Evaluating Intracranial Atherosclerosis Rather Than Intracranial Stenosis

TCD and Transcranial Color-Coded Duplex

TCD is a safe, noninvasive, and inexpensive method to diagnose ICAS. Although the accuracy of TCD in grading percent stenosis has varied among prior studies, it is superior in providing real-time flow information and evidence for direction of flow, collateralization, embolization, and steal phenomenon, as compared with static images of CTA and MRA.¹⁰ For instance, TCD detection of microembolic signals has been associated with specific infarct patterns on diffusion-weighted MR images.¹¹ Persistence of microembolic signals may also indicate subsequent worsening of neurological deficits during the acute phase of ischemic stroke.¹² Furthermore, microembolic signals have been reported as an independent predictor for stroke recurrence in patients with symptomatic ICAS.¹² In addition, TCD vasomotor reactivity quantification may reflect the capacity of cerebral autoregulation, often globally impaired in patients with ICAS and a possible risk factor for stroke.¹² Most recently, the use of transcranial color-coded duplex has advanced the diagnostic accuracy of ICAS by incorporating anatomic definition of the arterial lumen and has been applied increasingly to evaluate cerebral arteries in studies of revascularization therapies.¹³

Magnetic Resonance Angiography

TOF-MRA and contrast-enhanced MRA are now commonly used for assessment of the intracranial vasculature. TOF-MRA, based on the contrast mechanism known as flow-related enhancement, accentuates hemodynamic features and therefore generally overestimates the degree of stenosis, especially in cases with low flow distal to the ICAS. But the flow information carried by TOF-MRA is likely to play a role in the assessment of hemodynamic impact of the lesion, which is discussed in detail below.^8,14 Contrast-enhanced MRA, acquired with a combined head and neck coil, permits simultaneous imaging of the entire supra-aortic vasculature, from extracranial segments to distal intracranial branches.¹⁵ It may provide better morphological visualization as compared with TOF-MRA, especially for a high-degree stenosis with low flow, but its sensitivity to detect intracranial lesions is lower than that of extracranial lesions.¹⁵ Quantitative MRA, based on a phase-contrast technique and using TOF-MRA to facilitate vessel localization, is a relatively novel application of MRA to measure blood flow through vessels of interest.¹⁶ Quantification of blood flow by quantitative MRA has been found promising in identifying patients at high risk of stroke recurrence, assessing intracranial in-stent stenosis and revealing pathophysiology in various cerebrovascular disorders.^17,18

High-Resolution MRI

Recent research using modern MRI techniques, such as high-resolution MRI (HR-MRI), and other advanced technology, such as the computational fluid dynamics (CFD) technique as detailed below, produce important information to improve understanding of pathophysiology and diagnosis of intracranial atherosclerotic disease. Translation of HR-MRI from the depiction of coronary and carotid plaques to intracranial applications (Figure 1) has enabled imaging of intracranial plaque and the adjacent arterial wall, possibly identifying intracranial plaques because of atherosclerosis or other causes, revealing plaque morphology and constituents, including intraplaque hemorrhage, lipid core, and fibrous cap.⁹ Imaging features of intracranial plaque on HR-MRI (7 Tesla) have been reported to be closely correlated with plaque components by histopathologic analysis in a postmortem, in vitro study.¹⁹ HR-MRI was also been reported recently to be helpful in guiding endovascular intervention of atherosclerotic disease of the basilar artery.⁹ Further studies exploring the relationships between plaque features by HR-MRI and subsequent stroke risk may provide additional insight on the plaque stability and corresponding intervention strategy in patients with symptomatic ICAS.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Intracranial plaque and arterial wall imaging by high-resolution MRI. An intracranial atherosclerosis lesion located at proximal basilar artery with severe luminal stenosis was identified on time-of-flight magnetic resonance angiography (white arrow, A). High-resolution MRI revealed an eccentric atherosclerotic plaque along the anterolateral and posterolateral walls of basilar artery (black arrows, B–D). Courtesy of Professor WH Xu of Peking Union Medical College Hospital, Beijing, China.

