Abstract
Background
The purpose of this study was to determine the performance of computed tomography angiography (CTA) by using a scoring system to predict anterograde and collateral blood flow status in patients with symptomatic middle cerebral artery (MCA) stenosis with use of conventional angiography as standard reference.
Methods
We retrospectively identified all consecutive patients with unilateral symptomatic MCA stenosis in our center who underwent conventional angiography and CTA within 1 month. The anterograde and collateral blood flow (AnCo) scoring system consisted of anterograde score (AnS) and collateral score (CoS). Evaluation of the CTA images was done independently by two readers, based on the AnCo scoring system. The conventional angiography was assessed by using the Thrombolysis in Cerebral Infarction (TICI) and American Society of Interventional and Therapeutic Neuroradiology (ASITN/SIR) scoring system to determine the status of anterograde and collateral blood flow. Diagnostic performance of AnCo was evaluated by using the area under the receiver operating characteristic (ROC) curve.
Results
A total of 61 patients were included in the analysis with mean age of 53.4 ± 11.0 years. AnS demonstrated a strong correlation with TICI with statistical significance (r = 0.786; p < 0.001). CoS had a modest yet statistically significant correlation with ASITN/SIR (r = 0.574; p < 0.001). The ROC curve analysis for AnS demonstrated an area under the curve (AUC) of 0.894 (p < 0.001) and the ROC curve analysis for CoS showed an AUC of 0.824 (p < 0.001).
Conclusions
CTA was a potential method to evaluate anterograde and collateral blood flow status in patients with symptomatic unilateral MCA stenosis.
Keywords: Intracranial atherosclerosis, collaterals, computed tomography angiography
Introduction
Intracranial atherosclerosis (ICAS) is one of the most common causes of ischemic stroke, with a high recurrent stroke rate.1 In the Stenting and Aggressive Medical Management for Preventing Recurrent stroke in Intracranial Stenosis (SAMMPRIS) trial, 12.2% of patients had recurrent stroke within 1 year after the index event even with aggressive medical treatment.2 Identifying patients with high risk of recurrent stroke was crucial in the management of ICAS patients. Previous studies have shown that status of anterograde blood flow through stenosis lesion and retrograde blood flow from collaterals significantly correlated with recurrent stroke risk and clinical prognosis in patients with ICAS.3, 4 Therefore, assessment of the anterograde and collateral blood flow would play a prominent role in the management of patients with symptomatic ICAS.5
Conventional angiography is considered to be the standard reference to assess cerebral blood flow status as well as vessel anatomy. Scales such as Thrombolysis in Cerebral Infarction (TICI) and American Society of Interventional and Therapeutic Neuroradiology (ASITN/SIR) based on angiogram are commonly used for anterograde and collateral blood flow evaluation, respectively.6 However, conventional angiography is not routinely performed in clinical practice due to its invasive nature and requirement for expertise.7 Advanced noninvasive techniques, such as computational fluid dynamics (CFD)8 may be used to estimate hemodynamic status of cerebral blood flow. However, these techniques are still not widely available in clinical practice.
Computed tomography (CT) angiography (CTA) is a widely available and commonly used imaging modality in ischemic stroke patients.9 For patients with acute ischemic stroke caused by proximal arterial occlusion, CTA has been routinely used to evaluate collateral status to predict clinical outcomes and determine further management.10 A variety of collateral scoring systems based on CTA have been used for those patients.11–15 However, an equivalent CTA scoring system was not available for patients with intracranial stenosis and the diagnostic performance of CTA to assess collaterals has not been evaluated.
The objective of this study was to assess the performance of CTA using a novel scoring system, the anterograde and collateral blood flow score (AnCo score), which was developed on the basis of previous scoring systems for acute arterial occlusion to evaluate anterograde as well as collateral status in patients with symptomatic unilateral middle cerebral artery (MCA) stenosis with conventional angiography as standard reference.
