IRD Clinical Trial Endpoints: Lessons Learned
Choosing the right outcome measures for inherited retinal disease trials requires matching endpoints to disease biology, therapeutic mechanism, and patient experience.
Melissa Yuan, MD, and Ahmad Al Moujahed, MD, PhD
Retina Today 
KEY TAKEAWAYS BCVA as a trial endpoint is most responsive when the disease disrupts central function early but tells little about progression in rod-cone dystrophies. The multi-luminance mobility test, Freiburg Visual Acuity Test, low-luminance visual acuity, full-field stimulus threshold testing, fundus autofluorescence, OCT-derived metrics, composite endpoints, and patient-reported outcome measures may all serve as more accurate endpoints in specific inherited retinal disease trials. With more than two dozen gene therapy trials in phase 2 or 3, endpoint strategy is integral to whether an effective therapy succeeds or fails on paper. Conventional ophthalmic endpoints using BCVA were developed for conditions such as cataract surgery, AMD, and diabetic retinopathy, for which large sample sizes and central BCVA changes are readily captured. Inherited retinal diseases (IRDs) challenge these conventional endpoints because patients with conditions such as retinitis pigmentosa (RP) may retain a VA of 20/20 for years while losing peripheral and night vision—changes that BCVA cannot detect. At the same time, patients with end-stage disease may already be at the measurement threshold, leaving no room to capture further visual decline.1,2 Phenotypic heterogeneity, slow progression, and small patient populations compound these difficulties.3 The expanding diversity of potential therapeutic strategies means endpoints must be matched to both disease phenotype and mechanism of intervention.4 With numerous candidates advancing through clinical development, endpoint selection has emerged as a critical determinant of trial success or failure. For example, the phase 3 STAR trial of timrepigene emparvovec (BIIB111, Biogen) for choroideremia investigated a primary endpoint of a gain of 15 ETDRS letters or more, which was borrowed from wet AMD trials where anti-VEGF therapy routinely produces such gains. Although a meaningful proportion of treated patients showed 10-letter gains versus controls, the rigid primary endpoint did not capture this signal, given the disease’s slow progression.5 Borrowing endpoints from other retinal diseases without adapting them to the natural history of the specific IRD risks masking a genuine treatment effect.6 ESTABLISHED CLINICAL ENDPOINTS BCVA, a component of almost all IRD trials, offers regulatory familiarity but alone may be insufficient.1,4,7 In the pivotal trial for voretigene neparvovec-rzyl (Luxturna, Spark Therapeutics) for RPE65-associated retinal dystrophy, BCVA did not reach statistical significance between arms; efficacy was instead demonstrated through the multi-luminance mobility test (MLMT), which measured navigation ability through a maze and past obstacles in dim light—a clinically meaningful endpoint. Electronic alternatives such as the Freiburg Visual Acuity Test extend measurement into the hand-motions range, serving as the primary endpoint in the phase 2b RESTORE trial of MCO-010 (Nanoscope) for advanced RP, in which patients’ vision was so poor, it was “off chart.”8 BCVA is most responsive when the disease disrupts central function early (as in achromatopsia or Leber congenital amaurosis [LCA] genotypes, including GUCY2D, CEP290, and RPE65), but it tells little about progression in rod-cone dystrophies where patients may continue to read 20/25 while their peripheral fields collapse.4 Contrast sensitivity remains underused in IRD trials. The standard Pelli-Robson chart samples a limited range and bottoms out quickly.
Eyewire headlines are now available as a weekly podcast! Listen Now
