Abstract
Purpose :
The DRCR Network Protocol V randomized trial demonstrated that among eyes with center-involved diabetic macular edema (CI-DME) and good visual acuity (VA; Snellen equivalent 20/25 or better), there were no significant differences at 2 years in 5-letter or more loss of visual acuity comparing initial observation with aflibercept only if VA decreased versus initial focal/grid laser with aflibercept only if VA decreased versus aflibercept. Further exploration of outcomes using the observation approach is relevant for managing patients
Methods :
Post hoc analyses of the initial observation group from Protocol V were undertaken. Participants were adults with type 1 or 2 diabetes with one study eye having CI-DME and good VA randomly assigned to initial observation. Eyes were carefully followed and given aflibercept if VA decreased by 10 or more letters from baseline at any time or 5 to 9 letters at 2 consecutive visits. Visits occurred at 8 and then every 16 weeks unless there was worsening of VA or OCT central subfield thickness (CST).
Results :
Over 2 years, 80 of 236 (34%) eyes randomly assigned to initial observation received aflibercept. Overall, median 2-year VA was 86.0 letters (Snellen equivalent 20/20). Participants with a fellow eye receiving DME treatment within 4 months of randomization were more likely to receive aflibercept (52% vs 25%, hazard ratio = 2.55 (95% CI, 1.64-3.99), P < .001). Receipt of aflibercept was also more likely among eyes with moderately severe nonproliferative diabetic retinopathy (ETDRS retinopathy severity level 47) or worse disease versus moderate nonproliferative diabetic retinopathy (retinopathy severity level 43) or less severe retinopathy (51% vs. 27%, hazard ratio = 2.22 (95% CI, 1.42-3.47), ordinal P < .001) and eyes with CST of at least 300 µm (Zeiss-Stratus equivalent) versus CST less than 300 µm (45% vs. 26%, hazard ratio = 1.98 (95% CI, 1.26-3.13), continuous P = .005).
Conclusions :
Most eyes randomly assigned to initial observation with aflibercept only if VA worsened did not need aflibercept for VA loss and still had good VA at 2 years. The likelihood of receiving aflibercept for VA loss was about double among participants with a fellow eye receiving treatment for DME, eyes with more advanced DR, and eyes with greater CST.
This is a 2020 ARVO Annual Meeting abstract.