Proton pump inhibitors (PPIs) are widely used all over the world. First introduced in the late 1980s, they are widely prescribed as a first-line treatment in patients affected by acid-related disorders, such as gastroesophageal reflux disease (GERD), erosive disease, peptic strictures, and Barrett’s esophagus. In addition, PPIs are used for preventing gastrointestinal bleeding while on nonsteroidal anti-inflammatory drugs, in eosinophilic esophagitis, and non-ulcer dyspepsia (1, 2), successfully relieving symptoms and further complications in many cases. Available PPIs include esomeprazole, omeprazole, pantoprazole, dexlansoprazole, lansoprazole, and rabeprazole. They inhibit active parietal cell acid secretion acting by the irreversible inhibition of H⁺/K⁺ ATPase, the proton pump present in the gastric parietal cells, responsible for acid production. Since their introduction, remarkable evidence arose on possible associations between PPI use and adverse events, including the role of PPIs in the development of gastric cancer (GC) (3). Recently, in 2021, the role of PPI was investigated as a risk factor for GC, in a large population-based cohort of study (4), showing that the use of PPIs was associated with a 45% increased risk of GC compared with the use of Histamine-2 receptor antagonists (H2RAs). Similarly, in the same year, another large population-based cohort study (5), showed how PPI use was associated with a 2.37-fold increase in the general population. The association between PPIs and GC is biologically plausible and may be mediated by several different factors: PPIs, by inhibiting acid production, cause hypergastrinemia. Gastrin, secreted by antral G-cells, is considered a potent growth factor that may lead to enterochromaffin-like (ECL) cell hyperplasia (6). Second, non-acid gastric pH may contribute to gastric dysbiosis with an eventual overgrowth of strains promoting carcinogenesis (7, 8). Third, chronic suppression of acid secretion by PPIs may accelerate or induce the progression of gastric corpus atrophy (9).

Autoimmune atrophic gastritis (AAG), a chronic inflammatory disorder affecting the oxyntic mucosa, leads to progressive mucosal atrophy, reduced gastric acid secretion, and eventually to intestinal metaplasia (10), up to the possible development of epithelial lesions such as dysplasia and intestinal-type GC. AAG, which manifests itself most commonly with anemia or long-standing dyspepsia, is considered a condition at increased risk for GC (10). Whether PPIs may increase the risk of GC in AAG patients has not yet been investigated. This study aimed to investigate the relationship between PPI use before AAG diagnosis and the subsequent development of gastric neoplastic lesions (GNLs) at follow-up (FU) in a prospective cohort of AAG patients.

This article was drafted according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (Supplementary Table 1) to ensure the quality of reporting (11).

Patients with AAG who developed GNLs at FU were 35 (cases); 57.1% of patients were female gender and the median age was 67.6 (range 44–84) years. Cases were followed up for a median period of 2.3 (range 1–13) years, and the following GNLs were detected: 21 (60.0%) low-grade dysplasia, 3 (8.6%) high-grade dysplasia, and 11 (31.4%) intestinal-type GC. Regarding patients who presented low-grade dysplasia lesions at actual FU, none progressed to more invasive lesions, such as high-grade dysplasia or GC. The main features of the patients with GNLs are reported in Supplementary Tables 2, 3 .

Controls were 70 AAG patients without any GNLs at FU, and they were matched considering the age ( ± 3 years) at diagnosis, gender, and time of FU; 60% of controls were female, and the median age was 67.7 (range 42–86) years. The comparison of the main features between cases and controls is given in Table 2 .

The use of PPIs before the diagnosis of AAG was significantly more frequent in cases than in controls: 19 (54.3%) versus 13 (18.6%), (p<0.001).

Smoking habit was significantly more frequent in controls than in cases (48.6% vs. 11.4%, p<0.001). No significant difference was observed between cases and controls regarding the presence of severe corpus atrophy (37.1% vs. 38.6%, p=1.000).

The use of antiplatelet or anticoagulant drugs was reported in 14 (40.0%) cases vs. 11 (15.7%) controls; in particular, 3 cases reported the use of anticoagulant drugs, instead, none of the controls were on anticoagulant therapy. The use of antiplatelet or anticoagulant drugs was significantly more frequent in cases (p=0.008) (see Figure 2 ). At logistic regression, as reported in Table 3 , a positive association with GNLs in AAG was shown for PPI use before AAG diagnosis with an odds ratio (OR) of 9.6 (95%CI 2.3–40.3) while the other independent variables as the use of antiplatelets/anticoagulants (OR 2.8, 95%CI 0.7–12.0), 1st-degree family history for GC (OR 2.4, 95%CI 0.4–15.2), age ≥50 years (OR 2.0, 95%CI 0.2–18.1), and smoking habit (OR 0.4, 95%CI 0.1–2.1) were not associated with the development of GNLs at FU.

