Current ESC guidelines require anticoagulation with VKA in patients with “valvular atrial fibrillation (AF)”, which includes all types of prosthetic valve. We evaluated the value of the CHA₂DS₂ -VASc score for thromboembolism risk assessment in AF patients with biological valve replacement.

Among 8962 patients with AF seen between 2000 and 2010, 8053 (90%) had "non-valvular AF” and 909 (10%) had valvular AF. Patients with valvular AF had a biological prosthesis in 59% (n=549), among which 77% (n=426) had a single aortic prosthesis and 64% (n=309) received a VKA. Patients with aortic bioprosthesis were older and had a higher CHA₂DS₂-VASc score than those with a mitral prosthesis or a double valve replacement. During a follow up of 876±1048 days, 681 stroke/thromboembolic events were recorded. The occurrence of events were similar in patients with bioprosthesis compared to the patients without prosthesis: (hazard ratio HR 1.10, 95% CI 0.83-1.45, p=0.52). Patients with aortic bioprosthesis tended to have a higher risk of embolic events vs other AF patients with bioprosthesis (HR 1.73, 95% CI 0.87-3.45), p=0.12). In multivariate analysis, older age and higher CHA₂DS_2-VASC score were the only predictors of embolic events whilst the presence of a bioprosthesis was not an independent predictor of events. The CHA₂DS₂-VASC score predicted the embolism risk in AF patients with a bioprosthesis (c-statistic 0.55 95% CI 0.47-0.63) but was less efficient than in “non-valvular” AF patient (c-statistic 0.66, 95% CI 0.64-0.67).

This “real world” results support the use of oral anticoagulation in AF patients who have bioprosthetic heart valves. It also supports the use of CHA₂DS₂-VASc scoring for the risk evaluation of AF patients with bioprostheses, albeit with reduced reliability. Patients with aortic bioprosthesis had a non significant higher risk of embolic events. A higher CHA2DS2-VASc score in these patients is likely to explain these results.