Does Barrett's Esophagus Regress after Surgery (or Proton Pump Inhibitors)Spechler S.J.
Department of Medicine, VA North Texas Healthcare System, and the University of Texas Southwestern Medical Center at Dallas, Dallas, Tex., USA Corresponding Author
Stuart Jon Spechler, MD, Professor of Medicine
Division of Gastroenterology (111B1)
Dallas VA Medical Center
4500 South Lancaster Road, Dallas, TX 75216 (USA)
Barrett's esophagus, the condition in which metaplastic columnar epithelium that predisposes to cancer development replaces the squamous epithelium that normally lines the distal esophagus, is a complication of gastroesophageal reflux disease (GERD). Metaplasia is a potentially reversible condition, and partial regression of Barrett's metaplasia has been documented with effective medical or surgical therapy for GERD. The important issue for patient management is not whether antireflux treatment causes Barrett's esophagus to regress, but rather whether antireflux therapy prevents cancer in Barrett's esophagus. Proton pump inhibitors (PPIs) would be expected to prevent this cancer because they heal reflux esophagitis, reduce exposure to a potential carcinogen (acid), and might prevent acid-induced proliferation and cancer-promoting cytokine secretion by esophageal epithelial cells. Furthermore, observational studies have shown that PPI use is associated with a decreased incidence of neoplasia in Barrett's esophagus. In theory, successful antireflux surgery, which eliminates the reflux of both acid and bile, should be better for cancer prevention than medical therapy, which only decreases the reflux of acid. However, high-quality studies show no significant difference in cancer incidence between medically and surgically treated patients with GERD and Barrett's esophagus. Furthermore, for individual patients with nondysplastic Barrett's metaplasia, the cancer risk is so small and the number needed to treat for cancer prevention with surgery so large, that it does not matter whether or not surgery provides a tiny margin of extra protection against cancer beyond that provided by medical therapy. The cost and risks of the operation overwhelm any small, additional cancer protective benefit. Antireflux surgery is very effective at controlling the endoscopic signs and symptoms of GERD, but the operation should not be recommended to patients solely with the rationale that it protects against cancer better than medical therapy.
© 2014 S. Karger AG, Basel
Barrett's esophagus, a major risk factor for esophageal adenocarcinoma, is a metaplasia that develops as a consequence of gastroesophageal reflux disease (GERD) . Metaplasia is the condition in which one adult cell type replaces another and, in a number of organs, it develops as a response to chronic inflammation. In the esophagus, GERD inflames the esophageal squamous epithelium and provides the abnormal intraluminal milieu that leads to the healing of the reflux esophagitis through columnar metaplasia rather than through the regeneration of more squamous cells . Metaplasia is a potentially reversible condition if the inciting inflammatory process can be treated effectively. After the effective treatment of GERD with fundoplication or proton pump inhibitors (PPIs), a number of studies (mostly observational) have documented partial and, infrequently, complete regression of Barrett's metaplasia in some patients [3,4,5]. In many of the reported cases of ‘regression', it is difficult to exclude the possibility that Barrett's metaplasia did not regress, but instead was missed due to biopsy sampling error. It is especially difficult to exclude biopsy sampling error in cases of short-segment Barrett's esophagus, which is both patchy and limited in extent. Nevertheless, it seems highly likely that at least partial regression of Barrett's metaplasia occurs in some patients with effective medical or surgical antireflux therapy. What remains unclear is whether partial regression is clinically important.
Interest in regression of Barrett's esophagus with antireflux therapy has waned in recent years with the rising use of endoscopic ablative techniques like radiofrequency ablation, which can eradicate the metaplastic mucosa directly. Even before the availability of safe and effective ablative techniques, however, I would submit that the question of whether antireflux therapy causes partial regression of Barrett's metaplasia was never a key issue in patient management. The important issue has been whether antireflux therapy can prevent cancer in Barrett's esophagus. Partial regression of Barrett's metaplasia was considered a surrogate for decreased cancer risk, but it was never clear that mucosal regression was a valid surrogate. Rather than focus on the issue of metaplasia regression, which is of dubious importance, I will focus the remainder of this discussion on why antireflux therapy might be expected to prevent cancer in Barrett's esophagus.
Table 1 lists four reasons why gastric acid suppression might be expected to prevent cancer in Barrett's esophagus.
