Purchase this article with an account.
Tiarnan D L Keenan, Neal L Oden, Elvira Agron, Traci E Clemons, Alice Henning, Lars Fritsche, Wai T Wong, Emily Y Chew; Cluster analysis and phenotype-genotype assessment of geographic atrophy secondary to age-related macular degeneration in the Age-Related Eye Disease Study 2. Invest. Ophthalmol. Vis. Sci. 2021;62(8):2942.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To explore whether geographic atrophy (GA) in age-related macular degeneration (AMD) consists of two or more partially distinct phenotypic subtypes and to test genetic associations. This is important since GA subtypes driven by different genetic factors might require customized therapeutic approaches.
AREDS2 participants with incident GA were eligible (one eye per participant). Phenotypic features from reading center grading of fundus photographs were subjected to cluster analysis (Table), by both k-means and hierarchical methods. In pre-specified hypothesis tests, identified clusters were compared by four pathway-based genetic risk scores: complement (19 SNPs/6 loci), extracellular matrix (6 SNPs/5 loci), lipid (7 SNPs/4 loci), and ARMS2 (1 SNP/locus).
The cohort comprised 598 individuals (mean age 74y). In cross-sectional phenotypic analyses, k-means identified two clusters: A and B (367/231 members, respectively), while hierarchical clustering identified four (Figure): C-F (451/112/12/5). In longitudinal phenotypic analyses, k-means identified two: G and H (310/288); hierarchical clustering identified none. The groups of clusters were not correlated with one other by membership (Pearson’s r ≤ 0.20). In cross-sectional analyses, GA configuration was the predominant factor for A-E membership. In longitudinal analyses, smoking status determined G vs H. Despite adequate power, pairwise cluster comparison by the four genetic risk scores demonstrated no significant differences (p>0.05 for all).
In this large study, cross-sectional cluster analyses revealed GA subtypes defined principally by GA configuration. However, these subdivisions were not replicated in longitudinal analyses. Given the absence of significant cluster-genotype associations, for any eye with GA, physicians are unlikely to infer the main genetic driver of GA from phenotype alone. The inconsistencies in optimal cluster numbers and characteristics suggest that GA may show continuous phenotypic variation across a spectrum, rather than consisting of phenotypic subtypes that remain partially distinct over time, with separate genetic etiologies.
This is a 2021 ARVO Annual Meeting abstract.
Phenotypic characteristics used as input for clustering.
Dendrogram: hierarchical clustering (cross-sectional dataset). Phenotypic dissimilarity is plotted (y-axis); each participant is shown (x-axis).
This PDF is available to Subscribers Only