Methods
Patients
We retrospectively identified all consecutive patients admitted to our center due to symptomatic MCA stenosis from January 2014 to June 2015 who met the following criteria: (1) CTA and conventional angiography performed within 1 month; (2) conventional angiography confirmed moderate-to-severe M1 or proximal M2 stenosis or chronic occlusion on one side; (3) CT criteria for adequacy of scan met: included complete coverage from skull base to vertex, clear visualization of arterial phase, and internal cerebral veins and dural sinuses in the normal hemisphere; (4) adequate spatial and temporal information on conventional angiography with four-vessel run to evaluate anterograde and collateral status were available; (5) no other major intracranial artery stenosis greater than 50%. Symptomatic MCA stenosis was defined as moderate-to-severe MCA stenosis which was associated with transient ischemic attack or ischemic stroke within 90 days. Patients with acute ischemic stroke caused by MCA occlusion were excluded. The study was approved by the institutional review board and the requirement to obtain informed consent was waived. Demographic and clinical data were abstracted from the medical records.
Imaging protocol
CT scans were obtained using a multi-detector scanner (GE Medical Systems, Discovery CT750 HD, USA). The CTA was performed using the following parameters: 100 kVp, 320 mAs, slice thickness 0.625 mm, slice acquisition interval 0.62 mm, intravenous administration of 90 ml iodinated contrast material at a rate of 5 ml/sec followed by 30 ml of saline at the same rate. Data acquisition was performed from the base of skull to the vertex. The scanning was auto-triggered by using bolus-tracking technique to monitor the concentration of the contrast agent. The ROI (Region of interest) was placed in the common carotid. CTA and invasive angiography images were downloaded from the picture archiving and communication (PACS) system in the form of DICOM. Osirix version 7.5.1 (http://www.osirix-viewer.com Pixmeo SARL, Swiss), an image-processing software was used to reconstruct 2D maximal intensity projection (MIP) images in axial and coronal planes from CTA source images by using 30 mm-thick slabs.
Anterograde and collateral (AnCo) scoring system
Overall, two cuts of MIP images in each of the axial and coronal plane were selected for the scoring: (1) the level of the thalamus and basal ganglia in the axial plane; (2) the middle of the lateral ventricle in the axial plane; (3) the cut with MCA stem and its major branches in the coronal plane; (4) the cut with MCA branches in the Sylvian sulcus in the coronal plane (Figure 1, 2).
Figure 1.
Case example demonstrating correlation of AnCo score and TICI and ASITN/SIR. There was a severe stenosis in the M1 segment of left MCA. On all four cuts of CT angiography (a), the vessel contrast filling distal to the stenosis lesion was obviously compromised compared with the contralateral hemisphere (↓). On the axial plane, PCA contrast filling was slightly greater than the contralateral PCA. (*) On the coronal plan, ACA contrast filling was exuberant compared with the opposite ACA (Δ). On angiogram (b), the anterograde flow was compromised (TICI = 1). There are robust collaterals from ACA and moderate collateral from the PCA (ASITN/SIR = 2). The table (c) showed the AnS and CoS assigned by the two raters (rater 1/rater 2).
ACA: anterior cerebral artery; AnCo: anterograde and collateral; ASITN/SIR: American Society of Interventional and Therapeutic Neuroradiology; CT: computed tomography; MCA: middle cerebral artery; PCA: posterior cerebral artery; TICI: Thrombolysis in Cerebral Infarction
Figure 2.
Case example of moderate stenosis in the M1 segment of left MCA demonstrating correlation of AnCo score and TICI and ASITN/SIR. CTA MIPs (a) reveal that the vessel contrast filling distal to the stenosis lesion was equivalent to the contralateral hemisphere (↓). On the axial plane, PCA contrast filling was similar to the contralateral PCA (*). On the coronal plane, ACA contrast filling was similar to the opposite ACA (Δ). On the angiogram (b), the anterograde flow was preserved (TICI = 3) with no collaterals from the PCA or ACA (ASITN/SIR = 0). The table (c) showed the AnS and CoS assigned by the two raters (rater 1/rater 2).
