Journal of Neurology, Neurosurgery, and Psychiatry 1992;55:98-104 Non-invasive diagnosis of internal carotid artery dissections W Muillges, E B Ringelstein, M Leibold Department of Neurology, Klinikum der RWTH, Aachen W Muillges E B Ringelstein M Leibold Correspondence to: Dr Mullges, Neurologische Klinik, Klinikum der RWTH, Pauwelsstral3e, D- 5100 (W) Aachen, Germany. Received 5 December 1990 and in revised form 28 May 1991 Accepted 11 June 1991 Abstract Arteriography is thought to be manda- tory for the diagnosis of internal carotid artery (ICA) dissection. With the intro- duction of transcranial Doppler sono- graphy (TCD) and magnetic resonance imaging (MRI), however, this is no longer the case. In 13 consecutive patients with ICA dissections the diagnosis was made by means of non-invasive tests including extracranial and transcranial Doppler sonography, contrast enhanced com- puted tomography (ceCT), and, in five patients, MRI. Intra-arterial digital sub- traction angiography used as the gold standard in all cases was confirmative. Extracranial and transcranial ultrasound findings indicative of the diagnosis could be identified. MRI directly demonstrated the intramural haematoma and the false lumen of the dissected artery. These non- invasive techniques also allowed for re- petitive follow up examinations. They were, however, unable to demonstrate false aneurysms in the chronic state. Results show that the diagnosis of dissection can be made by mt cerebrovascular ultrasound and N O. ; ::...... ~ : ... Figure I Submandibular approach to distal internal carotid artery by transcra Doppler sonography. (a) Schematic diagram of dissected ICA. Probe is placed i submental triangle. Three different positions of sample volume are shown (black silhouette indicates false lumen). (b) Practical application. Spontaneous dissecting aneurysms of the inter- nal carotid artery (ICA) and other brain- supplying arteries are an essential cause of acute stroke (1%) particularly in the young (5%).' Clinical signs leading to the diagnosis are well known.' Up to now, cerebral intra- arterial angiography was thought to be man- datory for diagnosing dissecting aneurysms, 13 while only a few reports referred to the non- invasive diagnosis of ICA dissections by means of ultrasound Doppler sonography,4-0 contrast enhanced computed tomography (ceCT)311-14 and magentic resonance imaging (MRI).'° 12 15 16 In the past three years we have used these non- invasive methods in 13 consecutive patients with ICA dissections to make the initial diag- nosis and to perform close-meshed follow up studies and compared these findings with those of subsequent intra-arterial selective cerebral arteriography. This was done to evaluate non- invasive techniques by the conventional gold standard and to clarify how far angiography may be dispensable. carotid Patients and methods eans of Clinical evaluation-History of premonitory IRI. complaints and clinical findings were evaluated on admission, and outcome was checked for at least three months. The longest observation period was 28 months after the stroke. Charac- teristic clinical signs for presumptive diagnosis were ischaemic strokes, either transient or completed, and accessory symptoms due to direct damage of neighbouring structures by the dissected vessel wall such as carotidynia, ipsilateral head pain, Horner's syndrome, cran- ial nerve lesions, and pulsatile tinnitus. Sonography-All ultrasound examinations were done before arteriography-that is, inves- tigators were not aware of the final diagnosis. Extracranial cw-Doppler sonography of all brain-supplying arteries in the neck was per- formed initially in 11 of the 13 patients with a zero-crossing flowmeter with a bidirectional 4 MHz Doppler device (Delalande D-800R). In 10 patients, an initial transcranial pulsed Doppler ultrasound examination was also per- formed (EME TC 2-64R, 2 MHz) with insona- tion of the retromandibular ICA segment from a submandibular approach (fig 1), transorbital insonation of the carotid siphon, and transtem- poral access to the middle, anterior, and pos- terior cerebral arteries (MCA, ACA, PCA), as inial well as the very distal part of the ICA (Cl- segment). Transcranial insonation of the ver- tebrobasilar system from suboccipital 98 on 2 March 2019 by guest. Protected by copyright. http://jnnp.bmj.com/ J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.55.2.98 on 1 February 1992. Downloaded from
