Can perfusion CT unmask postictal stroke mimics? A case-control study of 133 patients

Objective To study the diagnostic value of volume perfusion CT (VPCT) in patients with transient focal neurologic deficits following and during epileptic seizures, that mimic symptoms of stroke. Methods A retrospective case-control study was performed on 159 patients who presented with a seizure and received an emergency VPCT within the first 3.5 hours of admission, after being misjudged to have an acute stroke. The reference test was a clinical-based, EEG-supported diagnostic algorithm for seizure. Results We included 133 patients: 94 stroke-mimicking cases with postictal focal neurologic deficits (Todd phenomenon, n = 67) or ongoing seizure on hospital admission Octal patients, n = 27), and 39 postictal controls without focal neurologic deficits. Patients with Todd phenomenon showed normal perfusion (64%), hypoperfusion (21%), and hyperperfusion (14%) on early VPCT. Ictal patients displayed more hyperperfusion compared to postictal patients (p = 0.015). Test sensitivity of hyperperfusion for ictal patients is 38% (95% confidence interval [CI] 20.7%-57.7%), specificity 86% (95% CI 77.3%-91.7%), positive predictive value is 42% (95% CI 27.5%-58.7%), and the negative predictive value 83% (95% CI 78.6%-86.9%). A cortical distribution was seen in all hyperperfusion scans, compared to a cortico-subcortical pattern in hypoperfusion (p < 0.001). A history of complex focal seizure and age were associated with hyperperfusion (p = 0.046 and 0.038, respectively). Conclusion VPCT can differentiate ictal stroke mimics with hyperperfusion from acute ischemic stroke, but not postictal patients who display perfusion patterns overlapping with ischemic stroke. Classification of evidence This study provides Class IV evidence that VPCT accurately differentiates ictal stroke mimics from acute ischemic stroke.