An accurate quantification of photoreceptor degeneration was facilitated by fully automated segmentation making evident a progressive degeneration of photoreceptor inner segments and outer nuclear layer, even when accounting for age or distance to the boundary of geographic atrophy (GA), according to findings published in JAMA Ophthalmology.

For treatment trials in GA that are secondary to age-related macular degeneration (AMD), sensitive outcome measures are needed to show disease progression. Investigators combined a monocenter cohort study (Directional Spread in Geographic Atrophy [NCT02051998]) with an analysis of data from a tertiary referral center normative data study to quantify photoreceptor degeneration outside GA regions in eyes with nonexudative AMD, evaluate associations with GA progression, and characterize spatiotemporal progression.

A total of 158 eyes of 89 participants (51 women and 38 men, mean age 77.7 years [7.1; range, 64.1 years-93.7 years]) with a median follow-up time of 1.1 years (interquartile range [IQR], 0.52-1.7) and a median of 2 follow-up visits (IQR 1-4) had a median GA area of 8.87 mm2 (IQR 4.09-15.60). Ninety-three normal eyes from 93 control participants (age range, 18.0 years-84.5 years) were included for patient data standardization. The primary outcome measure was the association of outer nuclear layer (ONL) thinning with follow-up time after adjusting for retinal topography [z-score], age, and distance to GA boundary.

The fully automated B-scan segmentation proved accurate as an overlap measure between the human readers (dice coefficient, 0.82; 95% CI, 0.80-0.85; compared with manual markings), revealing marked interpatient variability in photoreceptor degeneration. The ellipsoid zone (EZ) loss-to-GA boundary distance and the normalized thicknesses of the outer segment (OS), ONL, and inner segment (IS) were all significantly associated with future progression rates of GA.

Continue Reading

Thicknesses for all 3 photoreceptor layers revealed a clear association with GA boundary distance, with marked thinning toward the GA boundary, confirmed by mixed models with GA boundary distance as a fixed effect. Adding follow-up time as a fixed effect resulted in all 3 photoreceptor layer models showing significantly better goodness of fit (likelihood ratio tests; P <.001 for ONL, P <.001 for IS, and P =.01 for OS). In participants, thickness of ONL decreased in a macula-wide manner by an average of -0.03 normal SD/y (95% CI, -0.05 to -0.01), corresponding to approximately -0.16 μm/y; IS thickness by -0.17 normal SD/y (95% CI, -0.17 to -0.06), approximately -0.18 μm/y; and OS by -0.04 normal SD/y (95% CI, -0.10 to -0.01), approximately -0.03 μm/y.  

Despite limitations, such as axial length measurements not being available for analysis, study authors still conclude, “this study outlines a fully automated pipeline to obtain photoreceptor degeneration related biomarkers in eyes with GA and provides accuracy estimates. A large between eye variability in photoreceptor degeneration is evident, which is prognostic for future progression rates. Lastly, adjusting for age and retinal position (z score transformation) and accounting for the distance to the GA boundary, it is evident that outer retinal degeneration is a function of a junctional zone component, and an independent macula wide component over time. This framework could be used to detect beneficial treatment effects beyond mere GA lesion size progression.”

Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.


Pfau M, von der Emde L, de Sisternes L, et al. Progression of photoreceptor degeneration in geographic atrophy secondary to age-related macular degeneration.  JAMA Ophthalmol. [Published online August 13, 2020]. doi: 10.1001/jamaophthalmol.2020.2914