Review finds no clear increase in digestive-tract tumors with SGLT2 drugs
A pooled analysis of 49 randomized trials found no statistically significant link, but long-term risk remains uncertain.
High evidenceReviewSome caution advised
Medical disclaimer: This article summarizes research findings and is for informational purposes only. It is not medical advice.
Editorial illustration — not from the study.
The review examined trials in which people with type 2 diabetes received an SGLT2 inhibitor or a placebo or another diabetes medicine. The researchers combined the trial results and assessed overall gastrointestinal neoplasms as well as esophageal, gastric, liver, pancreatic, colonic, colorectal, rectal, and biliary outcomes specified in the review methods.
Across the combined trials, the researchers did not find a statistically significant increase in gastrointestinal neoplasms among people receiving SGLT2 inhibitors. Results were also not statistically significant when the researchers compared different drugs, doses, treatment durations, ages, body mass indexes, or hemoglobin A1c levels. These findings indicate that this analysis did not detect an increased risk during the follow-up periods studied; they do not establish that any possible long-term risk is absent.
The question reviewed
The researchers conducted a systematic review and meta-analysis of randomized controlled trials in people with type 2 diabetes. They searched five medical databases through March 17, 2025, for trials comparing an SGLT2 inhibitor with a placebo or another active comparator. The analysis included 49 trials and 49,054 participants. The main outcome was gastrointestinal neoplasms overall. Prespecified additional outcomes included esophageal, gastric, hepatic, pancreatic, colonic, colorectal, rectal, and biliary neoplasms. The researchers also examined whether results differed by drug, dose, treatment duration, age, body mass index, or hemoglobin A1c.
Key conclusions
For gastrointestinal neoplasms overall, the pooled odds ratio was 1.10, with a 95% confidence interval of 0.84 to 1.44. This was not statistically significant. Site-specific results were also not statistically significant: esophageal neoplasms had an odds ratio of 1.12, gastric 1.20, hepatic 0.62, pancreatic 0.93, colonic 1.28, colorectal 0.76, and rectal 0.98. The reported confidence intervals were wide for several sites, and all reported p-values were above 0.05. Subgroup analyses by individual SGLT2 inhibitor, dose, treatment duration, age, body mass index, and hemoglobin A1c were also not statistically significant. These results show no detected association in the included trials, rather than proving that SGLT2 inhibitors cannot affect tumor risk.
Where this may apply
These findings may be most relevant to adults with type 2 diabetes who resemble participants in the included randomized trials and to the treatment durations represented there. They provide limited information about people without type 2 diabetes, children, people with substantially different health conditions, or risks after many years of exposure. The analysis also does not determine an individual person's risk or establish that long-term gastrointestinal tumor risk is absent.
Why it matters
Gastrointestinal neoplasms are uncommon outcomes that may be difficult to assess in individual clinical trials. Combining many randomized trials can provide a broader assessment than one trial alone and may be reassuring regarding an increased risk detectable during the studied follow-up periods. The main remaining question is whether the available trials were long enough and large enough to identify rare tumors or effects that might emerge only after many years.
Limitations & evidence assessment
The authors specifically noted that about half of the trials had follow-up of one year or less, limiting assessment of long-term risk. The abstract does not report the number of tumor events, which makes it difficult to judge how much information was available for each outcome. Several site-specific confidence intervals were wide, consistent with limited precision for rare outcomes. As a meta-analysis, the findings depend on the quality, follow-up, outcome definitions, and reporting of the underlying trials. The abstract states that publication bias was evaluated with funnel plots and Egger’s test, but it does not provide those results.
Why this evidence level: Meta-analysis pooling multiple studies sits at the top of common evidence hierarchies.
Evidence levels are editorial estimates derived from study metadata — they are not clinical appraisals.
// Source
Clinical and Experimental Medicine · 2026 · DOI: 10.1007/s10238-026-02300-6
This review examined human clinical evidence about GLP-1 receptor agonists and neovascular age-related macular degeneration, sometimes called wet age-related macular degeneration, in people with type 2 diabetes. The reviewed studies had conflicting results: one reported higher risk, another lower risk, and two large comparisons found no significant increase. The evidence is observational and does not show that these medicines cause or prevent the eye condition.
A prospective, multicenter phase II trial examined whether removing only the treated tumor area through the anus could preserve the rectum in people with locally advanced rectal cancer who responded well to chemoradiotherapy. Among 80 participants, three-year disease-free survival was 84.8%, and local recurrence occurred in 5.0% during the reported follow-up. Because everyone received local excision and there was no comparison group, the findings do not establish that this approach is as effective as more extensive surgery.
This retrospective human study compared a visual CT scoring method with an automated computer analysis for tracking fibrosing interstitial lung disease. Among 92 people followed with CT scans and lung-function tests, both approaches were associated with changes in lung function and signs of disease progression. The study does not show that either method improves treatment or patient outcomes.
