entire study cohort. Few patients (4.5%) underwent an antireflux surgery procedure more than once. Risk of Esophageal Adenocarcinoma The estimates representing the relative risk of developing esophageal are presented in Table 2. The overall risk of being diagnosed with esophageal after antireflux surgery was increased 12-fold compared with the background population (SIR, 12.3; 95% CI, ). The risk did not decrease with time after surgery (P for trend.86), and among the cohort members who were followed up for at least 15 years the SIR was 14.6 (95% CI, ). No material differences between the sexes were found. A

3 April 2010 ANTIREFLUX SURGERY AND ESOPHAGEAL ADENOCARCINOMA 1299 Table 1. Characteristics of the Total Study Participants and of the Patients Developing Adenocarcinoma of the Esophagus, Cardia, and the Noncardia Part of the Stomach Characteristics Antireflux surgery cohort Esophageal Cardia Noncardia gastric Patients, n 14, Males, n (%) 7927 (56) 31 (79) 19 (90) 26 (70) Mean age at surgery (y) antireflux procedure, number of patients (%) 629 (4) 4 (10) 2 (10) 4 (11) higher point estimate was indicated among the limited number of persons who underwent an antireflux procedure more than once. Risk of Cardia Adenocarcinoma As seen in Table 2, the risk of developing cardia was 4 times higher in the antireflux surgery cohort than in the corresponding background population (SIR, 4.4; 95% CI, ). No statistically significant decrease in risk with longer follow-up time after surgery was revealed (P for trend.20). For cohort members who had been followed up for at least 15 years the SIR was 3.1 (95% CI, ). Risk of Comparison Tumors The number of persons who developed noncardia gastric cancer after antireflux surgery (n 37) was not statistically significantly greater than expected (SIR, 1.3; 95% CI, ) (Table 2), and no change with increased follow-up time after surgery was found (P for trend.27). Only 3 patients developed esophageal squamous cell carcinoma after antireflux surgery, suggesting that this cancer had no association with this surgical procedure (SIR, 0.6; 95% CI, ). Discussion Antireflux surgery was not found to have a protective effect against the development of of the esophagus or cardia in this study. Some methodologic issues of the study deserve attention. The cohort design used is probably the best available. The low incidence of esophageal makes it unfeasible to conduct a randomized controlled trial comparing antireflux surgery with antireflux medication or no treatment, even if the study participants are selected on the basis of a known Barrett s esophagus (a premalignant metaplasia of the esophagus representing an intermediate step between reflux and ). 12 Because antireflux surgery is typically a onetime therapy, assessment of long-term effects of this surgery circumvents the inherent problem of misleading influence of the indication for the therapy (ie, confounding by indication) in research addressing antireflux medication in relation to risk of esophageal. We did not have data regarding the onset or frequency of reflux before the antireflux surgery, but any remaining problems with confounding by reflux severity should influence the short-term rather than the long-term effects of antireflux surgery. The design with unselected and Table 2. Risk of Adenocarcinoma of the Esophagus, Cardia, and Noncardia Part of the Stomach in an Antireflux Surgery Cohort, Compared With a Corresponding Background Population Esophageal Cardia Noncardia gastric Variable N SIR 95% CI N SIR 95% CI N SIR 95% CI Total Sex Male Female Follow-up time (y) Antireflux surgery, n

4 1300 LAGERGREN ET AL GASTROENTEROLOGY Vol. 138, No. 4 complete sampling of the population reduces selection and facilitates generalization to other populations. Centers with a special interest in antireflux surgery might achieve better overall control of reflux and so might have shown a protective effect against cancer. Our unselected cohort, however, reflects the routine clinical practice; and, moreover, the Swedish population-based standards of esophageal surgery seem to be high, as reflected, for example, by the lower than 5% postoperative mortality rate within 30 days of surgery for esophageal cancer, a result comparable with that of international high-volume centers. 