Computed Tomographic Angiography

As a minimally invasive imaging modality, CTA provides better delineation of the anatomy of intracranial arteries, thus yielding higher diagnostic accuracy of the luminal stenosis of ICAS as compared with TCD and MRA, with DSA as the reference standard, although the visualization of petrous and cavernous segments of ICA by CTA may be affected by bony artifacts. Unlike TOF-MRA, the nature of CTA is not based on flow in the vessel, which is an advantage in depicting the vessel morphology but a disadvantage in the way that it tends to eliminate any temporal information about blood flow. Recently, CTA has been used increasingly in evaluating collateralization in ICAS, including leptomeningeal collateral routes, which have been correlated with risk of recurrent events.^20,21 CTA images are a good source for geometric reconstruction in preparation for blood flow simulation by CFD techniques, which is discussed below.

Digital Subtraction Angiography

DSA is currently considered as the reference standard for diagnosing intracranial vascular diseases including ICAS, because of its superb spatial and contrast resolution to depict the vessels, as well as its ability to reveal temporal information on antegrade and collateral flow.¹⁰ It is almost always used as the reference standard in studies testing the accuracy of other imaging modalities to grade the luminal stenosis of ICAS. For the evaluation of collaterals, DSA may clearly reveal patent segments and the direction of flow across segments of the circle of Willis, and it has been demonstrated of good to very good inter/intraobserver agreement to grade leptomeningeal collateralization although grading methods have varied.²¹ However, as an invasive method, DSA could lead to periprocedural complications, including transient or even permanent neurological deficits.

Perfusion Imaging

Hypoperfusion serves as a common cause for ischemic stroke in patients with ICAS.²² CT perfusion and perfusion-weighted MR imaging have been used to detect hypoperfused territories in acute ischemic stroke during the past decade, which allows for the identification of potentially salvageable tissue or the ischemic penumbra.²³ Mismatch between the hypoperfused tissue and the infarct core on multimodal CT and MR images has been used as an indicator for reperfusion therapies in clinical trials, which is altering the traditional concept of time windows.²³ Recently, selective arterial spin-labeling MR imaging has been used to reveal the perfusion territories and measure cerebral blood flow of individual cerebral arteries.²⁴ This technique also enables visualization and quantification of the actual collateral flow information in the setting of ICAS.²⁴ These perfusion imaging methods may help reveal the underlining pathophysiology of stroke and reflect the hemodynamic impact of ICAS, which may aid in clinical decision making.

CFD of ICAS

Beyond the methods detailed above to evaluate the presence of ICAS, CFD techniques can also be applied to the study of ICAS to investigate hemodynamic impact of a specific lesion.^25–27 Three-dimensional (3D) geometry of the diseased vessels may be reconstructed from angiographic images for simulation of blood flow, through which hemodynamic features of the lesion, including pressure gradients across the atherosclerotic plaque or fractional flow, may be analyzed.²⁵ CFD simulation based on biplanar DSA images revealed low fractional flow in only 40% of the severe (70%–99%) stenoses in Stenting and Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis (SAMMPRIS) trial,²⁵ again underscoring the considerable need for more comprehensive evaluation of ICAS rather than arbitrarily grading it as moderate or severe based on the maximal luminal stenosis.

Overall Problem With Studies on Diagnostic Tests Focusing on Stenosis Rather Than Other Measures

Noninvasive imaging modalities of TCD, MRA, and CTA, although widely used to detect ICAS, their diagnostic abilities have not been rigorously tested in large, prospective studies. The Stroke Outcomes and Neuroimaging of Intracranial Atherosclerosis (SONIA) trial, as a companion study to WASID and the first prospective, multicenter study to test the diagnostic ability of these imaging modalities against DSA, focused solely on the maximal degree of luminal stenosis as many other preceding studies did, which is only one among many aspects to properly characterize the lesions.¹⁰ This might be one of the most important reasons why this study showed poor consistency across these imaging modalities. More recently, the predictive ability of percent stenosis measured via TOF-MRA for the risk of stroke recurrence has been surpassed by that of the hemodynamic impact of ICAS derived from TOF-MRA as detailed below, in subsequent analyses of the MRA archives of SONIA study.⁸ Thus, the degree of stenosis may not be an appropriate target for studies investigating diagnostic tests for ICAS, which needs to be addressed in future studies.

