J7ournal ofNeurology, Neurosurgery, and Psychiatry 1993;56:1191-1199

Cranial nerve palsies in spontaneous carotidartery dissection

Matthias Sturzenegger, P Huber

AbstractTwo patients had isolated unilateral cra-nial nerve palsies due to spontaneousinternal carotid artery (ICA) dissectionwithout ischaemic cerebral involvement.One had acute glossopharyngeal andvagal, the other isolated hypoglossalnerve palsy. Reviewing all reported casesof angiographically confirmed ICA dis-section in the literature, 36 additionalcases with unequivocal ipsilateral cranialnerve palsies were analysed. While anisolated palsy ofthe IXth and Xth has notbeen reported previously, palsies of theXIIth nerve or the IXth to XIIth nerveswere most frequently found. In thesepatients, lower cranial nerve palsies areprobably the result of compression by anenlarging ICA due to mural haematoma.Symptoms and signs indicative of carotiddissection were concurrently presentonly in some reported cases. This raisesthe question of unrecognised carotid dis-section as a cause of isolated cranialnerve palsies. When the dissection occursin the subadventitial layer without rele-vant narrowing of the arterial lumen andwhen an aneurysm is thrombosed,angiography does not reliably yield thediagnosis. Therefore, carotid dissectionmight have been underestimated as acause of isolated lower cranial nervepalsies before the advent of MRI. MRIdemonstrates directly the extension ofthe wall haematoma in the axial and lon-gitudinal planes. Some arteriopathiessuch as fibromuscular dysplasia and tor-tuosity make a vessel predisposed to dis-section.

Department ofNeurology, UniversityofBerne, Inselspital,CH-3010 Berne,SwitzerlandM SturzeneggerDepartment ofNeuroradiology,University ofBerne,Inselspital, CH-3010Berne, SwitzerlandP HuberCorrespondence to:M Sturzenegger, MD,Department of Neurology,University of Beme,Inselspital, CH-3010 Berne,Switzerland.

Received 2 November 1992in revised form31 December 1992Accepted 14 January 1993

( Neurol Neurosurg Psychiatry 1993;56:1 191-1199)

The clinical presentation of spontaneousinternal carotid artery (ICA) dissection ishighly variable. Sudden, intense unilateralpain in the neck and face, ipsilateral ocu-

losympathetic paresis, and transient monocu-lar blindness (amaurosis fugax) associatedwith a contralateral transient ischaemic attackor stroke form the typical presenting triadsuggesting extracranial carotid dissection.'23Lower cranial nerve palsies seem to be a rare

finding, with the hypoglossal nerve being themost frequently affected (5%)4: However,when ICA dissection occurs without CNSinvolvement, it is not readily recognised as

the cause of a cranial nerve palsy. Some of

the reported "idiopathic" jugular foramen(Vernet) and similar syndromes (Villaret,Collet Siccard) in the older literature mightwell have been caused by spontaneous- carotiddissection.5 In those times ICA dissection as acause of cranial nerve palsy was poorlyknown, angiography quite risky, and ultra-sound and MRI not available. Out of 31 con-secutive patients with ICA dissectiondiagnosed at our department within fouryears, two presented with cranial nerve palsieswithout cerebral involvement.

Case reportsPATEENT 1A 42-year-old man had been in excellenthealth. He had no vascular risk factors. Afterstrenuous activity (cutting wood) in militaryservice he suddenly felt pain in the rightupper jaw which later extended to the ear andthe whole right side of his face. He sufferedan episode of hazy vision on the right eye last-ing about five minutes. The following morn-ing pain had almost disappeared, but he haddifficulty swallowing and could not drinkproperly. His voice had a nasal sound.Otherwise he felt well. In the hospital, threedays after first symptoms had occurred, hisvoice was severely hoarse, almost aphonic,and nasal. On examination he had a droopingright palate arc. On phonation the palate aswell as the dorsal wall of the pharynx deviatedto the left. Sensation to touch of the rightposterior pharyngeal wall was reduced andthe gag reflex could not be elicited on theright side. No other signs were found. Chestx-ray, ECG, and routine blood laboratorytests were normal. Initially brainstem infarc-tion was suspected and a posterior fossa CTscan was performed which was normal.Because of the right sided facial pain and theinitial episode of presumed right amaurosisfugax, right ICA dissection was suspected.

