Key points

Introduction

Barrett’s esophagus is a major risk factor for the development of esophageal adenocarcinoma, a tumor whose incidence has increased profoundly over the last 40 years in Western countries. The pathogenesis of Barrett’s esophagus involves chronic gastroesophageal reflux disease (GERD) wherein the reflux of acid and bile into the esophagus damages the esophageal mucosa and leads to its repair through the process of metaplasia. The specialized intestinal metaplasia of Barrett’s esophagus is predisposed to malignancy, and ongoing GERD is likely to contribute to that carcinogenesis. Because chronic GERD plays a role in the pathogenesis of Barrett’s metaplasia and in its malignant progression, it makes sense that aggressive treatment of GERD might prevent adenocarcinoma in Barrett’s esophagus. The modern medical therapy for GERD is directed primarily at decreasing gastric acid production, and proton pump inhibitors (PPIs), which were introduced into clinical practice in the United States in 1989, are the best medications available for that purpose. This article reviews the effects of PPIs that might impact on the neoplastic progression of Barrett’s metaplasia, and the clinical evidence that PPIs may prevent the development of dysplasia and cancer in patients with Barrett’s esophagus.

Introduction

Barrett’s esophagus is a major risk factor for the development of esophageal adenocarcinoma, a tumor whose incidence has increased profoundly over the last 40 years in Western countries. The pathogenesis of Barrett’s esophagus involves chronic gastroesophageal reflux disease (GERD) wherein the reflux of acid and bile into the esophagus damages the esophageal mucosa and leads to its repair through the process of metaplasia. The specialized intestinal metaplasia of Barrett’s esophagus is predisposed to malignancy, and ongoing GERD is likely to contribute to that carcinogenesis. Because chronic GERD plays a role in the pathogenesis of Barrett’s metaplasia and in its malignant progression, it makes sense that aggressive treatment of GERD might prevent adenocarcinoma in Barrett’s esophagus. The modern medical therapy for GERD is directed primarily at decreasing gastric acid production, and proton pump inhibitors (PPIs), which were introduced into clinical practice in the United States in 1989, are the best medications available for that purpose. This article reviews the effects of PPIs that might impact on the neoplastic progression of Barrett’s metaplasia, and the clinical evidence that PPIs may prevent the development of dysplasia and cancer in patients with Barrett’s esophagus.

Cellular effects of acid reflux and proton pump inhibitors in Barrett’s esophagus

There are several broad categories of PPI effects that might be expected to protect against carcinogenesis in Barrett’s esophagus. First, PPIs heal reflux esophagitis. Chronic inflammation is known to predispose to cancer in several organs, and the elimination of chronic esophageal inflammation by PPIs might protect against malignancy. Next, PPIs decrease esophageal exposure to acid, which can cause cancer-promoting DNA damage and increase proliferation in Barrett’s metaplasia. Finally, PPIs can prevent the release of cancer-promoting cytokines by esophageal epithelial cells through mechanisms independent of their acid-suppressive effects. Numerous studies have documented the efficacy of PPIs in healing reflux esophagitis; these data are not reviewed here. Rather, we focus on the latter two mechanisms whereby PPIs might prevent cancer in Barrett’s esophagus.

In one study, acid exposure of nondysplastic Barrett’s epithelial cells led to the production of reactive oxygen species (ROS) with double-strand breaks in DNA, which can result in genomic instability and carcinogenesis. Those acid-induced DNA double-strand breaks could be prevented by pretreating the Barrett’s cells with an ROS scavenger or a compound that inhibited intracellular acidification. These data suggest that refluxed acid can enter Barrett’s epithelial cells, leading to the generation of ROS that cause DNA damage. Agents that induce DNA double-strand breaks are considered carcinogens and, thus, PPIs might reduce the risk of cancer by limiting exposure to carcinogenic gastric acid.

Acid also may contribute to cancer by causing increased cellular proliferation in Barrett’s esophagus. This has been suggested by studies using Barrett’s biopsies maintained in organ culture and using Barrett’s biopsies taken before and after acid perfusion of the esophagus. Taken together, these studies suggest that acid exposure can cause increased expression of cyclooxygenase-2 and activation of the protein kinase-C and mitogen-activated protein kinase pathways in Barrett’s metaplasia, resulting in increased proliferation and decreased apoptosis.

