Affiliations 

  • 1 Department of Neurology, Haeundae Paik Hospital, Inje University College of Medicine, Busan, Republic of Korea
  • 2 Department of Neurology, Haeundae Paik Hospital, Inje University College of Medicine, Busan, Republic of Korea, epidoc@inje.ac.kr
Eur Neurol, 2020;83(1):56-64.
PMID: 32320976 DOI: 10.1159/000506591

Abstract

INTRODUCTION: Seizures as acute stroke mimics are a diagnostic challenge.

OBJECTIVE: The aim of the study was to characterize the perfusion patterns on perfusion computed tomography (PCT) in patients with seizures masquerading as acute stroke.

METHODS: We conducted a study on patients with acute seizures as stroke mimics. The inclusion criteria for this study were patients (1) initially presenting with stroke-like symptoms but finally diagnosed to have seizures and (2) with PCT performed within 72 h of seizures. The PCT of seizure patients (n = 27) was compared with that of revascularized stroke patients (n = 20) as the control group.

RESULTS: Among the 27 patients with seizures as stroke mimics, 70.4% (n = 19) showed characteristic PCT findings compared with the revascularized stroke patients, which were as follows: (1) multi-territorial cortical hyperperfusion {(73.7% [14/19] vs. 0% [0/20], p = 0.002), sensitivity of 73.7%, negative predictive value (NPV) of 80%}, (2) involvement of the ipsilateral thalamus {(57.9% [11/19] vs. 0% [0/20], p = 0.007), sensitivity of 57.9%, NPV of 71.4%}, and (3) reduced perfusion time {(84.2% [16/19] vs. 0% [0/20], p = 0.001), sensitivity of 84.2%, NPV of 87%}. These 3 findings had 100% specificity and positive predictive value in predicting patients with acute seizures in comparison with reperfused stroke patients. Older age was strongly associated with abnormal perfusion changes (p = 0.038), with a mean age of 66.8 ± 14.5 years versus 49.2 ± 27.4 years (in seizure patients with normal perfusion scan).

CONCLUSIONS: PCT is a reliable tool to differentiate acute seizures from acute stroke in the emergency setting.

* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.