Session Title: Free Paper Session 19: Vascular Diseases & Diabetic Retinopathy V
Session Date/Time: Saturday 09/09/2017 | 16:30-18:00
Paper Time: 17:06
Venue: Room 120
First Author: : A.Augustin GERMANY
Co Author(s): : J. Campbell S. Sadda P. Dugel C. Wykoff N. Holekamp S. Kiss
The published literature is inconsistent regarding the relationship between the anatomic and functional outcomes of anti-vascular endothelial growth factor (anti-VEGF) therapy in diabetic macular edema (DME). This analysis assesses the association between a novel measure of anti-VEGF anatomic response - the extent of edema over the first 52 weeks of treatment - and average best corrected visual acuity (BCVA) improvement over 156 weeks of treatment, using an area under curve (AUC) approach.
Protocol I was a prospective Phase III multicentre trial of DME patients, enrolled between March 2007 and December 2008. The study was conducted by the DRCRnet at 52 sites in the USA. The current analysis is an independent, post-hoc analysis of the Protocol I data.
A post-hoc analysis of Protocol I data was conducted including DME patients randomized to ranibizumab plus deferred or prompt laser arms, with observed central retinal thickness (CRT) at baseline and at least one follow-up time point. Extent of edema was measured as the average central retinal thickness in excess of 250μm (AUC) over the first 52 weeks following treatment initiation. Eyes were then stratified into four cohorts (C1-C4) defined by quartiles of extent of edema. The association between the extent of edema at week 52, and the average change in BCVA over the 156-week study duration (AUC), was assessed in unadjusted bivariate analyses and after adjusting for potential confounders, such as baseline BCVA, using multivariate linear regression analysis.
367 patient-eyes met inclusion criteria. At week 52, mean average edema (AUC) varied from 3.6 μm for C1 (N=92, SD=2.8), to 18.2 μm for C2 (N=92, SD=5.8), 47.2 μm for C3 (N=92, SD=12.6), and 129.4 μm for C4 (N=91, SD=46.4). Relative cohort ranking persisted across years with a correlation coefficient of 0.67 (p<0.001). In unadjusted analyses, mean average (AUC) BCVA CFB at week 156 was not significantly different between patients with greatest vs. lowest amount edema at week 52 (7.5 vs. 9.3 letters in C4 vs C1; p=0.21). However, there were clinically important and statistically significant differences between cohorts with respect to baseline characteristics: patients with greater extent of edema at 12 months had worse vision at baseline (59.7 vs. 66.6 letters for C4 vs. C1, p<.001); and thicker retinas at baseline (505.5 vs.306.9 μm for C4 vs. C1, p <.001). After adjusting for these and other potential confounders, C4 eyes gained on average 8.0 fewer letters (95% CI: 11.3, 4.6 fewer letters; p<.001) over the first 156 weeks, compared with C1 eyes.
This post-hoc analysis assesses the association between a novel measure of anti-VEGF anatomic response - the extent of edema over the first 52 weeks of treatment - and average best corrected visual acuity (BCVA) improvement over 156 weeks of treatment, using an area under curve (AUC) approach. Further, this analysis uses multivariate regressions to control for potential confounders such as baseline BCVA. In this study, patient-eyes with the greatest average edema in the 52 weeks post-treatment initiation had significantly lower average BCVA improvement over 3 years, after adjustment for potential confounders. Based on this association between anatomic and functional response, timely consideration of alternative treatment regimens in patients with a persistently thicker macula may be considered to maximize potential vision improvement.