Polygenic risk-based prediction of heart failure in young patients with atrial fibrillation: an analysis from UK Biobank

Heart failure (HF) is the most concerning morbidity in atrial fibrillation (AF) through mutual influence on a poor prognosis. A polygenic risk score (PRS) has recently been proposed to improve the risk prediction for cardiovascular disease. The additive predictive role of PRS for incident HF in patients with AF who inherently carry a high risk of HF is unknown.

From the UK Biobank, we identified 21 167 White Caucasian participants with newly diagnosed AF without a prior history of HF. The PRS for HF was constructed using genetic instruments from previous genome-wide association studies. The primary outcome was the occurrence of incident HF. The prediction of incident HF was evaluated using the tertile categorization of PRS for HF (low vs. moderate-high PRS) across the entire AF cohort, as well as within age subgroups (young AF, age <60 years; old AF, age ≥60 years, respectively). The mean age was 69.0 ± 6.9 years in the total population (55.2 ± 3.9 years in age <60 years; 70.7 ± 5.0 years in age ≥60 years group). During a median follow-up of 3.8 (1.4–7.2) years, the incidence rate (1000-patient year) of HF was 29.9. In the total population, AF patients with moderate-high PRS for HF were associated with a higher risk of HF than those with low PRS for HF [adjusted hazard ratio (HR) 1.18 (95% confidence interval (CI), 1.05–1.32), P = 0.005]. The higher risk of HF in the moderate-high PRS group was particularly accentuated in young AF patients: adjusted HR, 2.14 (95% CI 1.29–3.57) in young AF, and 1.13 (95% CI 1.01–1.27) in old AF, P-for-interaction = 0.015. In young AF, the onset of incident HF was earlier in those with the moderate-high PRS group [median time from AF diagnosis to incident HF, 4.2 (0.8–7.1) years in low PRS vs. 1.5 (0.3–4.7) years in the moderate-high PRS group, P = 0.001]. The prediction of HF was significantly improved by adding PRS to the clinical risk factors for HF, especially in young AF patients, with a net reclassification improvement of 29.7% (P = 0.003).

PRS for HF can significantly improve the prediction of incident HF in patients with AF, especially in the young population, providing clinical utility of an individualized approach to integrated management of AF.