Developing an Asymmetry Method for Detecting Postictal Hyperperfusion in Poststroke Epilepsy

Abstract

Using dual single-photon emission computed tomography (SPECT) scanning, we recently found the postictal-interictal (P-I) subtraction method frequently detects prolonged postictal hyperperfusion in poststroke epilepsy (PSE) and thus may be valuable for auxiliary diagnosis. This study aimed to determine if the asymmetry method can localize hyperperfusion to reflect epileptic activity in PSE using a single postictal SPECT scan. Sixty-four patients with PSE who had undergone perfusion SPECT two times (postictal and interictal) were enrolled. We formulated a novel asymmetry method (subtraction analysis of reversed postictal SPECT from postictal SPECT, co-registered to magnetic resonance imaging) to identify paradoxical asymmetric increase, defined as a higher perfusion area adjacent to stroke lesions compared to the contralateral side. The postictal hyperperfusion area and detection rates were determined by the asymmetry and P-I subtraction methods independently. We subsequently calculated the sensitivity and specificity of the asymmetry method, compared to the gold standard P-I subtraction method. We also evaluated lateralization concordance between the asymmetry method and other clinical findings. Among 64 patients (median age, 75 years), prolonged postictal hyperperfusion was detected in 43 (67%) by the asymmetry, and 54 (84%) the P-I, method. The asymmetry method had high sensitivity (80%) and specificity (100%) in detecting postictal hyperperfusion, showing high lateralization concordance with seizure semiology (97%) and epileptiform electroencephalography findings (interictal/ictal epileptiform discharges or periodic discharges) (100%). The present study demonstrated the advantages of the objective asymmetry method for detecting prolonged hyperperfusion through using one postictal SPECT scan in PSE.