P53

SECOND WORLD CONGRESS OF STROKE, 1992

P54

S29

SERIALMRI IN NORMAL ELDERLY PEOPLEWITH WHITE MATIER HYPERINTENSITIES

G. A. Rosenberg, C. C. Ford, M. Wesley, B. L Hart,R. L Rhyne, P. J. Garry, Albuquerque, NM, USA

There is a high incidence of incidental white matterhyperintensities (WMH) in normal elderly people. Thesechanges have been attributed to cerebrovascular diseaseor to increased perivascular water. Earlier, we reportedthat 30% of normal elderly had moderate or severe WMH(Neurolo~, 38:371,1988) without concommitant intel­lectual changes. To determine the changes in WMH overtime, we performed repeat MRI in those with severeWMH. In 1988/89 we scanned 108 elderly people thatwere in a normal aging process study; of those, 7 wereidentified with severe WMH (mean age 86.8.±.4.8 years)and compared with an age-matched group of 7 peoplewithout WMH (mean age 86.8.±.3.7 years). Thirteen ofthem agreed to have a repeat MRI scan. Volumemeasurements of WMH showed an increase in allsubjects. Those with WMH increased from an average of35.0.±.13.3 to 52.5.±.21.4 ern? (mean .±. SEM) whilecontrols showed an increase of 2.8.±.1.1 to 6.1.±.1.9 crrr',These results argue against a cerebrovascular etiology asthe cause of the WMH since they increased uniformly inall subjects. Rather, the results suggest that WMHobserved on MRI in healthy elderly people are probablydue to an increase in brain water.

P55

SLOW BLOOD FLOW IN THE VERTEBROBASILAR SYSTEM:A CLINICAL-RADIOLOGICAL ENTITY DISCOVERED THROUGH MRI.

Rangel-Guerra Ricardo MD, FACP; Martinez R. Hector MDMarfil-Rivera Alejandro MD.Apdo. Postal 1-4469 Mitras. Monterrey, N.L. Mexico.

In order to define the predictive variables ofbrainstem ischemia with or without infarctiondiagnosed through MRI, we studied 47 consecutivecases. Inclusion criteria: 1.- Patients withbrainstem syndromes on clinical basis; 2.- Basilarartery visible on ~ffiI in saggital and axial views,on Tl and T2 weighted images. There were two groups:Patients with infarct (BSI, brainstem infarction),and without infarct (SFS, slow blood flow syndrome).Patients were either treated with heparin orantiplatelet drugs. MRI follow up was performed onemonth later in most cases. The variables wereanalyzed with Fisher's exact and Mann-Whitney test.Twenty six cases presented SFS and 21 BSI.Differences among SBS/BSI were: Age:72.3/62.2 years(p=0.015, 95% confidence interval [CII 2.08-18.21);pre-hospitalization time: 544/44 hours (p=0.008; •CI 141-858); smoking: 2/8 (p=0.02, CI 0.02-0.086);death:l/6 (p=0.03). All cases in SFS group recoveredwith exception of one which died of an unrelateddisorder. Our results support the existence of a newclinical-radiological entity. We believe that thisconstitutes the radiological proof of the oldconcept of vertebrobasilar insufficiency.A pathophysiological basis of these findings remainto be elucidated.

MRI Advances in Cerebrovascular Disease: FlowDirection and Collateral Flow Patterns using MRAngiography

Steven Warach, MD PhD and Robert R. Edelman, MDBoston, MA USA

The direction ofarterial flow and presence or absenceofcollateral circulation can be determined non-invasively byMR angiography. Presaturation pulse slabs selectively placedover cerebral arteries prior to the RF excitation pulses eliminatethe signal frolp vessels within the territory supplied by thepresaturated artery. By systematic, sequential placement ofpresaturation slabs over each vascular territory duringangiographic imaging the source of flow to and direction offlow within a vessel can be identified. A complete study offlow dynamics in the Circle ofWillis and its major branchescan be done in less than 15 minutes using a 2D FLASHsequence (TRifF/flip angle=31 m&'10 ms'30 degrees, matrix= 192 x 256, field-of-view = 23 x 17 em, 5-10 mm slicethickness, one excitation, 8 second acquisition time per slice).Examples ofabnormal flow that have been observed include:left-to-right flow across the anterior communicating artery intothe right MCA in a case of right internal carotid arteryocclusion, retrograde flow through the ophthalmic artery in acase of left internal artery occlusion, fetal type origin oftheposterior cerebral artery, retrograde flow in the vertebral arteryin a case ofsubclavian steal.

