Polygenic risk scores enhance prediction of atrial fibrillation disease progression: an analysis of the UK Biobank.
Source: PubMed, NCBI / U.S. National Library of Medicine
Progression from paroxysmal to persistent atrial fibrillation (AF) is associated with reduced efficacy of rhythm control therapies and poorer prognosis. While clinical predictors of AF progression have been described, the contribution of genetic predisposition remains unknown. To evaluate the incremental utility of AF polygenic risk scores (PRS) in predicting progression from paroxysmal to persistent AF in the United Kingdom Biobank (UKB). In individuals with paroxysmal AF (based on International Classification of Diseases, Tenth Revision [ICD-10] coding), AF-PRS was categorized into low, intermediate, and high tertiles. The primary endpoint was time to AF progression (defined by ICD-10 diagnosis of chronic or persistent atrial fibrillation, or requirement for electrical cardioversion). Multivariable Cox regression analyses examined the impact of AF-PRS after adjustment for clinical risk factors. Among 7,128 individuals with paroxysmal AF (median age at diagnosis 73 years; 47% women), AF progression occurred in 7.3% over median follow-up of 3.0 years. The yearly incidence of AF progression was 2.2% (95% CI, 1.9-2.6), 2.4% (2.1-2.8) and 3.3% (2.9-3.8) in low, intermediate and high AF-PRS tertiles, respectively (p<0.001). In multivariable analysis, high AF-PRS tertile independently predicted AF progression (HR compared to low AF-PRS tertile: 1.54; 95% CI 1.21-1.96; p<0.001) alongside alcohol intake, hypertension, heart failure, obstructive sleep apnoea, and mitral valve disease
Abstract
Progression from paroxysmal to persistent atrial fibrillation (AF) is associated with reduced efficacy of rhythm control therapies and poorer prognosis. While clinical predictors of AF progression have been described, the contribution of genetic predisposition remains unknown. To evaluate the incremental utility of AF polygenic risk scores (PRS) in predicting progression from paroxysmal to persistent AF in the United Kingdom Biobank (UKB). In individuals with paroxysmal AF (based on International Classification of Diseases, Tenth Revision [ICD-10] coding), AF-PRS was categorized into low, intermediate, and high tertiles. The primary endpoint was time to AF progression (defined by ICD-10 diagnosis of chronic or persistent atrial fibrillation, or requirement for electrical cardioversion). Multivariable Cox regression analyses examined the impact of AF-PRS after adjustment for clinical risk factors. Among 7,128 individuals with paroxysmal AF (median age at diagnosis 73 years; 47% women), AF progression occurred in 7.3% over median follow-up of 3.0 years. The yearly incidence of AF progression was 2.2% (95% CI, 1.9-2.6), 2.4% (2.1-2.8) and 3.3% (2.9-3.8) in low, intermediate and high AF-PRS tertiles, respectively (p<0.001). In multivariable analysis, high AF-PRS tertile independently predicted AF progression (HR compared to low AF-PRS tertile: 1.54; 95% CI 1.21-1.96; p<0.001) alongside alcohol intake, hypertension, heart failure, obstructive sleep apnoea, and mitral valve disease. Addition of AF-PRS to a clinical-only model significantly enhanced risk prediction (net reclassification improvement 15.1%; p=0.016). AF polygenic risk independently predicts progression from paroxysmal to persistent AF and enhances risk stratification beyond clinical factors.