Barrett's esophagus: now what?
J. Peters
DOI: https://doi.org/10.1097/00000658-200303000-00002
IF: 13.787
2003-03-01
Annals of Surgery
Abstract:It has been said that medicine is an art that must be practiced before it is completely understood. The clinician faced with the management of patients with Barrett’s esophagus realizes this fact all too well. Every once in a while, however, a badly needed, well-done study comes around that helps us in this dilemma, improving our understanding of a disease and its therapy. The study by Parrilla et al. in this issue of the Annals is such a study. Professor Parrilla provides us an update of an ongoing prospective randomized study, a portion of which was published in the British Journal of Surgery in 1996. 1 The study compares pharmacologic and surgical treatment of patients with both long- and short-segment Barrett’s esophagus and includes 101 patients enrolled over 18 years (1982–2000).
Barrett’s esophagus is an acquired abnormality occurring in 10% to 15% of patients with gastroesophageal reflux disease. It is defined by the presence of an endoscopic segment of columnar-lined esophagus that has intestinal metaplasia (goblet cells) on biopsy. For reasons that are not entirely clear, its prevalence has exploded in the past 25 years, increasing from an average of 1/1,000 upper endoscopies in the early 1980s, to 10/1,000 in late 1980s, and over 55 to 60/1,000 in the late 1990s. 2,3 Patients with Barrett’s esophagus are characterized by long-standing severe reflux symptoms, often beginning at a relatively young age 4 and accompanied by severe anatomic and physiologic defects, including high esophageal acid exposure, large hiatal hernias 5 and virtually absent lower esophageal sphincter and often poor esophageal body contractility. 6 Esophagitis, stricture, and ulceration commonly accompany the Barrett’s segment.
There are two treatment options in patients with Barrett’s esophagus (lifelong proton pump inhibitors [PPIs] and antireflux surgery) with success reflected by two distinct outcomes: the clinical control of reflux and the prevention of esophageal adenocarcinoma. Although most gastroenterologists and surgeons believe they know how to treat Barrett’s esophagus, there are surprisingly few data to indicate the optimal treatment. 7
When evaluating studies of Barrett’s esophagus, experience has shown that the details are important. In the Parrilla study, medical therapy consisted of 20 mg omeprazole (PPI) twice daily since 1992 in all medically treated patients. Surgical therapy consisted of an open 1.5- to 3.0-cm Nissen over a 48 to 50 French bougie with short gastric division in 39% of patients and crural closure in all. On the surface, outcome in the two groups was nearly identical, although esophagitis and/or stricture persisted in 20% of the medically treated patients compared to only 3% to 7% of those following antireflux surgery. Fifteen percent of patients had abnormal acid exposure after surgery, emphasizing the importance of reporting both mean/median and prevalence data. This is consistent with two other recent reports of the long-term outcome of antireflux surgery in patients with Barrett’s esophagus in which approximately 20% of patients continued to reflux despite antireflux surgery. 8,9 It is unfortunate that pH data were not routinely collected in patients on PPI therapy. In the subgroup of 12 patients who did have 24-hour monitoring on treatment, 3 of 12 (25%) had persistently high esophageal acid exposure and most (75%) had persistently high bilirubin exposure.
Perhaps most importantly, the paper addresses one of the major unanswered questions in the pathophysiology of esophageal adenocarcinoma: Does antireflux treatment alter the natural history of Barrett’s esophagus? Given that gastroesophageal reflux unquestionably causes Barrett’s esophagus, the possibility that stopping reflux will prevent progression, or even lead to regression, is an important question that must be answered. Although the development of dysplasia and adenocarcinoma was no different overall, the subgroup of surgical patients with normal postoperative pH studies developed significantly less dysplasia and had no adenocarcinoma. This observation should not be ignored. Furthermore, both surgical patients who developed cancer were shown to have recurrent reflux. The data strongly support the concept that elimination of reflux may indeed prevent progression of neoplasia and goes a step further than most by including data on the efficacy of antireflux surgery. In this regard it is an improvement over the Spechler et al. paper to which it can be compared. 10
One wonders if Norman Barrett 11 had any insight into the quagmire he was unleashing with his initial description of the esophageal condition that bears his name. Confusion surrounding the columnar-lined esophagus began with Barrett himself (he believed it to be a congenitally short esophagus) 12 and has grown greater and more clinically relevant with each passing decade. As we now stand, some 50 years later, virtually every aspect of the disease we refer to as Barrett’s esophagus is controversial and frequently debated. 13 This includes its definition, 14 pathophysiology, 15 prevalence, 16 significance, 17 and treatment, 18 up to and including the benefit of screening and surveillance. 19 This lack of knowledge would be of little concern were it not for the fact that increasing numbers of patients, all too often young men and women, 20 many with few symptoms of gastroesophageal reflux, develop difficulty swallowing and are found to have esophageal adenocarcinoma. It doesn’t take experience with many such patients for most of us to develop a deep desire to prevent it, if at all possible. The facts of the previous sentences belie major hurdles to achieving this goal.
The conclusion that control of reflux may prevent esophageal adenocarcinoma is a profound one. The study, regardless of its flaws, suggests that this is true. The skeptic will say that the study found no significant difference in either symptomatic or neoplastic outcomes between medical and surgical therapy of Barrett’s esophagus. A closer examination of the data, however, will reveal that the skeptics are likely wrong. Given the importance of the possibility of preventing cancer, it seems prudent to emphasize the positive rather than the negative. 21 We must minimize the focus on whether this is best done with medical or surgical therapy, and maximize the possibility of answering this question once and for all. Given that failures of both types of therapy can almost always be identified and corrected, if reflux control does prevent cancer, we should be able to significantly impact the occurrence of esophageal adenocarcinoma, one of the most lethal tumors in human disease.