Network meta-analysis accurately predicted the outcome of a subsequent randomised trial comparing once weekly dulaglutide 1.5 mg and once daily liraglutide 1.8 mg.

BACKGROUND: Glucagon-like peptide-1 (GLP-1) receptor agonists are an established treatment for people with type 2 diabetes (T2D). We aimed to indirectly compare two GLP-1 receptor agonists, once weekly dulaglutide 1.5 mg and once daily liraglutide 1.8 mg, as a part of clinical trial planning. METHODS: Studies for inclusion in the network meta-analysis (NMA) included all available dulaglutide and liraglutide data as of November 2011 as well as results from the exenatide once weekly registration programme. Change from baseline in haemoglobin A1c (A1c) was the primary endpoint, and a 26-week treatment effect was estimated. RESULTS: Data for 7135 people with T2D from 15 randomised controlled trials (RCTs) followed for 12-52 weeks were included in the quantitative analysis. Observed results from the NMA predicted an A1c change from baseline of -15.1 mmol/mol (-1.38%) in the dulaglutide 1.5 mg group and -14.7 mmol/mol (-1.34%) in the liraglutide 1.8 mg group, with a predicted treatment difference (dulaglutide-liraglutide) of -0.4 mmol/mol (-0.04%) [95% credible interval: -2.4 to 1.5 mmol/mol (-0.22% to 0.14%)]. CONCLUSIONS: The subsequent RCT primary result of a -0.7 mmol/mol (-0.06%) treatment difference (dulaglutide-liraglutide) in A1c demonstrated that once weekly dulaglutide 1.5 mg and once daily liraglutide 1.8 mg resulted in similar glycaemic control, which was consistent with the NMA-predicted treatment difference. NMA is a useful tool and should be considered during clinical trial planning.