Quantitative magnetic resonance angiography as a potential

CLINICAL ARTICLEJ Neurosurg 128:1006–1014, 2018

Cerebral hyperperfusion syndrome (CHS) is de-fined as a major increase in ipsilateral cerebral blood flow (CBF) relative to metabolic needs after

the repair of carotid artery stenosis.47 Poor autoregulatory capacities in the brain territories that gradually adapt to lower post-stenosis pressures are important factors that may lead to CHS,46,47 which is a recognized complication of carotid endarterectomy (CEA).29 The reported incidence

of CHS varies from 0.2% to 18.9%.13,22,47 Because CHS is associated with significant morbidity and mortality, knowl-edge of the predictive risk factors is crucial. However, data on the clinical, metabolic, and radiological predictors are highly variable. A recent prospective study identified dys-lipidemia, high diastolic blood pressure, and nonelective CEA as potential risk factors for CHS.29 Malondialdehyde-modified low-density lipoprotein, a biochemical marker of

ABBREVIATIONS ASA = acetylsalicylic acid; AUC = area under the receiver operating characteristic curve; CBF = cerebral blood flow; CEA = carotid endarterectomy; CHS = cerebral hyperperfusion syndrome; ICA = internal carotid artery; MCA = middle cerebral artery; mRS = modified Rankin Scale; NOVA = noninvasive optimal vessel analy-sis; QMRA = quantitative phase-contrast MR angiography; TCD = transcranial Doppler.SUBMITTED April 22, 2016. ACCEPTED November 7, 2016.INCLUDE WHEN CITING Published online April 14, 2017; DOI: 10.3171/2016.11.JNS161033.

Quantitative magnetic resonance angiography as a potential predictor for cerebral hyperperfusion syndrome: a preliminary studyLukas Andereggen, MD,1,4 Sepideh Amin-Hanjani, MD,5 Marwan El-Koussy, MD,2 Rajeev K. Verma, MD,2 Kenya Yuki, MD, PhD,4 Daniel Schoeni, MD,1 Kety Hsieh, MD,2 Jan Gralla, MD,2 Gerhard Schroth, MD,2 Juergen Beck, MD,1 Andreas Raabe, MD,1 Marcel Arnold, MD,3 Michael Reinert, MD,1,6 and Robert H. Andres, MD1

Departments of 1Neurosurgery, 2Neuroradiology, and 3Neurology, University Hospital of Bern, Inselspital, Bern, Switzerland; 4Department of Neurosurgery and F.M. Kirby Neurobiology Center, Boston Children’s Hospital, Harvard Medical School, Boston, Massachusetts; 5Department of Neurosurgery, University of Illinois at Chicago, Illinois; and 6Department of Neurosurgery, Neurocenter Lugano, Lugano, Switzerland

OBJECTIVE Cerebral hyperperfusion syndrome (CHS) is a rare but devastating complication of carotid endarterec-tomy (CEA). This study sought to determine whether quantitative hemodynamic assessment using MR angiography can stratify CHS risk.METHODS In this prospective trial, patients with internal carotid artery (ICA) stenosis were randomly selected for pre- and postoperative quantitative phase-contrast MR angiography (QMRA). Assessment was standardized according to a protocol and included Doppler/duplex sonography, MRI, and/or CT angiography and QMRA of the intra- and extracranial supplying arteries of the brain. Clinical and radiological data were analyzed to identify CHS risk factors.RESULTS Twenty-five of 153 patients who underwent CEA for ICA stenosis were randomly selected for pre- and post-operative QMRA. QMRA data showed a 2.2-fold postoperative increase in blood flow in the operated ICA (p < 0.001) and a 1.3-fold increase in the ipsilateral middle cerebral artery (MCA) (p = 0.01). Four patients had clinically manifested CHS. The mean flow increases in the patients with CHS were significantly higher than in the patients without CHS, both in the ICA and MCA (p < 0.001). Female sex and a low preoperative diastolic blood pressure were the clearest clinical risk factors for CHS, whereas the flow differences and absolute postoperative flow values in the ipsilateral ICA and MCA were identified as potential radiological predictors for CHS.CONCLUSIONS Cerebral blood flow in the ipsilateral ICA and MCA as assessed by QMRA significantly increased after CEA. Higher mean flow differences in ICA and MCA were associated with the development of CHS. QMRA might have the potential to become a noninvasive, operator-independent screening tool for identifying patients at risk for CHS.https://thejns.org/doi/abs/10.3171/2016.11.JNS161033KEY WORDS cerebral blood flow; carotid endarterectomy; hyperperfusion syndrome; quantitative phase-contrast MR angiography; transcranial Doppler; vascular disorders

J Neurosurg Volume 128 • April 20181006 ©AANS 2018, except where prohibited by US copyright law

Unauthenticated | Downloaded 11/23/21 05:07 AM UTC

Quantitative magnetic resonance angiography in CEA

J Neurosurg Volume 128 • April 2018 1007

oxidative damage, reportedly correlates with the develop-ment of CHS after CEA,44 whereas the role of radiological predictors remains undetermined.11,23,28,30 This prospective trial investigated whether quantitative MRI-based hemo-dynamic assessment can stratify CHS risk.

MethodsPatient Selection and Inclusion Criteria

Patients were scheduled for CEA because of symptom-atic (≥ 50%) or asymptomatic (≥ 70%) carotid stenosis. The degree of stenosis was measured according to the North American Symptomatic Carotid Endarterectomy Trial criteria.1

Following the introduction of quantitative phase-con-trast MR angiography (QMRA) at our institution in No-vember 2011, we prospectively assessed and collected data from patients scheduled to undergo CEA between Novem-ber 2011 and December 2013. Patients were randomly se-lected to receive MRI with or without QMRA before and after CEA (further details on the criteria for enrollment are provided in the Supplemental Materials). A stroke team was responsible for the admission and medical man-agement of all patients. Preoperative assessment consisted of neurological and physical examinations, assessment of stroke severity using the National Institutes of Health Stroke Scale, routine blood analysis, 12-lead electrocar-diography, and brain MRI. Assessment was standardized according to a protocol and included duplex sonography, MRI, and/or CTA and QMRA of the intra- and extracra-nial arteries.

