In reply

We agree with Smith that the subject of hyperautofluorescent changes in Stargardt disease is an interesting one that may be illuminating to understanding underlying disease pathophysiology. In our report, we had examined longitudinal changes in fundus autofluorescence over a series of time points in a number of affected eyes. Our observations describe a consistent pattern of temporal and spatial progression in the flecks associated with Stargardt disease. The general pattern was one in which brighter hyperautofluorescent flecks move centrifugally from the center of the macula. As these flecks move outward, they leave in their wake darker, hypoautofluorescent lesions. As a result, fundus autofluorescence at a particular point in the fleck’s path may be described to be first increasing with time, becoming increasingly hyperautofluorescent relative to background, then decreasing back to background levels, and subsequently decreasing further to become hypoautofluorescent relative to background. We agree with Smith that within the time of our study (11-57 months), we did not observe further decreases in fundus autofluorescence in the tracks left behind by progressing flecks that may be described as frank atrophy (ie, the confluent absence of autofluorescence signal suggestive of retinal pigment epithelial cell loss).

One possibility is that these fleck progressions may indeed predispose toward retinal pigment epithelial atrophy but do so over a longer time than examined in either our study or the study by Smith et al.¹ Indeed, it can be imagined that if a fleck has a long “life cycle” between fleck initiation and the ultimate development of hypoautofluorescence/atrophy, correlative analysis over a shorter time can conversely report an anticorrelation between hyperautofluorescent flecks and frank atrophy. The observation that hyperautofluorescent flecks progress spatially from the foveal region to more peripheral areas combined with the observation that retinal pigment epithelial cell loss and frank atrophy generally arise late in the macular region do lead to a supposed relationship between the 2 observations, but we agree with Smith that these events may also be unrelated. Autofluorescence changes in Stargardt disease are important both to understand disease pathogenesis and for the development of new outcome measurements for clinical trials, and they should be investigated further.