Impact of Fundus Autofluorescence Pattern on Geographic Atrophy Progression in Untreated Eyes: A Meta-analysis

Abstract

Purpose : 6 distinct patterns of increased fundus autofluorescence (FAF) in the junctional zone of geographic atrophy (GA) have been identified and reported in the literature. However, the exact impact of each FAF pattern on GA progression is unclear. The reported progression rates of GA area within the same FAF pattern vary widely across different clinical trials. Also, there is little knowledge of FAF patterns' impact on the pattern of GA lesion expansion. In this meta-analysis, we assessed the impact of each FAF pattern on GA progression in untreated eyes with nonexudative AMD.

Methods : We performed a systematic literature search of Ovid MEDLINE, Ovid Embase, Cochrane Library, clinicaltrials.gov, and PubMed. We included studies that classified subjects based on FAF patterns and measured GA size in untreated eyes over at least 6-month period. In the analyses, we included 6 FAF patterns: “None”, “Focal”, “Banded”, “Patchy”, “Diffuse Nontrickling”, and “Diffuse Trickling”. Random-effect meta-analysis was performed to compare different FAF groups. We also derived the lesion growth function for each FAF group. In order to account for the subjects’ different entry times, we introduced a horizontal translation factor to shift each data subset until r² was maximized for the cumulative trend line. Heterogeneity, study quality, and publication bias were evaluated in accordance with the guideline of meta-analysis of observational studies in epidemiology.

Results : Of 1704 articles reviewed, we identified 7 studies (935 eyes) that met our inclusion criteria. The 6 FAF patterns can be clustered into 4 groups based on GA radius growth rate (Table). Each group has a distinct radius growth rate with very high r² (Figure): Group 1 (“None” pattern, 0.061 mm/year, r²=0.996), Group 2 (“Focal” pattern, 0.105 mm/year, r²=0.987), Group 3 (“Banded”, “Patchy”, and “Diffuse Nontrickling”, 0.149 mm/year, r²=0.993) and Group 4 (“Diffuse Trickling”, 0.245 mm/year, r²=0.986).

Conclusions : The FAF patterns are true phenotypes that significantly affected the GA radius progression rate without altering the progression pattern. The differences in GA progression rates among FAF patterns may be associated with the area of FAF instead of the specific patterns.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

Table. Statistical Comparison in GA Radius Growth Rate between Each Pair of FAF Patterns.

 

Figure. Comparison in GA Lesion Progression among Different FAF Groups.