ACA: anterior cerebral artery; AnCo: anterograde and collateral; ASITN/SIR: American Society of Interventional and Therapeutic Neuroradiology; CTA: computed tomography angiography; DSA: ; MCA: middle cerebral artery; MIP: maximal intensity projection; PCA: posterior cerebral artery; TICI: Thrombolysis in Cerebral Infarction
Since the anterior cerebral artery (ACA) and posterior cerebral artery (PCA) were two primary routes of collateral supply when the original blood flow of MCA was compromised, ACA and PCA were assessed as collateral parts in the scoring system. The MCA, ACA and PCA in the affected hemisphere were assessed separately for contrast filling in the branches as compared with the normal hemisphere. Differentiating between MCA territory from ACA/PCA territory is based on cerebrovascular anatomy and characteristics of vessel branches. The gap with no contrast filling between two artery territory was considered as the threshold.
A score of 0–4 was assigned to each vessel.16 For MCA, the score was: 0 (absent vessel filling distal to the lesion or with vessel filling in the territory but can diagnose MCA stem occlusion); 1 (vessel filling less than 50% of the contralateral hemisphere); 2 (vessel filling more than 50% but less than 100% of the contralateral hemisphere); 3 (vessel contrast filling equal to the contralateral hemisphere); 4 (vessel contrast filling greater than the contralateral hemisphere). For ACA and PCA, the score was: 0 (absent vessel filling); 1 (vessel filling less than the opposite hemisphere); 2 (vessel filling equal to the opposite hemisphere); 3 (vessel filling greater than that of opposite hemisphere); 4 (exuberant) (Table 1).
Table 1.
The anterograde and collateral blood flow (AnCo) score.
| Score | MCA | ACA/PCA |
|---|---|---|
| 0 | Absent vessel contrast filling distal to the lesion or with vessel filling in the MCA territory but can diagnose MCA stem occlusion | Absent vessel filling |
| 1 | Vessel contrast filling less than 50% of the contralateral hemisphere | Vessel filling less than contralateral hemisphere |
| 2 | Vessel contrast filling more than 50% but less than 100% of contralateral hemisphere | Vessel filling equal to that of contralateral hemisphere |
| 3 | Vessel contrast filling equal to the contralateral hemisphere | Vessel filling greater than the contralateral hemisphere |
| 4 | Vessel filling greater than the contralateral hemisphere | Exuberant |
ACA: anterior cerebral artery; AnCo: anterograde and collateral; MCA: middle cerebral artery; PCA: posterior cerebral artery
MCA was assigned a score on all four cuts mentioned above, and the anterograde score (AnS) (0–16 points) was the sum. PCA was assessed on the two axial MIP images, and ACA on the two coronal MIP images. The collateral score (CoS) (0–16 points) was the sum of PCA score and ACA score (Figure 1, 2). A high AnS or CoS indicates good MCA contrast filling or PCA/ACA contrast filling.
Imaging review
The CTA images were analyzed independently by two readers according to the AnCo score system and were blinded to clinical data and angiogram scores. The window width and window level were set as 600 and 150, respectively. The anterograde and collateral flow on the conventional angiogram were evaluated by two experienced interventional neuroradiologists on consensus using the TICI, and the ASITN/SIR, respectively. The criteria of TICI and ASITN/SIR were described previously.4 A TICI grade of 0, 1, or 2a was considered as ‘compromised’ anterograde flow, and 2b or 3 was considered as ‘preserved’ anterograde flow. An ASITN/SIR of 0–1 was considered as poor collateral circulation, while 2–4 was considered as good collateral circulation.
