Objective  

To determine whether serum N-terminal pro-B-type natriuretic peptide (N-BNP), a biomarker of myocardial wall stress, is specific to acute heart failure (HF) in patients hospitalized with stroke.

Design  

Case-control study.

Setting  

Tertiary hospital, Neurosciences Critical Care Unit and Stroke Unit.

Patients  

Consecutive patients with acute ischemic or hemorrhagic stroke who were evaluated for HF.

Intervention  

None.

Measurements and results  

Cases and controls were classified, respectively, as patients with or without HF, defined according to modified Framingham criteria. Seventy-two patients were evaluated, 39 with ischemic stroke, 22 with aneurysmal subarachnoid hemorrhage (SAH), and 11 with intracerebral hemorrhage (ICH). Thirty-four patients (47%) met criteria for HF, and 47 patients (65%) had systolic or diastolic left ventricular (LV) dysfunction on echocardiogram. Serum N-BNP was measured a median of 48 h following the onset of stroke and was increased (> 900 pg/ml) in 56 patients (78%), with higher levels in non-survivors (11898 ± 12741 vs 4073 ± 5691; p = 0.001). In a multiple regression model, N-BNP elevation was not independently associated with HF (OR 5.4, 95% CI 0.8–36.0, p = 0.084). At a cut-off of 900 pg/ml, the sensitivity of N-BNP for HF was 94%, specificity 37%, positive predictive value (PPV) 57%, and negative predictive value (NPV) 88%. For systolic or diastolic LV dysfunction, the sensitivity of N-BNP was 89%, specificity 44%, PPV 75%, and NPV 69%.

Conclusions  

These results demonstrate that N-BNP elevation is not specific to HF or LV dysfunction in patients with acute ischemic stroke, SAH, and ICH.