Article Text

Original research
Reduced cerebrovascular reserve is associated with an increased risk of postoperative ischemic lesions during carotid artery stenting
  1. Masaomi Koyanagi1,2,
  2. Kazumichi Yoshida2,
  3. Yoshitaka Kurosaki1,
  4. Nobutake Sadamasa1,
  5. Osamu Narumi1,
  6. Tsukasa Sato1,
  7. Masaki Chin1,
  8. Akira Handa1,
  9. Sen Yamagata1,
  10. Susumu Miyamoto2
  1. 1Department of Neurosurgery, Kurashiki Central Hospital, Kurashiki-City, Japan
  2. 2Department of Neurosurgery, Kyoto University Graduate School of Medicine, Kyoto, Japan
  1. Correspondence to Dr Masaomi Koyanagi, Department of Neurosurgery, Kyoto University Graduate School of Medicine, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan; koyanagm{at}


Background Reduced cerebrovascular reserve (CVR) is associated with increased risk of ischemic events in carotid steno-occlusive diseases.

Objective To determine whether pretreatment CVR can predict postoperative ischemic lesions after carotid artery stenting (CAS) by retrospective analysis.

Methods We retrospectively reviewed the medical records of 46 patients (42 men; mean age 74.2±8.3 years) who underwent CAS and preprocedural cerebral blood flow measurement by quantitative single-photon emission CT. Ischemic lesions were evaluated by diffusion-weighted image (DWI) within 72 h after the intervention. We also evaluated plaque characteristics using black-blood MR plaque imaging.

Results New ipsilateral DWI-positive lesions were found in 11 cases (23.9%). Patients were classified into two groups based on the presence or absence of new DWI-positive lesions, and no significant differences in characteristics were found between the DWI-positive and DWI-negative groups, except for age and CVR of the ipsilateral middle cerebral artery (MCA) territory. The DWI-positive group was significantly older than the DWI-negative group (79.7±4.1 vs 72.5±8.6 years; p=0.0085) and had lower average regional CVR (1.4±18.2% vs 22.4±25.8%; p=0.016). MR plaque imaging showed no significant difference in relative overall plaque MR signal intensity between the two groups (1.53±0.37 vs 1.34±0.26; p=0.113). In multivariate logistic regression analysis, lower CVR of the ipsilateral MCA territory (<11%) was the only independent risk factor for new ischemic lesions following CAS (OR=6.99; 95% CI 1.17 to 41.80; p=0.033).

Conclusions Impaired pretreatment CVR was associated with increased incidence of new infarction after CAS.

Statistics from


Atherosclerotic carotid stenosis is an important risk factor for ischemic stroke.1 Prevention of stroke in patients with symptomatic carotid stenosis requires carotid revascularization. Carotid endarterectomy (CEA) and carotid artery stenting (CAS) are both effective treatments for carotid artery stenosis. Although there is a low risk of perioperative stroke with both procedures, stroke is more common after CAS than CEA.2 Most of these perioperative strokes are clinically minor; however, they contribute to the increasing disability associated with this procedure. A randomized trial comparing CAS with CEA found no significant difference in the primary clinical outcome of stroke, myocardial infarction, or death within the periprocedural period, or any ipsilateral stroke thereafter. Although there was a higher risk of myocardial infarction after CEA, this was offset by a higher risk of stroke following CAS.3

The preoperative risk assessment of stroke is essential to define patients who might be suitable for intervention with CAS. Intravascular manipulation during CAS could result in disruption of fragile plaque and cause distal embolism. Nonetheless, some studies found that the embolic burden detected by transcranial Doppler monitoring during CAS was not related to the number of acute ischemic lesions detected by diffusion-weighted image (DWI).4–6 Although the mechanism of this phenomenon remains unclear and controversial,7 many microemboli detected by transcranial Doppler are washed from the vascular bed by cerebral blood flow (CBF). Any remaining microemboli that are not washed from the vascular bed during CAS could result in periprocedural stroke.