Transcript
Journal ofNeurology, Neurosurgery, and Psychiatry 1992;55:98-104
Non-invasive diagnosis of internal carotid arterydissections
W Muillges, E B Ringelstein, M Leibold
Department ofNeurology, Klinikumder RWTH, AachenW MuillgesE B RingelsteinM LeiboldCorrespondence to:Dr Mullges, NeurologischeKlinik, Klinikum derRWTH, Pauwelsstral3e, D-5100 (W) Aachen, Germany.Received 5 December 1990and in revised form28 May 1991Accepted 11 June 1991
AbstractArteriography is thought to be manda-tory for the diagnosis of internal carotidartery (ICA) dissection. With the intro-duction of transcranial Doppler sono-graphy (TCD) and magnetic resonanceimaging (MRI), however, this is no longerthe case. In 13 consecutive patients withICA dissections the diagnosis was madeby means of non-invasive tests includingextracranial and transcranial Dopplersonography, contrast enhanced com-puted tomography (ceCT), and, in fivepatients, MRI. Intra-arterial digital sub-traction angiography used as the goldstandard in all cases was confirmative.Extracranial and transcranial ultrasoundfindings indicative of the diagnosis couldbe identified. MRI directly demonstratedthe intramural haematoma and the falselumen of the dissected artery. These non-invasive techniques also allowed for re-petitive follow up examinations. Theywere, however, unable to demonstratefalse aneurysms in the chronic state.Results show that the diagnosis ofdissection can be made by mtcerebrovascular ultrasound and N
O.;::......~
: ...
Figure I Submandibular approach to distal internal carotid artery by transcraDoppler sonography. (a) Schematic diagram of dissected ICA. Probe is placed isubmental triangle. Three different positions of sample volume are shown (blacksilhouette indicatesfalse lumen). (b) Practical application.
Spontaneous dissecting aneurysms ofthe inter-nal carotid artery (ICA) and other brain-supplying arteries are an essential cause ofacute stroke (1%) particularly in the young(5%).' Clinical signs leading to the diagnosisare well known.' Up to now, cerebral intra-arterial angiography was thought to be man-datory for diagnosing dissecting aneurysms, 13while only a few reports referred to the non-invasive diagnosis of ICA dissections by meansof ultrasound Doppler sonography,4-0 contrastenhanced computed tomography (ceCT)311-14and magentic resonance imaging (MRI).'° 12 15 16In the past three years we have used these non-invasive methods in 13 consecutive patientswith ICA dissections to make the initial diag-nosis and to perform close-meshed follow upstudies and compared these findings with thoseof subsequent intra-arterial selective cerebralarteriography. This was done to evaluate non-invasive techniques by the conventional goldstandard and to clarify how far angiographymay be dispensable.
carotid Patients and methodseans of Clinical evaluation-History of premonitoryIRI. complaints and clinical findings were evaluated
on admission, and outcome was checked for atleast three months. The longest observationperiod was 28 months after the stroke. Charac-teristic clinical signs for presumptive diagnosiswere ischaemic strokes, either transient orcompleted, and accessory symptoms due todirect damage of neighbouring structures bythe dissected vessel wall such as carotidynia,ipsilateral head pain, Horner's syndrome, cran-ial nerve lesions, and pulsatile tinnitus.