Acid reflux causes reflux esophagitis, a form of chronic inflammation, and chronic inflammation in a number of organs is well known to predispose to cancer development . There is abundant evidence that acid suppression with PPIs heals reflux esophagitis  and, therefore, the elimination of that chronic esophageal inflammation with acid suppression might be expected to have cancer-preventing effects.
There are in vitro and in vivo data to suggest that acid might be a carcinogen in Barrett's esophagus. Acid exposure causes nonneoplastic Barrett's epithelial (BAR-T) cells in vitro to produce reactive oxygen species, which are toxic molecules that can damage DNA . DNA damage in the form of double-strand breaks (DSBs) can lead to genomic instability and promote cancer formation, and agents that induce DSBs can be considered carcinogens [8,9]. Following acid exposure of BAR-T cells in vitro, immunofluorescence staining for phospho-H2AX reveals nuclear foci that indicate the presence of DSBs. The acid-induced formation of these DSBs can be prevented by pretreating the cells with either N-acetyl-L-cysteine (a reactive oxygen species scavenger) or disodium 4,4′-diisothiocyanatostilbine-2,2′-disulfonate, which prevents intracellular acidification by inhibiting the Cl-/HCO3- exchanger. These observations indicate that acid enters Barrett's epithelial cells and causes severe DNA damage through the generation of reactive oxygen species.
The in vitro observation that acid exposure causes DNA damage in Barrett's metaplastic cells was confirmed in a translational study in which endoscopic biopsy specimens of Barrett's metaplasia were taken from 6 patients before and after the esophagus was perfused for 5 min with 0.1 M hydrochloric acid . Western blots demonstrated that even this brief esophageal acid exposure increased phospho-H2AX expression in the Barrett's metaplasia of all 6 patients. These findings suggest that acid might be a carcinogen in Barrett's esophagus and, therefore, gastric acid suppression might protect against cancer by decreasing esophageal exposure to this carcinogen.
Biopsy specimens of Barrett's metaplasia maintained in organ culture and exposed to a pulse of acid show evidence of increased proliferation and increased expression of cyclooxygenase-2 [10,11]. Acid exposure also causes Barrett's-associated adenocarcinoma cell lines to activate the mitogen-activated protein kinase and protein kinase C pathways, and to increase proliferation [10,12]. In patients with Barrett's esophagus, esophageal acid perfusion for only 3 min causes activation of the mitogen-activated protein kinase pathways in Barrett's metaplasia . In a study in which patients had biopsies taken from their Barrett's esophagus at baseline and after 6 months of treatment with PPIs, the expression of PCNA (a proliferation marker) decreased significantly in biopsy specimens from 24 patients in whom PPIs had normalized esophageal acid exposure, but not in biopsies from the 15 with persistently abnormal acid reflux . Another study found no significant change in the proliferative activity of Barrett's metaplasia in 22 patients treated with PPIs for 2 years, whereas proliferative activity increased significantly in 23 patients treated for the same time with histamine H2-receptor antagonists, which are far less effective at reducing gastric acid secretion . In the latter two clinical studies, it is not possible to determine whether acid suppression decreased proliferation directly, i.e. by eliminating a direct proproliferative effect of acid on the epithelial cells, or indirectly through the anti-inflammatory effects of healing reflux esophagitis. Irrespective of the mechanism, however, it appears that acid suppression with PPIs decreases proliferation in Barrett's metaplasia and, therefore, might protect against cancer development.
In contrast to the above-mentioned studies suggesting that acid has proproliferative effects in Barrett's esophagus, one study using nonneoplastic Barrett's epithelial cells suggested that acid might have antiproliferative effects . In this study, BAR-T cells exposed to acid exhibited a decrease in cell number at 24 h. Initially, these findings suggested the possibility that acid might have beneficial effects on proliferation in nondysplastic Barrett's esophagus. Subsequently, the same investigators found that this antiproliferative effect was the result of acid-induced DNA damage causing cell cycle arrest . Although the aforementioned study did show that acid exposure has a temporary antiproliferative effect in benign Barrett's epithelial cells , this observation should not be construed as a beneficial effect. The cells merely stopped proliferating temporarily while they attempted to repair the DNA damage caused by the acid exposure.