Cerebrovascular reserve (CVR), defined as the increase in CBF in response to a vasodilatory stimulus, is known to reflect the capacity of the brain to maintain adequate blood flow in the face of decreased perfusion due to arterial stenosis. A recent meta-analysis showed that reduced CVR is a predictor of future stroke in patients with carotid artery occlusive disease.8 These observations suggest that CVR can be used to predict periprocedural stroke in patients with CAS.

Several methods can be used to measure CVR. Single-photon emission computed tomography (SPECT) is one of the techniques for direct measurement of CBF and CVR. Although it has been widely used, standardization among different centers has not been established. To solve this problem, a quantitative SPECT (QSPECT) image reconstruction method has been developed, which is a program that automatically and accurately corrects attenuated and scattered radiation to diminish the variability in results.9 QSPECT combined with the dual table autoradiography method was reported to be a standardized evaluation method for CBF and CVR, and might be useful in large multicenter studies, even though it is not truly quantitative because of the lack of continuous arterial blood sampling.10

The aim of this study was to investigate whether reduced CVR might be a risk factor for periprocedural ischemic lesions after CAS. For measurement of CVR, we used the QSPECT/dual table autoradiography method as a standardized technique.

Materials and methods

The study included 46 consecutive patients who underwent elective CAS between August 2008 and February 2013 in our department. All patients underwent both QSPECT/dual table autoradiography before CAS, and DWI within 72 h following CAS. Indications for CAS included a symptomatic carotid artery stenosis of at least 50% or an asymptomatic carotid artery stenosis of at least 80% with higher risk for CEA (figure 1A). We tended to exclude those patients who had soft plaque diagnosed with MRI plaque imaging, which will be described below. Patients gave written informed consent to undergo CAS. The study was approved by the institutional ethics committee, and informed consent was obtained from all patients or their nearest kin before enrollment.

Figure 1

Intraoperative left common carotid angiogram showed severe stenosis in the carotid bifurcation (A). Intraoperative left common carotid roadmap (B) showed that inflation of the balloon of the guiding catheter (black arrow) made it easier to pass the distal balloon protection device (black arrowhead) through the severely stenotic lesion. (C) The distal balloon protection device was positioned and inflated just below the skull base (white arrowhead). Final left common carotid angiogram (D) showed the successful placement of stent in the carotid bifurcation. Preoperative cerebrovascular reserve (CVR) map (E) showed the impairment of CVR reserve in the left cerebral hemisphere. Diffusion-weighted image on postoperative day 1 (F) showed the small hyperintensity spot (white arrow) in the posterior border-zone area.

The procedure for QSPECT/dual table autoradiography was performed as previously reported.10 In brief, two dynamic SPECT scans were acquired in quick succession, with a 2 min interval between the scans. The first scan covered the initial 0–28 min period, and the second was acquired from 30 to 58 min. 123I-iodoamphetamine was infused twice, at 0 and 30 min. Ten minutes after the first 123I-iodoamphetamine administration, arterial blood sampling was performed and an input function was obtained. Acetazolamide (17 mg/kg, 1000 mg maximum) was administered intravenously at 20 min after the first iodoamphetamine injection, corresponding to 10 min before the second iodoamphetamine injection. Images were reconstructed and CBF quantification was performed by the QSPECT image reconstruction package. CVR was defined as {(CBF after acetazolamide challenge−CBF at rest)/CBF at rest}×100 (figure 1E).

All patients were premedicated with dual antiplatelet therapy using a combination of aspirin (100 mg/day), clopidogrel (75 mg/day), or cilostazol (200 mg/day) at least 3 days before the procedure. At the beginning of the procedure, heparin was given by intravenous infusion to maintain an activated clotting time that was twice the baseline value, with a target of ≥250 s. CAS was performed using transfemoral catheterization under local anesthesia. A 9 Fr Optimo balloon-guiding catheter (Tokai Medical Products, Aichi, Japan) was positioned in the common carotid artery proximal to the stenosis. One reason that we routinely used the balloon–guiding catheter is to stabilize navigation of distal protection devices. When it was difficult for the devices to pass through the lesion, we transiently, usually <2 min, inflate the balloon to support their navigation to the distal area (figure 1B). The balloon-guiding catheter was also reported to be useful for those patients with carotid stenosis and intraluminal thrombi.11 The degree of carotid stenosis was assessed using cerebral digital subtraction angiography by the method reported in the North American Symptomatic Carotid Endarterectomy trial.12

