Network Meta-Analysis Ranks 15 Incretin Therapies for Type 2 Diabetes Efficacy and Safety

Type 2 diabetes mellitus (T2DM) is a complex, multisystem disease increasingly managed with incretin-based therapies (IBTs), a rapidly expanding drug class that includes mono-, dual-, and triple-receptor agonists targeting GLP-1, GIP, and glucagon receptors. As newer agents reach the clinic, head-to-head comparative data remain scarce, making network meta-analysis an essential tool for synthesizing indirect evidence and guiding treatment hierarchies.

This study, registered with PROSPERO and conducted per PRISMA 2020 and PRISMA-NMA guidelines, searched MEDLINE, Cochrane Library, Embase, and Chinese databases through August 2025. One hundred and two RCTs enrolling 98,693 T2DM patients were included. The 15 evaluated agents spanned GLP-1RAs (semaglutide, liraglutide, dulaglutide, exenatide, lixisenatide, orforglipron, polyethylene glycol loxenatide, efpeglenatide, ITCA 650), dual agonists (tirzepatide as GIP/GLP-1RA; mazdutide and survodutide as GLP-1/GCGR agonists), and the triple agonist retatrutide (GIP/GLP-1/GCGR). Bayesian network meta-analysis was performed using SUCRA ranking, alongside sensitivity, subgroup, and meta-regression analyses in R Studio 4.4.3. Risk of bias was assessed using Cochrane RoB 2.0 and the CINeMA framework, with the overall bias rated low.

All 15 IBTs significantly outperformed placebo across primary and secondary endpoints. For glycemic control, tirzepatide, orforglipron, and semaglutide ranked highest in HbA1c reduction and achievement of targets below 7.0% and 6.5%. For weight loss, the triple agonist retatrutide showed the greatest body weight reduction, consistent with its additional glucagon receptor activity enhancing energy expenditure. Favorable effects were also seen across lipid panels (LDL, TG, HDL, TC), systolic and diastolic blood pressure, renal function markers (eGFR, UACR), major adverse cardiovascular events (MACE), and insulin function indices (HOMA-β, HOMA-IR, fasting C-peptide). The study's authors suggest that cardiovascular protection likely arises from synergistic improvements across these interconnected metabolic pathways rather than a single mechanism.

Safety profiles were broadly favorable. The most common adverse events were mild, transient gastrointestinal symptoms (nausea, vomiting, diarrhea), consistent with the known mechanism of GLP-1 receptor activation. Hypoglycemia risk was low across the class. Higher doses generally improved efficacy outcomes but were associated with more adverse events; notably, orforglipron at 45 mg emerged as a particularly well-balanced dose offering meaningful efficacy with acceptable tolerability—a clinically useful finding for an oral GLP-1RA.

Subgroup analyses revealed that optimal treatment duration and combination regimen efficacy varied by agent, underscoring the importance of individualized therapy. Background medication combinations (with OADs or insulin) were systematically analyzed, addressing a gap in prior meta-analyses. The authors acknowledge that long-term, high-quality RCTs are still needed to confirm durability of cardiorenal outcomes and to better characterize rare adverse events across this expanding drug class.