Sage Journals: Discover world-class research

Abstract

Aim

To investigate whether lesions on diffusion-weighted imaging (DWI+) after carotid artery stenting (CAS) or endarterectomy (CEA) might provide a surrogate outcome measure for procedural stroke.

Materials and Methods

Systematic MedLine® database search with selection of all studies published up to the end of 2016 in which DWI scans were obtained before and within seven days after CAS or CEA. The correlation between the underlying log odds of stroke and of DWI+ across all treatment groups (i.e. CAS or CEA groups) from included studies was estimated using a bivariate random effects logistic regression model. Relative risks of DWI+ and stroke in studies comparing CAS vs. CEA were estimated using fixed-effect Mantel-Haenszel models.

Results

We included data of 4871 CAS and 2099 CEA procedures (85 studies). Across all treatment groups (CAS and CEA), the log odds for DWI+ was significantly associated with the log odds for clinically manifest stroke (correlation coefficient 0.61 (95% CI 0.27 to 0.87), p = 0.0012). Across all carotid artery stenting groups, the correlation coefficient was 0.19 (p = 0.074). There were too few CEA groups to reliably estimate a correlation coefficient in this subset alone. In 19 studies comparing CAS vs. CEA, the relative risks (95% confidence intervals) of DWI+ and stroke were 3.83 (3.17–4.63, p < 0.00001) and 2.38 (1.44–3.94, p = 0.0007), respectively.

Discussion

This systematic meta-analysis demonstrates a correlation between the occurrence of silent brain infarcts on diffusion-weighted imaging and the risk of clinically manifest stroke in carotid revascularisation procedures.

Conclusion

Our findings strengthen the evidence base for the use of DWI as a surrogate outcome measure for procedural stroke in carotid revascularisation procedures. Further randomised studies comparing treatment effects on DWI lesions and clinical stroke are needed to fully establish surrogacy.

Keywords

Carotid stenosis stenting endarterectomy stroke ischaemic lesions DWI surrogate marker

Introduction

Atherosclerotic stenosis of the internal carotid artery is a major cause of stroke. Early randomised trials demonstrated that carotid endarterectomy (CEA) reduces stroke risk in patients with symptomatic and asymptomatic carotid stenosis.^1–4 In recent years, carotid angioplasty and carotid artery stenting (CAS) has emerged as an alternative to CEA, avoiding general surgical complications and morbidity associated with neck incision, as well as shortening hospital stay. In a systematic review of 11 randomised trials including 5778 patients with symptomatic carotid stenosis, CAS was associated with a higher risk of procedural stroke or death than CEA (8.2% vs. 5.0%), although the risk difference between CEA and CAS mainly concerned minor strokes which did not lead to disability or death and was seen in patients older than 70 years only.^5,6 Conversely, CAS avoided cranial nerve palsy and was associated with lower risks of access site haematoma and myocardial infarction. Both treatments appear to be equally effective at preventing recurrent stroke after the procedural period.^7–10

Further development of surgical and interventional techniques and optimisation of patient selection may lead to a reduction in the risk of procedural stroke associated with carotid revascularisation. However, phase III trials of new carotid interventions designed to show improvements in treatment safety will require many patients and take a long time to complete. For example, a clinical trial would require about 2000 patients to detect a true reduction in the risk of procedural stroke from 6% to 3% with a new intervention compared to an established intervention, at a significance level of 5% and 90% power. The use of a surrogate outcome measure, defined as ‘a variable that provides an indirect measurement of effect in situations where direct measurement of clinical effect is not feasible or practical’,¹¹ may therefore be justified to investigate new interventions in proof-of-concept or phase II studies and inform the decision whether a large phase III trial should be initiated.¹²

A large number of small to medium sized studies have reported acute cerebral ischaemic lesions on diffusion-weighted magnetic resonance imaging (DWI) to occur in a substantial proportion of patients after treatment of carotid stenosis by CAS or CEA (Figure 1). Recent review articles highlighted the potential of DWI as a surrogate safety outcome measure in clinical trials of carotid interventions.^13–15 Furthermore, the existing evidence has never been systematically reviewed in the context of validating ischaemia on DWI as a surrogate outcome measure for procedural ischaemic stroke.

Figure 1.

Diffusion-weighted magnetic resonance imaging (DWI) of a patient following stenting of the right carotid artery. Multiple hyperintense signals representing acute ischaemic lesions in the territory of the right middle cerebral artery are present. The patient did not experience any symptoms. Copyright Department of Radiology, University Hospital Basel.

In general, the strength of evidence for a validated surrogate outcome measure depends on the following criteria: Firstly, the biological plausibility of a relationship between the surrogate and the clinical outcome measure; secondly, the demonstration of a correlation between the occurrence of the surrogate and the clinical outcome measure and thirdly, evidence from clinical trials that the effect of treatment on the surrogate outcome measure (e.g. the relative risk of cerebral ischaemia on DWI in CAS vs. CEA) correlates with the treatment effect on the clinical outcome measure (e.g. the relative risk of procedural stroke in CAS vs. CEA).¹¹ Based on its specificity to image processes in the brain directly linked to cellular hypoxia, and its sensitivity in detecting acute cerebral ischaemia,^16–18 the use of DWI as a surrogate outcome measure for procedural ischaemic stroke seems biologically plausible. In order to explore the second and third criterion for surrogacy, we performed a systematic literature review and meta-analysis. We investigated the association between the risk of procedural stroke and the risk of cerebral ischaemia on DWI in patients treated by CAS or CEA, and compared the effect of CAS vs. CEA on the risk of stroke and the risk of cerebral ischaemia on DWI.

Methods

Search strategy and study eligibility criteria

A systematic MedLine® search was conducted to identify all studies investigating patients undergoing treatment of carotid artery stenosis by CAS or CEA with DWI, published online with access to the full article or in-print version up to 31 December 2016. The search term (‘carotid endarterectomy’ OR ‘carotid stenting’ OR ‘carotid angioplasty’) AND (‘DWI’ OR ‘diffusion’ OR ‘MRI’ OR ‘embolism’ OR ‘ischemia’ OR ‘silent’ OR ‘ischemic lesion’ OR ‘emboli’) was entered on http://www.ncbi.nlm.nih.gov/PubMed. We also checked cited studies in retrieved articles and existing reviews on MRI in carotid interventions.^13–15,19

Studies were included in the analysis if they fulfilled the following criteria: (1) Patients were treated for atherosclerotic carotid stenosis (any degree of stenosis, symptomatic or asymptomatic stenosis) with either CEA or CAS (with or without use of cerebral protection devices); (2) patients were examined with DWI both before and within seven days after treatment and (3) the numbers or percentages of patients with new lesions on DWI after treatment and procedural ischaemic stroke were reported. If several publications resulted from the same cohort of patients, we chose the report which included the highest number of patients. Studies assessing brain ischaemia on magnetic resonance imaging (MRI) sequences other than DWI were not included. Studies with prospective or retrospective design were included as long as they fulfilled the inclusion criteria above. Reports on single cases were not included. Likewise, studies reporting on emergency procedures (i.e. including intracranial thrombectomy in acute ischemic stroke) were not included. Two researchers (CT and LHB) reviewed the abstracts of all publications identified by the search, retrieved the full publications where abstract data were consistent with the inclusion criteria and selected the studies for inclusion based on the full published data.

Data extraction

Data were extracted independently by three researchers (LHB, CT and STE). Any disagreement was resolved by consensus. For all eligible studies, we extracted: the number (N) of included CAS or CEA procedures; N patients with symptomatic carotid stenosis, usually defined as ischaemic ocular or cerebral symptoms having occurred in the dependent territory of the carotid artery within the past six months; magnetic field strength (tesla) used in MRI; latency between treatment and post-treatment DWI; N patients with a positive DWI (henceforward referred to as ‘DWI+’), defined as the presence of at least one hyperintense lesion on DWI after treatment which was not present on the pre-treatment DWI and N patients with procedural ischaemic cerebral stroke (henceforward referred to as ‘stroke’), defined as a focal neurological deficit of probable ischaemic vascular cause lasting >24 h occurring up to 30 days after treatment, with exclusion of intracranial haemorrhage, hyperperfusion syndrome and other structural brain disease on neuroimaging. Pure retinal infarction was not included. Transient ischaemic attack (TIA) was not consistently recorded in the studies and was therefore not extracted.

For studies involving CAS, we extracted information on whether a cerebral protection device (CPD) was used and which type of device. For studies involving CEA, we noted whether a standard or eversion technique, a shunt, or a patch was used, and if the procedure was done under local or general anaesthesia. For studies including two or more treatment groups (e.g. studies comparing CAS vs. CEA or studies comparing CAS with vs. without CPD use), data were extracted for each treatment group separately.

Corresponding authors were contacted via e-mail if the published data were not sufficient to determine whether the study was eligible for analysis or not, or to extract the relevant information. Additional data obtained in this way (n = 5 studies) were also considered in this review.

Statistical analysis

To explore the second criterion for surrogacy, we included all groups of patients treated with CAS or CEA from eligible studies for which the proportion or number of patients with DWI+ and stroke was reported (henceforth referred to as ‘treatment groups’) as separate data points. First, a crude correlation coefficient for the association between the observed risks of DWI+ and the observed risks of stroke was calculated across treatment groups. This correlation underestimates the extent of the association between the underlying risks because the imprecision in the observed risks dilutes the association as it does whenever an association is measured between two measurement error-prone variables. To account for this, we utilised a bivariate random effects logistic regression model to estimate the correlation between the underlying log-transformed odds (log odds) of stroke and the underlying log odds of DWI+. A detailed description of this model is given in the supplemental methods.