Duplex sonographic examination of thecarotid arteries performed the same dayshowed a patent carotid bifurcation, no ather-osclerotic changes, and symmetrical bloodflow velocities. Both common carotid arterieswere also normal. Pulsed Doppler sono-graphic examination of the upper cervicalsegments of both ICAs disclosed a coiling ofboth with adjacent antegrade and retrogradeflow in the high cervical portion. However,the blood flow velocities on the right sidewere twice as high as on the left side in bothflow directions. This finding supported thediagnosis of an ICA dissection and anticoagu-

1 191

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from

Sturzenegger, Huber

lation with heparin sodium was started. Thefollowing day, a right common carotid arteri-ogram was performed which also revealed theICA tortuosity. There were irregularities ofthe vessel wall about 3 cm distal of the bifur-cation and extending cranially beyond thecoiling. Equivocal slight stenosis on theanteroposterior projections was visible (fig 1).In addition, a pouch with trapped contrastmedia showed up on the late lateral views(fig 1 a2 and 1 a3) which is a quite distinctsign of dissection. These findings confirmedthe suspected diagnosis of ICA dissection.They could, however, not explain the glosso-pharyngeal and vagal nerve palsy.MR images of the upper cervical region

performed eight days after the first symptoms

::.; ... ..:.zg... -. .

....X. ..... rr F}w

.:.:

.. ::......

.: '?:. :?:::.

.. j.*:.:::.*; .,: ... ... ........* X : .... : : : : ::*:". ::. - ': : .. : . :. .

:;::: ....: .t..:L::* ;' . .g:B . .\.: ............ :.:. e ^ ;. :.: .v. .... .. ...::: :^ .: .: :.

_ ::: .: ::; ... .. .::::.. ::::.. :::.: : . : : . ... .: :: .:;.:.x; .... : ::: ..._ .. ; : ::: :: :.: : :

:::: .:: ..:.. ..,: .: ..

.'.5 .: h.r* :::.: : . :... : .: .. :0:

::*:::.

:.-. Z..:

Figure 1 (Patient 1) Right common carotid arteriogram (al to a3sequential lateral views, bI and b2 sequential anteroposterior projections). Lateral viewsdemonstrate only minimal wall irregularities (arrows) without recognisable stenosis. Onlate images (a2 and a3) a pouch can be seen wherein contrast material is trapped (doublearrow) situated in the ascending loop ofa carotid artery loop (open arrow). Wallirregularities are much more marked on anteroposterior projections (small arrows in bland b2), starting about 2 cm distal to carotid bifurcation and extending beyond the arteryloop (open arrow). There seems to be only slight stenosis in this segment ofICA.

yielded the explanation. A large ICA muralhaematoma was demonstrated as a hyperin-tense signal on the sagittal and axial sectionsof the Ti- and T2-weighted and proton den-sity spin echo sequences (fig 2). It extendedfrom beneath the coiling up to the petrosalcarotid canal. In the axial plane, a large sub-adventitial extension of the haematoma wasdemonstrated with expansion of the vesselwall in the carotid space. This led to displace-ment of the parapharyngeal space (fig 2 bland 2 b2) where the glossopharyngeal nervecrosses the ICA medially leading from thejugular foramen to the root of the tongue. Inthis region the Xth cranial nerve is also inclose contact with the ICA (fig 3). CranialMRI was normal.

After one week of heparin treatment, oralanticoagulation with a coumarin was started.No visual or TIA-like symptoms occurred.Six weeks after admission, hoarseness andnasal voice were still present, but swallowingwas possible again. Pharyngeal sensation wasnormal on both sides, no palate droop wasfound any more and pharyngeal wall shift onphonation was only minimal. Doppler sonog-raphy still showed a higher blood flow veloc-ity on the right side. A second MRI scan eightweeks after admission showed completeresolution of the mural haematoma. Anti-coagulants were continued and after a total of11 weeks all symptoms and signs had cleared.A persistent slight asymmetry of blood flowvelocities at Doppler examination was consid-ered to be due to possible slight residualstenosis. Coumarins were stopped and aspirinstarted.