Role of Collateralization in ICAS

Collateral status has been demonstrated to correlate with acute and final infarct volume and infarct volume expansion in patients with acute ischemic stroke.^28,29 Collateral perfusion is also predictive of response to intravenous thrombolysis and endovascular therapies, with good collaterals reducing hemorrhagic transformation while enhancing revascularization rate.^30,31 Most importantly, the degree of collateral circulation has been found to be significantly correlated with functional outcomes of patients with symptomatic ICAS.^20,32 In the WASID trial, collateral status was identified as an independent predictor for stroke recurrence in the symptomatic arterial territory, altering the stroke risk that otherwise would be different if only predicted based on the severity of luminal stenosis.⁷ More extensive collaterals reduced the risk of recurrent territorial stroke in 70% to 99% stenoses (none versus good collaterals: hazard ratio, 4.60; 95% confidence interval, 1.03–20.56; poor versus good collaterals: hazard ratio, 5.90; 95% confidence interval, 1.25–27.81; P=0.0427), whereas increased that risk in 50% to 69% stenoses (none versus good collaterals: hazards ratio, 0.18; 95% confidence interval, 0.04–0.82; poor versus good collaterals: hazards ratio, 1.78; 95% confidence interval, 0.37–8.57; P<0.0001). Collateral flow therefore is one of the most essential mediators in cerebral ischemia because of ICAS and hence an important indicator in risk prediction and treatment allocation in patients with symptomatic ICAS. However, it has not been systematically investigated, and little is known about the process of collateral recruitment or arteriogenesis in progressive ICAS, which warrant further studies.

Role of Fractional Flow in ICAS

Translation of recent advances in the cardiology field has inspired vascular neurologists to apply novel measures to characterization of ICAS. The paradigm shift in ischemia-related coronary artery disease (CAD), from anatomic measures of percent stenosis to hemodynamic impact of lesions, may be paralleled in the diagnostic approach and decision making of ICAS. Fractional flow reserve (FFR), measured as the ratio of pressures distal and proximal to a coronary lesion under induced hyperemia by floating a pressure wire during percutaneous coronary angiography, has become the gold standard to assess the hemodynamic significance of CAD.³³ Large clinical trials on FFR have demonstrated the unreliability of percent stenosis as an indicator to define a hemodynamically significant CAD, especially in cases with anatomically moderate stenoses.³⁴ For instance, in the Fractional Flow Reserve versus Angiography in Multivessel Evaluation study, 35% of lesions with an angiographically moderate severity (50%–70% stenosis) were found to be functionally significant.³⁵ Moreover, FFR-guided coronary revascularization has been demonstrated safe and superior to angiography-guided strategy in reducing major cardiac events and composite adverse events.³⁶ More recently, the CFD technique has been applied to coronary CTA to noninvasively quantify FFR.³⁴ The noninvasive FFR has been proven of good diagnostic accuracy for the hemodynamic significance of CAD, with invasive FFR as the reference.³⁴ Although cerebral arteries in many ways differ from coronary arteries, these findings in cardiology further aroused the need for neurologists to divert the solely focus on the degree of stenosis to more sound and reasonable ways to evaluate ICAS.

Similar to the application of noninvasive FFR in the coronary arteries, the CFD technique may also be used to assess fractional flow across ICAS lesions. As mentioned above, CFD modeling based on the SAMMPRIS angiography disclosed hemodynamic effects of ICAS, with decreased pressure identified distal to the ICAS lesions.²⁵ Besides, we have performed a pilot study of 10 cases on CFD modeling of ICAS based on intracranial CTA, which have confirmed the feasibility to reconstruct CFD models out of routinely obtained intracranial CTA source images (Figure 2). For CFD modeling of an ICAS lesion based on CTA, 3D geometry of a target arterial segment containing the lesion could be extracted and reconstructed from the CTA source images, which could then be meshed for simulation of the blood flow across the lesion. The simulated CFD models showed decreased pressure and increased flow velocity in situ and beyond the ICAS lesions, as shown in Figure 2. Although the CFD-based fractional flow in ICAS has yet to be correlated with subsequent stroke risk, it provides a fertile ground for next steps in the clinical research on the diagnosis and optimal treatment of ICAS.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Computed tomographic angiography source image (A) showing a right middle cerebral artery stenosis, and the reconstructed computational fluid dynamics (CFD) models illustrating pressure (B) and flow velocity (C) changes across the lesion. Decreased pressure (B) and increased flow velocity (C) in situ and downstream to the lesion are highlighted with arrows on the CFD models.

