Author: Frank G. Holz
Co-authors: Roger Goldberg, Caleb Bliss, Ramiro Ribeiro, Charles Wykoff
Abstract
Purpose: To report the 18-month efficacy from DERBY and OAKS, two Phase 3, randomised, double-masked, clinical trials in geographic atrophy (GA) due to age-related macular degeneration (AMD), using three different statistical models to compare the relative effect of intravitreal pegcetacoplan dosed monthly or every other month (EOM) versus sham on GA lesion growth and growth rate. The 18-month safety will also be summarised.Setting: DERBY (NCT03525600) and OAKS (NCT03525613) are Phase 3, randomised, double-masked, sham-controlled trials.
Methods: Included patients are ?60 years old, have best-corrected visual acuity ?24 letters, and GA area between 2.5 and 17.5 mm2 or, if multifocal at baseline, at least one focal lesion ?1.25 mm2. The primary endpoint was change from baseline to Month 12 in GA lesion size (mm2) measured by fundus autofluorescence, using least-square means estimated from a mixed-effects model for repeated measures (MMRM). The rate of change in GA area between each pegcetacoplan treatment arm and the sham arm was also compared using a linear mixed-effects model assuming time as continuous and linear from baseline to Month 18. Similarly, a piecewise linear model evaluated the mean rate of change with added knots from baseline to Month 6, Month 6 to Month 12, and Month 12 to Month 18. All analyses were performed using an MMRM. Safety outcomes at 18 months are also summarised.
Results: Of 621 and 637 patients enrolled in DERBY and OAKS, 515 (82.9%) and 516 (81.0%) patients completed the studies through Month 18, respectively. At 18 months, pegcetacoplan decreased GA lesion growth vs sham by 13% (p=0.0254; nominal) and 12% (p=0.0332; nominal) in DERBY, and by 22% (p<0.0001; nominal) and 16% (p=0.0018; nominal) in OAKS, in the respective monthly and EOM arms. Using the linear slope model, the mean rate of change from baseline to Month 18 was 13% (p=0.0242; nominal) and 14% (p=0.0129; nominal), and 21% (p<0.0001; nominal) and 16% (p=0.0024; nominal) in the monthly and EOM arms vs sham, for DERBY and OAKS, respectively. The results using the piecewise linear slope model demonstrated similar results to the other two analyses. Safety results at 18 months were consistent with those at 12 months.
Conclusions: Overall, pegcetacoplan demonstrated sustained reductions in GA lesion growth at Month 18 in DERBY and OAKS. Efficacy results using the slope models were consistent with those from the primary statistical model. The piecewise slope model found that the treatment effect in DERBY over the first 6 months was not consistent with treatment effect in the subsequent treatment periods, with the treatment effect between 6 and 18 months being more consistent with that observed in OAKS. These findings continue to support the efficacy of pegcetacoplan in slowing the progression of GA lesions.
Financial Disclosure: F.G. Holz reports research grants and personal fees from Acucela, Allergan, Apellis, Bayer, Bioeq/Formycon, Roche/Genentech, Geuder, Heidelberg Engineering, ivericBio, Pixium Vision, Novartis, Zeiss; personal fees from Alexion, Boehringer-Ingelheim, Grayburg Vision, LinBioscience, Stealth BioTherapeutics, Aerie, Oxurion