Sonography-All ultrasound examinationswere done before arteriography-that is, inves-tigators were not aware of the final diagnosis.Extracranial cw-Doppler sonography of allbrain-supplying arteries in the neck was per-formed initially in 11 of the 13 patients with azero-crossing flowmeter with a bidirectional4 MHz Doppler device (Delalande D-800R).In 10 patients, an initial transcranial pulsedDoppler ultrasound examination was also per-formed (EME TC 2-64R, 2 MHz) with insona-tion of the retromandibular ICA segment froma submandibular approach (fig 1), transorbitalinsonation of the carotid siphon, and transtem-poral access to the middle, anterior, and pos-terior cerebral arteries (MCA, ACA, PCA), as
inial well as the very distal part of the ICA (Cl-segment). Transcranial insonation of the ver-tebrobasilar system from suboccipital
98 on 2 M
arch 2019 by guest. Protected by copyright.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.55.2.98 on 1 February 1992. D
approach was also performed. Methodologicaldetails of these techniques are published else-where.7 1617Abnormal flowwas diagnosed ifflowvelocities were either two standard deviationslower or higher than normal values.'6 In fivepatients B-mode imaging of the carotid bifur-cation and proximal segment of the ICA wasalso performed (Biosound 2000 SAR,7-5 MHz).
Intra-arterial angiography was performed inall cases, in 11 on admission and in two aftertwo weeks. The common carotid artery wascatheterised selectively via a transfemoralapproach. The injection volume of a non-ioniccontrast medium was 5 to 7 ml. At least twoviews of the carotid axis were imaged on digitalsubtraction angiograms. In each case, a biplanevertebrobasilar imaging series was also perfor-
med by selectively injecting into the left (10) or
right (3) vertebral artery.
Cranial computed tomography (CT) was donein every patient with a Somatom DRH-scanner(SiemensR, matrix 512*512). The infratentorialslice thickness was 4 mm and 8 mm supraten-torially. Contrast enhancement was used tovisualise the carotid artery with its bifurcationand the siphon."'3 Cerebral infarctions visibleon CT were categorised according to a classi-fication system recently published else-where. 17-19 Territorial infarcts, which werethought thromboembolic in origin andhaemodynamically caused low flow infarcts,were differentiated.MRI with visualisation of the cervical
ICA was performed in six patients (SiemensMagnetomR, 1S5T). Images were Ti- and
99 on 2 M
arch 2019 by guest. Protected by copyright.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.55.2.98 on 1 February 1992. D
Figure 2 Typical transcranial Doppler sonographic finding in ICA dissection. Left and middle: submandibularinsonation reveals high flow velocities (266 and 162 cm/s) at 40 and 50 mm insonation depths, indicating high gradestenosis. Right: poststenotic lowflow (62 cm/s) at 60 mm depth.
proton-weighted, as well as T2-weighted. Slicethickness was 5 mm without gap. The internalcarotid artery was visualised transversally andalso longitudinally.
Sonographic follow up investigations wereperformed in 12 of the 13 cases by extracranialand transcranial Doppler sonography. In sixcases, repeat arteriograms or repeat MRI orboth were also taken.
ResultsPATIENTS AND CLINICAL SIGNSFive women and eight men (mean (range) age46 (20-58) years with isolated unilateral dissec-tion of the left (12) and right (1) ICA wereincluded in the study. Four patients had suf-fered an unequivocal whiplash injury of theneck or a slight trauma without direct in-volvement of head and neck one to nine daysbefore the occurrence of suggestive clinicalsymptoms. In the nine other patients dissec-tions had occurred spontaneously. Vascularrisk factors were present in eight cases; five weresmokers, two women took contraceptive pills,and one had diabetes mellitus. Three patientshad two vascular risk factors; three had com-mon migraine.
Clinical signs and symptoms were quite varia-ble. Six patients had suffered a complete stroke,
¢ !S -NEW
Figure 3 B-mode imaging ofICA dissection in longitudinal (left) and transversal(right) views. Intimalflap (arrows) oscillated within blood stream.
all of them with premonitory attacks. Onepatient had suffered transient ischaemic attacks(TIA), five had prolonged reversible ischaemicneurological deficits with preceeding TIA inthree of them. While eight patients only hadischaemic events without any accessory symp-tom, one presented with accessory signs only-that is, she had pulsatile tinnitus, carotidyniaand temporal headache. Thus the combinationof both was present in four of the 13 patients.Four patients complained only of pain locatedeither temporally, retro-orbitally, or in thesubmandibular and carotid region. Anotherpatient showed involvement of more than onecranial nerve presenting with trigeminal dys-aesthesia, pulsatile tinnitus, and Homer's syn-drome. All patients underwent therapeuticalheparinisation with PTT 2 to 3 times normalcontrol. Initial and follow up findings aresummarised in the table.
