Here's the article on reflux surgery I was referring to:
jama.ama-assn.org
Surgical Therapy for Reflux Disease
Peter J. Kahrilas, MD
With publication of a follow-up report on the Department of Veterans Affairs cooperative trial comparing long-term outcomes of medical and surgical therapy for gastroesophageal reflux disease (GERD) by Spechler et al in this issue of THE JOURNAL,1 it is clear that the decision to pursue antireflux surgery has not gotten any simpler. The status of the medical and surgical cohorts is reported 10 to 13 years after initial enrollment in the trial making this report the most complete, longest-term, and most carefully collected data available. By all standards, these patients had severe GERD, evident by the high proportion who had esophageal ulcers, strictures, or Barrett esophagus at baseline. Common wisdom argues that this is the group most likely to benefit from antireflux surgery: surgical correction is long-lasting, replaces antisecretory medications, and eliminates the subsequent cancer riskor so the argument goes.
The major findings reported by Spechler et al are that patients in the medical and surgical groups are equally likely to have developed strictures or adenocarcinoma of the esophagus or to have had subsequent antireflux surgery. Furthermore, 62% of the surgical patients and 92% of the medical patients reported ongoing use of medications for GERD symptom control. Although these treatments were not administered in a protocol fashion, the symptom control and esophagitis control achieved were equal between cohorts. The fact that both groups were using medications seemingly negates the slight advantage demonstrated by the surgical patients in the acute study phase, before the era of proton pump inhibitors. Most unexpectedly, mortality was greater among the surgical cohort (40% vs 28%; P = .047). No immediate explanation is forthcoming for this finding and it is unclear how the excess cardiovascular deaths could be linked to antireflux surgery.
The authors conclude that surgery should not be advised for treatment of GERD with the expectation that patients will no longer use antisecretory medications or that it will prevent esophageal adenocarcinoma. In fact, the authors suggest that the low incidence of and mortality from esophageal adenocarcinoma observed in this prospective study call for a reevaluation of current screening and surveillance guidelines for Barrett esophagus.
For chronic GERD, the most feared sequence of events is that it causes Barrett esophagus, which can then become dysplastic and neoplastic. Logical preventive strategies thus call for early identification of Barrett esophagus in high-risk patients (ie, symptomatic middle-aged white men) and endoscopic surveillance for dysplasia once Barrett esophagus has been detected. Some surgeons have even suggested that the ultimate preventive strategy is early intervention with antireflux surgery to eliminate reflux and prevent development of Barrett esophagus (or of adenocarcinoma if Barrett esophagus is already present).2 Although these arguments all have been aired frequently, supportive data are sparse. Accordingly, the findings of Spechler et al provide an opportunity to put the issues of esophageal adenocarcinoma and Barrett esophagus in perspective.
First, the epidemiology of the relationship between gastroesophageal reflux and esophageal adenocarcinoma must be considered. In a population-based case-control study from Sweden that compared risk factors of patients with esophageal adenocarcinoma (n = 189) with those of randomly selected age-matched controls (n = 820), Lagergren et al3 reported that persons with symptomatic reflux were at significantly increased risk of esophageal adenocarcinoma and that the more frequent, more severe, and longer the duration of symptoms, the greater the risk. Using population-based data on esophageal adenocarcinoma from the National Cancer Institute Surveillance, Epidemiology and End Results (SEER) program from 1994,4 extrapolating cancer incidence rates to 2000, and applying these rates to the entire US non-Hispanic white population results in an estimated 5444 annual cases of esophageal adenocarcinoma. Merging this calculation with the GERD severity distribution among esophageal adenocarcinoma cases and age-matched controls experienced in the Swedish study and applied to the US population suggests that of the anticipated 5444 annual US cases of esophageal adenocarcinoma, 1544 will have had severe GERD symptoms. Thus, the absolute risk can be estimated at 1544/7.88 million or 1/5117 (0.020%) per year for patients with severe symptomatic GERD, with lower risk estimates for individuals with less severe symptoms or without symptoms.5
Second, evaluation of the hypothesis that antireflux surgery can reduce the risk of death from esophageal adenocarcinoma requires that the risk of esophageal adenocarcinoma be contrasted with the risk of death from the fundoplication procedure. In a pooled analysis of reports on laparoscopic Nissen fundoplication published between 1991 and 1995, Perdikis et al6 calculated a mortality rate of 0.2% among 2453 patients. In a similar analysis for reports on conventional antireflux surgery published between 1980 and 1993, Dent7 identified 4 deaths among 1152 patients, yielding a mortality rate of 0.35%. The only available population-based data on morbidity and mortality of antireflux surgery are in a report from Finland that included 3993 open and 1162 laparoscopic procedures performed in 1987-1996 and found mortality rates of 0.18% and 0.09%, respectively.8 Thus, as suggested by Spechler et al,1 a reasonable estimate of surgical mortality from conventional or laparoscopic fundoplication is 0.2%. Clearly, this mortality rate exceeds the estimated annual absolute risk of esophageal cancer of 1/5117 (0.02%) derived for highly symptomatic GERD patients.5 Even when selecting white men, who have 10 times the adenocarcinoma risk of white women, the annual absolute risk is only 0.035%.4, 5