Patient characteristics, including age, sex, neurologi-cal and cardiovascular history, and vascular risk factors, are summarized in Table 1. The prospective data registry was approved by the Ethics Committee of the Canton of Bern, which is part of the Swiss Ethics Committee on re-search involving humans. This study was performed in ac-cordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Written, informed consent was obtained from all participants. The quantitative results of the preoperative and postoperative QMRA examinations were blinded until the completion of data entry into the database by an independent investigator.

Exclusion CriteriaPatients who did not tolerate the first preoperative or

postoperative QMRA examinations were excluded. Pa-tients in whom MR flow studies were not performed 1–2 days before and within 2 days after endarterectomy were also excluded. Contralateral carotid occlusion was not an exclusion criterion (see Supplemental Materials).

RandomizationFor patients who fulfilled the study inclusion criteria,

informed consent was obtained and randomization was performed on site by the treating neurosurgeon using sealed, sequentially numbered envelopes containing the allocation information. The QMRA data were blinded un-til the completion of data entry into the database by an independent investigator.

Transcranial Doppler ProtocolTranscranial Doppler (TCD) sonography (Acuson

S2000, Siemens) with a linear-array transducer (9 MHz) was used for the extracranial examination, and a low-frequency phased-array transducer (2 MHz) was used for

TABLE 1. Clinical and radiological predictors for CHS in the QMRA population

Characteristic CHS Non-CHS p Value

Clinical characteristic No. of patients 4 (16) 21 (84) Mean age ± SD, years 73 ± 8.5 69 ± 8.9 0.48 Female sex 3 (75) 4 (19) 0.05 Hypertension according to

AHA3 (75) 17 (81) 1

Diabetes mellitus 2 (50) 6 (29) 0.57 Dyslipidemia 3 (75) 17 (81) 1 Smoking 4 (100) 13 (62) 0.27 Obesity* 1 (25) 4 (19) 1 Mean BMI ± SD 26.6 ± 6.3 26.6 ± 3.7 0.99 Coronary artery disease 3 (75) 6 (29) 0.12 Peripheral vascular disease 1 (25) 3 (14) 0.53 Positive family history of

cerebrovascular disease1 (25) 4 (19) 1

TIA as presenting symptom 2 (50) 15 (71) 0.57 Stroke as presenting

symptom2 (50) 6 (29) 0.57

Contralateral occlusion/stenosis

2 (50) 5 (24) 0.55

Mean baseline preop sys-tolic BP ± SD, mm Hg

151 ± 22.3 155.1 ± 27.8 0.79

Mean baseline preop dia-stolic BP ± SD, mm Hg

59.5 ± 6.4 76.6 ± 2.9 0.02

Mean baseline postop systolic BP, mm Hg

132.5 ± 27.8 132.3 ± 5.1 0.99

Mean baseline postop dia-stolic BP ± SD, mm Hg

64.3 ± 18.2 61.3 ± 2.1 0.62

Shunt used 1 (25) 2 (10) 0.42 Emergency 0 (0) 5 (24) 0.55Radiological characteristics† Total cases, n (%) 4 (16) 21 (84) Flow differences ICA

ipsilateral413 ± 99 124 ± 17 <0.001

Flow differences MCA ipsilateral

133 ± 45 18 ± 9 <0.001

Flow ICA ipsilateral preop 132 ± 49 156 ± 22 0.66 Flow MCA ipsilateral preop 131 ± 12 132 ± 6 0.93 Flow ICA ipsilateral postop 546 ± 89 281 ± 19 <0.001 Flow MCA ipsilateral postop 264 ± 38 152 ± 11 0.001

AHA = American Heart Association; BMI = body mass index; BP = blood pres-sure; TIA = transient ischemic attack.Values are presented as the number of patients (%) unless indicated other-wise.* Defined as a body mass index > 30 kg/m2.† Values given as mean ± SEM.


https://thejns.org/doi/suppl/10.3171/2016.11.JNS161033


L. Andereggen et al.

J Neurosurg Volume 128 • April 20181008

transtemporal insonation in the axial plane. The mean flow velocities are reported in centimeters per second.

QMRA ProtocolQMRA was performed on a 3-T MRI machine (Mag-

netom Verio, Siemens) equipped with a 12-channel head coil and a 4-channel neck coil. Quantification of blood flow using gated fast 2D phase-contrast sequencing was done using noninvasive optimal vessel analysis (NOVA) software (VasSol) on a separate workstation (see Supple-mental Materials).3,5 This protocol has been routinely used for the evaluation of patients with cerebrovascular diseas-es. NOVA added 25–40 minutes to the MRI scan time, depending on the technical experience of the operator and the complexity of the case. Blood pressure and end-tidal partial pressure of carbon dioxide during the QMRA stud-ies were recorded.

SurgeryPatients were treated with antiplatelet agents (acetyl-

salicylic acid [ASA; Bayer], clopidogrel [Sanofi-Aventis], or both). Warfarin reversal was attempted in patients on a regimen of anticoagulation therapy, and they received intravenous heparin instead. Patients were placed un-der general anesthesia. Neuromonitoring was performed throughout the procedure using intraoperative TCD ul-trasonography and somatosensory evoked potentials. Endarterectomy was completed under a neurosurgical microscope (OPMI Pentero, Zeiss), and the arteriotomy was sutured with a continuous 6-0 monofilament suture (6-0 polypropylene, Covidien Plc.) as previously described (see Supplemental Materials).41 Six hours after surgery, 300 mg ASA was administered intravenously, followed by daily 100–300 mg ASA orally. For patients receiving clopidogrel preoperatively with or without ASA for car-diac reasons, therapy was resumed 1 day postoperatively.

