Quantitative assessment of collateral time on perfusion computed tomography in acute ischemic stroke patients

Abstract

Background and aim: Good collateral circulation is recognized to maintain perfusion and contribute to favorable clinical outcomes in acute ischemic stroke. This study aimed to derive and validate an optimal collateral time measurement on perfusion computed tomography imaging for patients with acute ischemic stroke.

Methods: This study included 106 acute ischemic stroke patients with complete large vessel occlusions. In deriving cohort of 23 patients, the parasagittal region of the ischemic hemisphere was divided into six pial arterial zones according to pial branches of the middle cerebral artery. Within the 85 arterial zones with collateral vessels, the receiver operating characteristic analysis was performed to derive the optimal collateral time threshold for fast collateral flow on perfusion computed tomography. The reference for fast collateral flow was the peak contrast delay on the collateral vessels within each ischemic arterial zone compared to its contralateral normal arterial zone on dynamic computed tomography angiography. The optimal perfusion collateral time threshold was then tested in predicting poor clinical outcomes (modified Rankin score of 5-6) and final infarct volume in the validation cohort of 83 patients.

Results: For the derivation cohort of 85 arterial zones, the optimal collateral time threshold for fast collateral flow on perfusion computed tomography was a delay time of 4.04 s [area under the curve = 0.78 (0.67, 0.89), sensitivity = 73%, and specificity = 77%]. Therefore, the delay time of 4 s was used to define the perfusion collateral time. In the validation cohort, the perfusion collateral time showed a slightly higher predicting power than dynamic computed tomography angiography collateral time in poor clinical outcomes (area under the curve = 0.72 vs. 0.67; P < 0.001). Compared to dynamic computed tomography angiography collateral time, the perfusion collateral time also had better performance in predicting final infarct volume (R-squared values = 0.55 vs. 0.23; P < 0.001).

Conclusion: Our results indicate that perfusion computed tomography can accurately quantify the collateral time after acute ischemic stroke.

Keywords: collateral velocity; computed tomography; perfusion imaging; quantitative assessment; stroke.

Figure 1

Flowchart of the inclusion of patients in the study. AIS, acute ischemic stroke; ICA, internal carotid artery; MCA, middle cerebral artery; ACA, anterior cerebral artery; PCA, posterior cerebral artery; BA, basilar artery.

Figure 2

Arterial zones of the middle cerebral artery. O, orbitofrontal arteries; OF, operculofrontal arteries; C, central arteries; P, anterior and posterior parietal arteries; G, gyrus angularis arteries; T, temporal arteries.

Figure 3

Contrast peak time by the time vs. contrast density curve. (A) Contrast peak time in the parietal arterial zone of the ischemic hemisphere; (B) contrast peak time in the same site in the contralateral arterial zone of the unaffected hemisphere.

Figure 4

CTP map (DT map) by ROC analysis. (A) DT map in each arterial zone of the ischemic hemisphere; (B) DT map in each arterial zone of the unaffected hemisphere. DT values in each arterial zone were measured on the DT map automatically, and so did the CBF, CBV, and MTT.

Figure 5

Distribution of CTP core volume (A) and delay time (B) across dCTA collateral time.

Figure 6

Receiver operating characteristic curve. (A) Predicting dCTA fast collateral flow by DT and Tmax; (B) predicting poor clinical outcomes by dCTA slow collateral flow and DT slow collateral flow.