An embolic protection device was introduced over one of the following guidewires: FilterWire EZ (n=17; Boston Scientific, Natick, Massachusetts, USA), CarotidGuardwire (n=14; Medtronic, Santa Rosa, California, USA), Angioguard Xp (n=10; Cordis, Miami, Florida, USA), Spider Xp (n=4; Covidien, Plymouth, Minnesota, USA), or MOMA (n=1; Invatec Corp, Brescia, Italy). We preferred to use the distal balloon protection device for unstable plaque diagnosed by MRI plaque imaging (figure 1C). The distal balloon was inflated for almost 5–10 min. A Carotidwall (n=21; Boston Scientific), Precise (n=20; Cordis), or Protégé (n=5; Covidien) was deployed over the residual stenosis. Pre- and postdilatation was routinely performed with a Starling (Boston Scientific) balloon catheter for predilatation with nominal pressure for 30 s, and Aviator (Cordis) balloon catheter for postdilatation with nominal pressure for 5 s (figure 1D).

The MR instrument used in this study was a 1.5 T system (Achiva-1.5 T, Philips Medical Systems, Best, The Netherlands). DWI was performed under the following conditions: slice thickness 5 mm, gap thickness 0 mm, matrix 128×256, field of view 270 mm, repetition time/echo time 5000/93.0, and b value 1200 s/mm2. Baseline DWI was obtained before CAS in all patients. The characteristics of carotid plaque were assessed with T1-weighted black-blood MR imaging, as previously reported.13 We measured relative overall plaque MR signal intensity with manual operator-defined regions of interest drawn over the entire plaque in the carotid area with the greater degree of stenosis. We used the sternocleidomastoid muscle as the reference tissue on the T1 weighted image. Relative overall plaque MR signal intensity was calculated as follows: signal intensity of the entire plaque/signal intensity of the reference tissue.

Values are presented as the mean±SD or the number (%) of patients. Fisher's exact test was used to compare categorical variables, and a Mann–Whitney U test was used to compare continuous variables. We used receiver operating characteristic curve analysis to determine the optimal cutoff values of age, degree of stenosis, relative overall plaque MR signal intensity, mean CBF of the ipsilateral middle cerebral artery (MCA) territory, and mean CVR of the ipsilateral MCA territory to predict postprocedural ischemic lesions. Univariate logistic regression analysis was used to determine the ability of each of the baseline characteristic to predict the occurrence of new DWI-positive lesions within 72 h after the procedure (figure 1F). All categorical variables with p<0.15 in univariate analysis were subsequently included in multivariate logistic regression analysis. A p value <0.05 was considered statistically significant. All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University), which is a graphical user interface for R (The R Foundation for Statistical Computing, V.2.13.0). More precisely, it is a modified version of R commander (V.1.8–4) that includes statistical functions that are frequently used in biostatistics.


The baseline characteristics of the 46 patients (42 men, 4 women; mean age 74.2±8.3 years) included in this study are shown in table 1. Twenty-two patients were symptomatic and 24 were asymptomatic. The average carotid artery stenosis was 79.9±12.4%, according to the North American Symptomatic Carotid Endarterectomy criteria. The quality of MR plaque imaging was sufficient for measurement of relative overall plaque MR signal intensity in all patients. The mean relative overall plaque MR signal intensity was 1.39±0.30.

Table 1

Baseline characteristics of the study population

The quality of QSPECT/dual table autoradiography imaging was sufficient for measurement of CBF and CVR in all patients. CBF of the ipsilateral MCA territory measured by QSPECT was 26.2±5.1 mL/min/100 g and CVR of the ipsilateral MCA territory was 17.4±25.6%.

