Editors' Note: Accuracy of CT Perfusion-Based Core Estimation of Follow-up Infarction: Effects of Time Since Last Known Well.

Abstract

In a prospective multicenter study of 101 patients, Dr. Sarraj and colleagues assessed the accuracy of baseline CT perfusion (CTP) estimates of ischemic core, relative to follow-up diffusion-weighted imaging (DWI) lesion infarct volumes, in patients with ischemic stroke who achieved complete reperfusion with endovascular thrombectomy. The authors found that significant overestimation of the volume of irreversibly injured ischemic core—by 10 mL or more—was rare and was only observed in patients who presented within 90 minutes of last known well and achieved reperfusion within 120 minutes of CTP acquisition. Such overestimation also primarily occurred in the white matter. The overestimation of ischemia was eliminated by using a more conservative threshold for relative cerebral flood flow (rCBF) of <20% as opposed to the more typically used threshold of <30% in patients presenting within 90 minutes. In response, Drs. Garcia-Camba and Jacobi note that the prevalence and temporal distribution of patients with a large ischemic core (≥50 mL) was not reported; this could affect the overestimation rate in the study and potentially limit the generalizability of its findings. Replying to this, the authors note that their cohort was derived from consecutively enrolled patients at 9 high-volume thrombectomy centers and did not exclude patients from treatment based on ischemic core estimates. They note that their cohort reflects contemporary clinical practice, reporting the number of patients with different ischemic core volume thresholds, and note that overestimation rates did not vary with core size. They point to ongoing trials of thrombectomy in patients with large estimated core volumes and to the SELECT2 study, which includes CTP and noncontrast CT imaging, and will help establish if there is a ceiling to thrombectomy benefit in patients with large vessel occlusion based on noncontrast CT or CTP findings. This exchange highlights our evolving understanding of the concept and measurement of ischemic core and its changing importance in patient selection decisions for thrombectomy. In a prospective multicenter study of 101 patients, Dr. Sarraj and colleagues assessed the accuracy of baseline CT perfusion (CTP) estimates of ischemic core, relative to follow-up diffusion-weighted imaging (DWI) lesion infarct volumes, in patients with ischemic stroke who achieved complete reperfusion with endovascular thrombectomy. The authors found that significant overestimation of the volume of irreversibly injured ischemic core—by 10 mL or more—was rare and was only observed in patients who presented within 90 minutes of last known well and achieved reperfusion within 120 minutes of CTP acquisition. Such overestimation also primarily occurred in the white matter. The overestimation of ischemia was eliminated by using a more conservative threshold for relative cerebral flood flow (rCBF) of <20% as opposed to the more typically used threshold of <30% in patients presenting within 90 minutes. In response, Drs. Garcia-Camba and Jacobi note that the prevalence and temporal distribution of patients with a large ischemic core (≥50 mL) was not reported; this could affect the overestimation rate in the study and potentially limit the generalizability of its findings. Replying to this, the authors note that their cohort was derived from consecutively enrolled patients at 9 high-volume thrombectomy centers and did not exclude patients from treatment based on ischemic core estimates. They note that their cohort reflects contemporary clinical practice, reporting the number of patients with different ischemic core volume thresholds, and note that overestimation rates did not vary with core size. They point to ongoing trials of thrombectomy in patients with large estimated core volumes and to the SELECT2 study, which includes CTP and noncontrast CT imaging, and will help establish if there is a ceiling to thrombectomy benefit in patients with large vessel occlusion based on noncontrast CT or CTP findings. This exchange highlights our evolving understanding of the concept and measurement of ischemic core and its changing importance in patient selection decisions for thrombectomy.