Postoperative CHS AssessmentCHS was defined as a postoperative hemorrhage or sei-

zure, deterioration of the consciousness level not resulting from metabolic disorders or thromboembolism, and/or se-vere unilateral headache unresponsive to conventional an-algesic medication (e.g., paracetamol, metamizol, or mor-phine).13,21,29,47 The definition of CHS excluded infarction.

Statistical AnalysisData were analyzed using IBM SPSS statistical soft-

ware V21.0 (IBM) and GraphPad Prism V6.03 (GraphPad Software). Values are expressed as the mean ± SEM or ± SD, as noted. Differences between the normally distrib-uted data of 2 groups were analyzed using the paired t-test, and the Wilcoxon signed-rank test was used to analyze nonparametric data. Due to the small number of patients, predictors of CHS were assessed using contingency tables, and differences between the CHS and non-CHS groups were calculated using the chi-square test or Fisher exact test for categorical variables and Student t-test or Mann-Whitney U-test for continuous variables. A significance level of p ≤ 0.05 was applied.

ResultsPatient Demographics and Surgical Variables

Between November 2011 and December 2013, 25 of 153 patients with CEA for internal carotid artery (ICA) stenosis were randomly selected for inclusion in this pro-spective trial. Patient demographics are summarized in Table 2. There were no statistically significant differences between the QMRA and non-QMRA groups with regard to the baseline characteristics. In the QMRA cohort, the mean age ± SD of the patients was 70 ± 8.8 years (range 50–81 years), and 14 (56%) patients were ≥ 70 years old. Males comprised 72% of the study sample. Cardiovascu-lar risk factors were common: hypertension (80% of pa-tients), diabetes mellitus (32%), dyslipidemia (80%), active smoking (68%), peripheral artery occlusive disease (13%), and positive family history of cerebrovascular diseases (20%). Obesity (body mass index ≥ 30 kg/m2) was present in 20% of patients. The majority of the patients presented with high-grade stenosis: 68% with 70%–90% stenosis and 32% with > 90% stenosis.

Two patients with a mean age ± SD of 75 ± 4.2 years fulfilled the aforementioned exclusion criteria due to non-compliance during the QMRA examination in the periop-erative setting. In one of these patients, the preoperative assessment of stroke using MRI and QMRA had to be ter-minated earlier due motion artifacts. The other patient did not tolerate the extended period of the scan due to severe back pain.

TABLE 2. Patient characteristics at baseline

Characteristics w/ QMRA w/o QMRA p Value

No. of patients 25 (16) 128 (84)Mean age ± SD, years 70 ± 8.8 71 ± 9.0 0.736Male sex 18 (72) 94 (73) 1Hypertension according to AHA 20 (80) 90 (87) 0.529Diabetes mellitus 8 (32) 26 (26) 0.618Dyslipidemia 20 (80) 65 (68) 0.327Active smoking 17 (68) 58 (59) 0.495Obesity* 5 (20) 32 (33) 0.328Peripheral artery occlusive disease 3 (13) 83 (15) 1Positive family history of cerebro-

vascular disease5 (20) 14 (22) 1

TIA &/or stroke symptomatic 25 (100) 113 (88) 0.473Tandem stenosis 0 (0) 8 (8) 0.605Shunt used 3 (12) 4 (3) 0.087Intraop plaque ulcerations 24 (96) 127 (100) 0.164Symptomatic stenosis 24 (96) 113 (88) 0.473Right-side stenosis 12 (48) 59 (47) 1Grade of stenosis 50%–70% 0 (0) 10 (8) 0.233 70%–90% 17 (68) 69 (54) 0.233 >90% 8 (32) 49 (38) 0.233

Values are presented as the number of patients (%) unless indicated other-wise.* Defined as a body mass index > 30 kg/m2.







Clinical Predictors of CHSThe patients’ clinical characteristics and risk factors for

CHS in the analyzed QMRA population are summarized in Table 1. CHS occurred in 4 patients (16%) within 3 ± 2.4 days postoperatively (mean ± SD) and within 1 ± 1.2 days post-QMRA assessment, which was characterized by tonic-clonic seizures in 2 patients, refractory headache in 1 patient, and deterioration of the level of consciousness in another patient. No intracerebral hemorrhage was noted, whereas localized edema on FLAIR/T2-weighted MRI was observed in 25% of patients. Complete recovery of the patients who were symptomatic with CHS was observed in all of these patients, and the functional outcome did not differ from the patients in the non-CHS group. Namely, a good functional outcome (i.e., modified Rankin Scale [mRS] score ≤ 2) at discharge was noted in 3 (75%) pa-tients in the CHS group compared with 17 (81%) patients in the non-CHS group (p = 0.99). At the last follow-up, a good functional outcome (mRS score ≤ 2) was noted in 3 (75%) patients in the CHS group compared with 19 (91%) patients in the non-CHS group (p = 0.42). Women were overrepresented in the CHS group (p = 0.053). Baseline preoperative diastolic pressure was significantly lower in the patients in the CHS group than in the non-CHS group (p = 0.02). The other clinical risk factors assessed were not significantly different between the cohorts.

QMRA and TCD Blood Flow Changes After CEAThe QMRA and TCD results are summarized in Tables

3 and 4. Blood pressure and end-tidal partial pressure of

carbon dioxide during QMRA examination remained sta-ble in all patients.

