Cardiology Fellow The University of Calgary, Canada
Background: Stroke is an important cause of morbidity and mortality, with a significant proportion due to cardioembolic events from atrial fibrillation (AF). Detection of AF post-stroke is critical to guide management. The PER DIEM trial randomized patients without known AF to 12-months of implantable loop recorder (ILR) monitoring versus 30-days of external loop recorder monitoring, with higher AF detection using ILR. Echocardiography (echo) is commonly performed post-stroke and may also have a role in identifying patients that benefit from prolonged ECG monitoring. This echo substudy of the PER DIEM trial investigated the association of multiple echo parameters with the detection of AF lasting 2 minutes or longer at 12 months (the primary endpoint of the PER DIEM trial) in patient within 6 months of ischemic stroke.
METHODS AND RESULTS: Patients enrolled in the PER DIEM trial with a complete transthoracic echo performed within 30-days of randomization were included. Echo analysis was performed blinded to treatment group or outcomes. Demographic, clinical, and echo parameters (conventional two-dimensional, Doppler and speckle-tracking strain) were compared between those with and without detected AF. Of 153 patients included in this echo substudy, 19 (12.4%) had AF detected following ischemic stroke. Baseline demographic clinical and echo data are reported in Table 1. Patients with post-stroke AF compared to those without were significantly older (median 74-years versus 64-years; p=0.002), with a higher proportion of underlying heart failure (16% versus 1%; p=0.014), and a higher CHA2DS2-VASc score (median 5 versus 4; p=0.013). With respect to echo parameters, patients with post-stroke AF had a significantly higher left atrial (LA) volume index (34.7±11.6 vs. 27.7±10.6; p=0.019), lower absolute left ventricular (LV) global longitudinal strain (-16.4 ± 5.1 vs. -19.0±3.2; p=0.007) and a lower LA strain (12.4±11.6 vs. 16.4±8.9; p=0.03) compared to those without AF. Echo parameters that did not demonstrate significant difference included LV end-diastolic and end-systolic dimensions, LV ejection fraction, diastolic function indices, and tricuspid annular plane systolic excursion. Univariable regression analysis showed that only a higher LA volume index (OR 1.30 per 5 mL/m2 increase, 95%CI 1.03-1.64, p=0.025) and lower absolute LV global longitudinal strain (OR 2.73 per 5% decrease, 95%CI 1.37-6.04, p=0.007) were significant predictors of post-stroke AF.
Conclusion: This echo substudy of the PER DIEM trial demonstrated that worse LA volume index and LV GLS were significantly associated with AF detection after ischemic stroke. These findings may contribute to the development of an echo-based risk score for AF screening post-ischemic stroke.
