颅内动脉粥样硬化的影像学检查—管腔狭窄与侧枝灌注(英文

Imaging of Intracranial Atherosclerosis:Stenoses and Collateral Perfusion颅内动脉粥样硬化的影像学检查管腔狭窄与侧枝灌注David S Liebeskind,MDAssociate Professor of Neurology&Director of Stroke ImagingCo-Director,UCLA Cerebral Blood Flow Laboratory Associate Neurology Director,UCLA Stroke CenterNo potential conflicts to discloseNIH/NINDS K23NS054084NIH/NINDS K23NS054084June 20,2009June 20,2009Tiantan International Stroke Conference,Beijing,ChinaTiantan International Stroke Conference,Beijing,China DSLObjectives 目的Overview of current imaging in intracranial atherosclerosisConsider arterial lesions beyond degree of luminal stenosisUnderscore the paramount role of collateral perfusionExplore how imaging of intracranial atherosclerosis may provide insight on many other aspects of ischemic stroke DSLIschemia 缺血Ischemia as principal pathophysiologic event,collaterals may avert any detrimental effect of thrombosis or plaqueVast majority of efforts focus on restoration of antegrade flow,anti-thrombotics or arterial manipulation DSLHemorrhagic transformation 出血性转化Severe ischemia and neurovascular compromiseReperfusion rate ischemic pre-and post-conditioning driven via collaterals DSL DSLTime is not brain!Across population of stroke cases studied from onset to chronic phase,not in a given patient during early phasesTime of symptom onset time of vascular occlusionCollaterals prone to failure over time DSL DSLMoyamoyaIdeal model of collaterals in chronic cerebral ischemiaStriking consistency in stenotic arterial featuresUniversal dependence on leptomeningeal collateralsStrokes are typically limited in extent,allowing for recoveryTIA without stroke should be goalRecurrent ischemia and temporal features easily studied DSLImaging Tools to understand physiology,mechanismsNot TCD vs.CTA vs.MRA vs.DSA etc.Distinct features emphasized by each modalityAnatomyFlow DSLTrialsSONIA in parallel with WASIDSAMMPRISOpportunity to discern mechanismsNot just%stenosisFlow at lesion QMRATCD with embolus detectionCollaterals on DSA DSLN of 1-intracranial atherosclerosisDiagnosis,rational selection of therapy Peri-procedural/serial imagingOpportunity to revise treatment strategy DSLLesion/stenosisLength and architecture of plaque DSLTCD beyond velocitiesMFV,PI,AUCCurrent TCD velocity criteria for diagnosis of intracranial stenosis are insensitive to age-and gender-related changes in cerebral blood flowTCD diagnosis based on published criteria may overestimate disease prevalenceLab-specific norms may need to be established to account for demographic features of population DSLCTA DSLCTA SpecificsLuminal contents identified by filtering of image data based on Hounsfield unitsContrast opacification errors due to administration,dosing,body habitus,injection rate,bolus trackingWindow width and center levels are critical Measurement of stenosesErroneous fusion of different structuresAutomated bone segmentationCalcifications DSLCTA SpecificsPredominantly anatomical angiographic imagesInsensitivity to flow physiologyLack of temporal resolution DSLFuture of CTAImproved image quality with increasing multidetector capabilityPossible temporal resolutionContrast modificationsAlternative agentsRefined dosing schemes DSLCTA of Stented Artery DSLMRAFlow-sensitive nature of MRIGradient echo imaging rapid scan acquisition with particular sensitivity to flow-related effectsMRA techniques include time-of-flight(TOF)and phase contrast(PC)imaging DSLTime-of-Flight(TOF)MRAInflow of fully magnetized spins into a slice saturated by a radiofrequency pulse2D TOF MRA thicker slices,lower signal-to-noise ratios3D TOF MRA thin slices,higher signal-to-noise ratiosSignal intensity dependent on slice thickness,flip angle,TE Magnetization transfer contrast for background suppression Multiple overlapping thin slab angiography(MOTSA)DSLTime-of-Flight(TOF)MRA DSLGadolinium-Enhanced MRAImproves vascular signal intensityReduces signal dropout due to slow flowDynamic or time-resolved MRATemporal resolution offset by limited spatial resolution DSLGadolinium-Enhanced MRA DSLPhase Contrast(PC)MRA相位对比MRAPC MRA employs gradients that induce phase shifts