Association between genetic predisposition of AF and the risk of ischemic stroke and thromboembolism

Abstract Background Ischemic stroke and thromboembolism are the major complications of atrial fibrillation (AF), however, its association with genetic background of AF remains unclear. To investigate the association between genetic predisposition of AF underlying common variants and the risk of ischemic stroke and thromboembolism. Methods We included unrelated individuals with non-valvular AF from UK biobank. The polygenic risk scores of AF (PRSAF) was calculated for each individual using the effect weight from a previous study which has been validated in both European and mixed-ancestry population[1, 2]. The SNPs from the original score were lifted over to GRCh38 and 6514708 of 6730541 were available in the UK biobank genotype data. The clinical data was defined by ICD-10 code. The primary endpoint is ischemic stroke or thromboembolism resulting in hospitalization after first AF diagnosis. Results A total of 28810 unrelated patients were included in the study. Among them, 4064 had PRSAF over 90th percentile of the general population, and were considered having high genetic risk. The rest 18746 with per SD increase in PRSAF lower than 90th percentile of the general population were considered having lower genetic risk. Generally, age and sex adjusted model demonstrated that per standard deviation change in PRSAF is associated with 9% increase in the risk of stroke or thromboembolism (HR=1.09, 95%CI: 1.03-1.14), and patients with high genetic risk was associated with 15% increase in the risk of stroke or thromboembolism (HR=1.15, 95%CI: 1.02-1.31).The result remained stable when further adjusted for clinical risk factors (1.09, 95%CI: 1.04-1.15 per SD increase in PRSAF) and competing risk of death (1.10, 95%CI: 1.05-1.16 per SD increase in PRSAF). There was no interaction between CHA2DS2-VASc score risk stratification and the per SD increase in PRSAF (P interaction = 0.751, Figure 1A). However, the association between per SD increase in PRSAF and the event tended to be affected by age at AF onset (HR = 1.02 95%CI: 0.94, 1.11 for early onset AF; HR = 1.12 95%CI: 1.05, 1.19 for non-early onset AF, P for interaction = 0.062, Figure 1B). Multi-variate COX regression including PRSAF, components of CHA2DS2-VASc score, and other AF risk factors revealed that history of stroke/thromboembolism, age over 65 years, hypertension, diabetes and high AF genetic risk were significantly associated with the outcome (Figure 1C). To evaluate the importance of these risk factors, average population attributable fraction for 10-year risk of stroke/thromboembolism in the study population was calculated, indicating that genetic risk of AF is responsible for 3.26% (0.89%-5.63%) of the events (Figure 1D). Conclusion Genetic predisposition of AF underlying common variants is associated with ischemic stroke and thromboembolism independent of clinical risk factors. But its role in different AF populations, like early onset AF, remains further investigation.Figure 1