Insulin secretion is a strong predictor for need of insulin therapy in patients with new-onset diabetes and HbA1c of more than 10%: A post hoc analysis of the EDICT study.

Abstract

To identify predictors of response to glucose-lowering therapy in patients with new-onset diabetes and very high HbA1c (>10%). The study included EDICT participants with an initial HbA1c of more than 10% (N = 104). All subjects received a 75-g oral glucose tolerance test (OGTT) before initiation of therapy, and then were randomized to receive: (a) initial triple therapy with metformin, pioglitazone and exenatide versus (b) stepwise conventional therapy with metformin followed by glipizide and then glargine insulin to reduce HbA1c to less than 6.5%. Insulin secretion and insulin resistance were calculated with OGTT-derived indices. Sixty-one per cent of participants in the conventional therapy group achieved HbA1c of less than 6.5% at 6 months without need of insulin therapy compared with 78% in the triple therapy group (P = NS). Insulin secretion at baseline was the strongest predictor of subjects who did not require insulin therapy; a cut point of CPEP120 /CPEP0 -the ratio between plasma C-peptide concentration at 120 minutes during the OGTT and fasting plasma C-peptide concentration-of more than 1.7 predicted subjects who achieved the treatment target without insulin, irrespective of the fasting plasma glucose (FPG) concentration and whether or not they were started on conventional or triple therapy. Subjects with a CPEP120 /CPEP0 of less than 1.7 plus FPG of 269 mg/dL or less (≤14.9 mmoL/L) also achieved the treatment goal with triple therapy. Insulin secretion in response to a 75-g OGTT predicts the need for insulin therapy at the time of type 2 diabetes (T2D) diagnosis. A cut point of 1.7 of CPEP120 /CPEP0 provides a useful clinical tool to individualize glucose-lowering therapy in patients with new-onset T2D and HbA1c of more than 10%.