in flowing blood with subsequent background subtractionGradient modifications allow for determination of flow direction and flow rateMapping of flow directionMeasurement of flow ratesDifferentiating slow flow from frank thrombosis DSLMRA SpecificsInferior spatial resolutionLimited depiction of arterial structures beyond the proximal cerebral vesselsFlow-related artifacts due to slow or turbulent flowSusceptibility artifact induced by metal clips or objectsPatient motion may limit diagnostic utility DSLMRA Artifacts DSLWillisian collateralization DSLDSA DSLMethodology/critical appraisal?Frequently encounter case series biased by other selection biasesSONIANot how we generally use techniquesRule-out disease DSLIntracranial vascular calcification DSLPlaque stabilityUSPIOVessel wall imaging T1-weighted post-contrast images DSLGRE and phase imagingGRE =background magnitude +phase =+DSLFVHSlow collateral flow distal to severe lesionsRelatively rare DSLAsymptomatic stenosesWASID,18.9%(n=14/74)on 4-vessel cerebral angiography and 27.3%(n=65/238)on MRADuring a mean follow-up period of 1.8 years,no ischemic strokes were attributable to an AIS on angiography and 5 ischemic strokes(5.9%,95%CI:2.1%to 12.3%)occurred in the AIS territory on MRA(risk at 1 year=3.5%,95%CI:0.8%to 9.0%)Adequate collateral flow may offset ischemia and help embolic washout DSLCollaterals in WASIDCollateral circulation assessed on 287/569(50%)angiograms performed at study entry demonstrating proximal arterial stenosis ranging from 50-99%ASITN/SIR collateral grade broadly distributed0(absent or no collaterals)in 69%1(slow,minimal)in 10%2(more rapid,yet incomplete perfusion of territory)in 7%3(complete but delayed perfusion)in 11%4(rapid and complete collateral perfusion)in 4%DSLCollaterals and WASID stenosesThe extent of collateral flow for all proximal arterial lesions(ICA,MCA,BA,VA)correlated with the percentage of stenosis(p0.001),with more severe stenoses exhibiting greater compensatory collateral flowThe full range of collateral grades,however,was evident at all percentages of stenoses DSLCollaterals determine stroke riskAcross all percentages of stenosis,extent of collateral circulation an independent predictor for subsequent stroke in territory of symptomatic artery(p0.0001)DSLDivergent risk patternsContradictory role of collaterals in cases with severe (70%)versus mild or moderate stenoses(50-69%)Protective influence of collaterals in severe stenosisOminous marker of future stroke in milder stenosesCox proportional hazards model confirmed this interaction(p=0.001)DSLWhy have collaterals at 54%?DSLSevere stenosesGood collateralsPoor collateralsNo collaterals DSLCollaterals as ominous markerConversely,at milder degrees of stenoses(50-69%)the presence of more extensive collaterals was associated with a greater likelihood of subsequent stroke(p0.0001)DSLMild or moderate stenosesGood collateralsNo collateralsPoor collaterals DSLCollateral perfusion on DSA DSLNoninvasive perfusion imagingNot CT versus MRI,but reflection of physiology!Must consider identical perfusion parametersMultiparametric approaches to understand hemodynamics and evolving ischemia DSLCBVMTTCBFCT perfusion DSLCT perfusion DSLMR perfusion or PWI DSLMeaningless delays?DSLFocal delay without symptoms DSL ChangeCBF OEF CMRO2 CBVMTT(CBV/CBF)VR(hypercapnia)VR(hypocapnia)Hemodynamics and CBV!DSL DSL DSLCBVIschemic stroke due to IA results in dramatically less severe neurological deficits and diminished lesion volumesTime-domain perfusion parameters(e.g.Tmax)may not accurately reflect ischemic severity unless stroke etiology is considered DSLLow perfusion hyperemiaHyperemia,or increases in relative CBV values,were evident in regions of“oligemia”in all cases(p0.001).CPP was decreased,exhibiting a gradient extending deep from the cortical surface in all cases(p0.001).DSLLessonsIntracranial atherosclerosis newest area,yet critical lessons about ischemic stroke in generalRisk/benefit estimates ultimate goal,especially with high-risk interventions such as stentingImaging for detection and triage based on mechanism,expected prognosis at serial timepoints!。