Several clinical studies in patients with Barrett’s esophagus have found that protracted treatment with PPIs can cause improvements in markers of proliferation and other potentially beneficial effects. In one study of patients who had biopsies of Barrett’s metaplasia taken before and after 6 months of PPI therapy, those patients who achieved normalization of esophageal acid exposure with PPIs showed a significant decrease in proliferation as determined by the biomarker proliferating cell nuclear antigen, unlike the patients who had persistently abnormal esophageal acid exposure despite PPI treatment. Another study compared the proliferative activity in Barrett’s mucosal biopsies between patients treated with PPIs and those treated with H ₂ -receptor antagonists (H ₂ RAs) for 2 years. Patients treated with H ₂ RAs showed increased proliferative activity, whereas there was no increase in proliferative activity in patients treated with PPIs.

Patients with Barrett’s esophagus often have abnormalities in the expression of certain cell cycle proteins including p16, p21, and cyclins D1 and E in their Barrett’s metaplasia. In one study, PPI treatment was associated with fewer abnormalities in mucosal expression of these cell cycle proteins. In another study, tissue samples from patients with Barrett’s esophagus were analyzed for levels of iNOS (an oxidant-generating enzyme) and Mn-SOD (an antioxidant enzyme), and DNA damage was measured. In patients who had been treated with PPIs for at least 3 months, iNOS expression was not affected, but Mn-SOD antioxidant levels were higher and DNA damage was reduced. The authors concluded that treatment with PPIs may help trigger the expression of antioxidant genes, potentially reducing oxidative DNA damage.

Gastroesophageal reflux can cause the esophageal mucosa to produce proinflammatory cytokines, such as interleukin (IL)-8, which can increase inflammation and cellular proliferation. In a surgical rat model of reflux esophagitis, for example, increased expression of IL-8 was seen in the esophagus of rats that developed erosive esophagitis. PPIs seem to have anti-inflammatory effects that are independent of their effects on gastric acid suppression, and that might reduce esophageal production of these proinflammatory cytokines. In one study, esophageal squamous cells exposed to acid and bile increased their expression of IL-8 mRNA through effects on nuclear factor-κB and activator protein-1. Treating these cells with omeprazole inhibited IL-8 expression by blocking nuclear translocation of p65, a nuclear factor-κB subunit, and by blocking the binding of activator protein-1 subunits to the IL-8 promoter. This suggests that, in addition to reducing gastric acid production and acid reflux, PPIs also might have beneficial effects in GERD by modulating chemokine expression.

In summary, studies suggest that acid reflux can lead to increased proliferation, decreased apoptosis, production of ROS, and DNA damage in Barrett’s metaplasia. Acid reflux also stimulates esophageal production of proinflammatory and proproliferative cytokines. All of these effects might promote cancer development. Through acid-suppressive and acid-independent anti-inflammatory effects, PPIs can prevent or reduce these reflux-induced abnormalities, and thus might protect against carcinogenesis in Barrett’s esophagus.

Proton pump inhibitors and control of acid reflux in patients with Barrett’s esophagus

PPIs are highly effective at reducing gastric acid secretion, and they are the mainstay of therapy for GERD. However, one study suggests that almost one-third of patients with GERD taking a PPI once daily continue to have abnormal acid reflux. Patients with long-segment Barrett’s esophagus have especially poor antireflux mechanisms and often have more persistent and difficult-to-control acid reflux than patients who have GERD without Barrett’s esophagus. Several studies have shown that pathologic acid reflux often persists despite PPI therapy in patients with Barrett’s esophagus. One study of patients with long-segment Barrett’s esophagus found that 23% had abnormal esophageal acid exposure documented by pH monitoring despite treatment with esomeprazole in high dosage. In these patients, esomeprazole achieved levels of gastric acid suppression similar to those in patients without Barrett’s esophagus, but even this reduced amount of gastric acid production resulted in abnormal acid reflux, presumably because antireflux mechanisms were so ineffective in the patients with Barrett’s esophagus. Other studies have shown that control of GERD symptoms with PPI treatment does not guarantee that esophageal acid exposure is controlled. One of these studies compared the effects of PPIs on acid reflux in patients with Barrett’s esophagus and in patients who had GERD without Barrett’s esophagus. Despite PPI therapy that controlled GERD symptoms, approximately half of the patients in both groups had abnormal acid reflux documented by esophageal pH monitoring. However, the patients with Barrett’s esophagus had higher DeMeester scores than the patients without Barrett’s esophagus. These studies suggest that many patients with Barrett’s esophagus continue to have abnormal acid reflux despite treatment with PPIs that controls GERD symptoms.

