Background: Hemorrhagic infarction (HI) is among the most severe complications that can occur following the administration of intravenous recombinant tissue plasminogen activator (rt-PA). In the present study, we aimed to determine the optimal cut-off points of blood pressure (BP) for HI after rt-PA treatment, and to compare our findings with those for other prediction models.
Methods: We analyzed data from 109 consecutive patients with stroke treated at our hospital between 2009 and 2016. HI was confirmed via computed tomography or magnetic resonance imaging. Patients were classified into a symptomatic HI group, an asymptomatic HI group, and a non-HI group. BP was measured on admission and before rt-PA treatment. Glucose Race Age Sex Pressure Stroke Severity (GRASPS) and Totaled Health Risks in Vascular Events (THRIVE) scores were also calculated. Receiver operating characteristic (ROC) analysis was used to determine factors associated with symptomatic and asymptomatic HI.
Results: Among the 109 total patients, 25 patients developed symptomatic HI, while 22 patients developed asymptomatic HI. ROC analysis for predicting symptomatic and asymptomatic HI revealed that the area under the curve for pretreatment systolic BP (SBP) was .88 (95% confidence interval[CI]: .83-.94), while those for GRASPS and THRIVE scores were .75 (95% CI: .66-.85) and .69 (95% CI: .59-.79), respectively. We identified an optimal cut-off point of 160 mm Hg (sensitivity: 82.3%; specificity: 76.6%; diagnostic accuracy: 80.0%; positive predictive value: 76.6%; negative predictive value: 82.5%).
Conclusions: Pre-treatment SBP may be a simple predictor of symptomatic and asymptomatic HI in patients with stroke undergoing rt-PA treatment.
Keywords: Cerebral infarction; acute stroke management; hemorrhagic infarction; magnetic resonance imaging; tissue plasminogen activator.
Copyright © 2019 Elsevier Inc. All rights reserved.