TECHNICAL FINDINGSExtracranial cw-Doppler sonography per-formed in 11 patients before cerebral arterio-graphy showed a high-grade ICA stenosisretromandibularly at least 2 cm distal to thebifurcation in six cases. This finding wasthought to be characteristic of an ICA dissec-tion in terms of type and location of the lesion.Four other patients showed to-and-fromovement of the blood column in the proximalICA, possibly indicative of a blind arterialstump due to complete occlusion of the distalICA.7"9 Only a tentative diagnosis of ICAdissection was made in these patients as similarfindings may also be seen in embolic ICAocclusions. In one case a 90% ICA stenosis wasdetectable immediately at the carotid bifurca-tion, suggesting arteriosclerotic disease ratherthan dissection. During submandibular in-sonation of the ICA by means of transcranialDoppler sonography each of the seven extra-cranial stenotic lesions could be traced furtherdistally, some ofthem even extending up to thepars petrosa of the carotid artery. The typicalsonographic finding was a high-flow velocitysignal and post-stenotic disturbed flow indi-cated by bidirectional low frequency compo-nents of the flow signal (fig 2). In three cases,pronounced harmonic post-stenotic covibra-tions of the affected vessel wall segments gen-erated so-called "musical murmurs".20 In twopatients an intracarotid valve mechanism withfunctional ICA occlusion and stoppage of flow
100
...
on 2 March 2019 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.55.2.98 on 1 F
Non-invasive diagnosis of internal carotid artery dissections
A B C
_.'Ap
.JI .
Figure 4 Typical angiographic appearance ofICA dissections. (a/b) Graduallytapering ICA occlusion (arrows) during early and late phase of angiogram. (c) Longfiliform ICA stenosis exceeding 2 cm in length at base of skull (arrows) andpseudoaneurysmatic dilatation of artery (arrow head).
during diastole but a short-lasting jet-like flowsignal during systole could be detected. Duringthe additional transtemporal insonation of themiddle cerebral artery in nine patients, low-flow became obvious in five cases.B-mode sonography of the ICA in the neck
was performed in five cases. A false lumen withan intimal flap could be shown in only one ofthem (fig 3). In the other cases the carotidbifurcation and the proximal 1 to 2 cm of theICA appeared normal. In two cases flow signalwithin the proximal ICA appeared abnormalwith lack of diastolic or even reversed bloodflow indicating high peripheral resistance.These patients turned out to have either a distalICA occlusion or a pseudo-occlusion. Aninitial arteriography was performed afterultrasound investigations in 11 patients within24 hours after admission and in two within thefirst two weeks. Angiograms showed five ICAocclusions, four of them with the typicalgradually tapering appearance of the proximalstump (fig 4a, b). Eight ICA stenoses ofvariableextent were also seen (fig 4c). Four ofthem hadan extraluminal pouch of contrast mediumclose to a tight stenosis beyond the base of theskull. In the other cases the appearance of thelesions was also suggestive of but less specificfor ICA dissection, and final diagnosis wasbased on the rapid resolution of the lesionduring follow up. Angiograms were in accor-dance with Doppler sonography findings inevery case. The type of vessel lesion (occlusionv stenosis) did not correspond to certain clinicalsymptoms or the severity of the deficits.
Brain imaging by means of cranial CTshowed low flow infarctions in five and throm-boembolically induced territorial infarctions insix cases,'8 but was normal in two. (Details are
published elsewhere2"). The type and extent ofbrain infarctions did not depend on the type ofvessel lesion.