Statistical analysis
The AnS and CoS ware analyzed as nominal variables. Inter-rater reliability was assessed by intraclass correlation coefficient analysis of AnS and CoS separately. The average of the two reviewers’ score was considered as the final score. Wilcoxon test was used to compare the AnS between patients with TICI 0–2a and 2b–4, and CoS between patients with ASITN/SIR 0–1 and 2–4. Spearman rank correlation coefficient was used to assess the correlation of AnS with TICI and CoS with ASITN/SIR. A p < 0.05 (two-sided) was considered to indicate statistically significant difference. The ability of both AnS and CoS for the prediction of anterograde and collateral blood flow was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC) with the dichotomous TICI and ASITN/SIR being the standard reference. All statistical analyses were performed using SPSS version 22.0 (IBM, USA).
Results
A total of 61 patients were included in the study, with a mean age of 53.4 ± 11.0 years. Among them, 68.9% were men. Among all 61 lesions, 58 were in the M1 segment and 3 were in the proximal M2 segment. The mean stenosis severity degree was 76.92 ± 18.65. Moderate stenosis (50–69%) was observed in 23 patients and severe stenosis (70–99%) in 34 patients. Chronic occlusion was identified in 4 patients. Anterograde flow was determined as compromise (TICI 0–2a) in 31 patients and 16 of those had robust collaterals (ASITN/SIR 2–4). A total of 30 patients were considered as having adequate anterograde flow and none of them had robust collaterals (ASITN/SIR 2–4).
The median AnS was significantly lower in lesions with TICI 0–2a compared with lesions with TICI 2b–3, 7 (interquartile range, 5.5–8.5) versus 10.5 (interquartile range, 9.375–11.5; p < 0.001). Median CoS was also significantly lower in lesions with ASITN/SIR 0–1 compared with lesions with ASITN/SIR 2–4, 9.5 (interquartile range, 9–11) versus 12.25 (interquartile range, 10.625–13; p < 0.001). AnS demonstrated a strong correlation with TICI with statistical significance (r = 0.786; p < 0.001). CoS had a modest yet statistically significant correlation with ASITN/SIR (r = 0.574; p < 0.001). Figure 3 illustrates the correlation between AnS and TICI as well as CoS and ASITN/SIR. The ROC curve analysis for AnS demonstrated an AUC of 0.894 (p < 0.001). The optimal cut-off value was found to be 9.25, with a sensitivity of 0.767 and specificity of 0.871. The ROC curve analysis for CoS showed an AUC of 0.824 (p < 0.001). The optimal cut-off value was 12.25. The sensitivity and specificity were 0.688 and 0.800, respectively (Figure 4).
Figure 3.
Box plots showed the relationship of AnS and TICI scale (r = 0.786; p < 0.001) (left) and CoS and ASITN/SIR (r = 0.574; p < 0.001) (right).
AnS: anterograde score; ASITN/SIR: American Society of Interventional and Therapeutic Neuroradiology; CoS: collateral score; TICI: Thrombolysis in Cerebral Infarction
Figure 4.
ROC curve of AnS with TICI (left, AUC 0.894; p < 0.001) and CoS with ASITN/SIR (right, AUC 0.824; p < 0.001) AnS: anterograde score; ASITN/SIR: American Society of Interventional and Therapeutic Neuroradiology; AUC: area under curve; CoS: collateral score; ROC: receiver operating characteristic; TICI: Thrombolysis in Cerebral Infarction
Assessment of inter-observer variability of AnS and CoS demonstrated an intraclass coefficient of 0.900 (95% confidence interval: 0.839, 0.939) and 0.763 (95% confidence interval: 0.635, 0.851), respectively.
Discussion
We assessed the diagnostic performance of CTA using a novel scoring system to assess anterograde (MCA) and collateral (ACA/PCA) status in patients with symptomatic unilateral MCA stenosis based on the opacification of distal artery branches compared with the contralateral hemisphere. The AnCo score is a semi-quantitative grading system developed on the basis of previous described collateral evaluation scale based on CTA for acute ischemic stroke patients caused by MCA occlusion. We demonstrated a strong correlation between the AnS with TICI, and a moderate correlation between the CoS with ASITN/SIR. In addition, the AnCo score had high inter-rater reliability.