A number of studies compared two or more different revascularisation procedures with each other. Owing to the relatively small size of these studies, the confidence interval (CI) surrounding the treatment effects with regard to procedural stroke and DWI+ was wide. For these reasons, we refrained from further exploring any correlation between treatment effects on stroke and DWI+ to evaluate the third criterion for surrogacy. Instead, we investigated in a meta-analysis of all studies comparing CAS vs. CEA, whether the treatment effect for DWI+ pointed in the same direction as the treatment effect for procedural stroke (i.e. whether the comparison between CAS and CEA using DWI+ as the outcome favoured the same treatment as the comparison using stroke as the outcome). Data were aggregated using fixed-effect Mantel-Haenszel models and treatment effects calculated as risk ratios with 95% CI, using CAS as the reference treatment. We quantified heterogeneity using the I² statistic. Review Manager Software (RevMan, version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) was used.

Reporting standards

Data presentation and reporting in this article is done according to the reporting standards of the PRISMA Statement for Reporting Systematic Reviews and Meta-Analyses.^20,21

Results

Included and excluded studies

As of 30 March 2017, the search yielded 2555 results (Figure 2). Full publications were retrieved of 164 studies published up to 31 December 2016. Seventy-nine studies were excluded for the following reasons: MRI to detect cerebral ischaemia did not include DWI sequences, or DWI was not done in all patients (n = 15);^22–36 no DWI was performed before treatment (n = 11);^37–47 post-procedural DWI was performed more than seven days after treatment (n = 3);^48–50 the number of DWI+ patients or procedures could not be extracted (n = 6);^51–56 or information on procedural stroke was lacking or patients with procedural stroke were excluded from the analysis (n = 6).^49,57–68 One study was excluded because the analysis was restricted to patients with available six months’ follow-up who constituted less than half of the study population (n = 1).⁶⁹ One study was excluded since only pre-selected groups of patients showing micro-embolic signals in transcranial Doppler during CEA were included in the final analysis.⁷⁰ Another study investigating the use of intravascular ultrasound for carotid plaque characterisation was excluded since the additional device by itself may have contributed to the occurrence of DWI+ or stroke.⁷¹ A further 28 studies were excluded because updated reports from the same cohort including a larger number of patients had subsequently been published and included, or because the studies were secondary analyses of previously included studies that contained no additional data relevant for the present review.^72–99

Figure 2.

Flow chart of selection of retrieved studies for analysis.

The remaining 85 studies including 6970 carotid revascularisation procedures (CAS: n = 4871; CEA: n = 2099) fulfilled all eligibility criteria and were included in the analysis. Fourteen studies investigated DWI lesions in CEA.^100–113 One of these studies retrospectively compared conventional with flow-control-CEA (i.e. clamping of the common carotid artery, the external carotid artery and the superior thyroid artery before dissection of the carotid sheath around the internal carotid artery).¹⁰⁸ Six studies investigated DWI lesions in CAS without CPD,^114–119 one of which using a membrane-covered stent.¹¹⁹ Thirty-eight studies investigated DWI lesions in CAS with CPD:^120–157 one study randomly compared membrane-covered vs. bare-metal stents;¹²⁹ one study randomly allocated patients to CAS using flow reversal or distal filter protection;¹⁵⁰ one study randomly assigned to either proximal or distal cerebral protection during CAS;¹⁴⁵ one study compared open vs. closed-cell stent type in a randomised design;¹⁴⁷ three studies compared CAS with CPD vs. CAS without CPD^158–160 and three compared CAS with filter CPD vs. CAS with balloon CPD.^161–163 Three randomised studies, one of which had a 2 × 2 factorial design;¹⁴⁸ investigated clinical and imaging outcomes of CAS according to different regimens of peri-interventional platelet-inhibition^148,164 and statin treatment^144,148 a further study investigated DWI lesions in CAS using different catheter techniques (use of 7F or 8F catheter or use of a coaxial system with use of a 7F or 8F catheter in conjunction with a 4F or 5F catheter), both with and without using CPD.¹⁶⁵ Nineteen studies compared DWI lesions and clinical outcomes in CEA vs. CAS,^166–184 only two of which were randomised.^176,184 Details of included studies are provided in supplementary Table 1.

Incidence of positive DWI and procedural ischaemic stroke in CAS and CEA

Data from 85 separate CAS treatment groups reporting on outcomes of 4871 CAS procedures were available from eligible single treatment group and multiple treatment group (comparative) studies combined. Carotid symptom status was available in 3829 CAS procedures, and 2023 of those (53%) were done in symptomatic carotid stenosis. New ischaemic lesions on DWI were found after 1805 procedures (average risk 37%), while procedural ischaemic strokes were reported to have occurred in 131 procedures (2.6%; Table 1).

Table 1.

Average risks of cerebral ischaemia on diffusion-weighted magnetic resonance imaging and procedural ischaemic stroke in carotid artery stenting and carotid endarterectomy.

	Treatment groups N	Included procedures N	Symptomatic carotid stenosis N (%)^a	DWI+ N (%)	Ischaemic stroke N (%)
Carotid artery stenting^b	85	4871	2023/3829 (53)	1805 (37.1)	131 (2.7)
Any protection device used	64	3000	1225/2674 (46)	1104 (36.8)	78 (2.6)
No protection device used	14	698	490/679 (72)	259 (37.1)	25 (3.6)
Carotid endarterectomy^c	34	2099	1238/1877 (66)	222 (10.6)	30 (1.4)
General anaesthesia	25	1418	750/1193 (62.9)	161 (11.4)	19 (1.3)
Local anaesthesia	5	458	318/433 (73.4)	39 (8.5)	6 (1.3)

Note: Data from treatment groups reported in 85 different studies are presented as numbers and percentages.

CAS: carotid artery stenting; CEA: carotid endarterectomy; DWI+: at least 1 new lesion on diffusion-weighted imaging after treatment.

^aCarotid symptom status was available in 3829 CAS and 1877 CEA procedures, respectively. Percentages of procedures done on symptomatic stenosis of procedures with known symptom status are presented.

^bIncluding all CAS treatment groups. In seven studies, patients were treated with or without the use of cerebral protection devices, but data were not reported separately.

^cIncluding all CEA treatment groups. In two studies, patients were operated under general or local anaesthesia, but data were not reported separately. Two further studies did not report on the type of used anaesthesia.

Eligible studies provided data for 34 separate CEA treatment groups including 2099 CEA procedures. For 1877 of these CEA procedures, carotid symptom status was available, and 1238 of them (66%) were done in patients with symptomatic carotid stenosis. New DWI lesions were detected after 222 procedures (10.8%) and 30 procedures (1.4%) were complicated by procedural ischaemic strokes (Table 1).

Correlation between risk of cerebral ischaemia on DWI and risk of procedural ischaemic stroke

The crude risks of DWI+ and stroke observed in all 119 treatment groups (CEA and CAS combined) are displayed in Figure 3(a). The crude correlation between these risks was 0.29 (p = 0.0014). Figure 4 illustrates why this crude correlation underestimates the magnitude of the association between the underlying risks. The correlation between the underlying log odds of DWI+ and the log odds of stroke corrected for this underestimation using the bivariate random effects logistic regression model was 0.61 (95% CI 0.27 to 0.87; p = 0.0012). The slope of this relationship was 1.20 (95% CI 0.50 to 2.39), this being an estimate of the change in log odds of DWI+ per 1 unit change in the log odds of stroke. Figure 3(b) converts this estimated linear relationship between the two log odds ratios into the non-linear relationship between the estimated risks. Within the range of procedural stroke risk of between 2% and 7% reported in previous clinical trials of CEA or CAS for symptomatic and asymptomatic carotid stenosis, new ischaemic lesions on post-procedural scans can be expected to occur in about 10 times this proportion. Figure 3(b) shows that for a particular drug or interventional technique in a future randomised controlled trial postulated to reduce procedural stroke risk from 6% to 3%, the predicted reduction in DWI+ risk would be from 56% to 35%. For a standard two-arm clinical trial, with 90% statistical power to detect a difference that is statistically significant at the 5% level using a two-sided test, this equates to a reduction in sample size from 2004 to 228, almost a 90% reduction.

Figure 3.

Crude risks of cerebral ischaemia on DWI and procedural ischaemic stroke and fitted association between the underlying risks of these outcomes in 119 groups of patients undergoing carotid revascularisation. (a) Observed crude risks of stroke and ischaemia on diffusion-weighted imaging (DWI) in all 119 treatment groups. Red triangles represent groups of patients treated with carotid endarterectomy and blue dots groups of patients treated with stents. (b) Fitted regression line relating ‘underlying’ (true) risk of DWI lesions to ‘underlying’ (true) risk of stroke in all included studies. Correlation between log odds of DWI lesions and log odds of stroke (coefficient 0.61 (95% CI 0.27, 0.87), p = 0.0012). Red lines show the magnitude of reduction in risk of DWI lesions that might be observed in a pilot trial expected to reduce the risk of procedural stroke from 6% to 3%.

Figure 4.

