Abstract 26: Predictors of Atrial Cardiopathy Among Patients in the Arcadia Trial: An Analysis of the First 924 Patients
Background: Atrial cardiopathy (AC) in absence of atrial fibrillation is a suspected cause of embolic stroke of undetermined source (ESUS). Predictors of AC remain incompletely characterized. We hypothesized that demographics, vascular risk factors, and time from stroke presentation to assessment for AC predict AC among ESUS patients. Methods: ARCADIA is an ongoing, investigator-initiated, NINDS-funded, multicenter, randomized trial of standard dose apixaban versus aspirin 81 mg daily among consented patients with ESUS and biomarker evidence of AC, defined as any of: N-terminal pro-brain natriuretic peptide (NT-proBNP) > 250 pg/ml; P wave terminal force velocity on ECG lead V1>5,000 μV*ms; or echocardiographic left atrial diameter index ≥3 cm/m 2 . Patients may be enrolled up to 120 days after stroke. We used multivariable regression modelling to estimate relative risks and 95% confidence intervals (RR, 95%CI) for association of predictors with AC. Results: Among 924 ESUS patients who met inclusion/exclusion criteria as of July 16, 2019, 251 met > 1 AC criterion (164 NT-proBNP, 114 PTFV1, 4 echo). Compared to those without AC (n=673), those with AC were older (69.0 + 14.5 versus 64.0 + 15.0 years) and more often women (52.2% versus 40.9%). Multivariable predictors of AC were age (RR per decade 1.20, 95%CI 1.09-1.32), race (black versus other RR 1.20, 95%CI 0.95-1.52), sex (RR for women 1.22, 95%CI 0.98-1.52), hemoglobin (per unit RR 0.94, 95%CI 0.88-1.01), and cardiovascular/peripheral arterial disease, RR 1.48, 95%CI 1.16-1.88). Prior history of stroke or TIA, hypertension, diabetes, smoking, cancer, sleep apnea, and time from stroke to consent and blood draw for NT-proBNP were not associated with AC. Conclusions: Older age, female sex, black race, relative anemia, and history of vascular disease associate with AC among ESUS patients. Absence of association with time to assessment, particularly testing for NT-proBNP, provides indirect evidence that AC is not a reaction to stroke, but a stable condition, and supports patients being assessed for ARCADIA participation up to 4 months post stroke. The moderate strength of demographic associations with AC further underscores the importance of enrolling unselected ESUS patients into ARCADIA.