Risk Stratification Tools to Guide a Personalized Approach for Cardiac Monitoring in Embolic Stroke of Undetermined Source

Abstract

Current recommendations support a personalized sequential approach for cardiac rhythm monitoring to detect atrial fibrillation after embolic stroke of undetermined source. Several risk stratification scores have been proposed to predict the likelihood of atrial fibrillation after embolic stroke of undetermined source. This systematic review aimed to provide a comprehensive overview of the field by identifying risk scores proposed for this purpose, assessing their characteristics and the cohorts in which they were developed and validated, and scrutinizing their predictive performance. We identified 11 risk scores, of which 4 were externally validated. The most frequent variables included were echocardiographic markers and demographics. The areas under the curve ranged between 0.70 and 0.94. The 3 scores with the highest area under the curve were the Decryptoring (0.94 [95% CI, 0.88-1.00]), newly diagnosed atrial fibrillation (0.87 [95% CI, 0.79-0.94]), and AF-ESUS (Atrial Fibrillation in Embolic Stroke of Undetermined Source) (0.85 [95% CI, 0.80-0.87]), of which only the latter was externally validated. Risk stratification scores can guide a personalized approach for cardiac rhythm monitoring after embolic stroke of undetermined source.

Keywords: AF‐ESUS; Decryptoring; ESUS; NDAF; atrial fibrillation; ischemic stroke.

Figure 1. Venn diagram of clinical risk score components.

The diagram presents the variable categories included by the clinical risk scores. The radius of each circle corresponds to the number of times the category is included by the risk scores. The union represents the overlap of these categories between studies. Demographics and comorbidities include age, coronary artery disease, peripheral artery disease, heart failure, hypertension, valvular disease, obesity, hypercholesterolemia, chronic obstructive pulmonary disease, obstructive sleep apnea, thyroid disease, and nonstenotic carotid plaque. Stroke characteristics include previous stroke, National Institutes of Health Stroke Scale, recurrent stroke on antiplatelets or multiterritory brain infarct, and vascular cause. Neuroimaging includes posterior or cortical and/or subcortical lesion, and intracranial large vessel occlusion. ECG includes supraventricular extrasystoles, P‐wave duration and/or morphology, and atrial runs. Laboratory markers include N‐terminal pro‐B‐type natriuretic peptide levels and troponin levels. Echocardiography includes LV ejection fraction, LA size, LV end‐diastolic volume, LV hypertrophy, LA strain conduct, LA strain reservoir, and LA volume index. ACTEL indicates Age, Cholesterol, Tricuspid, End diastolic volume, Left atrium; AF, atrial fibrillation; AF‐ESUS, Atrial Fibrillation in Embolic Stroke of Undetermined Source; ESUS, embolic stroke of undetermined source; E2AF, The Empoli ESUS Atrial Fibrillation; HAVOC, Hypertension, Age, Valvular heart disease, Obesity, Congestive heart failure, Coronary artery disease; LA, left atrial; LV, left ventricular; NDAF, newly diagnosed atrial fibrillation; PROACTIA, Prediction of occult atrial fibrillation in patients after cryptogenic stroke and transient ischemic attack; SAFE, the screening for atrial fibrillation scale; and STAF, score for the targeting of atrial fibrillation.

Figure 2. Areas under the curve of the risk scores for the prediction of AF.

The forest plot summarizes the performance of each risk score in predicting the likelihood of AF after ESUS using the calculated area under the curve and the corresponding 95% CI. ACTEL indicates Age, Cholesterol, Tricuspid, End diastolic volume, Left atrium; AF: atrial fibrillation; AF‐ESUS, Atrial Fibrillation in Embolic Stroke of Undetermined Source; ESUS, embolic stroke of undetermined source; E2AF, The Empoli ESUS Atrial Fibrillation; PROACTIA, Prediction of occult atrial fibrillation in patients after cryptogenic stroke and transient ischemic attack; and STAF, score for the targeting of atrial fibrillation.