Session Title: Vascular Diseases and Diabetic Retinopathy II
Session Date/Time: Friday 18/09/2015 | 16:30-18:00
Paper Time: 17:50
First Author: : Y.Xiong UNITED STATES
Co Author(s): : J. Laurent P. Margaron A. Racine
PURPOSE:Ranibizumab 0.5 mg, an anti-vascular endothelial growth factor (anti-VEGF) agent, has been licensed since 2011 for the treatment of diabetic macular edema (DME) in >90 countries. Multiple large randomized controlled studies and post-marketing studies have demonstrated the favorable efficacy profile of ranibizumab 0.5 mg pro re nata (PRN) in improving patients’ best corrected visual acuity (BCVA). Recently, the 1-year results of the Protocol T trial conducted by the Diabetic Retinopathy Clinical Research Network (DRCR.net) have been reported, where ranibizumab 0.3 mg in a PRN regimen, a regimen never before reported, was compared with other anti-VEGF agents in patients with DME.
This work aims to assess ranibizumab dose responses and better understand the efficacy results in the total Protocol T study population as well as the subgroup with poor VA at baseline, within a quantitative framework using a two-step modeling approach.
Summary data from the randomized controlled studies RESOLVE, RESTORE, REVEAL, RESPOND, RETAIN, Protocol I, and Protocol T with PRN or treat-and-extend (TE) arms (sham, sham + laser, ranibizumab, and ranibizumab + laser treatments), as well as the sham + laser arms from the RISE and RIDE studies, were included in this analysis. An exploratory analysis was performed to verify a dose-response signal and then a dose-response model was built to describe the relationship between average monthly doses of ranibizumab and the mean BCVA change from baseline at the end of 1 year. The same analysis was also performed in a subgroup of patients stratified by baseline BCVA, and was based on available data from RESOLVE, RESTORE, REVEAL, RESPOND, RETAIN, and Protocol T, specifically in patients with a baseline BCVA of less than 69 letters.
The exploratory analysis using the pooled summary data verified a dose-response signal. An average monthly dose-dependent mean BCVA response within the range that the ranibizumab average monthly dose from Protocol T falls into was observed. The exploratory analysis also suggested that the ranibizumab 0.3 mg PRN dose provides VA gains of approximately 2-3 letters less than the VA gains achieved with 0.5 mg PRN dosing reported in previous phase III trials and in Protocol I, another trial conducted by DRCR.net. The dose-response model was then used and, considering inter-study variability, it was predicted that a median average BCVA increase of 13 letters would be expected at the end of 1 year if a ranibizumab 0.5 mg PRN arm were studied in Protocol T, with an average number of injections ranging from 7 to 7.5, which was observed in previous phase III trials. In the subgroup analysis including patients with a baseline BCVA less than 69 letters, it was predicted that a median average BCVA increase of 19 letters would be expected at the end of 1 year if a ranibizumab 0.5 mg PRN arm had been studied in Protocol T and an average number of 9 injections were given.
The ranibizumab 0.3 mg PRN regimen studied in Protocol T may have yielded sub-optimal VA gains compared with the robust, well controlled 0.5 mg PRN data reported in multiple phase III trials. After adjusting for the variability caused by other factors such as study design and retreatment algorithms, a ranibizumab 0.5 mg PRN regimen is predicted to yield an average BCVA change from baseline of 2-3 letters more than the 0.3 mg PRN regimen used in Protocol T. Furthermore, in a population with baseline BCVA worse than 69 letters, this difference in efficacy between ranibizumab 0.5 mg PRN and 0.3 mg PRN regimens is expected to be further enlarged to 4-5 letters.