Zhao · Clinical journal of the American Society of Nephrology : CJASN 2020 · systematic review and network meta-analysis of randomized trials · n=20 trials

Network Meta-Analysis of Novel Glucose-Lowering Drugs on Risk of Acute Kidney Injury.

Cited 82 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and network meta-analysis of randomized controlled trials

PubMed 33376101 · doi:10.2215/CJN.11220720 · record verified 2026-08-27

What was done

A systematic review and frequentist network meta-analysis of 20 event-driven cardiovascular or kidney outcome trials searched through September 2020. The authors compared the comparative risk of acute kidney injury (AKI) among three classes of glucose-lowering drugs: DPP-4 inhibitors, GLP-1 receptor agonists (GLP-1RAs), and SGLT2 inhibitors. The primary analysis included 18 trials restricted to type 2 diabetes; a secondary analysis included all 20 trials.

What was found

In 18 trials with 156,690 patients with type 2 diabetes and 2,051 AKI events, SGLT2 inhibitors were associated with a lower risk of AKI compared with placebo (odds ratio 0.76; 95% CI, 0.66 to 0.88). DPP-4 inhibitors and GLP-1RAs showed neutral effects on AKI risk. SGLT2 inhibitors were associated with lower AKI risk than GLP-1RAs (odds ratio 0.79; 95% CI, 0.65 to 0.97) and DPP-4 inhibitors (odds ratio 0.68; 95% CI, 0.54 to 0.86), with an 84% probability of being the safest intervention. Results were similar in the secondary analysis.

Why it matters

This study provides direct comparative trial evidence that SGLT2 inhibitors provide superior protection against acute kidney injury compared to both placebo and alternative novel diabetes medication classes.

Limits

The findings rely on secondary safety or adverse event reporting of AKI across trials designed primarily for cardiovascular or kidney composite endpoints, which may have used heterogeneous diagnostic criteria. Most evidence is restricted to populations with established cardiovascular or renal risk.

Cited by