Reflux Symptom Reporting, Zastaprazan For EE, PPIs and CV Outcomes

---

Three interesting esophageal papers were published in the February issue of the American Journal of Gastroenterology.

The first study, by Guadagnoli et al, found that real-time reflux symptom reporting is multifactorial, such that both psychological and physiological processes appear to play an important role in the gastroesophageal reflux disease experience. Importantly, the study found that real-time esophageal symptom reporting can be influenced by reflux and both general and health-specific psychological variables.

The second study, by Oh et al, compared the therapeutic effect of a new potassium-competitive acid blocker (PCAB), zastaprazan, with esomeprazole for healing and controlling symptoms of patients with erosive esophagitis (EE). This new PCAB represents the growing number of PCABs emerging in the GERD market locally and globally. I am expecting many more drugs in this class soon. The study demonstrated noninferiority of zastaprazan to esomeprazole in all clinical end points except esophageal healing at four weeks.

The third study, by Mosholder et al, reevaluated the relationship between proton pump inhibitors and cardiovascular events using a meta-analysis. A unique aspect of this study is its inclusion of only randomized controlled trials with at least 100 patients, a treatment duration longer than 30 days and a non-PPI comparator. The study further cemented the safety profile of PPIs by demonstrating no association of cardiovascular events with PPI treatment.

---

Psychological and Physiological Factors and Real-Time Reflux Symptom Reporting

Am J Gastroenterol 2025;120(2):344-352

Patient-reported reflux symptoms are not well correlated with reflux episodes, suggesting that a complicated set of variables, likely including psychological factors, influences reflux symptoms. This prompted researchers to conduct a study on the association of physiological and psychological factors with real-time symptom reports in patients with refractory reflux.

Participants included consecutive adult patients treated at a tertiary gastroenterology clinic between 2009 and 2014 who remained symptomatic (heartburn and/or regurgitation) after at least 12 weeks of treatment with a PPI and who were then referred for 24-hour pH-multichannel intraluminal impedance (MII) testing. These patients’ symptoms were unexplained after esophagogastroduodenoscopy. Those who had prior esophageal surgery, a prior or current esophageal motility disorder, a primary symptom of dyspepsia, or only atypical symptoms were excluded.

During the pH-MII, patients reported symptoms in real time (button press within two minutes of experiencing the symptom). In addition, at the time of testing, patients completed 12 psychological questionnaires that assessed various domains (e.g., general mental health, health-specific anxiety, personality, etc.). The investigators used principal component analysis to generate five psychological variables (health anxiety, general psychological symptoms, personality, pain coping, social functioning) from across questionnaires and looked at the Childhood Trauma Questionnaire independently.

The 393 patients included in the study were, on average, 48 years of age, and 60% were female. Patients reported symptoms an average of eight times and had an average of 44 reflux episodes during the test. Among the 333 patients who reported any symptoms, 53% were considered symptom association probability (SAP)-positive (SAP =95%) and 59% were symptom index (SI)-positive (>50% of symptoms associated with a reflux episode).

A higher level of generalized psychological symptoms was associated with a greater likelihood of experiencing at least one reflux symptom. Other psychological symptoms, number of reflux episodes and PPI use were not associated with likelihood of reflux symptoms. Among patients who experienced at least one reflux symptom, greater health anxiety, greater general psychological symptoms, a higher total number of reflux episodes, being off a PPI during pH-MII and missing PPI information were associated with a higher reported number of reflux symptoms, while higher childhood trauma severity was associated with a lower reported number.

In machine learning analyses to predict reporting of reflux symptoms based on psychological and physiological indicators, the best model explained 53% of the variance in reflux symptom frequency during pH-MII.

When assessing which factors are associated with positive SAP and SI, the investigators found that having a higher number of reflux episodes and being off (vs. on) a PPI during pH-MII were associated with positive SAP. The same variables and having missing PPI info (vs. being on PPI during pH-MII) were associated with positive SI.

In a machine learning analysis to predict whether a patient is SAP-positive or SAP-negative, the best model misclassified 41% of patients.

---

Zastaprazan Versus Esomeprazole for EE

Am J Gastroenterol 2025;120(2):353-361

Investigators in Korea conducted a phase 3, randomized controlled trial to compare the efficacy and safety of zastaprazan, a PCAB, and esomeprazole, a PPI, for patients with EE. PCABs have several expected advantages over PPIs, including full efficacy from the initial dose and flexible dosing relative to mealtime.

Patients in this noninferiority trial were randomized 1:1 to 20 mg of zastaprazan or 40 mg of esomeprazole for eight weeks. Patients were assessed at baseline, week 4 and week 8, with the primary end point being the proportion who achieved complete mucosal healing by week 8. In addition, the investigators assessed change in Reflux Disease Questionnaire (RDQ) score, GERD Health-Related Quality of Life (GERD-HRQL) score, patient-reported proportion of days without major symptoms and proportion of patients with healed EE (absence of esophageal mucosal erosions or ulcers on endoscopy) by week 4. Analyses were conducted in both the full sample of patients who received at least one dose and among the per-protocol sample.