For the QMRA values, CEA significantly increased blood flow in both the intra- and extracranial vessel seg-ments on the operated side (mean difference ± SEM 175 ± 31 ml/min for the ICA, p < 0.001, Fig. 1A; and 37 ± 13 ml/min for the ipsilateral middle cerebral artery (MCA), p = 0.01, Fig. 1B). As can be seen in Table 3, in patients with CHS the perioperative flow differences (postopera-tive flow - preoperative flow) were 3.3-fold higher than in non-CHS patients in the surgically treated ICA (413 ± 99 ml/min vs 124 ± 17 ml/min, p < 0.001) and 7.4-fold higher in the ipsilateral MCA (133 ± 45 ml/min vs 18.9 ± 9 ml/min, p < 0.001, Fig. 1C). Furthermore, patients with mani-fest CHS had 1.9-fold higher postoperative absolute flow values in the ipsilateral ICA and 1.8-fold higher absolute flow values in the ipsilateral MCA than patients without CHS (546 ± 89 ml/minute vs 280 ± 20 ml/minute, p < 0.001; and 264 ± 39 ml/minute vs 152 ± 11 ml/minute, p = 0.001, respectively).

For the TCD values, CEA significantly altered the flow velocity in both the intra- and extracranial vessel seg-ments on the operated side (Table 4): velocity increased in the ICA (mean difference ± SEM) by 156 ± 16 cm/sec (p < 0.001) and decreased in the ipsilateral MCA by -14 ± 7 cm/sec (p = 0.046). In patients with CHS, the perioperative velocity differences were 1.2-fold higher than in non-CHS patients on the operated ICA (176 ± 24 cm/sec vs 152 ± 19 cm/sec, p = 0.60) and 11.6-fold greater in the ipsilateral MCA, respectively (-52 ± 17 ml/min vs -5 ± 5 ml/min, p = 0.02). Patients with manifest CHS had 2.2-fold higher postoperative absolute velocities in the ipsilateral ICA and 2.3-fold higher velocities in the ipsilateral MCA than pa-

TABLE 3. QMRA-assessed blood flow values before and after CEA

Characteristic Preop* Postop* Difference* p Value

Ipsilateral All patients ICA 152 ± 19 326 ± 30 175 ± 31 <0.001 MCA 133 ± 5 170 ± 14 37 ± 13 0.01 Non-CHS ICA 156 ± 22 280 ± 20 124 ± 17 <0.001 MCA 133 ± 6 152 ± 11 18 ± 9 0.06 CHS ICA 132 ± 49 546 ± 89 413 ± 99 0.03 MCA 131 ± 12 264 ± 39 133 ± 45 0.06Contralateral All patients ICA 269 ± 15 246 ± 17 23 ± 17 0.21 MCA 141 ± 6 159 ± 8 18 ± 7 0.02 Non-CHS ICA 267 ± 17 246 ± 18 73 ± 18 0.26 MCA 141 ± 7 147 ± 8 6 ± 6 0.28 CHS ICA 290 ± 27 251 ± 62 40 ± 89 0.73 MCA 140 ± 13 220 ± 7 79 ± 12 0.01

* Values are presented as the mean ± SEM in milliliters per minute.

TABLE 4. TCD-assessed blood flow values before and after CEA

Characteristic Preop* Postop* Difference* p Value

Ipsilateral All patients ICA 199 ± 17 42 ± 4 156 ± 16 <0.001 MCA 50 ± 6 64 ± 8 −14 ± 7 0.046 Non-CHS ICA 188 ± 19 35 ± 2 152 ± 19 <0.001 MCA 47 ± 6 51 ± 4 −5 ± 5 0.375 CHS ICA 253 ± 27 77 ± 3 176 ± 24 0.019 MCA 61 ± 9 113 ± 24 −52 ± 17 0.055Contralateral All patients ICA 87 ± 20 78 ± 17 8 ± 12 0.505 MCA 53 ± 4 66 ± 7 −13 ± 7 0.067 Non-CHS ICA 79 ± 16 72 ± 19 7 ± 10 0.493 MCA 50 ± 4 53 ± 3 −3 ± 3 0.343 CHS ICA 126 ± 113 111 ± 40 15 ± 73 0.871 MCA 63 ± 8 113 ± 21 −50 ± 23 0.117

* Values are presented as the mean ± SEM in centimeters per second.




tients without CHS (77 ± 3 cm/sec vs 35 ± 2 cm/sec, p < 0.001 and 113 ± 24 cm/sec vs 51 ± 4 cm/sec, p < 0.001, respectively). Bland-Altman plots comparing sonography and QMRA indicated only moderate agreement between the 2 methods for the extracranial vessels compared with intracranial vessels (Fig. 2).

QMRA Predictors of CHSFlow differences in the ipsilateral ICA and MCA (p <

0.001 for both) and the postoperative absolute flow values in the ICA (p < 0.001) and MCA (p = 0.001) were sig-nificantly higher in patients with CHS than in non-CHS patients (Table 1).

The QMRA thresholds for the discrimination of CHS and non-CHS were estimated to be 3.49 for the mean flow value ratio in the ICA (postoperative flow value/pre-operative flow value) (area under the receiver operating characteristic curve [AUC] 0.8; sensitivity 75%; specific-ity 84%; likelihood ratio 4.75) (Fig. 3A) and 1.53 for the mean flow value ratio in the MCA (AUC 0.94; sensitivity 100%; specificity 90%; likelihood ratio 10) (Fig. 3B). Us-ing QMRA compared with TCD assessment, MCA ratio (p = 0.01) tended to be more efficient than the ICA ratio for predicting CHS (p = 0.06). In the TCD assessment, the thresholds for the discrimination of CHS and non-CHS were estimated to be 0.26 for mean flow value ratio in the ICA (AUC 0.9; sensitivity 100%; specificity 80%; likeli-hood ratio 5.0) and 1.59 for the mean flow value ratio in the MCA (AUC 0.83; sensitivity 75%; specificity 80%; likelihood ratio 3.75) (Fig. 3C and D).

