# Gastric Neoplasm Risk with DPP-4 Inhibitors, GLP-1 Receptor Agonists, and SGLT2 Inhibitors: Network Meta-Analysis of Randomized Trials

## TL;DR

This study finds that GLP1RAs may lower gastric cancer risk while DPP4is may increase it in long-term diabetes patients.

## Contribution

First network meta-analysis comparing gastric neoplasm risks across three glucose-lowering drug classes using global RCT data.

## Key findings

- GLP1RAs showed 49% lower gastric neoplasm risk versus controls in long-term diabetes patients.
- DPP4is showed 77% higher gastric neoplasm risk versus controls in older populations.
- Risk associations were strongest in trials with duration ≥52 weeks and in participants with diabetes.

## Abstract

Whether the risk of gastric neoplasm is modified by newer glucose-lowering therapies—dipeptidyl peptidase-4 inhibitors (DPP4is), glucagon-like peptide-1 receptor agonists (GLP1RAs), and sodium–glucose cotransporter 2 inhibitors (SGLT2is)—remains uncertain. Given their global uptake and long-term use in populations already predisposed to malignancy, decision-grade comparative safety evidence is needed. We conducted a systematic review and network meta-analysis (NMA) of randomized controlled trials (RCTs) in adults without baseline gastric neoplasms. PubMed, Embase, Cochrane CENTRAL, Web of Science, ClinicalTrials.gov, ClinicalKey, ProQuest, and ScienceDirect were searched from inception to 10 January 2026, without language restrictions. The primary outcome was incident gastric neoplasms (benign or malignant). Random-effects frequentist NMA estimated risk ratios (RRs) with 95% confidence intervals (CIs); Bayesian NMA served as sensitivity analysis. Certainty of evidence was assessed using GRADE adapted for NMA (PROSPERO CRD420261282728). Fifty-two RCTs (171,165 participants; mean age 63.6 years; 36.9% women; mean follow-up 141.8 weeks) were included. At the class level, GLP1RAs were associated with lower gastric neoplasm risk versus controls (RR = 0.51, 95% CI = 0.28–0.92), whereas DPP4is were associated with higher risk (RR = 1.77, 95% CI = 1.09–2.85). These signals persisted in prespecified subgroup analyses among participants with diabetes mellitus, in trials with duration ≥52 weeks (GLP1RA: RR = 0.52, 95% CI = 0.28–0.95; DPP4i: RR = 2.05, 95% CI = 1.19–3.55), and in older populations (age ≥60 years; DPP4i: RR = 2.08, 95% CI = 1.15–3.77). No class showed a significant association in younger participants (<60 years) or shorter trials (<52 weeks). Across available RCT evidence, GLP1RA prescription generally had a relatively lower gastric neoplasm risk than controls. In contrast, among patients with diabetes mellitus receiving longer-term therapy, GLP1RAs may be the preferable option from the perspective of gastric neoplasm risk, while DPP4is warrant heightened vigilance and mechanistic clarification. These findings support improved neoplasms ascertainment in future trials rather than immediate prescribing changes.