P5G

Correlation of MagneticResonance(MR)Imagingof StrokeandClinicalHistoryof Stroke in the Cardiovascular HealthStudy

R. N. Bryan, A. Elster, C. Jungrcis,T. A. Manolio,V. Poirier,L. D. Schertz, U.S.A.

The specificaim of this study is to comparemanifestations ofstrokeon MR imagingwith patients'clinicalhistoryregardingstrokein order to evaluaterelativesensitivityof strokedetectionin a largeepidemiologiestudyof strokerisk factors. Weprospectively performedMR imagingof the brainon 297 menand womenselectedfroma cohortof 4000individuals age 65orolderparticipating in the Cardiovascular HealthStudy(CHS),alargemulticenter studyof riskfactorsforcardia-andcerebro­vasculardisease. 97 participantshad a clinicalhistoryof stroke.200 did not, Historyof strokewasobtainedby participantinterviewprior to MR imaging. MR was performedat four fieldcentersusinga standardexamination protocol. MRimageswereevaluatedfor strokewithoutclinicalhistoryby neuroradiologistsusinga structuredinterpretation protocol. Strokeswerearbitrarilydefinedas small« 3 mm)or large(~3 mm). Non­specificwhitemattersignalabnormalities werenotclassifiedasstroke. MR evidenceof strokewas presentin 75 of97participantswith a historyof stroke(77%),and in 45 of 200participantswith no historyof stroke (23%).89 of 120strokeswere large,32 of which were in partieipantswithouthistoryofstroke. Intra-and inter-readerreliability of MR imageinterpretation was high for largeinfarcts (KAPPA= 0.71 and0.78respectively). For small infarcts,intra-readerreliabilitywashigh(KAPPA=0.11), but inter-reader reliabilitywas low(KAPPA=0.32). We concludethat MRevidenceof strokeandclinicalhistoryare highly,but imperfectly correlated,especiallyfor smallstrokes. In this study, moresubjectshad stroke by MRcriteria thanclinical criteria. MR may improvethe sensitivityofstrokedetection.though thereremainsa questionof specificity.

S30

P57

SECOND WORLD CONGRESS OF STROKE, 1992

PS8

CLINICAL SIGNIFICANCE OF TC99 HMPAO SPECT INNONHEMORRHAGIC STROKE OBSERVED WITHIN THE FIRST ~HOURS.Censori B.Camerlingo M.°VirottaG. °Bertocchi C,Casto L,FerraroB,ServalliMC:Medoiago G.MamoliA.2nd Neurol.Dept. and °Nuclear Med.Dept.,OspedaliRiuniti,Bergamo,ltaly.

Functional imaging of the brain could be of major interest to indicate thebest therapeutic strategies for acute nonhemorrhagic stroke byproviding data on the actual condition of the damaged tissue. For thataim we have evaluated Tc99-HMPAOSPECT in 21 consecutive patientswith a nonhemorrhagic major stroke in the carotid artery territories withinthe 6th hour of the onset of symptoms and on the 7th day .Protocolincluded a clinical evaluation (with the Canadian scale.CNS)and a CTscan at the same times of SPECT,and angiography or ultrasound withinthe 8th hour of stroke.Outcome was measured withthe Rankin score onday 30 and 180.With the Fisher exact test,admission SPECT wassignificantly related to 1 month outcome (p=O.029),to both the CNSscores (p=0.017 and p:0.006) and to the demonstrabon of occluded ICAor MCA(p:0.017},whereas the 7th day SPECT was related to the 7th day,eNS score (p=O.017}and to vessel occlusion(p--o.032}.Norelationshipwas found between both SPECTs and long-term outcome and site andsize of the ischemic areas in early or delayed CT scans. In additiondelayed SPECT did not correlate with short term outcome as well.Ourdata suggest that Tc99-HMPAOSPECT at entry might be an useful toolfor the selection of patients to treat and the prediction of short-termoutcome of stroke.

