Purpose
To analyze if there is linearity in the growth of geographic atrophy (GA) using different units of measurement.
Methods
We reviewed all visits of patients included in a prospective natural history study of GA, the GAIN study (NCT01694095), and those who attended at 6-month intervals (+/- 1 month) during 18 months were selected. Fundus autofluorescence was used to measure the area of atrophy. The progression was analyzed using slopes and growth was expressed in each of the following 5 metrics: size of atrophy in mm2, difference of growth between visits (diff mm2), growth of lesion size in percentage relative to the previous visit (%), square root (sqr) and difference of growth between visits expressed in square root (diff sqr). Patients were stratified in tertiles of baseline area of atrophy, and those in the extremes (small and large lesions) were compared in terms of the slope in the first and last periods (0-6 and 12-18 months) for each metric. We used the patients in the extremes to avoid confounding data from the medium sized lesions and from the intermediate period.
Results
The final sample included 26 eyes of 22 patients (72.7% females, with a median of 79 years old and a baseline size of atrophy of 4.48 mm2). The median size of lesions in the 1st and 3rd tertiles was 1.44 and 9.94 mm2, respectively. Absolute measures increased, while % of growth decreased (p<0.05). When mm2 or sqr were considered, there was a slower growth of atrophy (as identified by progressively decreasing slopes) with time, which was particularly marked in smaller lesions (p<0.05). When growth of atrophy is evaluated globally, a very high linear correlation coefficient is found (r>0.97, p<0.001); this analysis precludes the detection of the aforementioned subtle decrease in the speed of growth when measures are segmented in short periods of time.
Conclusions
The metrics in which progression of GA is measured have a large impact in its characterization and it cannot be assumed to be linear in all cases. Absolute measures (mm2, sqr) show a decreasing growth with time especially in small lesions, whereas spread of atrophy expressed in % decreases with time. These results may be useful in clinical trials, since a slowing down of the progression of growth may be expected in the natural history of GA and should not be attributed to a potential beneficial effect from interventional therapies.
Keywords: 412 age-related macular degeneration •
461 clinical (human) or epidemiologic studies: natural history