13,14 The antireflux surgery cohort investigated in this study was large, with a follow-up time of up to 42 years, and loss to follow-up evaluation was negligible by virtue of the use of personal registration numbers, the completeness of the Cancer Register, and the availability of complete nationwide registers for correct elimination of person-time not at risk of cancer, by reasons of death or emigration. Inherent adjustments for potential confounding by age, sex, and calendar time were made in the analyses, but the lack of information on the established risk factors besides reflux (ie, obesity 15 and tobacco smoking 16,17 ) introduce a risk of confounding. It is unlikely, however, that obese persons were overrepresented in the antireflux surgery cohort because such persons usually are advised not to undergo surgery whenever a noninvasive and effective treatment alternative is readily available. 18 Tobacco smoking is only a weak risk factor, 16,17 and smokers have not been found to be overrepresented in this cohort. 19 Finally, the lack of association between antireflux surgery and risk of noncardia gastric cancer further indicates that the study has validity. The opinion that antireflux surgery reduces the risk of largely has been derived from uncontrolled case series. Nevertheless, the results of these studies have been greatly overinterpreted to an extent that has led to claims of cancer-preventive effects of antireflux surgery that are inappropriate, particularly because the available previous research taken together has failed to establish any preventive effect. 7,8,20 Neither a previous cohort study from the United States 21 nor our previous study of the antireflux surgery cohort extracted from the Swedish Inpatient Register up to possessed the statistical power to provide conclusive long-term results regarding whether or not antireflux surgery prevented the development of esophageal. The additional 9 years of follow-up evaluation and inclusion of more patients into the cohort provide more satisfactory statistical power, although the power of subanalyses also is limited in the present study. Studies addressing antireflux medication in relation to the risk of developing esophageal have encountered problems with statistical power, and the few more large-scale studies have failed to provide any evidence of a cancer-preventive effect of such medication. 5,6,22,23 These studies, however, were hampered by confounding of reflux severity. There are some tentative explanations for the lack of a preventive effect of antireflux surgery on the development of esophageal and cardia. Recurrent reflux does occur after this surgery, 18 and some data suggest that such recurrence might be particularly harmful. 24,25 Another explanation might be that the carcinogenic process begins early, and typically will have progressed to a stage beyond return before the antireflux surgery is conducted. People who are selected for antireflux surgery are likely to have a long history of severe reflux, and it is possible that the timing of the current routine surgery is too late to be able to prevent the carcinogenic process. It is possible, for example, that surgery is protective against the development of Barrett s esophagus, but not against the later progression to cancer. Once reflux disease comes to medical attention, it is rare for Barrett s to develop. This hypothesis is supported by the finding that those who developed esophageal were older when they underwent surgery than the overall cohort. The likely risk/benefit balance of performing antireflux surgery in non-barrett patients solely to prevent cancer appears unfavorable, however, even in the hands of expert surgeons. The association of antireflux surgery with risk of cardia was similar to that with esophageal (ie, no protective effect was revealed). The risk estimates, however, were lower. In Sweden, cardia predominantly is related to Barrett esophagus, which might explain our results. This might not be true in all populations. In parts of the world, cardia is of gastric origin, being related more strongly to infection with Helicobacter pylori infection. Gastroesophageal reflux disease is an often troublesome disorder, affecting about 20% of the adult Western population, 26 and is defined by its cardinal symptoms of heartburn and regurgitation. 27 Reflux should not be dismissed as a trivial disorder, but dealt with thoroughly. The scientific evidence does not, however, provide any reason to recommend antireflux therapy for the particular purpose of preventing. Moreover, any potentially beneficial endoscopic follow-up evaluation of Barrett s esophagus should not be less intense among persons who have been treated with antireflux surgery. In conclusion, this large and population-based cohort study with long and complete follow-up evaluation indicates that antireflux surgery does not prevent the development of esophageal or cardia among persons with reflux. Thus, antireflux surgery should not be considered a cancer-preventive therapy. References 1. Devesa SS, Blot WJ, Fraumeni JF Jr. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998;83:

