Genetically adjusted cup-to-disk ratio for detecting primary open angle glaucoma

Abstract

Purpose : Cup-to-disk ratio (CDR) is a commonly used parameter for identifying glaucomatous optic neuropathy. However, its utility is limited due to extensive physiological variation in cup and optic disk size. Accounting for genetic determinants of physiologic variation in CDR can potentially improve screening utility of CDR. The aim of the present study was to develop a genetically adjusted CDR using previously known genetic variants associated with CDR and evaluate this adjusted CDR as a screening measure for POAG detection.

Methods : POAG cases (n = 168) and controls (n = 170) without glaucoma were defined based on manual review of clinical records from European ancestry patients with genotype data in the Mass General Brigham Biobank. Weighted genetic risk scores were calculated based on POAG-independent loci associated with CDR in a prior genome-wide association study. Genetically adjusted CDRs were calculated with a multiplicative correction factor integrating individual CDR genetic risk and average CDR genetic risk across controls. Receiver operating characteristic analyses were used to contrast unadjusted and adjusted CDRs in identifying POAG cases. Additionally, associations of unadjusted and adjusted CDRs with visual field mean deviation and retinal nerve fiber layer thickness were evaluated with linear regression models. All regression models were controlled for age, sex, and intraocular pressure. Secondary analyses were done defining cases and controls via ICD codes in patients of European and African ancestry.

Results : Using a CDR cutoff of 0.60, adjusted CDR had improved precision (0.929) compared to unadjusted CDR (0.909) with equivalent recall (1.000). Associations with visual field mean deviation were greater for adjusted CDR (β = -3.49, p < 0.001) than unadjusted CDR (β = -3.44, p < 0.001). On the other hand, associations with retinal nerve fiber layer thickness were greater for unadjusted CDR (β = -8.49, p < 0.001) than adjusted CDR (β = -8.12, p < 0.001). Secondary analyses using ICD-derived POAG cases resulted in similar findings for those of European and African ancestry.

Conclusions : Genetic adjustment to CDR may improve precision in glaucoma screening, although further work on separating physiological baseline variation from glaucomatous damage is needed.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

Figure 1. GRS-Adjusted CDR Differentiates Between Cases and Controls of European Ancestry.