Age-related macular degeneration (AMD) is the most frequent cause of irreversible blindness in the elderly in developed countries. Our previous studies implicated activation of complement in the formation of drusen, the hallmark lesion of AMD. Here, we show that factor H (HF1), the major inhibitor of the alternative complement pathway, accumulates within drusen and is synthesized by the retinal pigmented epithelium. Because previous linkage analyses identified chromosome 1q25-32, which harbors the factor H gene ( HF1 / CFH ), as an AMD susceptibility locus, we analyzed HF1 for genetic variation in two independent cohorts comprised of ≈900 AMD cases and 400 matched controls. We found association of eight common HF1 SNPs with AMD; two common missense variants exhibit highly significant associations (I62V, χ ² = 26.1 and P = 3.2 × 10 ^-7 and Y402H, χ ² = 54.4 and P = 1.6 × 10 ^-13 ). Haplotype analysis reveals that multiple HF1 variants confer elevated or reduced risk of AMD. One common at-risk haplotype is present at a frequency of 50% in AMD cases and 29% in controls [odds ratio (OR) = 2.46, 95% confidence interval (1.95-3.11)]. Homozygotes for this haplotype account for 24% of cases and 8% of controls [OR = 3.51, 95% confidence interval (2.13-5.78)]. Several protective haplotypes are also identified (OR = 0.44-0.55), further implicating HF1 function in the pathogenetic mechanisms underlying AMD. We propose that genetic variation in a regulator of the alternative complement pathway, when combined with a triggering event, such as infection, underlie a major proportion of AMD in the human population.

With the introduction of spectral-domain optical coherence tomography (OCT), much larger image datasets are routinely acquired compared to what was possible using the previous generation of time-domain OCT. Thus, the need for 3-D segmentation methods for processing such data is becoming increasingly important. We report a graph-theoretic segmentation method for the simultaneous segmentation of multiple 3-D surfaces that is guaranteed to be optimal with respect to the cost function and that is directly applicable to the segmentation of 3-D spectral OCT image data. We present two extensions to the general layered graph segmentation method: the ability to incorporate varying feasibility constraints and the ability to incorporate true regional information. Appropriate feasibility constraints and cost functions were learned from a training set of 13 spectral-domain OCT images from 13 subjects. After training, our approach was tested on a test set of 28 images from 14 subjects. An overall mean unsigned border positioning error of 5.69 plusmn 2.41 mum was achieved when segmenting seven surfaces (six layers) and using the average of the manual tracings of two ophthalmologists as the reference standard. This result is very comparable to the measured interobserver variability of 5.71 plusmn 1.98 mum.

Research from JAMA Ophthalmology — Automated Analysis of Retinal Images for Detection of Referable Diabetic Retinopathy

To report the durability of voretigene neparvovec-rzyl (VN) adeno-associated viral vector-based gene therapy for RPE65 mutation-associated inherited retinal dystrophy (IRD), including results of a phase 1 follow-on study at year 4 and phase 3 study at year 2. Open-label phase 1 follow-on clinical trial and open-label, randomized, controlled phase 3 clinical trial. Forty subjects who received 1.5×1011 vector genomes (vg) of VN per eye in at least 1 eye during the trials, including 11 phase 1 follow-on subjects and 29 phase 3 subjects (20 original intervention [OI] and 9 control/intervention [CI]). Subretinal injection of VN in the second eye of phase 1 follow-on subjects and in both eyes of phase 3 subjects. End points common to the phase 1 and phase 3 studies included change in performance on the Multi-Luminance Mobility Test (MLMT) within the illuminance range evaluated, full-field light sensitivity threshold (FST) testing, and best-corrected visual acuity (BCVA). Safety end points included adverse event reporting, ophthalmic examination, physical examination, and laboratory testing. Mean (standard deviation) MLMT lux score change was 2.4 (1.3) at 4 years compared with 2.6 (1.6) at 1 year after administration in phase 1 follow-on subjects (n = 8), 1.9 (1.1) at 2 years, and 1.9 (1.0) at 1 year post-administration in OI subjects (n = 20), and 2.1 (1.6) at 1 year post-administration in CI subjects (n = 9). All 3 groups maintained an average improvement in FST, reflecting more than a 2 log10(cd.s/m2) improvement in light sensitivity at 1 year and subsequent available follow-up visits. The safety profile was consistent with vitrectomy and the subretinal injection procedure, and no deleterious immune responses occurred. After VN gene augmentation therapy, there was a favorable benefit-to-risk profile with similar improvement demonstrated in navigational ability and light sensitivity among 3 groups of subjects with RPE65 mutation-associated IRD, a degenerative disease that progresses to complete blindness. The safety profile is consistent with the administration procedure. These data suggest that this effect, which is nearly maximal by 30 days after VN administration, is durable for 4 years, with observation ongoing.