SGLT2 inhibitors and gastrointestinal cancer: reassuring, and not the same as safe

Design

Systematic review and meta-analysis of randomised controlled trials, PROSPERO-registered

Population

48,765 adults with type 2 diabetes across 48 randomised trials of SGLT2 inhibitors against placebo or active comparator

Primary outcome

Gastrointestinal neoplasms, mostly captured as reported adverse events rather than adjudicated cancer endpoints

Effect

Overall OR 1.10 (95% CI 0.84 to 1.44, p = 0.46, I² 0%); no significant association at any individual site

Forty-eight randomised trials in type 2 diabetes, 48,765 patients, were pooled for gastrointestinal neoplasms. There was no association overall, odds ratio 1.10 (95% CI 0.84 to 1.44, p = 0.46) with heterogeneity of 0%. Site-specific estimates were all non-significant: oesophageal 1.12 (0.37 to 3.45), gastric 1.20 (0.65 to 2.23), hepatic 0.62 (0.31 to 1.22), pancreatic 0.91 (0.51 to 1.64), colonic 1.28 (0.78 to 2.08), colorectal 0.76 (0.27 to 2.17), rectal 0.98 (0.49 to 1.97). Subgroup analyses by agent, age, body mass index, HbA1c, duration and dose found nothing either.

The limitations are structural rather than incidental, and the authors say so. About half the trials ran for a year or less; solid tumours take longer than that to become detectable after an exposure begins. Cancers were captured as reported adverse events or registry entries rather than adjudicated endpoints, so ascertainment is inconsistent across trials. Event counts at individual sites are small, which is why the oesophageal interval runs from 0.37 to 3.45.

For an oncologist this matters mostly as an answer to a question patients and physicians ask. A patient with type 2 diabetes and a cancer diagnosis, or a survivor on an SGLT2 inhibitor for heart failure or chronic kidney disease, does not have a reason from this evidence to stop the drug — and given what SGLT2 inhibitors do for those conditions, stopping on an unsupported cancer worry would be the larger harm. Say reassuring, say short follow-up, and do not say proven.

• Do not stop an SGLT2 inhibitor over gastrointestinal cancer risk; nothing here supports it.
• Frame the evidence as short-term — half the trials followed patients for a year or less.
• Cancers were recorded as adverse events, not adjudicated endpoints; ascertainment is uneven.
• Site-specific intervals are wide enough to be uninformative on their own.
• The cardiorenal indications for these drugs are unaffected by this analysis.

The statistics, in plain English

Heterogeneity of 0% means the trials agreed with each other, which makes the pooled odds ratio of 1.10 a stable estimate — but agreement between short trials cannot substitute for length. The interval, 0.84 to 1.44, still permits a 44% relative increase, and with cancer incidence being what it is that is not a small absolute possibility. The site-specific figures are weaker again: an interval from 0.37 to 3.45 for oesophageal cancer tells you only that very few events occurred. The correct summary is an absence of a detectable signal in trials that were mostly too short to detect one, which is different from an absence of risk.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.