AUTHOR=Bianconi Elisa , Del Freo Giulia , Salvadori Emilia , Barbato Carmen , Formelli Benedetta , Pescini Francesca , Pracucci Giovanni , Sarti Cristina , Cesari Francesca , Chiti Stefano , Diciotti Stefano , Gori Anna Maria , Marzi Chiara , Fainardi Enrico , Giusti Betti , Marcucci Rossella , Bertaccini Bruno , Poggesi Anna TITLE=Can CHA2DS2-VASc and HAS–BLED Foresee the Presence of Cerebral Microbleeds, Lacunar and Non-Lacunar Infarcts in Elderly Patients With Atrial Fibrillation? Data From Strat–AF Study JOURNAL=Frontiers in Neurology VOLUME=13 YEAR=2022 URL=https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2022.883786 DOI=10.3389/fneur.2022.883786 ISSN=1664-2295 ABSTRACT=

Anticoagulants reduce embolic risk in atrial fibrillation (AF), despite increasing hemorrhagic risk. In this context, validity of congestive heart failure, hypertension, age ≥ 75 years, diabetes, stroke, vascular disease, age 65–74 years and sex category (CHA2DS2-VASc) and hypertension, abnormal renal/liver function, stroke, bleeding history or predisposition, labile international normalized ratio, elderly, drugs/alcohol concomitantly (HAS–BLED) scales, used to respectively evaluate thrombotic and hemorrhagic risks, is incomplete. In patients with AF, brain MRI has led to the increased detection of “asymptomatic” brain changes, particularly those related to small vessel disease, which also represent the pathologic substrate of intracranial hemorrhage, and silent brain infarcts, which are considered risk factors for ischemic stroke. Routine brain MRI in asymptomatic patients with AF is not yet recommended. Our aim was to test predictive ability of risk stratification scales on the presence of cerebral microbleeds, lacunar, and non-lacunar infarcts in 170 elderly patients with AF on oral anticoagulants. Ad hoc developed R algorithms were used to evaluate CHA2DS2-VASc and HAS–BLED sensitivity and specificity on the prediction of cerebrovascular lesions: (1) Maintaining original items' weights; (2) augmenting weights' range; (3) adding cognitive, motor, and depressive scores. Accuracy was poor for each outcome considering both scales either in phase 1 or phase 2. Accuracy was never improved by the addition of cognitive scores. The addition of motor and depressive scores to CHA2DS2-VASc improved accuracy for non-lacunar infarcts (sensitivity = 0.70, specificity = 0.85), and sensitivity for lacunar–infarcts (sensitivity = 0.74, specificity = 0.61). Our results are a very first step toward the attempt to identify those elderly patients with AF who would benefit most from brain MRI in risk stratification.