Predictive factors associated with three years of response to HbA1c goals with exenatide QW or insulin glargine: Post-hoc analysis of the DURATION-3 study

This post-hoc analysis of the DURATION-3 study aimed to identify factors associated with sustained glycaemic response with exenatide once weekly (QW) or insulin glargine (IG) among patients with type 2 diabetes. Response was defined as achieving treatment target of HbA1c <7.0% (<53 mmol/mol) at Week 26; sustained responders maintained the treatment target for ≥80% of remaining visits, including one during the final 6 months. Of 467 patients, 287 (61.5%) completed 156 weeks of treatment. At Week 26, 175 patients (61.0%) (exenatide QW, n = 95; IG, n = 80) achieved an HbA1c response. At Week 156, 84 of 175 responders (48.0%) had sustained response, with more sustained responders with exenatide QW (22.7% vs 13.9% with IG; P < 0.03). Logistic regression identified three predictors of sustained response: (a) exenatide QW vs IG treatment (odds ratio, 2.584 [95% confidence interval, 1.288-5.187]; P = 0.0075), (b) lower HbA1c at Week 26 (0.139 [0.053-0.366]; P < 0.0001), and (c) lower fasting serum glucose at Week 26 (0.693 [0.541-0.888]; P = 0.0037). A regression model was used to estimate the likelihood of sustained response with either treatment. This analysis provides a helpful tool for predicting sustained response with exenatide QW or IG.