Page 486

Karger_ESC London_2013

22. European Stroke Conference 378 Heart and brain Brain MRI predictors of occult atrial fibrillation detected with implantable Holter in patients with cryptogenic stroke. E. Rojo Martínez1, M.G. Sandín Fuentes2, A.I. Calleja Sanz3, E. Cortijo García4, P. García Berme-jo5, E. García Morán6, J. Rubio Sanz7, L. Casadiego Matarranz8, M. Rodríguez Velasco9, J.F. Are-nillas Lara10 Hospital Clínico Universitario, Valladolid, SPAIN1, Hospital Clínico Universitario, Valladolid, SPAIN2, Hospital Clínico Universitario, Valladolid, SPAIN3, Hospital Clínico Universitario, Vall-adolid, SPAIN4, Hospital Clínico Universitario, Valladolid, SPAIN5, Hospital Clínico Universitario, Valladolid, SPAIN6, Hospital Clínico Universitario, Valladolid, SPAIN7, Hospital Clínico Universi-tario, Valladolid, SPAIN8, Hospital Clínico Universitario, Valladolid, SPAIN9,Hospital Clínico Uni-versitario, Valladolid, SPAIN10 Introduction: Implantable Holter allows detecting paroxysmal atrial fibrillation (PAF) in patients with cryptogenic stroke (CS). Identification of baseline predictors of an increased risk for occult PAF in CS patients is greatly needed. Objective: To evaluate the association between brain magnetic resonance imaging (MRI) findings at baseline and the detection of occult PAF in patients with CS studied with implant-able 486 © 2013 S. Karger AG, Basel Scientific Programme Holter. Method: We selected consecutive CS patients with a high suspicion of embolic cerebral ischemia. Imaging protocol during admission in the stroke unit included brain MRI with diffusion (DWI) and FLAIR sequences. Several patterns of DWI acute lesion were defined, and finally categorized in the pres-ence of single vs multiple lesions. The presence of chronic territorial infarcts was evaluated on FLAIR sequences. Subcutaneous Holter Reveal-XT® was implanted after full diagnostic workup and patients were trained to perform transmissions monthly or when symptoms occurred. Long-term online and onsite monitoring for occult PAF was conducted thereafter. Results We implanted Reveal-XT® to 86 patients with CS: mean age 67 years, 45 (52.3%) women. DWI sequence showed multiple acute lesions in 48 (55.8%) patients, and 11 patients (12.8%) showed chronic territorial infarcts in FLAIR. Mean follow-up after implantation was 314 ± 206 days. PAF was detected in 26 (30%) patients. The probability of PAF was not associated with DWI pattern, whereas the presence of chronic infarcts on MRI was significantly associated with a lower event free of PAF survival in a Kaplan Meier curve (log-rank p=0.02) and a multivariate-ajusted Cox regres-sion model (HR 2.8, CI 95% 1.8-13.9, p=0.02). Conclusions: The presence of chronic territorial infarcts on FLAIR sequence emerged as a predictor of occult PAF in patients with CS. These findings may have practical implications for the choice of antithrombotic treatment in CS patients. 379 Heart and brain Brain Natriuretic Peptide (BNP): a relevant biomarker in atrial fibrillation (AF) diagnostic strategy after stroke? L. SUISSA1, S. LACHAUD2, M.H. MAHAGNE3 Centre Hospitalier Universitaire de Nice - Stroke unit, NICE, FRANCE1, Centre Hospitalier Universitaire de Nice - Stroke unit, NICE, FRANCE2, Centre Hospitalier Universitaire de Nice - Stroke unit, NICE, FRANCE3 Background: There is now a body of evidence to believe that prolonged continuous ECG monitoring (CEM) after stroke can increase occult AF detection rate. However, cost incurred limits CEM gener-alization in stroke units. Recent studies suggest that BNP level might rule out delayed AF in stroke patient. We hypothesized the usefulness of admission BNP level in AF algorithm diagnosis. Methods: We studied a prospective cohort including 711 consecutive ischemic stroke patients admit-ted in stroke unit between November 2010 and November 2012. Plasma BNP level was measured and CEM was started systematically at admission during hospitalization for 587 AF-naive patients. Using McNemar test, we compared in this cohort the AF detection rate obtained 1) with CEM per-formed systematically and 2) with a strategy based on admission BNP level (Figure 1). AF detection rate (and number of CEM days performed) was calculated according to different initial BNP values. Results: 587 AF-naïve patients (median age: 6553-75 years, sex ratio: 1.6) received CEM with a median duration of 5.83.0-10.7 days. Systematic CEM detected 88 (15.0%) new AF including 42 (7.2%) in the first 24 hours. There were no significant differences for AF detection rate with sys-tematic versus non-systematic CEM strategy with an initial BNP cut-off between 0 and 132 pg/ml (p=0.12). In other words, in this cohort, for BNP values below 132 pg/ml, simulated strategy based on BNP level, when compared with systematic CEM strategy, has saved 2299/3964 (58.0%) CEM days (Figure 2). Conclusions: Based on initial BNP level, non-systematic CEM strategy saves more than half CEM days with the same diagnostic profitability. We suggest that BNP testing could take place in the de-cision algorithm and requires a medical and economic assessment. Further studies should be carried out to validate plasma BNP threshold value and evaluate the cost/effectiveness of this strategy.


Karger_ESC London_2013
To see the actual publication please follow the link above