DiscussionTo the best of our knowledge, this is the first study to

characterize CHS in a subgroup of patients before and af-ter CEA using QMRA measurements. CEA is one of the most thoroughly investigated surgical procedures,1,12,26,41,48 and QMRA has been validated for the assessment of ste-nosis and measurement of flow impairment.3,5,8,15,27,40 Our data show that QMRA reflected the expected CBF chang-es before and after CEA in all patients. We found that the higher mean flow increases in the ICA and MCA were associated with the development of CHS.

Multiple studies identified radiological variables that potentially predict CHS. Newman et al. assessed mean and peak MCA velocity changes on TCD before and after CEA, but found no significant association between flow velocity increases on TCD and increased risk for CHS,33 in contrast to the findings of an older study by Dalman et al.14

Although the TCD technique is widely available and 3D reconstructions of the vascular territory can be depict-ed using duplex sonography,49 the results are highly op-erator dependent and quantitative measurement of the ef-fective blood flow is limited. In these cases, near-infrared spectroscopy could potentially be an alternative method to use when the TCD signal cannot be obtained.39 Other technologies, like xenon-CT or SPECT, have been used to quantify effective blood flow.20,34,38 However, these tech-niques are complex, time consuming, and of limited use for determining blood flow measurements in single ves-sels. Iwata et al. showed that SPECT and TCD before and after carotid artery stenting predicted CHS occurrence in 14.1% of patients.22 SPECT was assessed by Kaku et al., who showed that patients with < 20% preinterventional va-soreactivity to acetazolamide were more likely to develop CHS.23 However, SPECT requires the use of gamma rays and the delivery of a gamma-emitting radionucleotide to patients, which is invasive and associated with the risks of

FIG. 1. Blood flow values of the CEA patients as assessed by QMRA. Nonpatch CEA significantly increased blood flow in the operated ICA and ipsilateral MCA in all patients. A: A significant difference in blood flow was evident in the ipsilateral ICA in patients with or without CHS. Additionally, the postoperative flow values in the CHS patients were significantly higher than in non-CHS patients. B: MCA flow velocities were significantly higher in CHS patients than in non-CHS patients. C: Perioperative flow differences in CHS patients were significantly higher in both the ipsilateral ICA and MCA than in non-CHS patients. Data are given as the mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001. ns = not significant; post = postoperative; pre = preoperative.




radiation exposure. Perfusion-weighted MRI, on the other hand, does not permit the quantification of absolute CBF differences.24 Chang et al. demonstrated that perfusion CT, specifically regional cerebral blood volume and time-to-peak values in combination with clinical parameters, im-proves prediction in patients with unilateral high-grade ICA stenosis who are at risk for developing CHS.11 Intracranial flow values, e.g., in the MCA, include collateral perfusion from the circle of Willis and extracranial branches proxi-mal to the affected brain parenchyma. Anatomical varia-tions in the circle of Willis and patient age influence flow in the ICA and basilar artery,4,9,18 and, as a result, assess-ment of intracranial flow values by QMRA might better re-flect the effective blood flow increase attained in the brain territory after CEA and potentially better predict CHS risk. We observed a tendency for the mean flow values ra-tios in the MCA to be more efficient at predicting CHS than the ICA ratio (Fig. 3). Concordantly, we also found significantly higher postoperative blood flow values in pa-tients with CHS than in non-CHS patients (p < 0.001). The ultimate goal of the QMRA-assessed blood flow increase is to provide a screening tool to perioperatively determine those patients who are at risk for CHS so that they may be more closely monitored postoperatively. In this pilot study, the QMRA values had to first be assessed in non-CHS and CHS patients, and we did not adapt the care plans for the CHS patients or apply any additional preventive measures

for these patients (e.g., more rigorous blood pressure limits over an extended period). Nevertheless, the thresholds es-tablished by QMRA for the discrimination between CHS and non-CHS, particularly for the mean flow value ratios in the MCA, might provide threshold values that could be used for indicating patients at risk for whom closer moni-toring in the perioperative setting is warranted to possibly improve patient outcome. While treatment strategies are primarily directed toward the reduction of blood pressure, the early institution of adequate treatment is of paramount importance.47 Specifically, adequately lowering blood pres-sure using labetalol and clonidine over an extended period of time until cerebral autoregulation is restored,25,16,32, 35,45 treatment of cerebral edema using corticosteroids and bar-biturates,37 and anticonvulsant therapy have been proposed to form the basis for CHS therapy.31

Furthermore, our results show that apart from baseline low diastolic blood pressure, no significant clinical predic-tors for CHS could be identified. This finding contrasts with a prospective study by Maas et al., where preoperative high diastolic blood pressure, a brief time interval between ischemic symptoms, and nonelective endarterectomy were the clearest preoperative risk factors for CHS.29 In the postoperative period, poor control of blood pressure has been associated with the development of intracranial hem-orrhage in patients with CHS.47 Whereas in our study the majority of the patients with CHS were women (p = 0.053),

FIG. 2. Bland-Altman plots comparing sonography and QMRA. The narrower ± 1.96 (± SD) limits (dotted lines) indicate only moderate agreement for extracranial vessels (A and B) compared with intracranial vessels (C and D). There is a slight tendency toward better agreement between the 2 methods for smaller values. SD LOA = standard deviation with limits of agreement.




Ascher et al. found that female sex was not predictive for CHS risk.7 In contrast, Hines et al. reported a significant association between CHS and female sex.19 The influence of contralateral stenosis or the use of an intraoperative shunt remains controversial.25,29 The diversity of study results comparing the clinical predictors for CHS might arise because the definition of CHS includes a variety of clinical presentations.13 Furthermore, there is no defined time period for the development of CHS after CEA. CHS has been reported to occur from the immediate postopera-tive period until 1 month later, with most patients develop-ing CHS around postoperative Day 5.36 Since our protocol included postoperative QMRA assessment within the first 2 days after CEA, we were able to register high flow values during the time before the clinical manifestations of CHS are usually noted.

