Introduction: Diffusion-Weighted magnetic resonance imaging (DWI) studies have demonstrated sub-acute ischemic lesions in up to 41% of patients with intracerebral hemorrhage (ICH). Retrospective studies indicated an association with acute blood pressure (BP) reduction. We conducted a prospective study of DWI lesion incidence in ICH patients.
Methods: Patients with computed tomography (CT) scan confirmed ICH were prospectively recruited within 6 hours of symptom onset. Acute systolic BP (SBP) targets were randomly assigned as part of an ongoing trial. The SBP target remained blinded for this analysis. Patients were assessed with serial DWI at 48 hours.
Results: Eighty-three patients (48.2% male, mean±SD age 71.9±13.4) were enrolled. MRI was obtained in 42 patients. (11 patients were medically unstable, 11 refused consent, 19 died or were treated palliatively prior to the first MRI). Median (IQR) acute hematoma volume of the patients undergoing magnetic resonance imaging was 14.11 (4.40, 38.06) ml. Median baseline NIHSS score was 12.50 (9, 20). Median (IQR) 48 hour hematoma volume was 23.30 (9.10, 65.30) ml. DWI lesions were present in 13 (31%) of patients at 48 hours (median time to scan 54.22 (39.83, 65.5) hours). Mean SBP at 48 hours in patients with lesions (154.55±27.02 mmHg) was similar to those without (149.04±27.02; p= 0.514). Systolic BP at the time of the MRI was not a predictor of ischemic lesion development at 48 hours (OR=1.015, [0.972, 1.059], p=0.507), nor was the change in BP between baseline and the scan (OR=0.994, [0.962, 1.027], p=0.716). The only predictor of 48 hour DWI incidence was hematoma volume at 48 hours (OR= 1.025, 95% CI [1.001, 1.050], p=0.044).
Discussion: When enrolled prospectively within 6 hours of onset, DWI was obtained in approximately half of all ICH patients. This selection bias may be relevant to the results of both retrospective and prospective DWI studies. This study demonstrated no evidence of a relationship between actual BP achieved and DWI incidence after ICH.