Illustrative example of statistical method using simulated data. (a) Simulated ‘underlying’ (true) log odds of DWI+ (i.e. presence of at least one new DWI brain lesion after treatment) and log odds of procedural ischaemic stroke for 30 studies, with regression line. Simulated data were drawn from a bivariate normal distribution with means, SDs and correlation similar to those estimated using the actual data by the bivariate random effects logistic regression model. This plot therefore represents the ‘corrected’ association between log odds of DWI+ and log odds stroke. (b) Conversion of simulated ‘underlying’ log odds of DWI+ and stroke to ‘underlying’ risks of DWI+ and stroke for all 30 studies, with fitted regression line (risk=exp(log odds)/(exp(log odds)+1). (c) Simulation of observed risks (hollow circles) by addition of sampling error (random error; green arrows) to both the ‘underlying’ (true) risks of DWI+ and stroke (solid circles, as in (b)) through sampling subjects in each of the 30 studies (which have sizes typical of the studies included in our analysis). First observed stroke risk was simulated, and then it was assumed that everyone who has a stroke is DWI+, and finally the number of DWI+ cases amongst those without stroke was simulated. Size of hollow circles for observed risks is relative to the number of patients in that study. Smaller studies tend to be subject to greater sampling error (as indicated by typically longer arrows). (d) Simulated observed risks of DWI+ and stroke for 30 studies, with fitted regression line. This plot represents the crude association between study-specific DWI+ and stroke risks and demonstrates its underestimation compared with the association of ‘underlying’ (true) DWI+ and stroke risks as in Figure b (as both factors are subject to sampling error).

Including only CAS treatment groups, the crude correlation between DWI+ and stroke was 0.19 and the corrected correlation was 0.27 (95% CI –0.19, 0.66, p = 0.24). There were too few studies done on CEA and too little variance in stroke risk between studies to be able to accurately estimate the between study variance in stroke risk (with this actually being estimated as zero). Therefore, the corrected correlation between DWI+ and stroke could not be calculated for CEA.

Comparison of treatment effects on cerebral ischaemia on DWI and procedural ischaemic stroke

Nineteen eligible studies compared DWI findings and clinical outcomes between CAS and CEA. Across these studies, the risk of being DWI+ after CAS was 3.83 times that after CEA (95% 3.17 to 4.63; p < 0.00001; Figure 5), while the risk of ischaemic procedural stroke after CAS was 2.38 times (95%CI 1.44 to 3.94; p = 0.0007; Figure 6) that after CEA.

Figure 5.

Meta-analysis of studies comparing the risk of new ischaemic brain lesions on diffusion-weighted imaging after carotid artery stenting and carotid endarterectomy. Data are numbers of patients with DWI lesions (‘events’), total numbers of patients and Mantel-Haenszel fixed-effects risk ratios including 95% confidence intervals (CI), with endarterectomy as the reference treatment. Squares on the right represent point estimates of risk ratios at trial level, with 95% CI as horizontal bars. The diamond at the bottom represents the summary risk ratio and 95% CI.

Figure 6.

Meta-analysis of studies comparing the risk of procedural ischaemic stroke between carotid artery stenting and carotid endarterectomy. The same studies as in Figure 5 are included. Data are numbers of patients with strokes (‘events’), total numbers of patients and Mantel-Haenszel fixed-effects risk ratios including 95% confidence intervals (CI), with endarterectomy as the reference treatment. Squares on the right represent point estimates of risk ratios at trial level, with 95% CI as horizontal bars. The diamond at the bottom represents the summary risk ratio and 95% CI.

Discussion

In our systematic review of the literature, new ischaemic lesions on DWI were present on average in 37% of patients following CAS compared to 10.8% of patients after CEA; in contrast, the reported risks of procedural ischaemic stroke were only 2.6% and 1.4%, respectively. Across all groups of patients treated with CAS or CEA, the risk of a positive DWI scan after treatment significantly correlated with the risk of procedural stroke. Among those studies comparing CAS vs. CEA, summary treatment effects on the occurrence of DWI lesions pointed in the same direction as summary treatment effects on ischaemic stroke, i.e. the risks of both clinical and radiological cerebral ischaemia were increased in CAS compared to CEA. Does the current evidence therefore support the use of DWI as a surrogate outcome measure for procedural safety in trials of carotid interventions?

The statistical methodology involved in validating surrogate markers is complex and controversial.¹⁸⁵ A set of formal statistical rules for validating surrogate outcome measures has been proposed to reduce observation time in prospective trials of progressive diseases.¹⁸⁶ These rules state among other criteria that the surrogate must predict clinically manifest disease in the future and that the full effect of treatment on the clinical outcome must be explained by the effect of treatment on the surrogate outcome. It is evident that we cannot use these formal criteria to validate DWI lesions as a surrogate outcome measure of procedural stroke in carotid interventions: both the potential surrogate – DWI lesions – and the clinical endpoint – procedural stroke – are short-term outcome measures characterising risk of procedure; the major advantage of DWI as a surrogate outcome is that ischaemic brain lesions are much more common than clinically manifest stroke, allowing reductions in the sample size of pilot trials investigating novel approaches in reducing treatment risks (e.g. surgical technique, stent design, CPDs and peri-procedural medication). Hence, we evaluated the potential surrogacy of DWI against a set of more general rules defined in the ICH Harmonised Tripartite Guideline: Statistical Principals for Clinical Trials.¹¹

Is cerebral ischaemia on DWI a biologically plausible surrogate outcome measure for procedural ischaemic stroke?

Stroke is the most common serious adverse event occurring in carotid revascularisation procedures. In a meta-analysis of randomised trials comparing CAS vs. CEA for symptomatic carotid stenosis, 94% of all strokes occurring within 30 days of CAS and 81% of all strokes occurring within 30 days of CEA were attributable to ischaemic cerebral infarction and the remaining events to intracerebral haemorrhage.⁶

Acute cerebral ischaemia leads to a cascade of events on the biological level, including break-down of electrolyte transport across the cellular membrane with subsequent shift of water from the extracellular to the intracellular space (cytotoxic oedema). The reduction in extracellular water causes a decrease in random translational motion of water molecules. Diffusion-weighted images are generated by measuring this random motion of water molecules by the effect of magnetic gradients on protons. Areas of reduced water diffusion are shown hyperintense in relation to surrounding normal brain tissue. DWI detects brain ischaemia in more than 90% of patients with the final clinical diagnosis of acute ischaemic stroke, and sensitivity and inter-rater reliability are superior to computer tomography or conventional (T2-weighted or FLAIR) MRI.^18,187,188 However, DWI may also show ischaemic lesions in the absence of a stroke: hyperintense brain lesions on DWI may be found in about a third patients with TIAs.^189,190 Furthermore, brain ischaemia on DWI without associated focal neurological deficit has been shown in patients with carotid stenosis¹⁹¹ and following therapeutic or investigational procedures on the heart or the carotid artery.¹³ DWI therefore identifies a spectrum of acute cerebral ischaemia encompassing asymptomatic lesions, TIA and stroke. Based on its specificity to image processes in the brain directly linked to cellular hypoxia, and its sensitivity in detecting acute cerebral ischaemia, the use of DWI as a surrogate outcome measure for procedural ischaemic stroke seems biologically plausible.

Does ischaemia on DWI correlate with procedural stroke during carotid interventions?

The result of our analysis including all populations treated with CAS or CEA demonstrates a significant correlation between the true risk of stroke and the true risk of cerebral ischaemia (p = 0.0012). A statistically significant association of the potential surrogate and clinical outcome constitutes the basis for a use of the surrogate as an endpoint in clinical trials. Our statistical methodology does not allow us to fully establish surrogacy; however, it does go part of the way towards this since it allows the magnitude of the association between the risk of a stroke and the risk of DWI lesions to be estimated, taking account of the fact that imprecision in the estimates of these two risks will tend to dilute the magnitude of the underlying association. The fact that the observed correlation between the underlying log odds ratio is strong (albeit with a wide 95% CI) is encouraging. The steepness of the association, coupled with the fact that the prevalence of DWI lesions is much greater than that of stroke illustrates that substantial sample size reductions for clinical trials may be possible by switching from a clinically based stroke outcome to MRI-based DWI lesions. Returning to the example given in the Introduction section, instead of 2000 patients (1000 in each treatment group) needed to detect an underlying 3% absolute difference in procedural stroke risk, just over 200 patients would be needed to detect the corresponding difference in DWI lesion risk. The proof of a significant correlation of a potential surrogate outcome measure and the true clinical outcome is crucial but not sufficient by itself to fully validate surrogacy.

Does the effect of treatment on ischaemic lesions on DWI correspond to the effect of treatment on procedural stroke?

Despite the corresponding summary treatment effects on DWI lesions and procedural stroke in studies comparing CAS vs. CEA, there was no clear relationship between the relative risks of DWI positivity and stroke across these studies. The observed number of strokes was very small in most of these comparative studies and estimated relative risks were surrounded by high CIs. Therefore, despite a strong theoretical background supporting the use of DWI as a surrogate outcome measure for procedural stroke in carotid interventions, and evidence for a correlation of DWI lesions with procedural stroke, the existing evidence is insufficient to assess the relationship between the effect of treatment on DWI lesions and on procedural stroke. More data from randomised clinical trials of carotid revascularisation are required to test this criterion. Apart from comparisons of CAS vs. CEA, such randomised trials may also compare various surgical or interventional techniques, such as cerebral protection devices, access routes, or medication. They need to incorporate serial MR imaging including DWI and be of adequate size to detect a sufficient number of clinically manifest strokes. In our experience, a time window of one to seven days before and one to three days after the carotid revascularisation procedure for the pre- and post-procedural scan is suitable to detect new ischaemic brain lesions caused by the procedure while at the same time allowing for enough flexibility in scheduling the scans.

