Meta-analysis of head-to-head clinical trials comparing incretin-based glucose-lowering medications and basal insulin: An update including recently developed glucagon-like peptide-1 (GLP-1) receptor agonists and the glucose-dependent insulinotropic polypeptide/GLP-1 receptor co-agonist tirzepatide.

OBJECTIVE: To assess comparative efficacy, safety and tolerability of injectable incretin-based glucose-lowering medications (IBGLMs) versus basal insulin treatment in patients with type 2 diabetes. RESEARCH DESIGN AND METHODS: We performed an updated meta-analysis of randomized clinical trials of head-to-head comparisons of IBGLMs (short- and long-acting glucagon-like peptide-1 [GLP-1] receptor agonists [GLP-1RAs] and glucose-dependent insulinotropic polypeptide [GIP]/GLP-1 receptor co-agonist tirzepatide) versus basal insulin using a PubMed database search (April 2022). The primary endpoint was difference in reduction of glycated haemoglobin (HbA) versus baseline between pooled IBGLMs (fixed-effects meta-analysis) and their subgroups (random-effects meta-analysis) versus basal insulin treatment (mean differences). Secondary endpoints were fasting plasma glucose, body weight, HbAtarget achievement, hypoglycaemia, blood pressure and lipids. Risk of bias assessment was performed using Jadad scores and the Risk of Bias tool 2.0. RESULTS: In all, 20 studies, representing 47 study arms and 11 843 patients, were eligible. Compared with basal insulin, IGBLMs lowered HbAby 0.48 (0.45-0.52)% more than did basal insulin treatment. This effect was driven by pooled long-acting GLP-1RAs (ΔHbA-0.25 [-0.38; -0.11]%) and the only GIP/GLP-1 receptor co-agonist, tirzepatide (pooled doses; ΔHbA-0.90 [-1.06; -0.75]%), while short-acting GLP-1RAs were equally effective compared with basal insulin (P = 0.90). All IBGLM subgroups achieved significantly lower body weight versus insulin treatment (-4.6 [-4.7; -4.4] kg), in particular tirzepatide (-12.0 [-13.8; -10.1] kg). IBGLMs significantly reduced hypoglycaemia and blood pressure and improved lipid variables. Risk of bias was low. IBGLM treatment was associated with more nausea, vomiting and diarrhoea and study medication discontinuation. CONCLUSIONS: Recently introduced, highly effective IBGLMs were superior to basal insulin treatment, reinforcing the recommendation that IBGLMs should be considered as the first injectable treatment for most patients with type 2 diabetes.