Besides the CFD-based evaluation of fractional flow, we developed another method including use of the TOF-MRA termed signal intensity ratio, based on its flow-related signal contrast mechanism, to systematically gauge the hemodynamic effects of an ICAS. Signal intensity ratio of an ICAS was measured as the ratio of distal and proximal signal intensities within the vessel lumen, adjusted by the background signal intensity on the maximum intensity projection images (Figure 3).¹⁴ We have preliminarily explored the clinical significance of signal intensity ratio.^8,37 It was found to be significantly related to acute infarct volume on diffusion-weighted MR images in a preliminary study.³⁷ Moreover, it was identified as an independent predictor for recurrent stroke in the territory of the diseased artery in the SONIA-WASID cohort.⁸ Therefore, signal intensity ratio by TOF-MRA, as a noninvasive, easy-to-perform, and highly reproducible method, may be a useful tool to differentiate high-risk ICAS.^26,38

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

The method for measurement of signal intensity ratio (SIR) of an intracranial atherosclerosis on a magnetic resonance angiography maximum intensity projection. SIR of the lesion at right middle cerebral artery is calculated as the ratio of mean signal intensities distal (1039.6) and proximal (1340.3) to the lesion, adjusted by the mean background signal intensity (401.1; mean of 409.5 and 392.6), which is (1039.6−401.1)/(1340.3−401.1)=0.68.

Need for Noninvasive Detection of ICAS

Despite the use and accuracy in evaluating anatomic severity, antegrade and collateral flow in ICAS, the invasive nature and potential periprocedural risk of DSA prevent its extensive and repeated use in indicated patients, and the high costs and dependence on experienced operators further limit its use as a routine examination for intracranial arteries in all patients with ischemic stroke. Therefore, noninvasive methods, such as TCD, MRA, and CTA, are still used commonly for the diagnosis of ICAS in clinical scenarios. Although these currently available noninvasive imaging modalities also have limitations, detailed below, comprehensive interpretation of noninvasively obtained intracranial vascular images, in combination with perfusion imaging and other methods reflecting different aspects of ICAS, may avoid the need to proceed with an invasive angiography procedure.

Limitations of Noninvasive Diagnostic Modalities and Need for Reasonable Interpretation

Based on the inherent nature of TCD, TOF-MRA, and CTA, they address different aspects of the lesion when characterizing ICAS, yet the severity of stenosis remains the primary focus despite enormous information on flow and other aspects of ICAS often ignored by clinicians. Paradoxically, the percentage of stenosis itself is not concordant among these imaging modalities.¹⁰ Furthermore, each of these noninvasive methods has its specific limitations. TCD, as a low-cost and useful tool to provide real-time cerebral flow information, requires thorough skill training and is highly operator-dependent, which therefore is still underused throughout the world ≈30 years after its first use in cerebrovascular diseases.¹² TOF-MRA and CTA, as the most commonly used noninvasive methods to depict the morphology of major intracranial arteries, respectively, based on the blood flow and vascular geometry, could complement each other with respect to the evaluation of hemodynamic and lumen-narrowing effects of ICAS, yet may incorrectly reflect severity of the lesion if interpreted alone in the evaluation of ICAS. In addition, although CTA provides a reliable method to assess the leptomeningeal collateral circulation, inconsistency in the scaling methods impedes generalization of recent findings on the correlations between noninvasive leptomeningeal collateral grading and the clinical outcomes.²¹ Therefore, application of these imaging methods and interpretation of the results should be based on the unique characteristics of each modality, to permit a comprehensive assessment of the lesion.