Summarising data from several studies does imply limitations, most importantly the heterogeneity of the included studies. Firstly, the included studies commonly included both patients with symptomatic and asymptomatic carotid stenosis with the proportion of patients having a symptomatic stenosis ranging between 0% and 100% within subgroups (symptomatic stenosis across studies, 53% of all CAS procedures, and 66% of all CEA procedures). Recently, symptomatic carotid stenosis may be more prone to embolisation and subsequent DWI+ during CAS or CEA due to instability of the atherosclerotic plaque. Secondly, the included studies did use different time windows for pre- and post-treatment MRI scanning, which may have influenced the rate of post-procedural ischaemic brain lesions detected on DWI. Thirdly, definitions of what constituted a new ischaemic lesion after treatment differed: some studies purely relied on new hyperintense lesions on post-treatment DWI which had not been present before treatment. In other studies, a corresponding hypointense signal on ADC maps was additionally required. However, very small acute ischaemic lesions appearing hyperintense on DWI may not be identified on ADC maps due to limited spatial resolution. Finally, less than half of the included 85 studies clearly stated that either clinical assessment for procedural complications was done by an independent neurologist or outcome event adjudication was performed; lower reported incidence of procedural stroke in the absence of clinical evaluation by a neurologist is a well-known phenomenon. However, the average risk of procedural stroke in studies with reported event confirmation by a neurologist was only slightly higher than in studies without (2.6% vs. 2.1%).

The importance of DWI lesions occurring during treatment of carotid stenosis may go beyond being a surrogate of procedural stroke. In a prospective, population-based study investigating the association of silent brain infarcts and neurocognitive decline, silent brain infarcts were associated with a higher risk of dementia and a steeper decline in cognitive function.¹⁹² Silent ischaemic brain lesions following coronary artery bypass grafting¹⁹³ and intra-cardiac surgery¹⁹⁴ have also been associated with cognitive decline. Only a few small studies performed both cognitive testing and DWI before and after carotid revascularisation.^{38,86,105,195} These studies identified few patients with post-treatment ischaemia on DWI and were unable to demonstrate or refute a relationship between subclinical ischaemia and cognitive decline. An MRI-based sub-study within the International Carotid Stenting Study investigated multiple aspects of both the occurrence of new DWI lesions and ischaemic events in the procedural period after randomised assignment to CAS or CEA for symptomatic carotid stenosis.⁶ After CAS treatment, there was a significantly higher rate of recurrent stroke or TIA during follow-up among patients with new ischaemic brain lesions on DWI after treatment than among those without new DWI lesions.⁹⁸ These findings may suggest that new DWI lesions after CAS could further serve as a marker to identify patients at higher risk for future events.

Conclusion

Our findings prove a strong correlation between DWI+ and procedural stroke as clinical outcome measure in carotid interventions. The results of our analyses therefore strengthen the evidence base for the use of DWI+ as a surrogate outcome measure for procedural risk in carotid revascularisation. However, despite including a large number of studies with over 6000 carotid interventions, criteria for validation of a surrogate marker are still only partly satisfied. Further, randomised studies comparing treatment effects on DWI lesions and clinical stroke are needed to fully establish surrogacy.

Supplemental Material

Supplemental material for Silent brain infarcts on diffusion-weighted imaging after carotid revascularisation: A surrogate outcome measure for procedural stroke? A systematic review and meta-analysis

Supplemental Material for Silent brain infarcts on diffusion-weighted imaging after carotid revascularisation: A surrogate outcome measure for procedural stroke? A systematic review and meta-analysis by Christopher Traenka, Stefan T Engelter, Martin M Brown, Joanna Dobson, Chris Frost and Leo H Bonati in European Stroke Journal

Footnotes

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: CT has received funding for travel from Bayer. LHB has received funding for travel and served on scientific advisory boards for Bayer. STE has received funding for travel or speaker honoraria from Bayer and Boehringer Ingelheim. He has served on scientific advisory boards for Bayer, Boehringer Ingelheim, Pfizer/BMS and MindMaze and on the editorial board of Stroke. He has received an educational grant from Pfizer and compensation for educational efforts by Stago. The other authors have no conflicts of interest.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: CT was supported by the University of Basel and the Swiss National Science Foundation. LHB was supported by grants from the Swiss National Science Foundation (PBBSB-116873) and the University of Basel. STE has received an educational grant from Pfizer and compensation for educational efforts by Stago and research support from the Science Funds [Wissenschaftsfonds] of the University Hospital Basel, the University Basel, the Swiss Heart Foundation, and the Swiss National Science Foundation; MMB’s Chair in Stroke Medicine is supported by The Reta Lila Weston Trust for Medical Research.

This work was partly done at University College London Hospital and University College London, who received a proportion of funding from the Department of Health’s National Institute for Health Research Biomedical Research Centres funding scheme.

Ethical approval

Not applicable.

Informed consent

Not applicable.

Guarantor

CT and LHB.

Contributorship

LHB and CT performed the systematic review of the literature. CT designed/conceptualised the study, analysed/interpreted the data, collected data and drafted the article. JD, CF and CT performed statistical analyses. LHB initiated, designed, conceptualised and supervised the study, analysed/interpreted the data, revised the article, and collected data. STE extracted the data. All authors: data collection, critical review of the article, editing article for content. All authors agreed on submission of the present version of the article and reviewed the article and revised it critically for important intellectual content.

References

Author list according to pubmed is: Barnett HJM, Taylor DW, Haynes RB, Sackett DL, Peerless SJ, Ferguson GG, Fox AJ, Rankin RN, Hachinski VC, Wiebers DO, Eliasziw M. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med 1991; 325: 445–453.

Randomised trial of endarterectomy for recently symptomatic carotid stenosis: final results of the MRC European Carotid Surgery Trial (ECST). Lancet 1998; 351: 1379–1387.

Endarterectomy for asymptomatic carotid artery stenosis. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. JAMA 1995; 273: 1421–1428.

Halliday

Mansfield

Marro

et al . Prevention of disabling and fatal strokes by successful carotid endarterectomy in patients without recent neurological symptoms: randomised controlled trial. Lancet 2004; 363: 1491–1502.

Bonati

Lyrer

Ederle

et al . Percutaneous transluminal balloon angioplasty and stenting for carotid artery stenosis. Cochrane Database Syst Rev 2012; 9: CD000515.

Bonati

Dobson

Algra

et al .; Carotid Stenting Trialists' Collaboration. Short-term outcome after stenting versus endarterectomy for symptomatic carotid stenosis: a preplanned meta-analysis of individual patient data. Lancet 2010; 376: 1062–1073.

Eckstein

Ringleb

Allenberg

et al . Results of the Stent-Protected Angioplasty versus Carotid Endarterectomy (SPACE) study to treat symptomatic stenoses at 2 years: a multinational, prospective, randomised trial. Lancet Neurol 2008; 7: 893–902.

Brott

Hobson

Howard

et al . Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med 2010; 363: 11–23.

Mas

Arquizan

Calvet

et al . Long-term follow-up study of endarterectomy versus angioplasty in patients with symptomatic severe carotid stenosis trial. Stroke 2014; 45: 2750–2756.

10.

Bonati

Dobson

Featherstone

et al . Long-term outcomes after stenting versus endarterectomy for treatment of symptomatic carotid stenosis: the International Carotid Stenting Study (ICSS) randomised trial. Lancet 2015; 385: 529–538.

11.

Use ICoHoTRfRoPfH ICH Harmonised Tripartite Guideline: Statistical Principals for Clinical Trials, 1998.

12.

Fleming

DeMets

DL.

Surrogate end points in clinical trials: are we being misled?

Ann Intern Med 1996; 125: 605–613.

13.

Bendszus

Stoll

Silent cerebral ischaemia: hidden fingerprints of invasive medical procedures. Lancet Neurol 2006; 5: 364–372.

14.

Schnaudigel

Groschel

Pilgram

et al . New brain lesions after carotid stenting versus carotid endarterectomy: a systematic review of the literature. Stroke 2008; 39: 1911–1919.

15.

Cho

Deshpande

Pasupuleti

et al . Radiographic and symptomatic brain ischemia in CEA and CAS: a systematic review and meta-analysis. Neurology 2017; 89: 1977–1984.

16.

Moseley

Cohen

Mintorovitch

et al . Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med 1990; 14: 330–346.

17.

Warach

Chien

et al . Fast magnetic resonance diffusion-weighted imaging of acute human stroke. Neurology 1992; 42: 1717–1723.

18.

Lansberg

Norbash

Marks

et al . Advantages of adding diffusion-weighted magnetic resonance imaging to conventional magnetic resonance imaging for evaluating acute stroke. Arch Neurol 2000; 57: 1311–1316.

19.

Vermeer

Longstreth

Jr Koudstaal

PJ.

Silent brain infarcts: a systematic review. Lancet Neurol 2007; 6: 611–619.

20.

Moher

Shamseer

Clarke

et al . Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015; 4: 1.

21.

Moher

Liberati

Tetzlaff

et al . Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009; 6: e1000097.

