The effects of ertugliflozin on β-cell function: Pooled analysis from four phase 3 randomized controlled studies.

Abstract

To identify potential predictors and mediators of changes in β-cell function in response to ertugliflozin treatment in people with type 2 diabetes mellitus (T2DM). Data from patients with T2DM randomized to ertugliflozin (5 or 15 mg; observations from both doses were pooled) or placebo in four phase 3 clinical studies (clinicaltrials.gov: NCT01958671, NCT02226003, NCT02036515, NCT02099110) were pooled and analysed. Change from baseline in β-cell function at week 26 was assessed, and its potential predictors and mediators were analysed using linear and multiple regression analyses. Compared with placebo, ertugliflozin improved β-cell function when assessed by mean percent change from baseline in homeostatic model assessment of β-cell function (HOMA-%β; ertugliflozin: 14.7%, 95% confidence interval [CI] 12.3, 17.1; placebo: -0.4%, 95% CI -3.4, 2.5], but not when assessed by change in C-peptide index following a mixed meal tolerance test. Change in HOMA-%β correlated with change from baseline in glycated haemoglobin (HbA1c) and treatment with ertugliflozin, and weakly with change from baseline in body weight. In the ertugliflozin group, change in HOMA-%β correlated with baseline fasting plasma glucose (FPG; r = 0.235, P < 0.001), baseline HbA1c (r = 0.138, P < 0.001), baseline homeostatic model assessment of insulin resistance (HOMA-IR; r = 0.162, P < 0.01), and baseline HOMA-%β (r = -0.321, P < 0.001) in linear regression analyses. Multiple regression analyses yielded similar results. In people with T2DM, ertugliflozin treatment improved fasting β-cell function, but no effect on postprandial β-cell function was observed in this analysis. Improvement in HOMA-%β was predicted by high baseline FPG, HbA1c, HOMA-IR, and low baseline HOMA-%β, and mediated by ertugliflozin treatment, and improved HbA1c and body weight.