In summary, QMRA enabled the noninvasive detec-

tion of significant blood flow increases in the treated ICA and ipsilateral MCA after CEA, with significantly higher perioperative flow differences seen in patients with clini-cally manifest CHS than in patients without CHS. Fur-thermore, QMRA documents a successful intervention by quantifying flow improvement. These findings supports the revascularization strategy to increase CBF, especially in the more distal vessels, which has been associated with greater cognitive improvement in attention and executive functioning following CEA.17 In addition, QMRA pro-vides a solid postsurgical baseline for noninvasive follow-up.10 Assessment of distal flow impairment using QMRA might further stratify the risk for subsequent stroke in asymptomatic but potentially high-risk patients, with im-plications for more aggressive interventions or medical therapies, as has been shown in patients with vertebrobasi-lar diseases.5,6

FIG. 3. Receiver operating characteristic curves for the discrimination of CHS and non-CHS patients assessed by QMRA. QMRA thresholds for dis-crimination between CHS and non-CHS patients were estimated to be approximately 3.49 for the mean flow value ratio in the ICA (AUC 0.8; sensitivity 75%; specificity 84%; likelihood ratio 4.75) (A) and 1.53 for the mean flow value ratio in the MCA (AUC 0.94; sensitivity 100%; specificity 90%; likelihood ratio 10) (B). The MCA ratio (p = 0.01) tended to be more efficient at predicting CHS than ICA ratio (p = 0.06). TCD thresholds for discrimination between CHS and non-CHS were estimated to be 0.26 for the mean flow value ratio in the ICA (AUC 0.89; sensitivity 100%; specificity 80%; likelihood ratio 5.0) (C) and 1.59 for the mean flow value ratio in the MCA (AUC 0.83; sensitivity 75%; specificity 80%; likelihood ratio 3.75) (D).




Our study has certain limitations. The number of pa-tients who underwent prospective blood flow assessment before and after CEA is relatively small, and the number of patients who developed CHS is even smaller. Therefore, assessment of the independent predictors of CHS using multiple logistic regressions is not feasible. Not all patients who underwent CEA were studied, but the fact that the baseline characteristics of the study cohort were compa-rable to those of the patients who did not undergo QMRA suggests that there was no major selection bias. Addition-ally, bias in outcome determination was avoided by pro-spective data acquisition with blinding of the QMRA data. However, the potential drawbacks of QMRA in this setting are the sensitivity of the technique to patient movement, which can be more pronounced during the postoperative recovery of older, less compliant patients, and the related costs. Given that NOVA added 25–40 minutes to the MRI scan time, an altered mental status or comorbidity is often confounding on QMRA examination. The rate of incom-plete QMRA examinations in our study likely reflects the strict application of the quality control criteria, as image quality can be severely affected by patient movements.2,42,43 Nevertheless, this study adds to existing data by demon-strating the ability of QMRA to predict CHS after CEA.

ConclusionsIncrease in QMRA-assessed blood flow after CEA was

significant and may provide a noninvasive, operator-inde-pendent screening tool to identify patients at risk for CHS.

AcknowledgmentsWe are grateful for the support of the Swiss National Science

Foundation (PBBEB-146099 and -155299 awarded to L.A.). We wish to thank Dr. Kush Kapur, Clinical Research Center, Boston Children’s Hospital, for providing statistical advice, and the radiol-ogy technologists at Bern University Hospital for performing the QMRA examinations.

References 1. Alamowitch S, Eliasziw M, Algra A, Meldrum H, Barnett

HJ: Risk, causes, and prevention of ischaemic stroke in elder-ly patients with symptomatic internal-carotid-artery stenosis. Lancet 357:1154–1160, 2001

2. Ali SH, Modic ME, Mahmoud SY, Jones SE: Reducing clini-cal MRI motion degradation using a prescan patient informa-tion pamphlet. AJR Am J Roentgenol 200:630–634, 2013

3. Amin-Hanjani S, Alaraj A, Calderon-Arnulphi M, Aletich VA, Thulborn KR, Charbel FT: Detection of intracranial in-stent restenosis using quantitative magnetic resonance angiography. Stroke 41:2534–2538, 2010

4. Amin-Hanjani S, Du X, Pandey DK, Thulborn KR, Charbel FT: Effect of age and vascular anatomy on blood flow in ma-jor cerebral vessels. J Cereb Blood Flow Metab 35:312–318, 2015

5. Amin-Hanjani S, Du X, Zhao M, Walsh K, Malisch TW, Charbel FT: Use of quantitative magnetic resonance angiog-raphy to stratify stroke risk in symptomatic vertebrobasilar disease. Stroke 36:1140–1145, 2005

6. Amin-Hanjani S, Pandey DK, Rose-Finnell L, Du X, Rich-ardson D, Thulborn KR, et al: Effect of hemodynamics on stroke risk in symptomatic atherosclerotic vertebrobasilar occlusive disease. JAMA Neurol 73:178–185, 2016

7. Ascher E, Markevich N, Schutzer RW, Kallakuri S, Jacob

T, Hingorani AP: Cerebral hyperperfusion syndrome after carotid endarterectomy: predictive factors and hemodynamic changes. J Vasc Surg 37:769–777, 2003

8. Bauer AM, Amin-Hanjani S, Alaraj A, Charbel FT: Quantita-tive magnetic resonance angiography in the evaluation of the subclavian steal syndrome: report of 5 patients. J Neuroim-aging 19:250–252, 2009

9. Buijs PC, Krabbe-Hartkamp MJ, Bakker CJ, de Lange EE, Ramos LM, Breteler MM, et al: Effect of age on cerebral blood flow: measurement with ungated two-dimensional phase-contrast MR angiography in 250 adults. Radiology 209:667–674, 1998