CAS was completed successfully in all patients. New hyperintense DWI-positive lesions were recognized on the post-treatment scan in 11 patients (23.9%) (DWI-positive group), and no new lesions were seen in 35 patients (DWI-negative group). Most of the new lesions were small round spots located in the watershed zone. Among the 11 patients in whom new DWI-positive lesions were recognized, one or two lesions were recognized in nine patients and three or more lesions were recognized in the other two patients. All but one DWI-positive patient was asymptomatic. There was one transient ischemic attack at 5 days after the procedure.

In univariate logistic regression analysis, demographic and clinical characteristics did not discriminate between the DWI-negative and DWI-positive groups, except for age and CVR of the ipsilateral MCA territory (table 2). The DWI-positive group was significantly older than the DWI-negative group (79.7±4.1 vs 72.5±8.6 years; p=0.009). The CVR of the ipsilateral MCA territory was significantly lower in the DWI-positive group than in the DWI-negative group (1.4±18.2 vs 22.4±25.8; p=0.016). However, MR plaque imaging showed that relative overall plaque MR signal intensity was not significantly different between the two groups (1.53±0.37 vs 1.34±0.26; p=0.113). There was also no difference in CBF of the ipsilateral MCA territory between the DWI-positive and DWI-negative groups.

Table 2

Ability of each baseline characteristic to predict new postprocedural DWI-positive lesions in a univariate logistic regression analysis

We determined the optimal cutoff values for age, relative overall plaque MR signal intensity and CVR of the ipsilateral MCA territory using receiver operating characteristic curve analysis. For age, the optimal cutoff value to detect postprocedural ischemic lesions was 75 years, and the sensitivity and specificity were 54.3% and 90.9%, respectively. The optimal cutoff value for relative overall plaque MR signal intensity was 1.34, and the sensitivity and specificity were 60.0% and 72.7%, respectively. Furthermore, the optimal cutoff value for CVR of the ipsilateral MCA territory was 11%, and the sensitivity and specificity were 65.7% and 81.8%, respectively.

In multivariate logistic regression analysis, the only independent risk factor for new postprocedural ischemic lesions after CAS was low CVR of the ipsilateral MCA territory (<11%) (OR=6.99; 95% CI 1.17 to 41.80; p=0.033; table 3).

Table 3

Independent risk factors for new ischemic lesions after CAS based on multivariate logistic regression analysis


In this study, we demonstrated that lower preoperative CVR of the ipsilateral MCA territory (<11%) measured by QSPECT and acetazolamide injection was an independent risk factor for new ipsilateral ischemic lesions following CAS.

Our study focused on the relationships between CVR in response to acetazolamide measured by QSPECT and new ipsilateral ischemic lesions detected by DWI within 72 h after CAS. A previous study also showed that lower CVR evaluated by QSPECT was a risk factor for the development of ipsilateral DWI-positive lesions caused by microemboli during CEA.14 It has been hypothesized that in cases of lower cerebral perfusion, the microembolic load is less likely to clear from the vascular bed.15 In CAS, Orlandi et al16 reported that the mean blood flow velocity in the ipsilateral MCA was significantly lower in the group with, than in the group without, ischemic events after CAS. They concluded that lower perfusion in the affected territory might impair the clearance of microemboli during the CAS procedure. Although CBF will recover soon after carotid revascularization, CVR will take more time to become normal. This discrepancy might cause not the postprocedural hyperperfusion17 and also the impairment of clearance of the small emboli generated during the procedure.

Although impairment of CVR is one of the risk factors for new ischemic lesions after both CAS and CEA, there are some differences in the characteristics of postintervention ischemic lesions between CAS and CEA. Subanalysis in a randomized trial18 showed that patients treated with CAS had about ninefold more new ischemic lesions and the areas of the individual lesions were smaller than in CEA-treated patients. As a result, there was no significant difference in total lesion volume per patient between the CEA and CAS groups. These findings suggest that smaller thrombi are more likely to cause small infarction following CAS and larger thrombi are more likely to cause large infarction after CEA. Further investigation is needed to explore the difference in mechanisms responsible for ischemic lesions induced by CAS and CEA.