22.

Jansen

Ramos

van Heesewijk

et al . Impact of microembolism and hemodynamic changes in the brain during carotid endarterectomy. Stroke 1994; 25: 992–997.

23.

Cantelmo

Babikian

Samaraweera

et al . Cerebral microembolism and ischemic changes associated with carotid endarterectomy. J Vasc Surg 1998; 27: 1024–1030.

24.

van Heesewijk

Vos

Louwerse

et al . New brain lesions at MR imaging after carotid angioplasty and stent placement. Radiology 2002; 224: 361–365.

25.

Terada

Tsuura

Matsumoto

et al . Results of endovascular treatment of internal carotid artery stenoses with a newly developed balloon protection catheter. Neurosurgery 2003; 53: 617–623.

26.

Madycki

Staszkiewicz

Gabrusiewicz

Carotid plaque texture analysis can predict the incidence of silent brain infarcts among patients undergoing carotid endarterectomy. Eur J Vasc Endovasc Surg 2006; 31: 373–380.

27.

Henry

Polydorou

Henry

et al . New distal embolic protection device the FiberNet 3 dimensional filter: first carotid human study. Catheter Cardiovasc Interv 2007; 69: 1026–1035.

28.

Wong

Chen

et al . Carotid stenting of asymptomatic and symptomatic carotid artery stenoses with and without the use of a distal embolic protection device. Acta Cardiol 2011; 66: 453–458.

29.

Huang

Chang

et al . Impact of silent ischemic lesions on cognition following carotid artery stenting. Eur Neurol 2011; 66: 351–358.

30.

Almekhlafi

Demchuk

Mishra

et al . Malignant emboli on transcranial Doppler during carotid stenting predict postprocedure diffusion-weighted imaging lesions. Stroke 2013; 44: 1317–1322.

31.

Inoue

Tsutsumi

Ohwaki

et al . Stratification of intraoperative ischemic impact by somatosensory evoked potential monitoring, diffusion-weighted imaging and magnetic resonance angiography in carotid endarterectomy with routine shunt use. Acta Neurochir 2013; 155: 2085–2096.

32.

Enevoldsen

Torfing

Kjeldsen

et al . Cerebral infarct following carotid endarterectomy. Frequency, clinical and hemodynamic significance evaluated by MRI and TCD. Acta Neurologica Scand 1999; 100: 106–11000.

33.

Akkaya

Vuruskan

Gul

et al . Cerebral microemboli and neurocognitive change after carotid artery stenting with different embolic protection devices. Int J Cardiol 2014; 176: 478–483.

34.

Berman

Wang

Mitchell

et al . The relationship between carotid artery plaque stability and white matter ischemic injury. Neuroimage Clin 2015; 9: 216–222.

35.

Corriere

Edwards

Geer

et al . Longitudinal evaluation of neurobehavioral outcomes after carotid revascularization. Ann Vasc Surg 2014; 28: 874–881.

36.

Montorsi

Caputi

Galli

et al . Microembolization during carotid artery stenting in patients with high-risk, lipid-rich plaque. A randomized trial of proximal versus distal cerebral protection. J Am Coll Cardiol 2011; 58: 1656–1663.

37.

Muller

Reiche

Langenscheidt

et al . Ischemia after carotid endarterectomy: comparison between transcranial Doppler sonography and diffusion-weighted MR imaging. AJNR Am J Neuroradiol 2000; 21: 47–54.

38.

Heyer

DeLaPaz

Halazun

et al . Neuropsychological dysfunction in the absence of structural evidence for cerebral ischemia after uncomplicated carotid endarterectomy. Neurosurgery 2006; 58: 474–480.

39.

Gossetti

Gattuso

Irace

et al . Embolism to the brain during carotid stenting and surgery. Acta Chir Belg 2007; 107: 151–154.

40.

Barbato

Dillavou

Horowitz

et al . A randomized trial of carotid artery stenting with and without cerebral protection. J Vasc Surg 2008; 47: 760–765.

41.

Takayama

Nakagawa

Iwasaki

et al . Initial experience of using the filter protection device during carotid artery stenting in Japan. Radiat Med 2008; 26: 348–354.

42.

Gattuso

Martinelli

Alunno

et al . Carotid stenting and transcranial Doppler monitoring: indications for carotid stenosis treatment. Vasc Endovascular Surg 2010; 44: 535–538.

43.

Yamatogi

Furukawa

Iida

et al . Evaluation of small ischemic lesions after carotid artery stenting: the usefulness of thin-slice diffusion-weighted MR imaging. Neuroradiology 2011; 53: 255–260.

44.

Chung

Jeong

Kwak

et al . Associations between cerebral embolism and carotid intraplaque hemorrhage during protected carotid artery stenting. AJNR Am J Neuroradiol 2016; 37: 686–691.

45.

Lee

Cho

Lee

et al.

Hemodynamic tandem intracranial lesions on magnetic resonance angiography in patients undergoing carotid endarterectomy. J Am Heart Assoc 2016; 5.

46.

Maruyama

Fukuda

Kataoka

et al . Evaluation of carotid artery outward remodeling by T1-weighted magnetic resonance imaging in carotid endarterectomy and stenting. J Vasc Surg 2015; 61: 1464–1471.e1.

47.

Mazzalai

Piatto

Toniato

et al . Using protamine can significantly reduce the incidence of bleeding complications after carotid endarterectomy without increasing the risk of ischemic cerebral events. World J Surg 2014; 38: 1227–1232.

48.

Falkensammer

Oldenburg

Hendrzak

et al . Evaluation of subclinical cerebral injury and neuropsychologic function in patients undergoing carotid endarterectomy. Ann Vasc Surg 2008; 22: 497–504.

49.

Takemoto

Ueba

Takano

et al . Quantitative evaluation using the plaque/muscle ratio index panels predicts plaque type and risk of embolism in patients undergoing carotid artery stenting. Clin Neurol Neurosurg 2013; 115: 1298–1303.

50.

Wang

Zhou

et al . Effects of carotid endarterectomy on cerebral reperfusion and cognitive function in patients with high grade carotid stenosis: a perfusion weighted magnetic resonance imaging study. Eur J Vasc Endovasc Surg 2015; 50: 5–12.

51.

Posacioglu

Engin

Cinar

et al . Carotid endarterectomy versus carotid artery stenting: findings in regard to neuroclinical outcomes and diffusion-weighted imaging. Tex Heart Inst J 2008; 35: 395–401.

52.

MacDonald

Evans

Griffiths

et al . Filter-protected versus unprotected carotid artery stenting: a randomised trial. Cerebrovasc Dis 2010; 29: 282–289.

53.

Timaran

Rosero

Higuera

et al . Randomized clinical trial of open-cell vs closed-cell stents for carotid stenting and effects of stent design on cerebral embolization. J Vasc Surg 2011; 54: 1310–1316.

54.

Pinero Gonzalez de la Pena

Gonzalez Garcia

Moniche Alvarez

et al . Filter content after carotid angioplasty and stenting: relation to ischemic lesions in diffusion-weighted imaging. Radiologia 2012; 54: 155–164.

55.

Grunwald

Reith

Kuhn

et al . Proximal protection with the Gore PAES can reduce DWI lesion size in high-grade stenosis during carotid stenting. EuroIntervention 2014; 10: 271–276.

56.

Sakamoto

Taoka

Nakagawa

et al . Magnetic resonance plaque imaging to predict the occurrence of the slow-flow phenomenon in carotid artery stenting procedures. Neuroradiology 2010; 52: 275–283.

57.

Barth

Remonda

Lovblad

et al . Silent cerebral ischemia detected by diffusion-weighted MRI after carotid endarterectomy. Stroke 2000; 31: 1824–1828.

58.

Tiemann

Reidt

Esposito

et al . Neuropsychological sequelae of carotid angioplasty with stent placement: correlation with ischemic lesions in diffusion weighted imaging. PLoS One 2009; 4: e7001.

59.

Blasel

Hattingen

Berkefeld

et al . Evaluation of angiographic and technical aspects of carotid stenting with diffusion-weighted magnetic resonance imaging. Cardiovasc Intervent Radiol 2009; 32: 666–671.

60.

Kurata

Okura

Kumon

et al . Plasma thrombin-cleaved osteopontin elevation after carotid artery stenting in symptomatic ischemic stroke patients. Hypertens Res 2012; 35: 207–212.

61.

Nanba

Ogasawara

Nishimoto

et al . Postoperative cerebral white matter damage associated with cerebral hyperperfusion and cognitive impairment after carotid endarterectomy: a diffusion tensor magnetic resonance imaging study. Cerebrovasc Dis 2012; 34: 358–367.

62.

Cho

Lee

et al . Reperfusion injury on magnetic resonance imaging after carotid revascularization. Stroke 2014; 45: 602–604.

63.

Hitchner

Baughman

Soman

et al . Microembolization is associated with transient cognitive decline in patients undergoing carotid interventions. J Vasc Surg 2016; 64: 1719–1725.

64.

Lee

Suh

BY.

Risk factor analysis of new brain lesions associated with carotid endarterectmy. Ann Surg Treat Res 2014; 86: 39–44.

65.

Orlicky

Vachata

Bartos

et al . A selective carotid artery shunting for carotid endarterectomy: prospective MR DWI monitoring of embolization in a group of 754 patients. J Neurol Surg A Cent Eur Neurosurg 2015; 76: 89–92.