10. Calderon-Arnulphi M, Amin-Hanjani S, Alaraj A, Zhao M, Du X, Ruland S, et al: In vivo evaluation of quantitative MR angiography in a canine carotid artery stenosis model. AJNR Am J Neuroradiol 32:1552–1559, 2011

11. Chang CH, Chang TY, Chang YJ, Huang KL, Chin SC, Ryu SJ, et al: The role of perfusion computed tomography in the prediction of cerebral hyperperfusion syndrome. PLoS One 6:e19886, 2011

12. Cohen DJ, Stolker JM, Wang K, Magnuson EA, Clark WM, Demaerschalk BM, et al: Health-related quality of life after carotid stenting versus carotid endarterectomy: results from CREST (Carotid Revascularization Endarterectomy Versus Stenting Trial). J Am Coll Cardiol 58:1557–1565, 2011

13. Coutts SB, Hill MD, Hu WY: Hyperperfusion syndrome: toward a stricter definition. Neurosurgery 53:1053–1060, 2003

14. Dalman JE, Beenakkers IC, Moll FL, Leusink JA, Ackerstaff RG: Transcranial Doppler monitoring during carotid endar-terectomy helps to identify patients at risk of postoperative hyperperfusion. Eur J Vasc Endovasc Surg 18:222–227, 1999

15. Douglas AF, Christopher S, Amankulor N, Din R, Poullis M, Amin-Hanjani S, et al: Extracranial carotid plaque length and parent vessel diameter significantly affect baseline ipsilateral intracranial blood flow. Neurosurgery 69:767–773, 2011

16. Faivre JM, Freysz M, Coulon C, Jaboeuf R, David M: Post carotid endarterectomy hypertension: clonidine therapy. Anaesth Intensive Care 9:179–180, 1981

17. Ghogawala Z, Amin-Hanjani S, Curran J, Ciarleglio M, Be-renstein A, Stabile L, et al: The effect of carotid endarterec-tomy on cerebral blood flow and cognitive function. J Stroke Cerebrovasc Dis 22:1029–1037, 2013

18. Hendrikse J, van Raamt AF, van der Graaf Y, Mali WP, van der Grond J: Distribution of cerebral blood flow in the circle of Willis. Radiology 235:184–189, 2005

19. Hines GL, Oleske A, Feuerman M: Post-carotid endarterec-tomy hyperperfusion syndrome-is it predictable by lack of cerebral reserve? Ann Vasc Surg 25:502–507, 2011

20. Hirooka R, Ogasawara K, Inoue T, Fujiwara S, Sasaki M, Chi-da K, et al: Simple assessment of cerebral hemodynamics us-ing single-slab 3D time-of-flight MR angiography in patients with cervical internal carotid artery steno-occlusive diseases: comparison with quantitative perfusion single-photon emis-sion CT. AJNR Am J Neuroradiol 30:559–563, 2009

21. Iwata T, Mori T, Miyazaki Y, Tanno Y, Kasakura S, Aoyagi Y: Global oxygen extraction fraction by blood sampling to anticipate cerebral hyperperfusion phenomenon after carotid artery stenting. Neurosurgery 75:546–551, 2014

22. Iwata T, Mori T, Tajiri H, Nakazaki M: Predictors of hyper-perfusion syndrome before and immediately after carotid artery stenting in single-photon emission computed tomogra-phy and transcranial color-coded real-time sonography stud-ies. Neurosurgery 68:649–656, 2011

23. Kaku Y, Yoshimura S, Kokuzawa J: Factors predictive of cerebral hyperperfusion after carotid angioplasty and stent placement. AJNR Am J Neuroradiol 25:1403–1408, 2004

24. Karapanayiotides T, Meuli R, Devuyst G, Piechowski-Jozwiak B, Dewarrat A, Ruchat P, et al: Postcarotid endar-




terectomy hyperperfusion or reperfusion syndrome. Stroke 36:21–26, 2005

25. Kawamata T, Okada Y, Kawashima A, Yoneyama T, Yama-guchi K, Ono Y, et al: Postcarotid endarterectomy cerebral hyperperfusion can be prevented by minimizing intraopera-tive cerebral ischemia and strict postoperative blood pressure control under continuous sedation. Neurosurgery 64:447–454, 2009

26. Lal BK, Beach KW, Roubin GS, Lutsep HL, Moore WS, Malas MB, et al: Restenosis after carotid artery stenting and endarterectomy: a secondary analysis of CREST, a ran-domised controlled trial. Lancet Neurol 11:755–763, 2012

27. Langer DJ, Lefton DR, Ostergren L, Brockington CD, Song J, Niimi Y, et al: Hemispheric revascularization in the setting of carotid occlusion and subclavian steal: a diagnostic and management role for quantitative magnetic resonance angiog-raphy? Neurosurgery 58:528–533, 2006

28. Lin CJ, Chang FC, Tsai FY, Guo WY, Hung SC, Chen DY, et al: Stenotic transverse sinus predisposes to poststenting hy-perperfusion syndrome as evidenced by quantitative analysis of peritherapeutic cerebral circulation time. AJNR Am J Neuroradiol 35:1132–1136, 2014

29. Maas MB, Kwolek CJ, Hirsch JA, Jaff MR, Rordorf GA: Clinical risk predictors for cerebral hyperperfusion syndrome after carotid endarterectomy. J Neurol Neurosurg Psychia-try 84:569–572, 2013

30. Matsubara S, Moroi J, Suzuki A, Sasaki M, Nagata K, Kanno I, et al: Analysis of cerebral perfusion and metabolism as-sessed with positron emission tomography before and after carotid artery stenting. Clinical article. J Neurosurg 111:28–36, 2009

31. Moulakakis KG, Mylonas SN, Sfyroeras GS, Andrikopoulos V: Hyperperfusion syndrome after carotid revascularization. J Vasc Surg 49:1060–1068, 2009