P59

serial MRI and Proton MRS in Acute Completeand incomplete cerebral Ischemia in the Baboon

L.H. Monsein, V.P. Mathews,P.B. Barker, SoJ. Blackband, J.C. Chatham, R. N.Bryan

USA

purpose: This study compares acute changes in MRIs1gnal 1ntensity and proton MRS in the basal gangliausing a baboon model of complete or incompleteischemia.Materials and methods: Complete ischemia wasprOduced 1n 6 adult baboons by occluding the middlecerebral artery (MCA) with glue. Incomplete ischemiawas produced in 4 animals by partially occluding theMCA with a catheter. rCBF was documented withmicrosphere techniques. Serial MRI and proton MRSwere performed on a 1.5 T clinical instrument usingroutine spin echo and IVIM sequences and the "STEAM"sequence to obtain spectra from 2cm3 voxels from thebasal ganglia.Results: Increased lactate preceded MRI signalchanges and was seen in all animals from the earliestmeasurement (10 minutes) to the latest measurement'(12 hours) after initiation of ischemia. Lactatelevels were higher and plateaued later in completeischemia than in incomplete ischemia. MRI signalchanges were seen earliest on T2WI. These changesoccurred earlier and involved a larger area inincomplete ischemia than complete ischemia.Conclusions: Complete and incomplete cerebralischemia result in temporally and quantitativelydifferent changes in MRi signal proton MRS spectrareflecting their distinctly different pathophysiology.

Morphologic sequels of migraine: A MRI study

Payer F., Fazekas F., Koch 1.1., Schmidt R., Offenbacher H., Freidl W..Lechner H., Dept of Neurology, Karl-Franzens-University, Graz, Austria

Studies on the prevalence of MRI signal abnormalities in the brains ofmigraineurs yielded controversial results. In order to provide further dataon this issue we reviewed the MRI scans of 38 migraine patients withoutcurrent neurologic symptoms (mean age 35.8 +/. I 1.9 years). In addition,we compared the findings in those 24 migraineurs under 50 years withoutmajor cerebrovascular risk factors (mean age 30.1 +/-9.0 years) to that in14 headache and risk factor free volunteers (mean age 37.8 +/-5.3 years).Overall. focal ares of hyperintense signal were seen in 15 (39%) patients.Lesion prevalence varied according to the type of headache (18%) inmigraine without aurea, 53% in migraine with typical aura, 38% inbasilar migraine). The subset of migraine patients under 50 yearsexhibited MRI signal abnormalities more than twice as often thancontrols (33% vs. 14%). Punctate white matter hyperintensities were thepredominant finding and were seen in 10 of 15 individuals with MRIlesions. More striking signal abnormalities consisted of symmetricalareas of hypcrintcnsity lateral to the posterior horns in two 24 year oldpatients and of extensive white matter damage with lacunar infarcts in a59 year old woman.Our findings confirm a higher prevalence of MRI lesions in a mixedgroup of migraineurs than in headache free individuals. Signalabnormalities are most often uncharacteristic. however, and theiroccurence relates to the type of migraine.

P60

MRI Advances in Cerebrovascular Disease: DiffusionImaging ofAcute Human Stroke

Steven Warach, MD PhD and Robert R. Edelman, MDBoston, MA USA

The molecular diffusion of water can be imaged usingMRI. The work of Moseley and others in animal models ofce.reJ:ral ~schemia have shown decreases in diffusion by MRIWithin rmnutes after the onset of focal cerebral ischemia; thisdecrease is evident before changes on T2-weighted images(T2WI). We studied 32 patients 1-4 times (as early as 105minu~es) after stroke onset. A turboSTEAM (Merboldt, et al.)technique was used on a 1.5 T whole body imaging systemwith standard hardware (Siemens) with a TR=12ms,TM~8(}(}ms, flip angleel Iv, gradient duration=4ms, diffusiongradient strength =9~T/m in the x and y directions, matrix=128x128, field ofvlew=280, a minimum of 4 acquisitions.The b value was 294.1 sec/mm2.

Diffusion-weighted imaging revealed infarcts soonerthan T2WI; four hyperacute infarcts four hours or less wereonly shown by diffusion-weighted imaging. Acute infarctshad lower apparent diffusion coefficients (ADCs) than non­infarcted regions. This difference reached a nadir in the first24 hours and rose progressively thereafter. Chronic infarctsshow~d an increase in dU:fusion and were distinguishable fromacute infarcts. The technique takes less than two minutes toapply in the clinical setting. Diffusion-weighted imaging hasthe potential to playa role in improving the early anatomicaldiagnosis of stroke and therefore in the development of clinicaltrials for and implementation of early stroke interventions.