Spatial Decomposition of Longitudinal RNFL Maps Reveals Distinct Modes of Glaucomatous Progression With Structure–Function and Genetic Signatures

Purpose: To determine whether spatial decomposition of longitudinal retinal nerve fiber layer (RNFL) change maps reveals distinct modes of glaucomatous progression masked by conventional averaging, and to validate these modes through structure-function mapping and genetic association analysis. Methods: Pixel-wise RNFL rates of change were computed from longitudinal optic disc OCT scans of 15,242 eyes (8419 adults with POAG; Massachusetts Eye and Ear, 1998-2023). A loss-only constraint zeroed all thickening values, reflecting the biological prior that adult RNFL does not regenerate. Non-negative matrix factorization decomposed these maps into spatial progression components (80% training set). Components were evaluated in a held-out set (20%) for retinotopic structure-function concordance, visual field progressor classification against global and quadrant RNFL rates, and enrichment of genetic association signals at established POAG loci. Results: Six anatomically distinct progression patterns emerged, including diffuse circumferential loss, focal peripapillary defects, and arcuate bundle degeneration. Pattern-based models provided significant incremental predictive information over global RNFL rate (integrated discrimination improvement, 0.049; P = 0.003), with concordant gains in discrimination (area under the curve, 0.795 vs. 0.768) and variance explained (pseudo-R², 0.382 vs. 0.311). Structure-function mapping confirmed retinotopic coherence. In an exploratory genetic analysis, spatial pattern weights showed stronger association signals than the global RNFL rate at the majority of established POAG susceptibility loci, consistent with the hypothesis that spatial decomposition may reduce phenotypic heterogeneity. Conclusions: Glaucomatous structural progression occurs through spatially distinct modes with independent structure-function and preliminary genetic signatures that conventional RNFL averaging obscures.