Nonneoplastic human esophageal squamous cells exposed to acid and bile salts dramatically increase their production of IL-8, a chemokine that is a major mediator of inflammation and that has proproliferative effects . In a rat model of reflux esophagitis produced by esophagoduodenostomy, immunohistochemical staining shows robust expression of IL-8 in the epithelial cells of the reflux-inflamed esophagus . Thus, acid reflux appears to stimulate the production of a proinflammatory, proproliferative chemokine (IL-8) that conceivably could contribute to esophageal carcinogenesis. Gastric acid suppression with PPIs might prevent this reflux-stimulated production of IL-8.
A recent report has shown that the PPIs also might block esophageal chemokine production through mechanisms that are independent of their antisecretory effects . In this study, the investigators exposed human esophageal squamous cells to an acidic bile salt medium, which caused the cells to express IL-8 mRNA and protein by activating the IL-8 promoter through NF-κB and activator protein (AP)-1 binding. Omeprazole inhibited that acidic bile salt-stimulated IL-8 expression by blocking the nuclear translocation of p65 (an NF-κB subunit) and by blocking the binding of p65, c-Jun and c-Fos (AP-1 subunits) to the IL-8 promoter. These findings indicate that omeprazole inhibits IL-8 expression through effects on NF-κB and AP-1 that are entirely independent of PPI effects on gastric acid secretion. Conceivably, these acid-independent anti-inflammatory effects also could help to prevent esophageal cancer.
Omeprazole was the first PPI introduced into clinical practice in the USA in 1989. Figure 1 shows that the rising incidence of esophageal adenocarcinoma clearly preceded the introduction of PPIs . Although figure 1 suggests that the time trend of the rise in this cancer might be leveling off somewhat, the figure also shows that the incidence of esophageal adenocarcinoma has not declined dramatically since the introduction of PPIs. Since PPIs are very effective at controlling gastric acid, this suggests that perhaps the reflux of something other than acid might be contributing to carcinogenesis in Barrett's esophagus. One of the prime suspects for that other potentially carcinogenic material is bile acid.
In a recent translational study, the investigators took biopsies of Barrett's metaplasia from 3 patients before and after perfusing the esophagus for 5 min with a physiologic solution of the bile acid DCA (deoxycholic acid) . Western blots revealed increased expression of phospho-H2AX, a marker of DNA damage, after DCA perfusion in biopsy specimens from all 3 patients. Severe and irreparable DNA damage usually triggers apoptosis (programmed cell death) that destroys cells with potentially carcinogenic genetic alterations. In addition to causing DNA damage, however, the investigators found that DCA also activated NF-κB proteins, which can prevent apoptosis. This study shows that, in Barrett's metaplasia, bile acids cause DNA damage while simultaneously activating NF-κB, which can prevent the apoptosis that should be induced by DNA damage. This might enable the survival of Barrett's cells that have sustained potentially carcinogenic genetic alterations. Unfortunately, PPIs do not prevent bile reflux.
PPIs do not entirely eliminate acid reflux in Barrett's esophagus either, and a number of studies have documented that patients with Barrett's esophagus often continue to have pathological levels of acid reflux despite PPI therapy. In one study of 31 patients with long-segment Barrett's esophagus treated with esomeprazole 40 mg twice a day, pH monitoring revealed that the gastric pH remained above 4 for 81% of the day, a level of gastric acid suppression similar to that found in studies of healthy volunteers on a similar PPI dosage . This showed that Barrett's esophagus patients had no unusual resistance to the acid-suppressing effects of PPIs. However, simultaneous esophageal pH monitoring showed that 7 of the 31 patients (23%) had abnormal acid reflux despite the good gastric acid suppression. This suggests that the so-called ‘PPI resistance' of Barrett's esophagus patients is due to their strong propensity for reflux. In other words, the antireflux mechanism in patients with long-segment Barrett's esophagus is so poor that, even during PPI treatment, the small amount of acid that remains in the stomach still refluxes into the esophagus.
The PPIs have four broad categories of potential side effects . It has been proposed that PPIs might: (1) increase the risk of upper gastrointestinal tract cancer development, (2) increase susceptibility to a variety of infections, (3) affect the absorption or metabolism of certain vitamins and minerals, and (4) cause miscellaneous problems.