Various collateral scoring systems for acute ischemic stroke caused by MCA occlusion were introduced and demonstrated to have good performance for predicting imaging and clinical prognosis.11 The assessment of collaterals for artery stenosis is more complicated than that of artery occlusion because of the presence of anterograde flow. For patients with ‘preserved' anterograde flow, distal perfusion across the lesion was adequate, and the collaterals played a less important role. On the contrary, for patients with compromised anterograde flow across the lesion, collaterals would play a prominent role in sustaining distal territory perfusion.4 Therefore, both anterograde and collateral blood flow should be evaluated in these patients. The AnCo score combined the two routes into one scoring system and evaluated the them separately, which enables the discrimination between contribution of anterograde and collateral blood flow.
One concern of CT angiography to evaluate collaterals for MCA stenosis was whether it is reliable to distinguish anterograde blood flow from collateral blood flow. In our experience, the borderline could be identified on CT angiography based on the anatomy of cerebrovascular and characteristics of vessel branches. Our results also showed good correlation between AnCo score on CT angiography and TICI/ASITN on angiography. Although the confidence intervals of AnCo for patients with different TICI/ASITN levels are overlapping, an obvious trend of increase could be observed along the level change, which indicates that CT angiography could to some extent reflect hemodynamic information.
The anterograde flow is more likely to be compromised compared with the contralateral hemisphere due to the luminal stenosis. While collateral flow tends to be more robust than the opposite hemisphere, particularly in patients with compromised anterograde flow, possibly because of diminished blood pressure in downstream vessels. Therefore, the grading systems for MCA and ACA/PCA were designed to be different. For MCA, the compromise status (less than contralateral hemisphere) was divided into two grades (<50% and 50–100%), while the strength of ACA/PCA flow was more detailed and graded into two scales (greater than the contralateral hemisphere and exuberant). Such a design may make the scoring system more specific and targeted to evaluate the two routes.
Novel techniques have been developed to evaluate blood flow status of ICAS patients, such as CFD,8 arterial spin labeling (ASL) MRI,17 or invasive fractional flow reserve18. However, the application of these novel techniques in clinical practice still needs some time due to their limited availability. Clinicians may estimate the blood flow status as well as vessel anatomy at the same time on the basis of commonly accessible CTA.
The shortcoming of single-phase CTA was considered to be lack of temporal information.19 However, the contrast filling or not filling at a certain time could reflect the cerebral blood flow information. Our results demonstrated good correlation between the hemodynamic information derived from CTA images and those from dynamically acquired angiograms. In addition, the AnCo score could also be applied to multiphase or dynamic CTA images for medical centers that have access to such techniques.
Limitations of the study should be considered, including the retrospective nature that may lead to bias due to patient selection and the relatively small sample size. The scoring system was relatively crude by assessing the extent of ‘vessel contrast filling’ without further evaluation of characteristics of the contrast filling. The scoring system could be improved in the future by using more detailed information derived from CTA. In addition, prospective studies which assess the correlation between the AnCo score with clinical prognosis will provide more evidence for the clinical implications of this scoring system.
Conclusion
We established that CTA with a novel scoring system (AnCo score) was a potential method to evaluate anterograde as well as collateral status in patients with symptomatic MCA stenosis. Once it was verified in prospective studies, it may have potential in clinical use for evaluation of symptomatic MCA stenosis.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (contract grant number: 81371290) and the Beijing High-level Personnel Funds, China (contract grant number: 2013-2-19).