Examining these risks, interventionists still might argue that if the annual risk of adenocarcinoma were additive over 10 years, survival may be improved by antireflux surgery. That optimism assumes that fundoplication prevents esophageal adenocarcinoma and that adenocarcinoma of the esophagus is almost universally fatal. Data presented by Spechler et al dispute both assumptions; 4 of 166 (2.4%) patients in the medical cohort and 1 of 82 (1.2%) patients in the surgical cohort developed esophageal adenocarcinoma in the subsequent 10 to 13 years.1 This difference was not significant, although the authors acknowledge that the study did not have power to detect differences between groups for cancer incidence. Moreover, Spechler et al found that adenocarcinoma of the esophagus was not a major cause of death in either treatment group. Of the 79 patients who died during the 10- to 13-year follow-up period, only 2 deaths (2.5%), both in the medical cohort, were attributable to esophageal adenocarcinoma, and no deaths were otherwise related to GERD. Similar data were reported from a study of continuous long-term therapy with omeprazole that showed significant overall cohort attrition but no disease- or treatment-related mortality after up to 11 years of follow-up.9
Herein lies the good news for patients with GERD. Despite the heavy and long-term symptom burden it imparts on the population, GERD has a remarkably low mortality rate. In fact, it can be argued that GERD mortality is too low to be significantly affected by any intervention in a controlled trial.
Another implication of the low GERD mortality rate is that the yield of any preventive strategy for esophageal adenocarcinoma will be low at best. Even among patients with Barrett esophagus, generally accepted to be at greatest risk, the annual incidence of esophageal adenocarcinoma was only 0.4%. Furthermore, 1 of the 5 cases observed by Spechler et al occurred in an individual without recognized Barrett esophagus. This is consistent with the observation in the study by Lagergren et al3 that only 62% of patients with esophageal adenocarcinoma had coexistent Barrett esophagus identifiable in resected specimens. Although Barrett esophagus is technically considered a "premalignant condition," its malignant potential appears to be much less than widely reported or feared. Further diluting the risk burden of the population with Barrett esophagus, at least some cases of esophageal adenocarcinoma arise in patients without recognized Barrett esophagus.3 These observations, coupled with a recent analysis suggesting that previous estimates of the cancer risk associated with Barrett esophagus were exaggerated by publication bias,10 provide evidence to suggest that current practice guidelines focused on screening endoscopy to detect Barrett esophagus and subsequent biannual surveillance endoscopy in patients with uncomplicated Barrett esophagus should be reassessed.11, 12
Thus, it has become a bit easier for physicians to offer patients with GERD reasons not to have antireflux surgery rather than proceed with fundoplication. For a patient with reflux symptoms, with or without a history of esophagitis, fundoplication introduces a risk of death that is not present in patients undergoing maintenance treatment with proton pump inhibitors.13 From a risk-benefit viewpoint, if such individuals are satisfied with the symptom relief gained from maintenance proton pump inhibitor therapy, they should be advised to continue with this therapy indefinitely. The basic tenets of the surgical argumentpermanence, cancer prevention, and freedom from use of antisecretory medicationshave all been seriously challenged by the report by Spechler et al.
Alternative rationales to use antireflux surgery remain, but the focus should shift away from a global disease management strategy to an individualized one. Rather than trying to compete with a proven safe and effective medical approach for most patients, antireflux surgery should be reserved for patients with unique circumstances who might selectively benefit. Examples include patients who are intolerant of proton pump inhibitor therapy because of adverse effects or those who are inadequately responsive to proton pump inhibitor therapy. Extensive clinical data suggest that this is exceedingly rare with respect to healing erosive esophagitis.9
More commonly, treatment failure occurs in patients with symptoms such as postural regurgitation, cough, chest pain, or laryngopulmonary manifestations of GERD. Available data support the belief that antireflux surgery can be effective in such individuals, although neither medical nor surgical therapy in these circumstances approaches the 90% efficacy rates achieved for patients with typical esophageal symptoms of GERD.14-16 Clearly, this is the direction in which the therapeutic frontier of GERD has shifted; the problem of peptic esophagitis was largely solved with the introduction of highly effective inhibitors of gastric acid secretion.
Author Affiliation: Department of Medicine, Northwestern University Medical School, Chicago, Ill.
Peter |