Gastric acid secretion is stimulated by the hormone gastrin, which is made in the gastric antrum. In a classic, feedback loop arrangement, gastrin stimulates acid production in the gastric fundus, and acid inhibits antral gastrin production. When PPIs block acid secretion, antral gastrin production continues unabated and serum gastrin levels rise . Without acid, furthermore, bacteria can colonize the stomach . Gastrin is a growth factor that has been shown to increase proliferation in Barrett's metaplasia . In addition, bacteria can deconjugate bile acids, which can injure the esophagus at neutral pH levels, and bacteria can convert dietary nitrates into potentially carcinogenic N-nitroso compounds . In addition, some studies suggest that PPIs might promote gastric atrophy in patients who are infected with Helicobacter pylori . All this could increase the risk for developing cancers of the upper gastrointestinal tract.
Gastric acid plays an important role in killing ingested bacteria, and some reports allege that acid suppression with PPIs might increase the risk of enteric infections , pneumonias , Clostridium difficile colitis , small intestinal bacterial overgrowth , and spontaneous bacterial peritonitis in patients with cirrhosis . PPIs have been shown to interfere with the absorption of vitamin B12 and to cause hypomagnesemia, and PPI effects on magnesium and calcium metabolism might increase the risk for bone fractures [32,33]. Finally, there are reports of miscellaneous problems associated with PPI use, including interstitial nephritis , microscopic colitis , food allergy development , decreased efficacy of clopidogrel , and increased frequency of cardiovascular events in patients with acute coronary syndromes .
Most of the problems described above are more theoretical than real, and the PPIs have an outstanding track record of safety over the past 25 years. However, GERD and Barrett's esophagus are lifelong diseases, and these potential problems should be considered when prescribing chronic PPI therapy.
Despite the reasons to suspect that PPIs could increase the risk for upper gastrointestinal tract malignancy discussed above, there are no credible events that substantiate this concern. Furthermore, a number of observational studies suggest that PPIs protect against cancer development in patients with Barrett's esophagus. In a long-term follow-up study of 236 veteran patients with Barrett's esophagus, the cumulative incidence of dysplasia at 10 years was 21% for the patients who received PPI therapy compared to 58% for the patients who did not . In an update of this study that included 344 veteran patients, patients who were given a prescription for PPIs had a significant reduction in the risk of developing high-grade dysplasia and cancer, with a hazard ratio of 0.43 (95% CI: 0.21-0.83) . In an Australian study of 350 Barrett's patients followed for a median of 4.7 years, patients who delayed using PPIs for ≥2 years after the diagnosis of Barrett's esophagus had a significantly increased risk for developing low-grade dysplasia (hazard ratio 5.6, 95% CI: 2.0-15.7) and for developing high-grade dysplasia or cancer (hazard ratio 20.9, 95% CI: 2.8-158) . Finally, a study of 540 Dutch patients with Barrett's esophagus followed for a median of 5.2 years found that PPI use was associated with a 75% reduction in the risk of neoplastic progression .
In summary, there is a plausible rationale for using PPIs in chemoprevention for patients with Barrett's esophagus. PPIs heal reflux esophagitis, reduce exposure to a potential carcinogen (acid), and might prevent acid-induced proliferation and cancer-promoting cytokine secretion by esophageal epithelial cells. A number of observational studies have shown that PPI use protects against neoplasia in Barrett's esophagus, and the theoretical cancer-promoting effects of PPIs have not been substantiated in any observational study. Consequently, it seems prudent to control esophageal acid exposure in patients with Barrett's esophagus, which can be done either with PPIs or with antireflux surgery.
One can make excellent theoretical arguments about why surgery should be better than medical therapy for cancer prevention in Barrett's esophagus . As discussed, medical therapy for GERD with PPIs targets acid exclusively, but acid clearly is not the only potentially harmful agent in refluxed gastric juice. An effective antireflux operation stops the reflux of all noxious material, including the potentially carcinogenic bile acids, and a successful operation obviates the potential side effects of PPIs. Furthermore, a number of reports of uncontrolled observational studies have suggested that surgically treated patients with Barrett's esophagus develop less dysplasia and less cancer than medically treated patients [44,45]. Unfortunately, antireflux surgery can also have serious complications including dysphagia, gas-bloat syndrome, diarrhea, and (rarely) death . In addition, high-quality studies, which include two prospective randomized trials [47,48], two meta-analyses [49,50], and three cohort studies using very large databases [51,52,53], have found no significant differences in cancer incidence between medically and surgically treated GERD patients with Barrett's esophagus.