MR imaging is one of the emerging modalities for evaluation of CBF because of its non-invasive nature. CO2 blood oxygen level-dependent MR imaging has been reported to be useful for the detection of impaired CVR in patients with carotid stenosis.19 Because the application of a vasoactive agent such as CO2 or acetazolamide induces a response dependent on blood oxygen level, this method could be useful for monitoring CVR. With CO2 blood oxygen level-dependent MR imaging, patients who developed new peri-interventional infarcts during CAS or CEA had a greater reduction in CVR of the ipsilateral MCA territory before treatment than those who did not develop new infarcts. However, that study only reported the value of CVR relative to the contralateral hemisphere. Because patients with severe carotid steno-occlusive disease often have bilateral lesions, this method might not have accurately evaluated CVR on the ipsilateral side.

Previous studies have also reported a relationship between preprocedural imaging characteristics of carotid plaque and new ischemic lesions after CAS.20–22 MR imaging is reported to be useful for evaluation of plaque characteristics. We previously demonstrated a relationship between MR signal intensity of plaque using the black-blood MR imaging technique and histopathological findings of plaque specimens obtained after CEA.13 In that study, a higher signal intensity on T1 weighted imaging (T1WI) was found in soft plaque than in non-soft plaque. In addition, another study showed that new ipsilateral ischemic lesions after CAS were more common in those who had higher signal intensity plaque on T1WI.22 However, in our study, there was no significant difference in relative T1 signal intensity of plaque between the DWI-positive and DWI-negative groups. This result might reflect our strategy that CEA should be the first treatment chosen for patients with carotid stenosis.

The association between older age and increased risk of ischemic events following CAS compared with CEA has been seen in many multicenter randomized controlled trials.3 ,23 ,24 The underlying mechanisms of increasing risk with CAS in elderly patients might include vascular tortuosity, severe calcification, or unstable plaque, which may increase the risk of new thromboemboli during the procedure. However, older age was also shown to be associated with lower CVR in patients with severe carotid stenosis.25 Nine hundred and sixteen studies of CBF evaluated using xenon CT showed that CVR, measured by the CBF response to acetazolamide, was significantly lower in patients aged >70 years with severe carotid stenosis than in younger patients. Reduced CVR is a marker of impaired cerebral collateral circulation, and older patients may therefore have poor collateral circulations. Combined with the results of this study, the higher incidence of stroke following CAS in elderly patients could be explained by their lower CVR.

CVR impairment has also been reported to be strongly associated with the future risk of ischemic events in carotid steno-occlusive disease. A meta-analysis of 13 studies, which included a total of 1061 independent CVR tests in 991 patients, showed a positive association between lower CVR and future cerebral ischemic events (OR=3.86).8 Furthermore, impaired CVR is also a known risk factor for hyperperfusion syndrome after CAS.17 These findings suggest that it may be important to measure CVR before CAS in patients with carotid steno-occlusive disease.

This study has several limitations. First, the number of patients in our study was relatively small, and the study was performed retrospectively. Prospective studies are needed. Second, the clinical availability of SPECT is limited, since SPECT is technically complex, has a higher cost, and a risk of radiation exposure. Thus, SPECT is not suitable for repeated observations, which might provide information about the periprocedural time course of CVR. Third, we used many types of device for CAS. The availability of devices was limited depending on when CAS was performed, and this might have influenced the clinical results. Finally, this study was not randomized. Thus, there was a bias for the selection of candidates for CAS. As mentioned above, we tend to select CAS for those patients with relatively non-soft plaque, as evaluated by preintervention MR imaging. Therefore, our results might not apply to CAS in patients with vulnerable plaque.


The patients who developed new ischemic lesions after CAS had significantly lower pretreatment CVR. QSPECT, used for preprocedural evaluation of CVR, could be a valuable tool for the prediction of new ischemic lesions after CAS.


View Abstract


  • Contributors MK: conception and design, analysis, and drafting of the article, KY: revising the article critically for important intellectual content, YK, NS, ON, TS, MC, and AH: interpretation of data, SY and SM: final approval of the version to be published.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Institutional review board and local ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.