Gene therapy for leber congenital amaurosis caused by RPE65 mutations: safety and efficacy in 15 children and adults followed up to 3 years

Abstract

Objective: To determine the safety and efficacy of subretinal gene therapy in the RPE65 form of Leber congenital amaurosis using recombinant adeno-associated virus 2 (rAAV2) carrying the RPE65 gene.

Design: Open-label, dose-escalation phase I study of 15 patients (range, 11-30 years of age) evaluated after subretinal injection of the rAAV2- RPE65 vector into the worse-functioning eye. Five cohorts represented 4 dose levels and 2 different injection strategies.

Main outcome measures: Primary outcomes were systemic and ocular safety. Secondary outcomes assayed visual function with dark-adapted full-field sensitivity testing and visual acuity with Early Treatment Diabetic Retinopathy Study charts. Further assays included immune responses to the vector, static visual fields, pupillometry, mobility performance, and optical coherence tomography.

Results: No systemic toxicity was detected; ocular adverse events were related to surgery. Visual function improved in all patients to different degrees; improvements were localized to treated areas. Cone and rod sensitivities increased significantly in the study eyes but not in the control eyes. Minor acuity improvements were recorded in many study and control eyes. Major acuity improvements occurred in study eyes with the lowest entry acuities and parafoveal fixation loci treated with subretinal injections. Other patients with better foveal structure lost retinal thickness and acuity after subfoveal injections.

Conclusions: Gene therapy for Leber congenital amaurosis caused by RPE65 mutations is sufficiently safe and substantially efficacious in the extrafoveal retina. There is no benefit and some risk in treating the fovea. No evidence of age-dependent effects was found. Our results point to specific treatment strategies for subsequent phases.

Application to clinical practice: Gene therapy for inherited retinal disease has the potential to become a future part of clinical practice.

Trial registration: clinicaltrials.gov Identifier: NCT00481546.

Figure 1

Fundus images with near-infrared illumination and sites of retinal detachments from subretinal injections of vector-gene in the 15 patients (P1-P15) with RPE65-LCA. Dotted circles on the images of individual patients represent the estimated areas of retinal detachment from drawings at time of surgery. The tip of each white ‘syringe’ indicates the retinotomy site that produced the detachment. All images are depicted as left eyes for comparability. F indicates fovea.

Figure 2

Visual function in all clinical trial participants analyzed with full-field stimulus testing (FST) and the transient pupillary light reflex (TPLR). A, FST sensitivity (mean±SD) to blue stimuli measured under dark-adapted conditions in each eye of each subject at four baseline visits (white bars) and all postoperative visits to date (black bars). Normal range for the FST is shown along the horizontal axis. B, Changes in FST sensitivity from mean baseline value in control and study eyes. White bars depict the intervisit variability at 4 baseline visits. Black bars represent all available postoperative timepoints to date ordered in groups for each patient. p values refer to two-sided paired t-test statistics between indicated groups. C, TPLR sensitivity to green flashes under dark-adapted conditions in each eye of each subject measured at 1, 3 or 6 months postoperatively (black bars) compared to baseline (white bars). Normal range for the TPLR is shown along the horizontal axis. D, Postoperative changes in TPLR sensitivity from baseline in control and study eyes. p value refers to two-sided paired t-test statistics performed between postoperative and baseline timepoints.

Figure 3

Dark-adapted visual field maps in study eyes to localize regions of improved sensitivity after treatment. All maps are depicted as left eyes for comparability and with an overlaid schematic of retinal features (optic nerve and posterior pole vessels) for reference to fundus images (Figure 1). Loci in the visual field that consistently showed ≥8 dB of sensitivity change during the postoperative period are highlighted (green). Estimated boundaries of blebs resulting from subretinal injections are depicted (dotted circles, see Figure 1) on the visual field maps to ask whether there is any correspondence between locations of injection and the responding loci. I indicates inferior visual field; N nasal field; P, patient; S, superior field; T, temporal field.

Figure 4

Mobility performance of clinical trial participants as measured by the number of navigation incidents experienced while traveling an indoor course of fixed length, for five ambient illumination levels. A, Change from baseline performance for the study (first row) and control (second row) eyes of patients in Cohorts 4 and 5, as a function of ambient illumination. At lower ambient illuminations, mobility performance was better with the study eye after treatment; at the highest illumination, patients were able to navigate almost without errors with either eye before and after treatment. There where changes in interocular differences with treatment (third row) with most cases showing a difference in performance of the study eyes relative to the control eyes. Values are averages of 4 repetitions performed for each illumination level except for 100 lx, at which two runs were performed. B, Postoperative performance difference between eyes (study minus control) for each patient. Most patients tended to show better relative mobility performance when using the study eye for lower illumination levels; there was no such effect at 100 lx. C, Difference in performance grouped by illumination. As a group, patients show less incidents when navigating with the study eye postoperatively. Symbols at the top of each panel indicate the group mean interocular difference (study minus control); surrounding brackets are 95% confidence levels. P-values (t-test) indicate significance of the departure of IOD from zero for each ambient illumination level.

Figure 5

Retinal location and instability of fixation in RPE65-LCA eyes at baseline and postoperatively, and its relation to changes in visual acuity. A, D, Fixation clouds of all study eyes during a 10 sec epoch recorded while gazing to a 1 deg diameter stationary target adjusted to be visible to each eye. Foveally fixating eyes shown in A and extrafoveally fixating eyes shown in D. Circular patterns show the standard grid centered on the anatomical fovea extending to radii of 1.65, 5 and 10 degrees. All panels are shown in equivalent left retina representation. N and T refer to nasal and temporal retina, respectively. B, Fixation instability values are shown as change from mean baseline values at all postoperative visits in control and study eyes with foveal fixation. VAs are shown as change from mean baseline at all postoperative visits in control and study eyes with foveal (C) or extrafoveal (E) fixation. Limits for 0.30 logMAR (15 letter or 3 line) gain or loss shown with vertical dashed lines.

Figure 6

Foveal structure and quantitation of thickness using OCT scans in control and study eyes. A, Foveal thickness measurements in control and study eyes at baseline, and at short-term and long-term postoperative timepoints. Changes from baseline are displayed adjacent to the foveal thickness measurements. FD, the eyes that had foveal detachments as part of their subretinal injection procedure. B, OCT scans along the horizontal meridian in four representative patients – one without foveal detachment (P7) and three with foveas detached at the time of the procedure (P1, P13, P6). Ellipses denote the central retinal region of interest that shows changes in the IS/OS lamination in 3 of the 4 study eyes at early timepoints but with some resolution at later times. C, Longitudinal reflectivity profiles (LRPs) through the fovea in the patients compared with a normal LRP (upper row, left). The LRPs are color coded and labeled for ONL, IS+OS and RPE to illustrate the postoperative changes. ONL and IS+OS measurements are shown to the right of the LRPs.