Abstract
Purpose :
Numerous clinical trials have compared anti-VEGF agents, dosing, and/or injection frequencies, but no single trial can compare all available treatment regimens. We used a network meta-analysis (NMA), which includes only those studies that have a comparator in common, to evaluate outcomes between anti-VEGF regimens for nAMD across clinical trials.
Methods :
A systematic literature review identified randomized controlled trials (RCTs) of anti-VEGF therapy for nAMD (Table). Interventions were defined by drug, dose, and regimen (fixed Q4W, Q8W, Q12W, pro re nata [PRN], or treat-and-extend [T&E]). Over 20 outcomes were assessed to determine if they were reported with sufficient consistency to allow inclusion in the NMA (feasibility analysis). The literature and expert opinion were consulted to assess whether differences in specific baseline characteristics between studies would impact the outcomes included; if so, that impact was factored into the NMA.
Results :
Feasibility analysis determined that the outcomes that could be compared using NMA were 3 best-corrected visual acuity (BCVA) outcomes (measured in Early Treatment Diabetic Retinopathy Study letters): mean change from baseline at month (M)12 and M24 and patients (%) gaining ≥15 letters at M12. All studies reported similar baseline characteristics, with few exceptions; no effect modification. For BCVA change at M12, 21 RCTs were included (Table). Baseline VA varied from 52−65 letters across all trials. Ranibizumab (RAN) Q4W and PRN (the most commonly compared treatment regimens) were used as reference cases. The NMA found no statistically significant differences in any of the evaluable outcomes in any pairwise comparison with RAN Q4W or PRN (comparisons with RAN PRN shown in Figure).
Conclusions :
This NMA suggests that the different regimens evaluated have not resulted in superior vision benefits compared with RAN PRN or Q4W. If confirmed, these results suggest that greater visual acuity improvements may require a different approach to treatment, such as a different drug delivery paradigm or the development of drugs with different or multiple mechanisms of action.
This is a 2020 ARVO Annual Meeting abstract.