5 April 2010 ANTIREFLUX SURGERY AND ESOPHAGEAL ADENOCARCINOMA Bollschweiler E, Wolfgarten E, Gutschow C, et al. Demographic variations in the rising incidence of esophageal in white males. Cancer 2001;92: Jemal A, Murray T, Ward E, et al. Cancer statistics, CA Cancer J Clin 2005;55: Sundelof M, Ye W, Dickman PW, et al. Improved survival in both histologic types of oesophageal cancer in Sweden. Int J Cancer 2002;99: Lagergren J, Bergstrom R, Lindgren A, et al. Symptomatic gastroesophageal reflux as a risk factor for esophageal. N Engl J Med 1999;340: Chow WH, Finkle WD, McLaughlin JK, et al. The relation of gastroesophageal reflux disease and its treatment to s of the esophagus and gastric cardia. JAMA 1995;274: Chang EY, Morris CD, Seltman AK, et al. The effect of antireflux surgery on esophageal carcinogenesis in patients with Barrett esophagus: a systematic review. Ann Surg 2007;246: Shaheen NJ. Does fundoplication change the risk of esophageal cancer in the setting of GERD? Am J Gastroenterol 2005;100: Catarci M, Gentileschi P, Papi C, et al. Evidence-based appraisal of antireflux fundoplication. Ann Surg 2004;239: Lindblad M, Ye W, Lindgren A, et al. Disparities in the classification of esophageal and cardia s and their influence on reported incidence rates. Ann Surg 2006;243: Breslow NE, Day NE. The design and analysis of cohort studies. IARC scientific publications no 82. Lyon, France: International Agency for Research on Cancer, Spechler SJ. Clinical practice. Barrett s esophagus. N Engl J Med 2002;346: Dimick JB, Wainess RM, Upchurch GR Jr, et al. National trends in outcomes for esophageal resection. Ann Thorac Surg 2005;79: Rouvelas I, Zeng W, Lindblad M, et al. Survival after surgery for oesophageal cancer: a population-based study. Lancet Oncol 2005;6: Hampel H, Abraham NS, El-Serag HB. Meta-analysis: obesity and the risk for gastroesophageal reflux disease and its complications. Ann Intern Med 2005;143: Gammon MD, Schoenberg JB, Ahsan H, et al. Tobacco, alcohol, and socioeconomic status and s of the esophagus and gastric cardia. J Natl Cancer Inst 1997;89: Lagergren J, Bergstrom R, Lindgren A, et al. The role of tobacco, snuff and alcohol use in the aetiology of cancer of the oesophagus and gastric cardia. Int J Cancer 2000;85: Spechler SJ, Lee E, Ahnen D, et al. Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease: follow-up of a randomized controlled trial. JAMA 2001;285: Ye W, Chow WH, Lagergren J, et al. Risk of s of the esophagus and gastric cardia in patients with gastroesophageal reflux diseases and after antireflux surgery. Gastroenterology 2001; 121: Corey KE, Schmitz SM, Shaheen NJ. Does a surgical antireflux procedure decrease the incidence of esophageal in Barrett s esophagus? A meta-analysis. Am J Gastroenterol 2003;98: Tran T, Spechler SJ, Richardson P, et al. Fundoplication and the risk of esophageal cancer in gastroesophageal reflux disease: a Veterans Affairs cohort study. Am J Gastroenterol 2005;100: Farrow DC, Vaughan TL, Sweeney C, et al. Gastroesophageal reflux disease, use of H2 receptor antagonists, and risk of esophageal and gastric cancer. Cancer Causes Control 2000;11: Garcia Rodriguez LA, Lagergren J, Lindblad M. Gastric acid suppression and risk of oesophageal and gastric : a nested case control study in the UK. Gut 2006;55: Csendes A, Burdiles P, Braghetto I, et al. Adenocarcinoma appearing very late after antireflux surgery for Barrett s esophagus: long-term follow-up, review of the literature, and addition of six patients. J Gastrointest Surg 2004;8: Lagergren J, Viklund P. Is esophageal occurring late after antireflux surgery due to persistent postoperative reflux? World J Surg 2007;31: Locke GR 3rd, Talley NJ, Fett SL, et al. Prevalence and clinical spectrum of gastroesophageal reflux: a population-based study in Olmsted County, Minnesota. Gastroenterology 1997;112: Vakil N, van Zanten SV, Kahrilas P, et al. The Montreal definition and classification of gastroesophageal reflux disease: a global evidencebased consensus. Am J Gastroenterol 2006;101: ; quiz Received October 6, Accepted January 6, Reprint requests Address requests for reprints to: Jesper Lagergren, MD, PhD, Professor of Surgery, Upper Gastrointestinal Research, Department of Molecular Medicine and Surgery, Karolinska Institutet, SE Stockholm, Sweden. jesper.lagergren@ki.se; fax: (46) Conflicts of interest The authors disclose no conflicts. Funding Financial support was provided by the Swedish Cancer Society and the Swedish Research Council. These funders were not involved in the design or implementation of the study; the collection, management, analysis, or interpretation of the data; or the preparation, review, or approval of the manuscript.