Direct imaging of the affected carotid arteryby means of contrast enhanced CT revealedhighgrade lumen narrowing in only two cases.Both ultrasound and arteriography were confir-mative. In five patients dissection of the artery
..... t.}4
FigureS T2-weightedparaxial image ofICAdissection. Hyperintense vessel wall signal (arrow heads)beyondpars petrosa of skull (asterisk) is abnormallybroad and indicates intramural haematoma.
101
ibb,..}* a.
n..<-
on 2 March 2019 by guest. P
rotected by copyright.http://jnnp.bm
j.com/
J Neurol N
eurosurg Psychiatry: first published as 10.1136/jnnp.55.2.98 on 1 F
Figure 6 Follow up ofICA dissection by means ofMRI imaging (T2-weighted, horizontal cut). (a) Initialexamination showed hyperintense sickle-like vessel wall haematoma (arrow) leaving only small residual, hypointenselumen with bloodflow (arrowhead). (b) Two weeks later, ICA lumen (arrowhead) nearly restored by resorption ofintramural haematoma (arrow). 1 = internal carotid artery, 2 = vertebral artery, 3 = jugular vein, 4 = spinalfluid,5 = dens axis, 6 = pharynx, 7 = medulla.
could be pinpointed directly by MRI. Theintramural haematoma presented as a hyper-intense vessel wall signal in both Ti- and T2-weighted images (fig 5). Again, findings were inaccordance to Doppler sonography andarteriography.
CLINICAL AND LABORATORY FOLLOW UPFINDINGSFour TIA patients subsequently remainedasymptomatic, another four patients with com-plete stroke had only minimal residual deficits.Five patients remained considerably handi-capped by their ischaemic deficit. Duringthe course of the treatment, only one patientshowed worsening of the neurological deficitdue to complete occlusion of an initially sten-osed ICA. Transcranical Doppler sonographyshowed normalisation in seven cases or, at least,severe reduction ofthe lesions in five cases. Thedegree of improvement did not depend on theinitial grade of the occluding lesions. In onepatient the initial typical occlusion proved byarteriography did not improve. In only onepatient did the initial 80% stenosis progress toan occlusion. There was no relation betweenclinical outcome or angiological follow up find-ings and the initial severity of the ICA lesion.Day by day transcranial Doppler studies alsorevealed rapidly improving intracerebralhaemodynamics paralleling ICA blood flownormalisation. Repeat angiography in fourpatients showed normalisation of the vessellumina. In four other cases follow up studies byMRI also showed normalisation of the vesselwall (fig 6) corresponding to the transcranialDoppler findings.
DiscussionWe studied 13 patients with isolated ICAdissection by means of a battery of non-invasive and invasive tests. The causes of thevessel lesion were whiplash injury of the neckand questionable blunt trauma in four patients,the remainder suffering from idiopathic spon-taneous ICA dissections. In no patient didarteriography deliver indications for fibro-muscular dysplasia, Marfan's syndrome, orcystic medionecrosis as the underlying vesseldisease.2227 Hypertension, cigarette smoking,migraine, and contraceptive pills are thought tobe risk factors for ICA dissections228 and werepresent in eight of our patients. In one patient,however, the dissection may have been athero-sclerotic in origin, as arteriograms showed anulcerated atherosclerotic plaque immediatelynext to the proximal end of the taperingstenosing dissection; while the latter dis-appeared, the plaque remained. Clinically thecombination of an acute stroke including TIAwith so called accessory symptoms, likeHomer's syndrome, pulsatile tinnitus, tem-poral pain, and carotidynia, strongly suggestsan underlying ICA dissection."'210 14 15 2831 Justone third of our patients, however, showed thiscombination of symptoms, suggesting why thetrue nature of the initial signs may beoverlooked and that technical investigations arenecessary to lead to the correct diagnosis.Up to now carotid arteriography was thought
to be mandatory for the diagnosis of ICA dis-section.'-3 Findings thought diagnostic32 arean extraluminal pouch filled with contrastmaterial; a double lumen; a smooth or scal-lopped narrowing of the ICA distal to the
102 on 2 M
arch 2019 by guest. Protected by copyright.