66.

Sahin

Solak

Genc

et al . Brain diffusion changes in unilateral carotid artery stenosis with non-shunt endarterectomy: correlation with white matter lesions. Clin Neurol Neurosurg 2015; 133: 24–29.

67.

Skoloudik

Kuliha

Hrbac

et al . Sonolysis in Prevention of Brain Infarction During Carotid Endarterectomy and Stenting (SONOBUSTER): a randomized, controlled trial. Eur Heart J 2016; 37: 3096–3102.

68.

Yoshimura

Yamada

Kawasaki

et al . High-intensity signal on time-of-flight magnetic resonance angiography indicates carotid plaques at high risk for cerebral embolism during stenting. Stroke 2011; 42: 3132–3137.

69.

Hauth

Jansen

Drescher

et al . MR and clinical follow-up of diffusion-weighted cerebral lesions after carotid artery stenting. AJNR Am J Neuroradiol 2005; 26: 2336–2341.

70.

Fujimoto

Matsumoto

Oikawa

et al . Cerebral hyperperfusion after revascularization inhibits development of cerebral ischemic lesions due to artery-to-artery emboli during carotid exposure in endarterectomy for patients with preoperative cerebral hemodynamic insufficiency: revisiting the “impaired clearance of emboli concept”. Int J Mol Sci 2016; 119: 1–14.

71.

Hitchner

Zayed

Lee

et al . Intravascular ultrasound as a clinical adjunct for carotid plaque characterization. J Vasc Surg 2014; 59: 774–780.

72.

Forbes

Shill

Britt

et al . Assessment of silent embolism from carotid endarterectomy by use of diffusion-weighted imaging: work in progress. AJNR Am J Neuroradiol 2001; 22: 650–653.

73.

Poppert

Wolf

Resch

et al . Differences in number, size and location of intracranial microembolic lesions after surgical versus endovascular treatment without protection device of carotid artery stenosis. J Neurol 2004; 251: 1198–1203.

74.

Wolf

Heider

Heinz

et al . Frequency, clinical significance and course of cerebral ischemic events after carotid endarterectomy evaluated by serial diffusion weighted imaging. Eur J Vasc Endovasc Surg 2004; 27: 167–171.

75.

Wolf

Heider

Heinz

et al . Microembolic signals detected by transcranial Doppler sonography during carotid endarterectomy and correlation with serial diffusion-weighted imaging. Stroke 2004; 35: e373–e375.

76.

Gonzalez

Pinero

Martinez

et al . Silent cerebral ischemic lesions after carotid artery stenting with distal cerebral protection. Neurol Res 2005; 27: S79–S83.

77.

Hammer

Lacroix

Duprez

et al . Cerebral microembolization after protected carotid artery stenting in surgical high-risk patients: results of a 2-year prospective study. J Vasc Surg 2005; 42: 847–853.

78.

Kastrup

Nagele

Groschel

et al . Incidence of new brain lesions after carotid stenting with and without cerebral protection. Stroke 2006; 37: 2312–2316.

79.

Krapf

Nagele

Kastrup

et al . Risk factors for periprocedural complications in carotid artery stenting without filter protection: a serial diffusion-weighted MRI study. J Neurol 2006; 253: 364–371.

80.

Tedesco

Lee

Dalman

et al . Postprocedural microembolic events following carotid surgery and carotid angioplasty and stenting. J Vasc Surg 2007; 46: 244–250.

81.

Aso

Ogasawara

Sasaki

et al.

Preoperative cerebrovascular reactivity to acetazolamide measured by brain perfusion SPECT predicts development of cerebral ischemic lesions caused by microemboli during carotid endarterectomy. Eur J Nucl Med Mol Imaging 2009; 36: 297–301.

82.

Heider

Poppert

Wolf

et al . Fibrinogen and high-sensitive C-reactive protein as serologic predictors for perioperative cerebral microembolic lesions after carotid endarterectomy. J Vasc Surg 2007; 46: 449–454.

83.

Groschel

Pilgram

Ernemann

et al . Aortic calcification on plain chest radiography predicts embolic complications during carotid artery stenting. Eur J Neurol 2008; 15: 730–736.

84.

Kastrup

Groschel

Schnaudigel

et al . Target lesion ulceration and arch calcification are associated with increased incidence of carotid stenting-associated ischemic lesions in octogenarians. J Vasc Surg 2008; 47: 88–95.

85.

Grunwald

Papanagiotou

Roth

et al . Lesion load in unprotected carotid artery stenting. Neuroradiology 2009; 51: 313–317.

86.

Grunwald

Papanagiotou

Reith

et al . Influence of carotid artery stenting on cognitive function. Neuroradiology 2010; 52: 61–66.

87.

Yamada

Kawasaki

Yoshimura

et al . Prediction of silent ischemic lesions after carotid artery stenting using integrated backscatter ultrasound and magnetic resonance imaging. Atherosclerosis 2010; 208: 161–166.

88.

Yamada

Yoshimura

Kawasaki

et al . Prediction of silent ischemic lesions after carotid artery stenting using virtual histology intravascular ultrasound. Cerebrovasc Dis 2011; 32: 106–113.

89.

Zhu

Wintermark

Saloner

et al . The distribution and size of ischemic lesions after carotid artery angioplasty and stenting: evidence for microembolization to terminal arteries. J Vasc Surg 2011; 53: 971–975.

90.

Russjan

Goebell

Havemeister

et al . Predictors of periprocedural brain lesions associated with carotid stenting. Cerebrovasc Dis 2012; 33: 30–36.

91.

Gensicke

Zumbrunn

Jongen

et al . Characteristics of ischemic brain lesions after stenting or endarterectomy for symptomatic carotid artery stenosis: results from the international carotid stenting study – magnetic resonance imaging substudy. Stroke 2013; 44: 80–86.

92.

Wasser

Hildebrandt

Groschel

et al . Age-dependent effects of carotid endarterectomy or stenting on cognitive performance. J Neurol 2012; 259: 2309–2318.

93.

Capoccia

Speziale

Gazzetti

et al . Comparative study on carotid revascularization (endarterectomy vs stenting) using markers of cellular brain injury, neuropsychometric tests, and diffusion-weighted magnetic resonance imaging. J Vasc Surg 2010; 51: 584–591, 591.

94.

Leal

Orgaz

Fontcuberta

et al . A prospective evaluation of cerebral infarction following transcervical carotid stenting with carotid flow reversal. Eur J Vasc Endovasc Surg 2010; 39: 661–666.

95.

Altinbas

Algra

Bonati

et al . Periprocedural hemodynamic depression is associated with a higher number of new ischemic brain lesions after stenting in the International Carotid Stenting Study – MRI Substudy. Stroke 2014; 45: 146–151.

96.

Burow

Lyrer

Nederkoorn

et al . Echographic risk index and cerebral ischemic brain lesions in patients randomized to stenting versus endarterectomy for symptomatic carotid artery stenosis. Ultraschall Med 2014; 35: 267–272.

97.

Doig

Hobson

Muller

et al . Carotid anatomy does not predict the risk of new ischaemic brain lesions on diffusion-weighted imaging after carotid artery stenting in the ICSS-MRI substudy. Eur J Vasc Endovasc Surg 2016; 51: 14–20.

98.

Gensicke

van der Worp

Nederkoorn

et al . Ischemic brain lesions after carotid artery stenting increase future cerebrovascular risk. J Am Coll Cardiol 2015; 65: 521–529.

99.

Rostamzadeh

Zumbrunn

Jongen

et al . Predictors of acute and persisting ischemic brain lesions in patients randomized to carotid stenting or endarterectomy. Stroke 2014; 45: 591–594.

100.

Feiwell

Besmertis

Sarkar

et al . Detection of clinically silent infarcts after carotid endarterectomy by use of diffusion-weighted imaging. AJNR Am J Neuroradiol 2001; 22: 646–649.

101.

Tomczak

Wunderlich

Liewald

et al . Diffusion-weighted MRI: detection of cerebral ischemia before and after carotid thromboendarterectomy. J Comput Assist Tomogr 2001; 25: 247–250.

102.

Muller

Ciccotti

Axmann

et al . Embolic cerebral ischemia in carotid surgery: a model for human embolic stroke? Med Sci Monit 2003; 9: CR411–CR416.

103.

Inoue

Tsutsumi

Maeda

et al . Incidence of ischemic lesions by diffusion-weighted imaging after carotid endarterectomy with routine shunt usage. Neurol Med Chir (Tokyo)) 2006; 46: 529–533.

104.

Ogasawara

Suga

Sasaki

et al . Intraoperative microemboli and low middle cerebral artery blood flow velocity are additive in predicting development of cerebral ischemic events after carotid endarterectomy. Stroke 2008; 39: 3088–3091.

105.

Soinne

Helenius

Tikkala

et al.

The effect of severe carotid occlusive disease and its surgical treatment on cognitive functions of the brain. Brain Cogn 2009; 69: 359–359.

106.

Hebb

Heiserman

Forbes

et al . Perioperative ischemic complications of the brain after carotid endarterectomy. Neurosurgery 2010; 67: 286–293.

107.

Oikawa

Ogasawara

Saito

et al . Combined measurement of cerebral and cerebellar blood flow on preoperative brain perfusion SPECT imaging predicts development of new cerebral ischemic events after endarterectomy for symptomatic unilateral cervical carotid stenosis. Clin Nucl Med 2013; 38: 957–961.

