Perfusion Profiles in Patients with Chronic and Acute Internal Carotid Artery Occlusion
Description
Background It is important to identify patients with acute internal carotid artery occlusion (ICAO) from chronic ICAO, and there is still no easily applicable and consistent method for differentiation. Aims To determine whether the MR perfusion profile differs between acute and chronic ICAO, and to propose a diagnostic algorithm based on MR perfusion findings. Methods We included 38 chronic and 48 acute ICAO patients. Perfusion parameters were compared quantitatively and qualitatively between the two groups. Based on the qualitative ratings, a decision tree was built to differentiate acute from chronic ICAO. Results Chronic ICAO patients had smaller Tmax delay lesions than acute patients. A cutoff of >2ml of tissue with Tmax>10s had the highest sensitivity (0.979, 95% CI 0.875-0.998) and specificity (0.974, 95% CI=0.845-0.998) for identifying a chronic occlusion. The perfusion profile of chronic ICAO patients was characterized by symmetric mean transit time (MTT) (97.4%) and symmetric cerebral blood flow (CBF) (94.7%). While acute ICAO patients was characterized by Tmax delay (100%), prolonged MTT (97.9%), and decreased CBF (81.2%). When provided with a decision tree based on these profiles, independent raters could differentiate between acute and chronic ICAO with 96.5-98.8% accuracy and excellent inter-rater reliability (agreement coefficient=0.89, 95% CI 0.82-0.97). Conclusions Perfusion imaging shows promise for distinguishing acute from chronic ICAO with high accuracy. These results need to be validated in an external dataset.
Files
Steps to reproduce
MR perfusion maps, including the time-to-maximum of the residue function (Tmax), mean transit time (MTT), cerebral blood flow (CBF), and cerebral blood volume (CBV), were generated using RAPID software (version 4.9, iSchemaView, Menlo Park). On the Tmax map, perfusion lesion volumes in the affected hemisphere were calculated at four thresholds (delays of >4, >6, >8, and >10 s). For each patient, the hypoperfusion intensity ratio (HIR), a measure of collateral perfusion status, was calculated as the volume of tissue with Tmax >10 s delay divided by the volume of tissue with Tmax >6 s delay.9 For each patient, the four perfusion maps (Tmax, MTT, CBF and CBV) were visually rated by a stroke neurologist and a neuroradiologist. They independently identified the affected hemisphere and rated the perfusion of the affected hemisphere relative to the unaffected hemisphere on each perfusion map as: decreased vs. symmetric vs. increased. Differences between raters were resolved during a consensus read. Based on the consensus results, a decision algorithm was generated to differentiate chronic from acute ICA occlusions. Three additional raters who are all stroke neurologists , blinded to clinical information, rated each ICA occlusion as acute vs chronic based on the decision algorithm.