Outcomes of Diabetic Macular Edema Patients by Baseline Hemoglobin A1c: Analyses from VISTA and VIVID

Abstract

To examine the relationship between glycemic control at baseline and response to anti-vascular endothelial growth factor treatment for diabetic macular edema (DME). Post hoc analysis of 2 similarly designed phase III trials, VISTA and VIVID. Patients with central-involved DME. Both VISTA and VIVID compared efficacy and safety of intravitreal aflibercept injection (IAI) with macular laser photocoagulation for DME. Current analysis focused on comparison within each treatment group in an integrated VISTA and VIVID dataset. Baseline hemoglobin A1c (HbA1c) was partitioned into 4 quartiles: 4.5% to <6.7% (n= 233), 6.7% to <7.4% (n= 206), 7.4% to <8.6% (n= 209), and 8.6% to <14.7% (n= 208). Outcomes were analyzed by mixed model for repeated measures. Intragroup differences were quantified by a regression model. Change from baseline best-corrected visual acuity (BCVA), central subfield thickness (CST), and HbA1c. In the IAI group, mean BCVA improvement from baseline did not depend on baseline HbA1c at week 52 (P= 0.1852), but seemed to be dependent at week 100 (P= 0.0425). The mean CST reduction from baseline was independent of baseline HbA1c at both weeks 52 (P= 0.1857) and 100 (P= 0.7346). Mean HbA1c change from baseline in IAI group was small across all HbA1c quartiles. In the laser group, the mean BCVA gain decreased with increasing baseline HbA1c at both weeks 52 (P= 0.0421) and 100 (P= 0.0001). Similarly, the mean CST decrease was greater with decreasing baseline HbA1c, at both weeks 52 (P= 0.0065) and 100 (P=0.0162). The mean HbA1c change from baseline in the laser group was minimal across HbA1c quartiles, although glycemic control tended to worsen in upper quartiles. The benefit of IAI in patients with DME was less dependent on their presenting glycemic status as opposed to laser.