Dose-response analysis of ranibizumab as-needed regimens for visual improvement in patients with diabetic macular edema using a modelling approach.

Background: Ranibizumab and aflibercept are anti-vascular endothelial growth factor therapies for diabetic macular edema (DME) but have only been directly compared in one study: the Protocol T study, a 24-month randomized controlled trial which compared the safety and efficacy of three anti-VEGF agents (ranibizumab 0.3 mg, aflibercept 2.0 mg and bevacizumab 1.25 mg). The ranibizumab dose used in Protocol T is not licensed for use outside of the US, where a higher ranibizumab dose of 0.5 mg is approved. Therefore, the relevance of the head-to-head Protocol T study findings to healthcare providers in Europe is limited. The purpose of this research was to predict the visual outcomes that may have been achieved in Protocol T with ranibizumab 0.5 mg.

Methods: A simplified dose-response model was constructed to describe the relationship between average monthly dose and one-year best corrected visual acuity (BCVA) change from baseline. A linear mixed effects model was evaluated and Bayesian Monte-Carlo Markov chains method was used to estimate the model parameters.

Results: If ranibizumab 0.5 mg PRN had been studied in Protocol T, it would have resulted in a BCVA gain of 14-15 early treatment diabetic retinopathy study (ETDRS) letters; 3-4 letters more than the actual BCVA gain reported with ranibizumab 0.3 mg PRN. In Protocol T patients with poor baseline BCVA (<69 letters), a similar additional letter gain would have been achieved.

Conclusion: The relevance of the Protocol T study findings are limited due to the use of ranibizumab 0.3 mg PRN which, based on the modelling approach reported herein, resulted in sub-optimal visual gains.