References
- 1.Qureshi AI, Caplan LR. Intracranial atherosclerosis. Lancet 2014; 383: 984–998. [DOI] [PubMed] [Google Scholar]
- 2.Chimowitz MI, Lynn MJ, Derdeyn CP, et al. Stenting versus aggressive medical therapy for intracranial arterial stenosis. N Engl J Med 2011; 365: 993–1003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Liebeskind DS, Cotsonis GA, Saver JL, et al. Collaterals dramatically alter stroke risk in intracranial atherosclerosis. Ann Neurol 2011; 69: 963–974. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lau AY, Wong EH, Wong A, et al. Significance of good collateral compensation in symptomatic intracranial atherosclerosis. Cerebrovasc Dis 2012; 33: 517–524. [DOI] [PubMed] [Google Scholar]
- 5.Leng X, Wong KS, Liebeskind DS. Evaluating intracranial atherosclerosis rather than intracranial stenosis. Stroke 2014; 45: 645–651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Higashida R, Furlan A, Roberts H, et al. Trial design and reporting standards for intra-arterial cerebral thrombolysis for acute ischemic stroke. J Vasc Interv Radiol 2003; 14: S493–S494. [DOI] [PubMed] [Google Scholar]
- 7.Kortman HG, Smit EJ, Oei MT, et al. 4d-cta in neurovascular disease: A review. AJNR Am J Neuroradiol 2015; 36: 1026–1033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Leng X, Scalzo F, Ip HL, et al. Computational fluid dynamics modeling of symptomatic intracranial atherosclerosis may predict risk of stroke recurrence. PLoS One 2014; 9: e97531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Power S, McEvoy SH, Cunningham J, et al. Value of CT angiography in anterior circulation large vessel occlusive stroke: Imaging findings, pearls, and pitfalls. Eur J Radiol 2015; 84: 1333–1344. [DOI] [PubMed] [Google Scholar]
- 10.Leng X, Fang H, Leung TW, et al. Impact of collaterals on the efficacy and safety of endovascular treatment in acute ischaemic stroke: A systematic review and meta-analysis. J Neurol Neurosurg Psychiatry 2016; 87: 537–544. [DOI] [PubMed] [Google Scholar]
- 11.McVerry F, Liebeskind DS, Muir KW. Systematic review of methods for assessing leptomeningeal collateral flow. AJNR Am J Neuroradiol 2012; 33: 576–582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Menon BK, Smith EE, Modi J, et al. Regional leptomeningeal score on CT angiography predicts clinical and imaging outcomes in patients with acute anterior circulation occlusions. AJNR Am J Neuroradiol 2011; 32: 1640–1645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Maas MB, Lev MH, Ay H, et al. Collateral vessels on CT angiography predict outcome in acute ischemic stroke. Stroke 2009; 40: 3001–3005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Miteff F, Levi CR, Bateman GA, et al. The independent predictive utility of computed tomography angiographic collateral status in acute ischaemic stroke. Brain 2009; 132: 2231–2238. [DOI] [PubMed] [Google Scholar]
- 15.Lee KH, Cho SJ, Byun HS, et al. Triphasic perfusion computed tomography in acute middle cerebral artery stroke: A correlation with angiographic findings. Arch Neurol 2000; 57: 990–999. [DOI] [PubMed] [Google Scholar]
- 16.Tan IY, Demchuk AM, Hopyan J, et al. CT angiography clot burden score and collateral score: Correlation with clinical and radiologic outcomes in acute middle cerebral artery infarct. AJNR Am J Neuroradiol 2009; 30: 525–531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lyu J, Ma N, Liebeskind DS, et al. Arterial spin labeling magnetic resonance imaging estimation of antegrade and collateral flow in unilateral middle cerebral artery stenosis. Stroke 2016; 47: 428–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Han YF, Liu WH, Chen XL, et al. Severity assessment of intracranial large artery stenosis by pressure gradient measurements: A feasibility study. Catheter Cardiovasc Interv 2016; 88: 255–261. [DOI] [PubMed] [Google Scholar]
- 19.Menon BK, O'Brien B, Bivard A, et al. Assessment of leptomeningeal collaterals using dynamic CT angiography in patients with acute ischemic stroke. J Cereb Blood Flow Metab 2013; 33: 365–371. [DOI] [PMC free article] [PubMed] [Google Scholar]