The current paper shows that in patients with AAG, the occurrence of GNLs was significantly related to the use of PPIs before the diagnosis of AAG. To the best of our knowledge, this is the first study focused on the association between the development of GNLs in patients with AAG. The association between GC and the use of PPIs has become a topic of considerable importance. In literature, several observational studies examined the association between PPIs and GC, all of which reported an increased risk with relative risks ranging between 1.1 and 3.6 (4, 5, 21–26). Salvo et al. (3), in 2020, performed an umbrella review to assess the associations between PPI use and adverse events. In addition to the significant association between PPIs and fractures, kidney disease, and infections (e.g., Clostridioides difficile), a significant pooled relationship was reported with GC development (OR 2.5, 95%CI 1.7–3.8). More recently, Seo SI et al. (5) analyzed two different Korean nationwide cohort databases and found that PPIs used for >30 days at a median FU of 4.3 years lead to a 2.4-fold GC risk when compared to non-PPI users, and GC incidence was associated with the duration of PPI use. Abrahami D et al. (4) investigated the GC risk associated with PPIs using a large population-based cohort from UK Clinical Practice Research Datalink showing that a PPI intake of median 5 years was associated with a 45% increased GC risk when compared to H₂-blockers (HR 1.4, 95%CI 1.1–2.0). According to our findings, the use of PPIs before AAG diagnosis seems to considerably increase the probability to develop GNLs in subsequent years in AAG patients (OR 9.6, 95%CI 2.3–40.3). The mechanisms by which this may occur are not easy to explain, as the drug was withdrawn at the time of diagnosis of AAG. The long-term effect of previous PPI use in AAG may deteriorate the labile balance of these patients already characterized by an increased GC risk, eventually acting as a promotor of some of the mechanisms, such as the progression of gastric oxyntic atrophy and gastric dysbiosis (7–9).

To this day, a specific definition of the long-term use of PPIs is not available since different studies consider different timings concerning the purpose of that specific study: a systematic review (15) explored the definitions of long-term PPI use in literature and reported how the definition of long-term PPI treatment varied substantially between studies and was seldom rationalized. The authors reported how the definition of long-term use may change, considering a clinical instead of a research context. For a research context, >6 months have been proposed as a possible definition for long-term PPI use. In the current study, a PPI use of at least 12 months was considered that can be viewed as a pretty- confident definition of the long-term use of PPIs for the purpose of this study, also taking into consideration that the GC risk seems related to the duration of PPI assumption (27)

The use of antiplatelet/anticoagulants was significantly more frequent in cases than in controls (40.0% vs. 15.7%, p=0.008), but the multivariate analysis did not confirm the use of these drugs as an independent predictor for the development of GNLs. Probably, the frequent use of antiplatelets among cases may be explained as one possible reason why many of these patients were on PPI treatment due to the well-known gastrolesive and bleeding-promoting action of these drugs contrasted by the blockade of gastric acid secretion (27).

PPIs are a widely accepted treatment for dyspepsia (28). However, dyspeptic symptoms, such as epigastric pain, early satiety, and/or postprandial fullness, were reported in 35.5%, 10%, and/or 7% of AAG patients, respectively (29), and may be clinical clues leading to the AAG diagnosis. Moreover, dyspepsia may sometimes hide more relevant organic conditions, such as GC. In our study, dyspepsia was also reported in 48.6% and 40.0% cases and controls, respectively. Relieving dyspeptic symptoms with PPIs could delay the AAG diagnosis and accelerate the undergoing oxyntic atrophic process. Instead, patients with persistent uninvestigated dyspepsia, in particular with long-term use of PPIs, should undergo gastroscopy with biopsies to rule out the presence of AAG or neoplastic complications (20).

Clinical practice guidelines on the appropriate use of PPIs and the duration of therapy when indicated have been described in several position papers (27, 30–32). Despite this, the misuse of PPIs is common. Taking together our findings and those of the previous papers (3–5, 21–26), clinicians’ awareness of the appropriateness of PPI prescription should be considered as a remarkable topic in daily clinical practice.

Our study suffers from some limitations: data on PPI classes were not recorded, but the GC risk seems not related to the PPI class used (33). The exact dosage of PPIs could also not be reliably collected. This case–control study was conducted at a single center and the sample size was obliged by the available number of AAG patients with GNLs. The serum gastrin values were not available for our AAG patients. Immunohistochemical staining to rule out the neuroendocrine differentiation in our GC series was not carried out; therefore, we are not able to rule out the idea that at least some of the found lesions might be linked to PPI use because of harboring a neuroendocrine differentiation not recognized by conventional immunohistochemistry. Qvigstad et al. (34) demonstrated in a series of GC the presence of neoplastic cells with neuroendocrine and probably ECL cell differentiation, using sensitive immunohistochemical techniques. This point would deserve particular attention, worthy of a further study specifically dedicated.

Nevertheless, our findings seem to add important pieces to the puzzle of the intriguing relationship between pharmacological acid suppression and GC risk, inviting us to reflect on the appropriate PPI use.

In conclusion, PPI use before the diagnosis of AAG appears to considerably increase the risk of subsequent GNL development. Considering the common misuse of PPIs, physicians should regularly reevaluate the appropriateness of ongoing PPI therapy, in particular in patients with a clinical suspicion of or already-diagnosed AAG.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by Sant’andrea Hospital ethical committee. The patients/participants provided their written informed consent to participate in this study.

EL and ED planned the study. EL, ED, MB, and MC conducted the study, collected and interpreted data, and drafted the article. ED and GE performed gastroscopies. EL and BA supervised the study. All authors have approved the final draft submitted.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.