http://jnnp.bmj.com
/J N
eurol Neurosurg P
sychiatry: first published as 10.1136/jnnp.55.2.98 on 1 February 1992. D
Non-invasive diagnosis of internal carotid artery dissections
carotid sinus; a long filiform stenosis exceeding2 cm in length; and a gradually taperedocclusion. Some of these signs, however, areless specific and do not unequivocally prove anICA dissection in every case. l 2 27 32 33 Only eightof our 13 patients showed arteriographicallypathognomonic tapering occlusion or longfiliform stenosis. Other findings were suspiciousbut in these cases follow up findings proved thediagnosis correct. There is a need for diagnostictools which are both non-invasive and reliableas arteriography during the acute phase bears asubstantial risk of embolic complications.35Extracranial Doppler sonography, which wasperformed in 11 cases, showed a retroman-dibular high grade stenosis of a typical findingin most of the cases. The site and type of thislesion is highly suggestive of ICA dissection,particularly in the young and if risk factors orsigns of atherosclerosis are lacking. Flowabnormalities indicating a blind ICA stumpsuggest ICA occlusion distal to the bifurcation.This finding is less conclusive as to the under-lying vascular disease" but rather suggestsdissection than distal embolic ICA occlusion.Atherothrombotic ICA occlusion at this site israre.35
Transcranial Doppler sonography turnedout to be a sensitive technique for detectingarterial disease inaccessible to conventionalDoppler tests' and provided quite characteris-tic findings for detecting ICA dissection. Sten-oses were identified in all cases. The location ofthe lesions, either stenosis or occlusion, wasalso highly suggestive for the diagnosis. Therewas no false positive finding, although therewere too few cases for statistical analysis.Definite proof of the diagnosis might beachieved during repetitive TCD tests in thatboth stenoses and occlusions had a strongtendency to improve over time.' Day by dayultrasound checks demonstrated rapid ICArecanalisation in all but one case. A systolicvalve mechanism with transient occlusion ofthe distal ICA during diastole also seems to be aspecific TCD finding as to our knowledge it hasnot been described as a transient phenomenonin atherosclerotic ICA stenoses or pseudo-occlusions.37 ICA dissections were also foundto have considerable impact on intracranialhaemodynamics in that reduced intracranialperfusion pressure led to severe slowing ofblood flow velocity.2' Duplex scanning allowsfor rapid visualisation of ICA in the neck by B-mode imaging. There are, however, limitationsof this technique. Due to the distal localisationofthe lesions' 7 32 33 the characteristic intimal flapmay be shown only in a limited number ofpatients; we succeeded in only one out of fivepatients. Sector scanners with small trans-ducers have been reported to be more sensitivefor detecting a dissected ICA wall.'9
Early MRI was performed in five cases andled to the correct diagnosis in all ofthem, as wasproven by simultaneous arteriography in fourcases and by angiological follow up investiga-tions in all. The intramural haematoma wasindicated by a hyperintensive arterial wallsignal on both T 1- or proton-weighted and T2-weighted images, as has been reported by
others." Other stenosing processes likefibromuscular dysplasia, vasculitis, or athero-sclerosis may also be differentiated by MRI.'4 15
The false vessel lumen can directly be visual-ised and separated from the flowing bloodcolumn in both the residual and false lumen."1In our study, five initial and four follow upstudies of ICA dissections exactly corre-sponded in site and severity to Doppler andangiographic findings, indicating the high diag-nostic reliability and validity of this technique.In contrast, enhanced CT imaging of ICA inthe neck, as a rule, did not prove to bediagnostic.The diagnosis of an ICA dissection is
undoubtedly confirmed by rapid day by dayregression of the arterial lesion.' 3 We showedthis by monitoring lesions by close-meshed,repetitive transcranial Doppler sonographystudies. This is also true for progression of astenosis, which eventually may occur. RepeatMRI scans in four cases also showed normal-ised or nearly normalised arteries. Whethertreatment with therapeutic heparinisation orantiplatelet agents will prompt a better out-come is still a matter of debate.' 2922233239Interestingly, infarctions due to carotid arterydissections are as commonly thromboembolicin origin as haemodynamic.2' This mayinfluence the type of treatment.
