Initiation of intravitreal aflibercept injection treatment in patients with diabetic macular edema: A review of VIVID-DME and VISTA-DME data

Background: Diabetic macular edema (DME) shows a gradual and sustained functional and morphologic response to anti-vascular endothelial growth factor (VEGF) drugs, but the optimal schedule for initiation of anti-VEGF therapy is not known. This study evaluates the treatment response behavior of DME in the Phase 3 trials of intravitreal aflibercept, with 5 initial intravitreal aflibercept injections (IAI), 2 mg every 4 weeks (2q4), in the upload phase. Methods: This post hoc pooled analysis of the VISTA-DME (NCT01363440) and VIVID-DME (NCT01331681) trials evaluated the change in best-corrected visual acuity (BCVA) and central retinal thickness (CRT) during the upload phase, using pooled data from both IAI treatment groups [2q4 and 2 mg every 8 weeks (2q8)]. The mean visit-to-visit change in BCVA and CRT, and the respective rate of gainers and losers was calculated for each successive visit. A secondary analysis compared the visit-to-visit change in BCVA between the 2q4 and 2q8 treatment arms during the upload period and the first year treatment period. Results: The majority of eyes showed a continuing improvement of BCVA after the first IAI. The proportions of eyes gaining BCVA (≥5 letters) at each visit compared with the previous visit during the IAI 2q4 upload phase were 60 (4-weeks), 19 (8-weeks), 16 (12-weeks), 15 (16-weeks), and 14 % (20-weeks). In contrast, the proportions of eyes losing BCVA (≥5 letters) were 3 (4-weeks), 7 (8-weeks), 7 (12-weeks), 9 (16-weeks), and 8 % (20-weeks), respectively. The odds of BCVA (≥5 letters) gain/loss exceeded 1.7 at each visit (range 1.7-20). Overall, the proportion of patients with BCVA gain ≥5 letters at week 20 (compared with baseline) was 76 and 80 % in the 2q4 and 2q8 groups, respectively. The proportions of eyes showing a visit-to-visit decrease in CRT of ≥30 μm during the first 5 IAI were 77 (4-weeks), 27 (8-weeks), 21 (12-weeks), 17 (16-weeks), and 12 % (20-weeks). In the secondary analysis, the BCVA outcomes were similar for the 2q8 and 2q4 treatment arms. Conclusions: The data presented here are consistent with continual functional and anatomic improvement following the fourth and fifth initial 2q4 injections, suggesting that an intensive and sufficiently long upload may be beneficial.