Author: Claudia Farinha (Portugal)
Co-authors: Patrícia Barreto, Rita Coimbra, Maria Helena Madeira, Maria Luz Cachulo, Carel B. Hoyng, Rufino Silva
The heritable component in AMD is estimated to be as high as 70% and to date, 52 variants at 34 genomic regions are known to be associated with increased risk of disease development and progression. Strategies such as calculating the genetic risk score (GRS) can also be useful if integrating the genetic information with environmental and demographic factors when assessing individual risk. The Coimbra Eye Study (CES) is an epidemiologic study for the estimation of prevalence and 6.5-year incidence of AMD in a Portuguese population (NCT01298674, NCT02748824). This report aims to explore the effect of common and rare genetic risk variants in the development of AMD in the CES population and to calculate the GRS.
Association for Innovation and Biomedical Research on Light and Image (AIBILI), Coimbra, Portugal
Participants in the CES underwent standardized interviews, ophthalmologic examination, and imaging in both visits. AMD staging was performed with Rotterdam classification in a centralized reading center. Two analyses for common variants were performed, a case-control analysis and a progression to AMD analysis. In the first, cases were AMD patients (stage 2,3 or4) and non-AMD cases were participants staged 0 (>60years-old) or 1 (>70years-old) at the follow-up visit. In the progression analysis, non-progressors were participants that remained in stage 0 or 1 through follow-up. Genomic DNA was isolated from blood samples. The genotyping assay was based on single-molecule molecular inversion probes for target selection and used next-generation sequencing to sequence 87 single nucleotide polymorphisms (SNPs),(EyeRisk Project,E3 Consortium). Successfully genotyped SNPs were tested for association under an additive model with presence/absence of AMD in the follow-up visit as a binary outcome, and progression/no progression in the longitudinal analysis. Logistic regression analysis was performed to assess allelic odds ratio at 95%CI for each variant, adjusted for age and sex, significance level at 0.05. The GRS for common AMD risk variants was calculated. For the rare variant analysis, we performed logistic regression analyses to assess the cumulative effect of rare variants with AMD.
Samples from 237 AMD cases and 638 non-AMD controls were successfully genotyped for a total of 69 SNPs in the cohort of the follow-up Incidence CES. The risk-variants associated to increased risk of AMD presence were: ARMS2 rs10490924, ARMS2_HTRA1 rs3750846, CFH rs35292876, SLC16A8 rs8135665 and TGFBR1 rs1626340. As for a protective effect we identified the variants: C2_CFB_SKIV2L rs429608, CFH rs10922109 and rs1410996, CNN2 rs10422209, CETP rs5817082, CB rs641153 and RDBP_CFBrs760070. When considering progression from no-AMD to AMD through the 6.5-year follow-up time of the complete epidemiological study, we obtained 630 samples from non-progressors and 137 from progressors. Variants associated to risk of progression were: ARMS2 rs10490924, ARMS2_HTRA1 rs3750846 and CFH rs35292876; and variants protective of progression were again C2_CFB_SKIV2L rs429608, CFH rs10922109 and rs1410996, CNN2 rs10422209, but also CFHR5 rs10922153, SYN3/TIMP3 rs5754227 and COL10A1 rs3812111. The GRS was significantly different between AMD vs non-AMD cases(0.535±1.47 vs 0.034±1.36,p<0.001), and between progressors vs non-progressors(0.647±1.48 vs 0.042±1.38,p<0.001). The rare variants analysis included 218 cases and 597 controls. For the CFH gene, a total of 78 rare variants were included with a known Polyphen score. Damaging rare variants in the CFH were significantly more present in AMD patients compared to controls(OR, 10.29; P= 4.762e-05).
Several variants were found to be associated with the presence and progression to AMD in our epidemiological longitudinal study, while others were protective. The genes associated with AMD act in different pathophysiologic pathways, sustaining the multifactorial etiology of AMD. Their effects on our population agree with major reports, including large GWAS studies. The GRS was significantly different between AMD and non-AMD cases and between progressors, confirming its utility in more complex multifactorial analysis when assessing individual risk. Furthermore, we also found that rare variants in the CFH gene with damaging effects were significantly associated with AMD in our cohort. This is the first genetic study in AMD in a Portuguese population, and in the future, we will also explore the correlation between genetics, phenotypic and environmental features in risk assessment. Genetic characterization is important to pursue in different populations, as the identification of potential genetic therapeutic targets is of major interest.