32. Muzzi DA, Black S, Losasso TJ, Cucchiara RF: Labetalol and esmolol in the control of hypertension after intracranial surgery. Anesth Analg 70:68–71, 1990

33. Newman JE, Ali M, Sharpe R, Bown MJ, Sayers RD, Naylor AR: Changes in middle cerebral artery velocity after carotid endarterectomy do not identify patients at high-risk of suffer-ing intracranial haemorrhage or stroke due to hyperperfusion syndrome. Eur J Vasc Endovasc Surg 45:562–571, 2013

34. Niibo T, Ohta H, Yonenaga K, Ikushima I, Miyata S, Takeshi-ma H: Arterial spin-labeled perfusion imaging to predict mismatch in acute ischemic stroke. Stroke 44:2601–2603, 2013

35. Ogasawara K, Mikami C, Inoue T, Ogawa A: Delayed cere-bral hyperperfusion syndrome caused by prolonged impair-ment of cerebrovascular autoregulation after carotid endar-terectomy: case report. Neurosurgery 54:1258–1262, 2004

36. Ogasawara K, Sakai N, Kuroiwa T, Hosoda K, Iihara K, Toyoda K, et al: Intracranial hemorrhage associated with cerebral hyperperfusion syndrome following carotid endar-terectomy and carotid artery stenting: retrospective review of 4494 patients. J Neurosurg 107:1130–1136, 2007

37. Ogasawara K, Yukawa H, Kobayashi M, Mikami C, Konno H, Terasaki K, et al: Prediction and monitoring of cerebral hyperperfusion after carotid endarterectomy by using single-photon emission computerized tomography scanning. J Neu-rosurg 99:504–510, 2003

38. Olivot JM, Mlynash M, Zaharchuk G, Straka M, Bammer R, Schwartz N, et al: Perfusion MRI (Tmax and MTT) correla-tion with xenon CT cerebral blood flow in stroke patients. Neurology 72:1140–1145, 2009

39. Pennekamp CW, Immink RV, den Ruijter HM, Kappelle LJ, Ferrier CM, Bots ML, et al: Near-infrared spectroscopy can predict the onset of cerebral hyperperfusion syndrome after carotid endarterectomy. Cerebrovasc Dis 34:314–321, 2012

40. Prabhakaran S, Warrior L, Wells KR, Jhaveri MD, Chen M,

Lopes DK: The utility of quantitative magnetic resonance angiography in the assessment of intracranial in-stent steno-sis. Stroke 40:991–993, 2009

41. Reinert M, Mono ML, Kuhlen D, Mariani L, Barth A, Beck J, et al: Restenosis after microsurgical non-patch carotid endarterectomy in 586 patients. Acta Neurochir (Wien) 154:423–431, 2012

42. Schültke E, Nanko N, Pinsker M, Katzev M, Sebastian A, Feige B, et al: Improving MRT image quality in patients with movement disorders. Acta Neurochir Suppl 117:13–17, 2013

43. Seeger A, Klose U, Poli S, Kramer U, Ernemann U, Hauser TK: Acute stroke imaging: feasibility and value of MR angi-ography with high spatial and temporal resolution for vessel assessment and perfusion analysis in patients with wake-up stroke. Acad Radiol 22:413–422, 2015

44. Suga Y, Ogasawara K, Saito H, Komoribayashi N, Kobayashi M, Inoue T, et al: Preoperative cerebral hemodynamic im-pairment and reactive oxygen species produced during ca-rotid endarterectomy correlate with development of postoper-ative cerebral hyperperfusion. Stroke 38:2712–2717, 2007

45. Tietjen CS, Hurn PD, Ulatowski JA, Kirsch JR: Treatment modalities for hypertensive patients with intracranial pathol-ogy: options and risks. Crit Care Med 24:311–322, 1996

46. Vagal AS, Leach JL, Fernandez-Ulloa M, Zuccarello M: The acetazolamide challenge: techniques and applications in the evaluation of chronic cerebral ischemia. AJNR Am J Neu-roradiol 30:876–884, 2009

47. van Mook WN, Rennenberg RJ, Schurink GW, van Oosten-brugge RJ, Mess WH, Hofman PA, et al: Cerebral hyperper-fusion syndrome. Lancet Neurol 4:877–888, 2005

48. Voeks JH, Howard G, Roubin GS, Malas MB, Cohen DJ, Sternbergh WC III, et al: Age and outcomes after carotid stenting and endarterectomy: the carotid revascularization endarterectomy versus stenting trial. Stroke 42:3484–3490, 2011

49. Yao J, van Sambeek MR, Dall’Agata A, van Dijk LC, Koza-kova M, Koudstaal PJ, et al: Three-dimensional ultrasound study of carotid arteries before and after endarterectomy; analysis of stenotic lesions and surgical impact on the vessel. Stroke 29:2026–2031, 1998

DisclosuresDr. Amin-Hanjani receives non–study-related research support from VasSol, Inc. Dr. Arnold is a consultant for and receives non–study-related research support from Bayer Schering, BMS, Pfizer, Boehringer, and Covidien.

Author ContributionsConception and design: Reinert, Andereggen, Amin-Hanjani. Acquisition of data: Reinert, Andereggen, Schoeni. Analysis and interpretation of data: Andereggen, Amin-Hanjani. Drafting the article: Reinert, Andereggen, Andres. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Reinert. Statistical analysis: Andereggen, Yuki.

Supplemental Information Online-Only ContentSupplemental material is available with the online version of the article.

Supplemental Materials. https://thejns.org/doi/suppl/10.3171/ 2016.11.JNS161033.

CorrespondenceMichael Reinert, Department of Neurosurgery, Neurocenter Luga-no, Lugano 6930, Switzerland. email: [email protected].




Documents

Quantitative magnetic resonance angiography as a potential