Effectiveness of dulaglutide vs liraglutide and exenatide once-weekly. A real-world study and meta-analysis of observational studies.

INTRODUCTION AND AIM: Real-word data on the head-to-head comparisons among glucagon-like peptide-1 receptor agonists (GLP-1RA) are scant. Therefore, we aimed to compare the effectiveness of dulaglutide versus liraglutide and exenatide once weekly (exeOW) in type 2 diabetic (T2D) patients under routine care. METHODS: This was a retrospective, multicenter, real-world study on patients with T2D (aged 18-80) initiating a GLP-1RA between 2010 and 2018 at specialist outpatient clinics. We compared the effectiveness of dulaglutide versus liraglutide and exeOW on the changes in HbA1c (primary outcome), body weight, blood pressure and fasting glucose (secondary outcomes). Average follow-up was 5.9 months. Channelling biases were addressed with propensity score matching or multivariable adjustment. Meta-analyses of observational studies, covering the same comparisons, are also presented. RESULTS: 849, 1371 and 198 patients were included in the dulaglutide, liraglutide and exeOW groups, respectively. The reduction of HbA1c was greater with dulaglutide than with liraglutide (-0.24 ± 0.08%; p = 0.003), and was confirmed in the meta-analysis of observational studies. In our study, dulaglutide showed similar effectiveness compared to exeOW. When these results were pooled with other observational studies, dulaglutide showed a greater reduction of HbA1c (-0.19%; p = 0.003) and body weight (-0.8 kg; p = 0.007). CONCLUSIONS: In a real-world scenario, dulaglutide reduced HbA1c more than liraglutide. Conversely, we found similar effect of dulaglutide and exeOW, with statistical differences arising solely when results were meta-analysed with those from other observational studies. Lack of up-titration for liraglutide and higher discontinuation rate for exeOW likely influenced the estimated treatment difference.