Does the CHA(2)DS(2)-VASc scale sufficiently predict the risk of left atrial appendage thrombus in patients with diagnosed atrial fibrillation treated with non-vitamin K oral anticoagulants?

The CHA(2)DS(2)-VASc scale does not include potential risk factors for left atrial appendage thrombus (LAAT) formation such as a form of atrial fibrillation (AF) and impaired kidney function. The real risk of thromboembolic complications in AF patients is still unclear as well as an optimal anticoagulant treatment in males with a CHA(2)DS(2)-VASc score of 1 and females with a CHA(2)DS(2)-VASc score of 2. The aim of this study was to compare the predictive value of the CHA(2)DS(2)-VASc scale and other scales to estimate the risk of LAAT formation in AF patients treated with non-vitamin K oral anticoagulants (NOACs) and to assess the prevalence of thrombi in patients at intermediate risk of stroke. The observational study included consecutive patients with a diagnosis of non-valvular AF treated with NOACs, admitted to 3 high-reference institutions between 2013 and 2018. All individuals underwent transoesophageal echocardiography before cardioversion or ablation. Out of 1163 enrolled AF patients (62.1% male, mean age 62 years) the LAAT had been detected in 50 individuals (4.3%). Among patients with LAAT, 1 patient (2.0%) was classified as a low-risk category, 9 (18.0%) were at intermediate-risk, and 40 (80.0%) were at high risk of thromboembolic complications according to CHA(2)DS(2)-VASc scale. All patients were treated with NOACs: 51.0% rivaroxaban, 47.1% dabigatran, and 1.9% apixaban. Patients at intermediate stroke-risk with detected LAAT had higher R(2)CHADS(2) score (2.1 ± 1.2 vs 1.2 ± 0.8, P = .007), higher CHA(2)DS(2)-VASc-RAF score (6.4 ± 4.4 vs 3.7 ± 2.6, P = .027) and more often had an estimated glomerular filtration rate below 56 mL/min/1.73 m(2) (44.4% vs 13.2%, P = .026) compared to patients without LAAT. The receiver operating characteristics revealed that the CHA(2)DS(2)-VASc-RAF scale had better predictive ability to distinguish between patients with and without LAAT in the study group than CHA(2)DS(2)-VASc (P = .0006), CHADS(2) (P = .0006) and R(2)CHADS(2) scale (P = .0140). The CHA(2)DS(2)-VASc scale should be supplemented with an assessment of renal function and form of AF to improve stroke risk estimation. The application of additional scales to estimate the risk of LAAT might be especially useful among males with a CHA(2)DS(2)-VASc score of 1 and females with a CHA(2)DS(2)-VASc score of 2.