Recurrent Risk in Patients With ICAS of Mild or Moderate Luminal Stenosis

In contrast to intracranial atherosclerotic lesions resulting in 70% to 99% reduction of the vessel caliber, which are considered severe and high-risk in relevant studies, ICAS of <50% and 50% to 69% luminal stenosis is usually regarded as nonsignificant, or defined as mild and moderate lesions, respectively.^2,3 However, despite the high risk of stroke recurrence in patients with symptomatic 70% to 99% stenosis, ICAS of nonsevere luminal stenosis may also be at risk of recurrent events. In the WASID and the Chinese ICAS (CICAS) studies, nearly half of the recurrent stroke occurred in patients with 50% to 69% stenosis.^5,6 Few data are available on the prognosis of those with ICAS of mild (<50%) stenosis. According to large parallel cardiovascular studies performed in patients with acute coronary syndromes, those with CAD of <50% luminal stenosis also faced a non-negligible risk of death and reinfarction, although relatively lower than that of those with obstructive lesions.³⁹ Among the limited data concerning outcomes of patients with ICAS of <50% luminal stenosis, a considerable risk of recurrent stroke was observed in the CICAS cohort, yet the percentages of luminal stenosis in these mild lesions were not specifically reported.⁶ Therefore, recurrent risks in patients with ICAS of <70% stenosis need to be fully appreciated in future studies, so that high-risk patients would not be missed because of the artificially graded severity of ICAS by the degree of luminal stenosis.

Need for Correlating With Subsequent Clinical Events, Not Percent Stenosis

Increasing evidence on potential determinants of subsequent stroke risk beyond the severity of stenosis calls for changes in the concept of how to diagnose ICAS. The traditional method to grade ICAS by the maximal percentage of luminal stenosis is irrational, in light of the considerable risk of stroke recurrence in patients with ICAS of <70% luminal stenosis and the complex effects of different aspects of ICAS on stroke risk. Thus, the currently defined high-risk ICAS based on the angiographic severity of stenosis may be misleading. For this reason, evaluation of ICAS based on its correlations with subsequent clinical events rather than the percent stenosis may be of higher clinical significance, supported by recent findings in the roles of collateral status, plaque characteristics, and hemodynamic features in determining recurrent risks in symptomatic ICAS.^7,9,26

Need to Develop Methods to Evaluate All Stroke Cases With ICAS

Because of the widely adopted philosophy of grading ICAS by the severity of luminal stenosis, successive observational and interventional studies tend to be performed in restricted populations with a certain degree of stenosis, for instance, the WASID-SONIA and the SAMMPRIS trials specifically focused on patients with symptomatic ICAS of 50% to 99% and 70% to 99% luminal stenosis.^2,3,10 Given the increasingly emerging evidence for other potential indicators altering subsequent stroke risks in symptomatic ICAS, it is in great need to establish more reasonable and generalizable methods for the evaluation of all stroke cases with ICAS, regardless of the degree of stenosis. Future studies with broader considerations of this patient subset will provide abundant information on the diagnosis and treatment of all symptomatic ICAS and embrace better understanding of this important cause of ischemic stroke.

Stroke Risk of Asymptomatic ICAS

Identifying asymptomatic intracranial atherosclerotic lesions with high risk of first-ever ischemic stroke or transient ischemic attack will be a further step in advancing the management of such a patient subset, yet data concerning the stroke risk and prognostic factors in asymptomatic ICAS have been scarce to date.⁴⁰ Although previous studies reported a relatively benign clinical course of asymptomatic ICAS as compared with symptomatic lesions, these findings have not been verified in large, prospective, population-based studies. In addition, in patients with symptomatic intracranial atherosclerotic lesions, recurrent stroke may also occur in the territories of concomitant asymptomatic ICAS, for which case little is known about the clinical course and mechanisms. Exploration of the natural history and stroke risks in individuals with asymptomatic ICAS in future studies will undoubtedly facilitate primary prevention of stroke.