108.

Yoshida

Kurosaki

Funaki

et al . Surgical dissection of the internal carotid artery under flow control by proximal vessel clamping reduces embolic infarcts during carotid endarterectomy. World Neurosurg 2014; 82: e229–e234.

109.

Cho

Lee

Yun

et al . Selective shunt during carotid endarterectomy using routine awake test with respect to a lower shunt rate. J Korean Surg Soc 2013; 84: 238–244.

110.

Sfyroeras

Bessias

Moulakakis

et al . New cerebral ischemic lesions after carotid endarterectomy. Ann Vasc Surg 2013; 27: 883–887.

111.

Akpinar

Sahin

et al . Previous chronic cerebral infarction is predictive for new cerebral ischemia after carotid endarterectomy. J Cardiothorac Surg 2015; 10: 141.

112.

Bourke

Beiles

CB.

Operative factors associated with the development of new brain lesions during awake carotid endarterectomy. Eur J Vasc Endovasc Surg 2016; 51: 167–173.

113.

Zhang

Chen

et al . Cognitive function after carotid endarterectomy: early decline and later recovery. Turk Neurosurg 2016; 26: 833–839.

114.

Grunwald

Reith

Karp

et al . Comparison of stent free cell area and cerebral lesions after unprotected carotid artery stent placement. EurJ Vasc Endovasc Surg 2012; 43: 10–14.

115.

Rosenkranz

Thomalla

Havemeister

et al . Older age and greater carotid intima-media thickness predict ischemic events associated with carotid-artery stenting. Cerebrovasc Dis 2010; 30: 567–572.

116.

Lovblad

Pluschke

Remonda

et al . Diffusion-weighted MRI for monitoring neurovascular interventions. Neuroradiology 2000; 42: 134–138.

117.

Jaeger

Mathias

Hauth

et al . Cerebral ischemia detected with diffusion-weighted MR imaging after stent implantation in the carotid artery. AJNR Am J Neuroradiol 2002; 23: 200–207.

118.

Rosenkranz

Fiehler

Niesen

et al . The amount of solid cerebral microemboli during carotid stenting does not relate to the frequency of silent ischemic lesions. AJNR Am J Neuroradiol 2006; 27: 157–161.

119.

Peynircioglu

Geyik

Yavuz

et al . Exclusion of atherosclerotic plaque from the circulation using stent-grafts: alternative to carotid stenting with a protection device? Cardiovasc Intervent Radiol 2007; 30: 854–860.

120.

Jaeger

Mathias

Drescher

et al . Clinical results of cerebral protection with a filter device during stent implantation of the carotid artery. Cardiovasc Intervent Radiol 2001; 24: 249–256.

121.

Schluter

Tubler

Steffens

et al . Focal ischemia of the brain after neuroprotected carotid artery stenting. J Am Coll Cardiol 2003; 42: 1007–1013.

122.

Du Mesnil de

Schneider

Yan

et al . Diffusion-weighted MR imaging lesions after filter-protected stenting of high-grade symptomatic carotid artery stenoses. AJNR Am J Neuroradiol 2006; 27: 1321–1325.

123.

Maleux

Demaerel

Verbeken

et al . Cerebral ischemia after filter-protected carotid artery stenting is common and cannot be predicted by the presence of substantial amount of debris captured by the filter device. AJNR Am J Neuroradiol 2006; 27: 1830–1833.

124.

McDonnell

Fearn

Baker

et al . Value of diffusion-weighted MRI during carotid angioplasty and stenting. Eur J Vasc Endovasc Surg 2006; 32: 46–50.

125.

Pinero

Gonzalez

Mayol

et al . Silent ischemia after neuroprotected percutaneous carotid stenting: a diffusion-weighted MRI study. AJNR Am J Neuroradiol 2006; 27: 1338–1345.

126.

Asakura

Kawaguchi

Sakaida

et al . Diffusion-weighted magnetic resonance imaging in carotid angioplasty and stenting with balloon embolic protection devices. Neuroradiology 2006; 48: 100–112.

127.

Asakura

Kawaguchi

Sakaida

et al . Diffusion-weighted MR imaging in carotid angioplasty and stenting with protection by the reversed carotid arterial flow. AJNR Am J Neuroradiol 2006; 27: 753–758.

128.

Grunwald

Supprian

Politi

et al.

Cognitive changes after carotid artery stenting. Neuroradiology 2006; 48: 315–323.

129.

Schillinger

Dick

Wiest

et al . Covered versus bare self-expanding stents for endovascular treatment of carotid artery stenosis: a stopped randomized trial. J Endovasc Ther 2006; 13: 312–319.

130.

Rapp

Wakil

Sawhney

et al . Subclinical embolization after carotid artery stenting: new lesions on diffusion-weighted magnetic resonance imaging occur postprocedure. J Vasc Surg 2007; 45: 867–872.

131.

Ghorab

Macian

Adoukounou

et al . Carotid angioplasty stenting revisited: clinical and radiological (MRI) outcome. Cerebrovasc Dis 2008; 25: 21–25.

132.

Palombo

Faraglia

Stella

et al . Late evaluation of silent cerebral ischemia detected by diffusion-weighted MR imaging after filter-protected carotid artery stenting. AJNR Am J Neuroradiol 2008; 29: 1340–1343.

133.

Schofer

Arendt

Tubler

et al . Late cerebral embolization after emboli-protected carotid artery stenting assessed by sequential diffusion-weighted magnetic resonance imaging. JACC Cardiovasc Interv 2008; 1: 571–577.

134.

Faggioli

Ferri

Rapezzi

et al.

Atherosclerotic aortic lesions increase the risk of cerebral embolism during carotid stenting in patients with complex aortic arch anatomy. J Vasc Surg 2009; 49: 80–85.

135.

Tedesco

Dalman

Zhou

et al . Reduction of postprocedure microemboli following retrospective quality assessment and practice improvement measures for carotid angioplasty and stenting. J Vasc Surg 2009; 49: 607–612.

136.

Taha

Maeda

Sakaida

et al . Cerebral ischemic lesions detected with diffusion-weighted magnetic resonance imaging after carotid artery stenting: Comparison of several anti-embolic protection devices. Neurol Med Chir (Tokyo)) 2009; 49: 386–393.

137.

Faraglia

Palombo

Stella

et al . Cerebral embolization during transcervical carotid stenting with flow reversal: a diffusion-weighted magnetic resonance study. Ann Vasc Surg 2009; 23: 429–435.

138.

Palombo

Stella

Faraglia

et al . Cervical access for filter-protected carotid artery stenting: a useful tool to reduce cerebral embolisation. EurJ Vasc Endovasc Surg 2010; 39: 252–257.

139.

Pinter

Ribo

Loh

et al . Safety and feasibility of a novel transcervical access neuroprotection system for carotid artery stenting in the PROOF study. J Vasc Surg 2011; 54: 1317–1323.

140.

Uchiyama

Misaki

Mohri

et al . Association between carotid plaque composition assessed by multidetector computed tomography and cerebral embolism after carotid stenting. Neuroradiology 2012; 54: 487–493.

141.

Tulip

Rosero

Higuera

et al . Cerebral embolization in asymptomatic versus symptomatic patients after carotid stenting. J Vasc Surg 2012; 56: 1579–1584; discussion 1584.

142.

Leal

Orgaz

Flores

et al . A diffusion-weighted magnetic resonance imaging-based study of transcervical carotid stenting with flow reversal versus transfemoral filter protection. J Vasc Surg 2012; 56: 1585–1590.

143.

Palombo

Stella

Fantozzi

et al.

Transcranial Doppler and diffusion-weighted magnetic resonance evaluation of cerebral embolization occurring during transfemoral carotid stenting with proximal flow blockage. J Cardiovasc Surg (Torino) 2012; 53: 1–6.

144.

Takayama

Taki

Toma

et al . Effect of pitavastatin on preventing ischemic complications with carotid artery stenting: a multicenter prospective study – EPOCH-CAS study. Cardiovasc Intervent Radiol 2014; 37: 1436–1443.

145.

Cano

Kambara

de Cano

et al . Randomized comparison of distal and proximal cerebral protection during carotid artery stenting. JACC Cardiovasc Intervent 2013; 6: 1203–1209.

146.

Pini

Faggioli

Fittipaldi

et al . Inflammatory mediators and cerebral embolism in carotid stenting: new markers of risk. J Endovasc Therapy 2013; 20: 684–694.

147.

Park

Kim

et al . Incidence of embolism associated with carotid artery stenting: open-cell versus closed-cell stents. J Neurosurg 2013; 119: 642–647.

148.

Patti

Tomai

Melfi

et al . Strategies of clopidogrel load and atorvastatin reload to prevent ischemic cerebral events in patients undergoing protected carotid stenting. Results of the randomized ARMYDA-9 CAROTID (Clopidogrel and Atorvastatin Treatment During Carotid Artery Stenting) study. J Am Coll Cardiol 2013; 61: 1379–1387.

149.

Bijuklic

Wandler

Varnakov

et al . Risk factors for cerebral embolization after carotid artery stenting with embolic protection: a diffusion-weighted magnetic resonance imaging study in 837 consecutive patients. Circ Cardiovasc Interv 2013; 6: 311–316.

150.

Castro-Afonso

Abud

Rolo

et al . Flow reversal versus filter protection: a pilot carotid artery stenting randomized trial. Circ Cardiovasc Interv 2013; 6: 552–559.

