Abstract 3398: Excessive Supraventricular Ectopic Activity On 24h-Holter Predicts Paroxysmal Atrial Fibrillation In Patients With Cerebral Ischemia Presenting In Sinus Rhythm

Abstract

Background Detection of paroxysmal atrial fibrillation (PAF) in patients with cerebral ischemia is challenging. Echocardiography (LAVI/á, an index of left atrial remodeling) can select a subgroup of patients for extended ECG-monitoring. Excessive supraventricular ectopic activity (ESVEA), measured as atrial premature beats (APB) and longest supraventricular run on 24h-Holter-ECG (SV-run 24h ) may further improve prediction of PAF. Methods Retrospective analysis from a prospective monocentric observational trial (Find-AF) in patients >18y with cerebral ischemia (ISRCTN 46104198). PAF was diagnosed by 7d-Holter monitoring. Transthoracic echocardiography parameters were prospectively assessed. ESVEA was quantified in one 24h interval, which had to be free from PAF. For the present analysis, 44 patients with atrial fibrillation at baseline and 20 patients with limited evaluability for ESVEA (9 with pacemaker rhythms, 4 with low-quality recordings and 7 with PAF who had no 24h interval free from PAF) were excluded. Results PAF was detected in 21 out of 215 patients (9.8%) of the analysis population. PAF was more prevalent in those with ESVEA above the median: 17.3 vs. 2.7% for APB > vs. <= 4/h (p<0.001 chi-square test); 15.9 vs. 3.7% for SV-run 24h > vs. <= 5 beats (p=0.003). ESVEA differed between PAF and Non-PAF: APB median 17/h [IQR 9; 143] vs. 4/h [1; 15] and SV-run 24h 10 [7; 18] vs. 0 [0; 8] beats (both p<0.001, Mann-Whitney-U test). Both parameters discriminated moderately between PAF and Non-PAF (area under the receiver-operator-characteristics curve 0.751 and 0.730, respectively). In multivariate analysis with age, sex, body-mass index, systolic blood pressure, hypertension, heart failure and LAVI/á as covariates, log(APB/h) (p=0.027) and log(SV-run 24h ) (p=0.022) were independently predictive for the presence of PAF. 0% of patients with normal LAVI/á and APB <=4/h, 5.6% of those with either abnormal LAVI/á or APB >4/h, and 27.1% of those with both abnormal LAVI/á and APB >4/h had PAF (p<0.001 chi-square). Results were similar when PAF-cases were included that had been diagnosed clinically until 1-year follow-up. Conclusions In patients with cerebral ischemia, ESVEA discriminates PAF from Non-PAF beyond clinical factors and LAVI/á. PAF is ruled out in patients with normal LAVI/á and APB <=4/h, while prevalence is high in those with abnormal LAVI/á and APB >4/h.