In the USA, the national incidence of esophageal adenocarcinoma has risen from 3.6 per million in 1973 to 25.6 per million in 2006 . That is a more than sevenfold increase over the course of only three decades. Ironically, however, although the incidence of this cancer has been rising, estimates of cancer risk for individual patients with nondysplastic Barrett's esophagus have been getting lower. In the 1990s, this risk was estimated at 1% per year . In 2000, Shaheen et al. [55 ]argued convincingly that this old estimate was exaggerated because it was based on reports that suffered from publication bias, and estimated the annual risk at 0.5% per year. Several reports of large studies published within the past 2 years have described an even lower risk. For example, a meta-analysis of 57 reports including more than 11,000 patients followed for more than 58,000 patient-years found a cancer incidence of only 0.33% per year . A multicenter US study with more than 6,000 patient-years of follow-up found an annual cancer incidence of only 0.27% . A large population-based study from Northern Ireland also found an annual cancer incidence of 0.27% . Moreover, the largest population-based study to date followed Danish patients who had Barrett's with nondysplastic intestinal metaplasia for more than 56,000 person-years, and found the overall cancer incidence to be only 0.12% . These recent reports suggest that the cancer risk for patients with nondysplastic Barrett's esophagus is only approximately 0.25% per year, a risk lower than had been thought for many years.
The low risk of cancer for individual patients with nondysplastic Barrett's esophagus has important implications for patient management, especially if one considers the number needed to treat (NNT) for cancer prevention. The formula for determining the NNT is 1 divided by the absolute risk reduction. For example, if treatment X results in an absolute reduction in the risk of a bad outcome by 20%, then the NNT is 1 divided by 20%, which equals 5. In this situation, one only needs to treat 5 patients with treatment X to prevent one bad outcome.
Now, consider the following question. If antireflux surgery prevents cancer in Barrett's esophagus, then how many patients would that therapy need to prevent 1 cancer case in 1 year? Data discussed above suggest that the annual risk for cancer in Barrett's esophagus is 0.25%. If we assume that antireflux surgery is highly effective and that it decreases the risk of cancer development by one half (note: there is no proof that any GERD therapy decreases the cancer risk at all, but for this discussion assume that antireflux surgery is a highly effective anticancer therapy), then antireflux surgery decreases the cancer risk from 0.25 to 0.125%, which is an absolute risk reduction of 0.125%. Therefore, the NNT is 1 divided by 0.125%, which equals 800. This means that one would need to treat 800 patients with antireflux surgery to prevent 1 cancer case in 1 year. Such a large NNT can only be justified for a very inexpensive procedure that has virtually no side effects. Clearly, antireflux surgery does not meet those criteria.
In conclusion, there is abundant indirect evidence to suggest that effective antireflux therapy, medical or surgical, protects against cancer in Barrett's esophagus. In theory, successful antireflux surgery, which eliminates the reflux of both acid and bile, should be better for cancer prevention than medical therapy, which only decreases the reflux of acid. But high-quality studies show no significant difference in cancer incidence between medically and surgically treated patients with GERD and Barrett's esophagus. Furthermore, for individual patients with nondysplastic Barrett's, the cancer risk is so small and the NNT to prevent cancer with surgery is so large, that it does not matter whether or not surgery provides a tiny margin of extra protection against cancer beyond that provided by medical therapy. The cost and risks of the operation totally overwhelm any small, additional cancer protective benefit. Antireflux surgery is very effective at controlling the endoscopic signs and symptoms of GERD, and there are a number of good reasons to recommend this operation. However, antireflux surgery should not be recommended to patients solely with the rationale that it protects against cancer better than medical therapy.
The author has been a consultant for Takeda Pharmaceuticals, Inc. and for Torax Medical, Inc.
Stuart Jon Spechler, MD, Professor of Medicine
Division of Gastroenterology (111B1)
Dallas VA Medical Center
4500 South Lancaster Road, Dallas, TX 75216 (USA)
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher or, in the case of photocopying, direct payment of a specified fee to the Copyright Clearance Center.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.