Multitrait analysis of glaucoma identifies new risk loci and enables polygenic prediction of disease susceptibility and progression

Variation in FOXC1 and PITX2 is associated with Axenfeld-Rieger syndrome, characterised by structural defects of the anterior chamber of the eye and a range of systemic features. Approximately half of all affected individuals will develop glaucoma, but the age at diagnosis and the phenotypic spectrum have not been well defined. As phenotypic heterogeneity is common, we aimed to delineate the age-related penetrance and the full phenotypic spectrum of glaucoma in FOXC1 or PITX2 carriers recruited through a national disease registry. All coding exons of FOXC1 and PITX2 were directly sequenced and multiplex ligation-dependent probe amplification was performed to detect copy number variation. The cohort included 53 individuals from 24 families with disease-associated FOXC1 or PITX2 variants, including one individual diagnosed with primary congenital glaucoma and five with primary open-angle glaucoma. The overall prevalence of glaucoma was 58.5% and was similar for both genes (53.3% for FOXC1 vs 60.9% for PITX2, P=0.59), however, the median age at glaucoma diagnosis was significantly lower in FOXC1 (6.0±13.0 years) compared with PITX2 carriers (18.0±10.6 years, P=0.04). The penetrance at 10 years old was significantly lower in PITX2 than FOXC1 carriers (13.0% vs 42.9%, P=0.03) but became comparable at 25 years old (71.4% vs 57.7%, P=0.38). These findings have important implications for the genetic counselling of families affected by Axenfeld-Rieger syndrome, and also suggest that FOXC1 and PITX2 contribute to the genetic architecture of primary glaucoma subtypes.

IMPORTANCE The p.Gln368Ter (rs74315329) risk allele in the myocilin gene (MYOC) was initially reported to have high penetrance in glaucoma registry-based studies, but much lower estimates were recently obtained from population-based studies. We investigated this disparity using data from Australia and the United Kingdom. OBJECTIVES To examine the penetrance and effect size of the MYOC p.Gln368Ter variant with glaucoma and ocular hypertension (OHT). DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study within the UK Biobank (UKBB) included participants of white British ancestry. Glaucoma cases were defined by International Classification of Diseases, Ninth Revision (ICD-9) and Tenth Revision (ICD-10) diagnoses and self-reported questionnaires. Carriers of the MYOC p.Gln368Ter variant were identified using genotype imputation from arrays. In contrast, 2 Australian registry-based studies, the Australian and New Zealand Registry of Advanced Glaucoma and the Glaucoma Inheritance Study in Tasmania, ascertained glaucoma cases referred by eye care clinicians, with historic control participants recruited from other Australian studies. Samples were either directly sequenced or had genotypes determined by imputation (for the Australian registry and historic control participants). Recruitment to the UKBB occurred between 2006 and 2010, and data analysis occurred from September 2017 to July 2018. MAIN OUTCOMES AND MEASURES The penetrance and odds ratio (OR) were estimated for the MYOC p.Gln368Ter variants in participants with glaucoma and OHT. RESULTS A total of 411 337 UKBB participants of white British ancestry (mean [SD] age, 56.6 [8.0] years) were included, plus 3071 Australian registry and 6750 historic control participants. In the UKBB, the minor allele frequency of the MYOC p.Gln368Ter variant was 1 in 786 individuals (0.13%). The odds ratio of p.Gln368Ter in patients with primary open-angle glaucoma (POAG) was 6.76 (95% CI, 4.05-11.29); glaucoma (POAG, self-reported glaucoma, and unspecified glaucoma), 4.40 (95% CI, 3.38-5.71); OHT, 3.56 (95% CI, 2.53-4.92); and OHT and glaucoma combined, 4.18 (95% CI, 3.05-5.67). The penetrance of the MYOC p.Gln368Ter variant was 7.6% in patients with glaucoma, 24.3% in patients with OHT, and 30.8% in patients with OHT and glaucoma combined. In the Australian registry studies, the odds of MYOC p.Gln368Ter variant were 12.16 (95% CI, 6.34-24.97) in patients with advanced glaucoma and 3.97 (95% CI, 1.55-9.75) in those with nonadvanced glaucoma; the penetrance of glaucoma was 56.1%, and penetrance in those considered to have glaucoma or be glaucoma suspects was 69.5%. CONCLUSIONS AND RELEVANCE The MYOC p.Gln368Ter variant confers a very high-risk effect size for advanced glaucoma; the risk is lower in nonadvanced glaucoma and OHT. In the general population sample, approximately 50% of MYOC p.Gln368Ter carriers 65 years and older had glaucoma or OHT, with higher prevalence in the Australian registry studies.

No abstract