In conclusion, patients with acute strokes,particularly the young, and those presentingwith accessory symptoms should be subjectedto intense extracranial and transcranialultrasound checks by means of ECD, TCD,and, if available, B-mode imaging. The sub-mandibular approach of transcranial Dopplersonography is most promising, the characteris-tic finding in ICA dissection being an extendeddistal stenosis. In the case of distal extracranialICA occlusion, with proximal "blind stump"phenomenon, the diagnosis is highly sugges-tive. Follow up ultrasound examinations oftendeliver further diagnostic clues as dissectionstend to regress (or progress) rapidly. Findingsmay be further clarified by means ofMRI as itmay demonstrate the underlying vessel walldisease. Both procedures are non-invasive,reliable, and quick. From our point of view,arteriography should be confined to certaincases in the post-acute period when regressionof the lesion has already taken place. At thistime, residual aneurysmata spuria may befound, which require surgical removal.30 Basedon our experiences, four further cases of ICAdissection have been diagnosed non-invasivelyby both transcranial Doppler sonography andMRI. Follow up findings strongly suggestedthe diagnosis to be correct.
We are greatly indebted to Christiane Goebel-Dizinger, GabiTietz, Waltraud Schumacher, and Vera Schappeler, vasculartechnicians, for considerable technical assistance.
and pathology. AJR 1987;148:393-8.4 deBray JM, Dubas F, Joseph PA, Causeret H, Pasquier JP,
Emile J. Etude ultrasonique de 22 dissections carotid-iennes. Rev Neurol 1989;145:702-9.
5 Kaps M, Dorndorf W, Damian MS, Agnoli L. Intracranialhemodynamics in patients with spontaneous carotid dis-section: transcranial ultrasound follow-up studies. EurArch Psychiatr Neurol Sci 1990; in press.
6 Kotval PS, Babu SC, Cozzi A, Barakat K. Role ofthe intimalflap in arterial dissection: sonographic demonstration.AJR 1988;15O:1181-2.
7 Ringelstein EB. Continuous-wave doppler sonography ofthe extracranial brain-supplying arteries. In: WeinbergerJ, ed. Noninvasive imaging of cerebrovascular disease. NewYork: Alan R Liss, 1989:27-48.
8 Rohr J, Gauthier G. Hematomes dissequants spontanessuccessifs a 3 ans et demi d'intervalle des deux arterescarotides internes. Rev Neurol 1987;143:115-20.
9 Steinke W, Aulich A, Hennerici M. Diagnose und Verlaufvon Carotisdissektionen. Dtsch Med WSchr1989;1 14:1869-75.
10 Bogousslavsky J, Despland PA, Regli F. Spontaneouscarotid dissection with acute stroke. Arch Neurol1987;44:137-40.
11 DalPozzo G, Mascalchi M, Fonda C, Cadelo M, Ronchi 0,
Inzitari D. Lower cranial nerve palsy due to dissection ofthe internal carotid artery: CT andMR imaging. J ComputAssist Tomogr 1989;13:989-95.
11 Heinz ER, Pizer SM, Fuchs H. Examination of theextracranial bifurcation by thin-section dynamic CT:direct visualization of intimal atheroma in man. AJNR1984;5:355-9.
15 Lieschke GJ, Davis S, Tress BM, Ebeling P. Spontaneousinternal carotid artery dissection presenting as hypoglos-sal nerve palsy. Stroke 1988;19:1151-5.