Potential carcinogenic effects of proton pump inhibitors in Barrett’s esophagus

The studies discussed in the previous sections have suggested plausible mechanisms whereby PPIs might decrease the risk of cancer in Barrett’s esophagus. However, PPIs also have effects that conceivably could increase the risk of carcinogenesis, especially PPI effects on serum gastrin levels. Gastrin, which is released from antral G cells in response to meals and other stimuli, causes parietal cells in the gastric body and fundus to secrete acid. This acid stimulates D cells in the antrum to release somatostatin, which then inhibits the further release of gastrin from antral G cells. Thus, in a negative feedback loop, gastrin release stimulates acid secretion, which inhibits further gastrin release. By inhibiting gastric acid secretion, PPIs interrupt this negative feedback loop and cause serum gastrin levels to rise, and gastrin is a growth hormone that can cause proliferation in Barrett’s metaplasia. Without gastric acid, furthermore, bacteria can colonize the stomach. These bacteria can deconjugate bile acids, which can injure the esophagus at neutral pH levels, and might convert primary bile acids in the stomach into toxic secondary bile acids, such as deoxycholic acid, which can cause DNA double-strand breaks in Barrett’s epithelial cells. Bacteria also can convert dietary nitrates into potentially carcinogenic N-nitroso compounds.

Studies using Barrett’s esophageal adenocarcinoma cell lines have found that gastrin activates Janus kinase (JAK2) to trigger STAT3 signaling that stimulates proliferation and also increases levels of several antiapoptotic proteins. In addition, gastrin activates the CCK2 receptor, which also increases proliferation and reduces apoptosis in Barrett’s metaplasia. One retrospective study focused on gastrin levels in patients with GERD, nondysplastic Barrett’s metaplasia, low-grade dysplasia, high-grade dysplasia, and adenocarcinoma in Barrett’s esophagus who were taking PPIs. Although there were no significant differences in gastrin levels among the patients with dysplasia and adenocarcinoma, the patients with the highest quartile of serum gastrin levels were more likely to have high-grade dysplasia or cancer (odds ratio [OR], 5.46; 95% confidence interval [CI], 1.2–24.8). These studies suggest that gastrin has proproliferative and antiapoptotic effects that could contribute to the neoplastic progression of Barrett’s metaplasia.

A prospective randomized trial of low- and high-dose PPI therapy was performed in patients with Barrett’s esophagus to address concerns about gastrin and proliferation in Barrett’s metaplasia. During a 2-year follow-up period, the investigators followed serum gastrin levels and measured the length of Barrett’s metaplasia, with the rationale that PPI-induced elevations of serum gastrin would stimulate the growth of Barrett’s tissues. Although serum gastrin levels increased significantly in the study participants, the length of Barrett’s metaplasia did not change in either the low- or high-dose PPI groups, suggesting that PPI effects on serum gastrin levels might not have important clinical effects in patients with Barrett’s esophagus.

Some epidemiologic studies have found an association between PPI use and esophageal adenocarcinoma, but this association is likely to be the spurious result of a confounding-by-indication bias. Because GERD and Barrett’s esophagus are risk factors for esophageal adenocarcinoma, and because PPIs are often prescribed for patients with GERD and Barrett’s esophagus, an association between PPIs and esophageal adenocarcinoma might not be caused by the drug, but rather by the underlying conditions for which the PPIs were prescribed. One group explored this issue using the large, general practitioners research database in the United Kingdom. They found that patients who were taking PPIs or H ₂ RAs for an “esophageal indication,” such as GERD, had a significantly increased risk of developing esophageal adenocarcinoma (OR, 5.42; 95% CI, 3.13–9.39), but patients taking those drugs for a “gastroduodenal indication,” such as peptic ulcer disease, had no significantly increased risk of adenocarcinoma (OR, 1.74; 95% CI, 0.90–3.34). This is strong evidence of confounding by indication in associating PPI use with esophageal cancer. Furthermore, most studies that have sought an association between PPI use and adenocarcinoma in patients with Barrett’s esophagus have found a cancer-protective effect for these agents.