A total of 149 and 151 patients were randomized to zastaprazan and esomeprazole, respectively. Patients in the zastaprazan and esomeprazole arms were comparable across baseline characteristics (mean age, 53 vs. 55 years; proportion female, 37% vs. 39%; body mass index, 25.0 vs. 25.1 kg/m²; current smokers, 20% vs. 17%; current alcohol consumption, 55% vs. 50%; Los Angeles classification Grade A, 66% vs. 67%; LA classification Grade B, 30% vs. 29%; LA classification Grade C/D, 4% vs. 4%, all respectively).

The proportion of patients with healing at week 4 was slightly higher in the zastaprazan arm than the esomeprazole arm but was comparable at week 8 (Table). The results were similar in the per-protocol sample analysis. In the subset of patients who were LA classification B/C/D only, there was some evidence of a greater benefit of zastaprazan for healing, but the limited sample size influenced estimate precision (Table).

Symptom improvement, as indicated by RDQ score and GERD-HRQL changes, was comparable in the zastaprazan and esomeprazole arms. Statistically significant improvements were seen in mean RDQ scores and GERD-HRQL scores at weeks 4 and 8 relative to baseline, and there were no statistically significant differences across arms. A subgroup analysis found that this was also true among the subgroup of patients with baseline LA classification B/C/D.

Safety data were available for 297 patients, 57 of whom reported 80 adverse events. A lower proportion of patients in the zastaprazan group than the esomeprazole group reported treatment-emergent adverse events (16% vs. 22%, respectively). There were no serious adverse events or deaths in either arm. In both groups, serum gastrin increased from baseline to week 8, although to a slightly greater extent in the zastaprazan group (zastaprazan: 45.3-114.9 pg/mL; esomeprazole: 42.7-92.2 pg/mL). Serum gastrin levels returned to baseline levels for both groups by two weeks after treatment. There were no other differences in vitals, ECG findings or laboratory tests between the two arms during the study.

---

Meta-Analysis of RCTs to Assess PPI Use and Adverse Cardiovascular Outcomes

Am J Gastroenterol 2025;120(2):362-369

To assess the association of PPI use with adverse cardiovascular outcomes suggested in some observational studies, investigators at the FDA conducted a meta-analysis of PPI clinical trial data.

Data from all randomized controlled trials assessing PPIs that had at least 100 participants, at least 30 days of randomization and included at least one non-PPI comparator were included in the analysis. The primary outcome assessed the incidence rates of major adverse cardiovascular events-plus (MACE+) (nonfatal myocardial infarction, nonfatal stroke, fatal cardiovascular adverse events, hospitalization for unstable angina or coronary revascularization) in those taking PPIs compared with those taking comparators.

Among 52 trials with a placebo comparator (14,998 patients taking PPIs, 8,323 patients taking placebo), the summary incidence rate ratio (IRR) of MACE+ from the random effects in those taking PPIs instead of placebo was 0.72 (95% CI, 0.42-1.26), and the incidence rate difference (IRD) was 0.62 per 100 person-years (95% CI, 0.04-1.20). Among 61 trials with any active comparator (12,505 patients taking PPIs, 8,566 taking active comparator), the summary IRR was 1.19 (95% CI, 0.75-1.90) and the IRD was 0.38 per 100 person-years (95% CI, 0.13-0.63).

Related Content

On-Demand Vonoprazan, GLP-1s in GERD Patients, Young-Onset BE and More

In this new column, I will focus on all things upper GI, from motility to Barrett’s esophagus and beyond, ...

Research Identifies Distinct Patterns and Predictors of PPI Adherence in Patients With Barrett’s Esophagus

Although PPIs are a mainstay in preventing cancer in patients with Barrett’s esophagus, 56% of these patients ...

Among 51 trials with H2 receptor antagonists as comparators (9,430 patients taking PPIs, 6,050 patients taking H2 receptor antagonists), the summary IRR was 1.04 (95% CI, 0.49-2.19) and the IRD was 0.56 per 100 person-years (95% CI, 0.24-0.88). Thus, in all trial subsets, the analyses indicated statistically significant associations on the additive but not multiplicative scale.

When analyzing results by indication (GERD, n=24 trials; peptic ulcer, n=11 trials; miscellaneous, n=17 trials), the investigators found an IRD of 1.04 per 100 person-years (95% CI, 0.58-1.50) in the GERD trials, but no association of PPIs with MACE+ in peptic ulcer or miscellaneous indication trials. The investigators also found no association between PPIs and MACE+ in the subset of three trials that included patients taking concomitant low-dose aspirin.

This article is from the May 2025 print issue.