16 Ringelstein EB. A practical guide to transcranial Doppler-sonography. In: Weinberger J, ed. Noninvasive imaging ofcerebrovascular disease. New York: Alan R Liss, 1989:75-121.
17 Ringelstein EB, Koschorke S, Holling A, Thron A, Lam-bertz H, Minale C. Computed tomographic patterns ofproven embolic brain infarctions. Ann Neurol1989;26:759-65.
18 Ringelstein EB, Zeumer H, Angelou D. The pathogenesis ofstrokes from internal carotid artery occlusion. Diagnosticand therapeutical implications. Stroke 1983;14:867-75.
19 Reutern von GM. Free hand Doppler techniques for theexamination of the extracranial arteries with continuouswave Doppler. In: Spencer MP, ed. Ultrasonic diagnosis ofcerebrovascular disease. Doppler techniques and pulse echoimaging. Dordrecht, Boston, Lancaster: Martinus Nijhoff,1987:157-77.
20 Aaslid R, Nornes H. Musical murmurs in human cerebralarteries after subarachnoid hemorrhage. J Neurosurg
Reiche W. Patterns of brain infarctions in internal carotidartery dissections. Neurosurg Rev 1991;14:111-13.
22 Anderson CA, Collins GJ jr, Rich NM, McDonald PT.Spontaneous dissection of the internal carotid arteryassociated with fibromuscular dysplasia. Am Surg1980;46:263-6.
23 Correll JW, Quest DO, Carpenter DB. Nonatheromatouslesions of the extracranial cerebral arteries. In: Smith RR,ed. Stroke and the extracranial vessels. New York: RavenPress, 1984:312-31.
24 Mettinger KL. Fibromuscular dysplasia and the brain:current concept of the disease. Stroke 1982;13:53-8.
25 Osborne AG, Anderson RE. Angiographic spectrum ofcervical and intracranial fibromuscular dysplasia. Stroke1977;8:617-26.
26 Austin MB, Schaefer RF. Marfan's syndrome with unusualblood vessel manifestations: primary medionecrosis dis-section of right innominate, right carotid, and left carotidarteries. Arch Pathol 1957;64:205-9.
27 Luken MG III, Ascherl GF jr, Correll JW, Hilal SK.Spontaneous dissecting aneurysms of the extracranialinternal carotid artery. Clin Neurosurg 1979;26:353-5.
28 Shuster E, Mokri B, Sundt TM, Houser OW, Piepgras D.Spontaneous dissection of the internal carotid arteries:clinical presentation, angiographic features and outcome.Neurology 1985;35:suppl 1:212.
31 West TET, Davies RJ, Kelly RE. Horner's syndrome andheadache due to carotid artery disease. Br Med J1976;1 :818-20,
32 Fisher CM, Ojemann RG, Robertson GH. Spontaneousdissection of cervico-cerebral arteries. Can Neurol Sci1978;5:9-19.
33 O'Dwyer JA, Moscow N, Trevor R, Ehrenfeld WK, New-ton TH. Spontaneous dissection of the carotid artery.Radiology 1980;137:379-85.
34 Rieger P, Piepgras U. Nebenwirkungen und Komplik-ationen bei neuroradiologischen Kontrastuntersuchun-gen. RoPraxis 1984;37:89-98.
35 Hass WK, Fields WS, North RR, Kricheff II, Chase NE,Bauer RB. Joint study of extracranial arterial occlusion.JAMA 1968;203:159-66.
36 Ley-Pozo JA, Ringelstein EB. Noninvasive detection ofocclusive disease ofthe carotid siphon and middle cerebralartery. Ann Neuroll990;20:640-7.
37 Ringelstein EB, Berg-Dammer E, Zeumer H. The so-calledatheromatous pseudoocclusion of the internal carotidartery. A diagnostic and therapeutical challenge.Neuroradiology 1983;25:147-55.
38 Kapp JP, Smith RR. Spontaneous resolution of occlusivelesions of the carotid artery. J Neurosurg 1982;56:73-9.
