Thromboembolic and bleeding risk stratification according to the EHRA valvular heart disease classification in patients with atrial fibrillation

Abstract

In 2017, a new definition of ‘valvular/non-valvular’ atrial fibrillation (AF) has been proposed. We compared thromboembolic (TE) and bleeding risks in patients with AF according to the new ‘Evaluated Heartvalves, Rheumatic or Artificial’ (EHRA) valve classification. Patients were divided into 3 categories: (i) EHRA type 1 corresponds to the previous ‘valvular’ AF patients, including those with either rheumatic mitral valve stenosis or mechanical prosthetic heart valves; (ii) EHRA type 2 includes AF patients with other valvular heart disease (VHD) and valve bioprosthesis or repair; and (iii) ‘non-VHD controls’ i.e. all AF patients with neither VHD nor post-surgical valve disease. Among 8962 AF patients seen between 2000 and 2010, 357 (4%) were EHRA type 1, 1754 (20%) were EHRA type 2 and 6851 (76%) non-VHD controls. Type 2 patients were older and had a higher CHA2DS2-VASc and HAS-BLED scores than either type 1 and non-VHD patients. After a mean follow-up of 1264 ± 1160 days (median 922, interquartile range 234–2083), 715 stroke/TE events and 274 major bleeding (≥ 3 in BARC definition) were recorded. The occurrence of TE events was significantly higher in EHRA type 2 than non-VHD patients [HR (95% CI): 1.30 1.09–1.54], P = 0.003; also, P = 0.31 for type 1 vs. 2, P = 0.68 for type 1 vs. non-VHD controls. The rate of major BARC bleeding events for AF patients was higher in either EHRA type 1 [HR (95% CI): 3.16(2.11–4.72), P < 0.0001] and type 2 [HR (95% CI): 2.19(1.69–2.84), P < 0.0001] than in non-VHD controls. This systematic analysis in real life conditions shows that distinguishing AF patients according to the new EHRA valve classification could be relevant for creating more homogenous groups of patients in terms of TE and bleeding risk. This clearer classification than the previous one should be useful as in clinical research for harmonization of studies, as well as in clinical practice for targeted choices of OAC therapy.