151.

Tanemura

Maeda

Ichikawa

et al . High-risk plaque for carotid artery stenting evaluated with 3-dimensional T1-weighted gradient echo sequence. Stroke 2013; 44: 105–110.

152.

Gunduz

Akdemir

Ayhan

et al . Can Doppler flow parameters of carotid stenosis predict the occurrence of new ischemic brain lesions detected by diffusion-weighted MR imaging after filter-protected internal carotid artery stenting? AJNR American Journal of Neuroradiology 2014; 35: 760–765.

153.

Huang

Chang

et al . Impact of coexisting coronary artery disease on the occurrence of cerebral ischemic lesions after carotid stenting. PLoS One 2014; 9: e94280.

154.

Matsukawa

Fujii

Uemura

et al . Pathology of embolic debris in carotid artery stenting. Acta Neurol Scand 2015; 131: 197–202.

155.

Adhikari

Takeda

Kolakshyapati

et al . Somatosensory evoked potentials in carotid artery stenting: Effectiveness in ascertaining cerebral ischemic events. J Clin Neurosci 2016; 30: 71–76.

156.

Kuliha

Roubec

Goldirova

et al . Laboratory-based markers as predictors of brain infarction during carotid stenting: a prospective study. J Atheroscler Thromb 2016; 23: 839–847.

157.

Ruffino

Faletti

Bergamasco

et al . Incidence of new ischaemic brain lesions after carotid artery stenting with the micromesh roadsaver carotid artery stent: a prospective single-centre study. Cardiovasc Intervent Radiol 2016; 39: 1541–1549.

158.

Gauvrit

Delmaire

Henon

et al . Diffusion/perfusion-weighted magnetic resonance imaging after carotid angioplasty and stenting. J Neurol 2004; 251: 1060–1067.

159.

Cosottini

Michelassi

Puglioli

et al . Silent cerebral ischemia detected with diffusion-weighted imaging in patients treated with protected and unprotected carotid artery stenting. Stroke 2005; 36: 2389–2393.

160.

Kastrup

Groschel

Nagele

et al . Effects of age and symptom status on silent ischemic lesions after carotid stenting with and without the use of distal filter devices. AJNR Am J Neuroradiol 2008; 29: 608–612.

161.

El-Koussy

Schroth

et al . Periprocedural embolic events related to carotid artery stenting detected by diffusion-weighted MRI: comparison between proximal and distal embolus protection devices. J Endovasc Ther 2007; 14: 293–303.

162.

Kim

Roh

Jeon

et al . Cerebral ischemia detected with diffusion-weighted MR imaging after protected carotid artery stenting: comparison of distal balloon and filter device. Korean J Radiol 2007; 8: 276–285.

163.

Bijuklic

Wandler

Hazizi

et al . The PROFI study (prevention of cerebral embolization by proximal balloon occlusion compared to filter protection during carotid artery stenting): a prospective randomized trial. J Am Coll Cardiol 2012; 59: 1383–1389.

164.

Kopp

Steiner

Nasel

et al . Abciximab reduces monocyte tissue factor in carotid angioplasty and stenting. Stroke 2003; 34: 2560–2567.

165.

Kim

Lee

Yang

et al . The influence of carotid artery catheterization technique on the incidence of thromboembolism during carotid artery stenting. AJNR Am J Neuroradiol 2010; 31: 1732–1736.

166.

Flach

Ouhlous

Hendriks

et al . Cerebral ischemia after carotid intervention. J Endovasc Ther 2004; 11: 251–257.

167.

Garcia-Sanchez

Millan-Torne

Capellades-Font

et al . Ischemic brain lesions following carotid revascularisation procedures: a comparative study using diffusion-weighted magnetic resonance imaging. Rev Neurol 2004; 38: 1013–1017.

168.

Roh

Byun

Ryoo

et al . Prospective analysis of cerebral infarction after carotid endarterectomy and carotid artery stent placement by using diffusion-weighted imaging. AJNR Am J Neuroradiol 2005; 26: 376–384.

169.

Iihara

Murao

Sakai

et al . Outcome of carotid endarterectomy and stent insertion based on grading of carotid endarterectomy risk: a 7-year prospective study. J Neurosurg 2006; 105: 546–554.

170.

Poppert

Wolf

Theiss

et al . MRI lesions after invasive therapy of carotid artery stenosis: a risk-modeling analysis. Neurol Res 2006; 28: 563–567.

171.

Lacroix

Hammer

Astarci

et al . Ischemic cerebral lesions after carotid surgery and carotid stenting. Eur J Vasc Endovasc Surg 2007; 33: 430–435.

172.

Faraglia

Palombo

Stella

et al . Cerebral embolization in patients undergoing protected carotid-artery stenting and carotid surgery. J Cardiovasc Surg (Torino)) 2007; 48: 683–688.

173.

Tedesco

Coogan

Dalman

et al . Risk factors for developing postprocedural microemboli following carotid interventions. J Endovasc Ther 2007; 14: 561–567.

174.

Skjelland

Krohg-Sorensen

Tennoe

et al . Cerebral microemboli and brain injury during carotid artery endarterectomy and stenting. Stroke 2009; 40: 230–234.

175.

Zhou

Dinishak

Lane

et al . Long-term radiographic outcomes of microemboli following carotid interventions. J Vasc Surg 2009; 50: 1314–1319.

176.

Bonati

Jongen

Haller

et al . New ischaemic brain lesions on MRI after stenting or endarterectomy for symptomatic carotid stenosis: a substudy of the International Carotid Stenting Study (ICSS). Lancet Neurol 2010; 9: 353–362.

177.

Wasser

Pilgram-Pastor

Schnaudigel

et al . New brain lesions after carotid revascularization are not associated with cognitive performance. J Vasc Surg 2011; 53: 61–70.

178.

Yamada

Yoshimura

Kawasaki

et al . Embolic complications after carotid artery stenting or carotid endarterectomy are associated with tissue characteristics of carotid plaques evaluated by magnetic resonance imaging. Atherosclerosis 2011; 215: 399–404.

179.

Felli

Alunno

Castiglione

et al . CEA versus CAS: short-term and mid-term results. Int Angiol 2012; 31: 420–426.

180.

Capoccia

Sbarigia

Rizzo

et al . Silent stroke and cognitive decline in asymptomatic carotid stenosis revascularization. Vascular 2012; 20: 181–187.

181.

Zhou

Hitchner

Gillis

et al . Prospective neurocognitive evaluation of patients undergoing carotid interventions. J Vasc Surg 2012; 56: 1571–1578.

182.

Akutsu

Hosoda

Fujita

et al . A preliminary prediction model with MR plaque imaging to estimate risk for new ischemic brain lesions on diffusion-weighted imaging after endarterectomy or stenting in patients with carotid stenosis. AJNR Am J Neuroradiol 2012; 33: 1557–1564.

183.

Mitsuoka

Shintani

Furuya

et al . Ultrasonographic character of carotid plaque and postprocedural brain embolisms in carotid artery stenting and carotid endarterectomy. Ann Vasc Dis 2011; 4: 106–109.

184.

Kuliha

Roubec

Prochazka

et al . Randomized clinical trial comparing neurological outcomes after carotid endarterectomy or stenting. Br J Surg 2015; 102: 194–201.

185.

Fleming

Powers

JH.

Biomarkers and surrogate endpoints in clinical trials. Stat Med 2012; 31: 2973–2984.

186.

Prentice

RL.

Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med 1989; 8: 431–440.

187.

Mullins

Schaefer

Sorensen

et al . CT and conventional and diffusion-weighted MR imaging in acute stroke: study in 691 patients at presentation to the emergency department. Radiology 2002; 224: 353–360.

188.

Fiebach

Schellinger

Jansen

et al . CT and diffusion-weighted MR imaging in randomized order: diffusion-weighted imaging results in higher accuracy and lower interrater variability in the diagnosis of hyperacute ischemic stroke. Stroke 2002; 33: 2206–2210.

189.

Giles

Albers

Amarenco

et al . Addition of brain infarction to the ABCD2 Score (ABCD2I): a collaborative analysis of unpublished data on 4574 patients. Stroke 2010; 41: 1907–1913.

190.

Al-Khaled

Matthis

Munte

et al . The incidence and clinical predictors of acute infarction in patients with transient ischemic attack using MRI including DWI. Neuroradiology 2013; 55: 157–163.

191.

Kastrup

Ernemann

Nagele

et al . Risk factors for early recurrent cerebral ischemia before treatment of symptomatic carotid stenosis. Stroke 2006; 37: 3032–3034.

192.

Vermeer

Prins

den

et al . Silent brain infarcts and the risk of dementia and cognitive decline. N Engl J Med 2003; 348: 1215–1222.

193.

Restrepo

Wityk

Grega

et al . Diffusion- and perfusion-weighted magnetic resonance imaging of the brain before and after coronary artery bypass grafting surgery. Stroke 2002; 33: 2909–2915.

194.

Barber

Hach

Tippett

et al . Cerebral ischemic lesions on diffusion-weighted imaging are associated with neurocognitive decline after cardiac surgery. Stroke 2008; 39: 1427–1433.

195.

Altinbas

van Zandvoort

van den

et al . Cognition after carotid endarterectomy or stenting: a randomized comparison. Neurology 2011; 77: 1084–1090.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.82 MB