Natural History of Autosomal Recessive Stargardt Disease in Untreated Eyes: A Systematic Review and Meta-analysis of Study- and Individual-Level Data.

TOPIC: Systematic review and meta-analysis of the natural history of autosomal recessive Stargardt disease (STGD1).

CLINICAL RELEVANCE: Controversy exists regarding the progression pattern of atrophic lesions secondary to STGD1, and the reported growth rates vary widely among studies.

METHODS: We searched in 6 literature databases up through March 15, 2019, to identify studies that monitored atrophy progression by fundus autofluorescence in untreated eyes with STGD1 for 6 months or more. We analyzed both study- and individual-level data from the included studies using 3 models: the area linear model (ALM), radius linear model (RLM), and area exponential model (AEM), in which the area, radius, and natural log-transformed area changes linearly with time, respectively. A horizontal translation factor was added to each dataset to correct for different participants' entry times into the studies. The best model was determined by the predicted age of lesion onset and dependence of growth rates on baseline lesion sizes. The risk of bias was assessed using the Newcastle-Ottawa scale.

RESULTS: Of 3158 articles screened, 7 studies (564 eyes) met our inclusion criteria. Cumulative study- and individual-level datasets fit along a straight line in the RLM after introducing horizontal translation factors to correct for different entry times (r2  = 0.99 and r2  = 0.93, respectively). The growth rate of effective lesion radius was 0.104 mm/year (95% confidence interval, 0.086-0.123 mm/year). The age of atrophy onset predicted by the RLM (22.7±5.0 years) was comparable to the reported age at onset of symptoms (22.1±3.1 years); in contrast, the predictions by the ALM and AEM deviated from this number by more than 5 years. Based on the individual-level data, the effective radius growth rate was independent of the baseline lesion size (r = 0.06); in comparison, the growth rates of area and natural log-transformed area were significantly dependent on the baseline lesion size (r = 0.47 and r = -0.33, respectively).

CONCLUSIONS: The progression of STGD1 lesions followed the RLM in both study- and individual-level data. The effective radius growth rate of atrophic lesions could serve as a reliable outcome measure to monitor STGD1 progression.