Lack of association between primary angle-closure glaucoma susceptibility loci and the phenotypes axial length to lens thickness ratio and lens thickness to axial length factor

Abstract

Purpose : The genetic background in the development of primary angle closure glaucoma (PACG) is still inconsistent and little is known in European populations. This study investigated the genetic overlap between the biometric risk predictors of PACG axial length to lens thickness ratio (ALR) or the lens thickness to axial length factor (LAF) and known genetic variants associated with PACG.

Methods : Within the Gutenberg Health Study (GHS), a genome-wide association study for ALR and LAF was conducted in two cohorts (GHS 1, n = 1725 and GHS 2, n = 832) of Caucasian individuals (age range, 40 - 79) with a subsequent meta-analysis of both cohorts. The GHS is a population-based prospective, observational single center study in the Rhine-Main-Region in mid-western Germany. Axial length (AL) and lens thickness (LT) were measured using optical low-coherence reflectometry (Lenstar LS 900). Any pseudophakic or aphakic eye was excluded from analysis. The ALR was calculated by dividing AL by LT and the LAF by dividing LT by AL x 10. The average of both eyes or the value of the remaining eye were used for association analysis.
Genotyping was performed using the Affymetrix Genome-Wide Human SNP Array 6.0 and genotyped data were imputed with MACH/Minimac to the 1000 Genomes reference panel (phase 1, March 2012). Genome-wide association was tested by a linear regression model, adjusting for age and sex (ProbABEL 0.4.5 software). Finally, a meta-analysis based on the effect and standard error of both single analyzed cohorts (GHS 1 and GHS 2) was performed using the METAL software.

Results : None of the known SNPs to be associated with PACG (rs1015213, rs3753841, rs11024102, rs1401999, rs4656461, rs1900004, rs17576 or rs2250880) showed an association with ALR or LAF (p>0.05 for all listed SNPs). However, we identified novel SNPs of borderline significance associated with ALR and LAF. Results of the ALR exhibited systematically lower p-values than for the LAF.

Conclusions : While ALR and LAF are clinical useful risk predictors for PACG, they showed no strong genetic overlap with genetic variants previously associated with PACG, but these clinical-based endophenotype might help to improve our genetic insight. However, our results emphasize that the knowledge of the genetic background of PACG is still limited.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.