PATENT 2A 45-year-old man had untreated mild arter-ial hypertension for several years. During thepast two years he suffered from intermittentslight nuchal pain. Two days before admis-sion nuchal pain intensified but did notchange its well known quality. There was nospecial event and no trauma. On the day ofadmission he noticed difficulty with articula-tion, a heavy tongue, and problems to eat.Slightly slurred speech was noted by his wife.Brainstem infarction was suspected and a CTscan performed which was normal. Onadmission he clearly stated that he could notproperly move the food with his tongue, whileoral sensibility, chewing, and swallowing werenot disturbed. Examination showed moderateleft hypoglossal palsy with tongue deviation tothe left when thrusting it out and deviation tothe right when phonating. No other signswere found. Chest and cervical spime x ray,electrocardiogram, and routine blood labora-tory tests were normal.

Doppler and Duplex sonographic examina-tion of the cervical carotid and vertebralarteries and transcranial Doppler sonographyon the day of admission were normal. Repeatultrasound study the following day revealedocclusion of left ICA with typical signs indi-cating dissection.6 Transcranial Dopplerexamination showed good collateral supplyfrom the right carotid and from the verte-

1192 on A

ugust 22, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 Novem

ber 1993. Dow

nloaded from

Cranial nerve palsy in carotid dissection

:I

Figure 2 (Patient 1) MR images at 1 5 tesla. The sagittal right paramedian Tiweighted (TR = 400 ms, TE = 18 ms) section (al) demonstrates mural haematoma"'0oping" along the ICA coiling as hyperintense signal (arrows). There is only minimalstenosis of the vessel lumen represented by the adjacent low signal ("flow void"). The Tlweighted (TR = 500 ms, TE = 16 ms) adjacent axial sections at the atlas level (bl andb2) demonstrate large expansion ofICA vessel wall by mural haematoma. Thehyperintense signal surrounds in a semilunar shaped way all three coils of the loop(arrows in bl). Also the contralateral left ICA shows a coiling (three sectioned lumina(flow voids)) without mural haematoma (see bl). Note the large posterior expansion of themural haematoma (arrows) into the region where the glossopharyngeal nerve passes onimage b2. The same, trefoil leaf-shaped, hyperintense signal is demonstrated on T2weighted (TR = 3,000 ms, TE = 90 ms) axial sections (c) at the atlas level (arrows)identifying the underlying pathology as blood.

Figure 3 Drawing ofananatomical axial sectionthrough the neck at thelevel of the first cervicalvertebral body (atlas). Theclose relationship betweenthe ICA and the IXth, XthXIth and XIIth cranialnerves is highlighted in themagnified cutting. (1)nasal cavity; (2) maxillarysinus; (3) facial muscles;(4) jaw muscles; (5) neckmuscles; (6) atlas body;(7) parotid gland; (8)parapharyngeal space.(Compare with axialMRimage: fig 2 bl).

brobasilar system across the circle of Willis.MR images of the upper cervical region wereperformed two days after the first symptoms.They showed a mural haematoma in theprepetrosal left ICA segment with severelynarrowed lumen (flow void). Caudal to itthere was an isodense mass adjacent to theartery, probably representing an aneurysm(fig 4). Three sectioned lumina (flow void) onboth sides indicated bilateral coiling. Therewere no signal changes in the brain, brain-stem, or cerebellum. Bilateral carotid angiog-raphy was performed three days after the firstsymptoms. It showed bilateral coiling beforethe ICAs entered the petrosal canal. On theleft side contrast filling was markedly delayedand a saccular aneurysm attached to the coil-ing filled. The segnent immediately distal tothe coiling was severely stenosed (fig 5). Onthe day of admission heparin treatment wasbegun and seven days later changed to oralanticoagulation. Control ultrasound examina-tion seven days later showed partial resolu-tion of the occlusion with a now present,though diminished flow signal in the left ICA.Hypoglossal palsy was still slightly presentbut subjective discomfort was minimal.

Review ofthe literatureAll reported series and cases of spontaneousICA dissection were reviewed looking forunequivocally documented symptoms andsigns of cranial nerve involvement.Swallowing difficulty alone was not consid-ered to indicate IXth or Xth nerve dysfunc-tion because local expansion of the aneurysmtowards the pharynx might be the cause.7Unilateral "scalp tenderness", a frequentlyencountered symptom in some series,8 wasnot considered to indicate trigeminal nerveinvolvement unless there was a clear sensorydeficit reported. "Dysgeusia" (unpleasant,strange taste) does, in our opinion, indicatefacial or glossopharyngeal nerve dysfunction.As Maitland did,9 we suspect a compressionof the IXth rather than of the VIIth nerve orof the chorda tympani to be the cause.810Detailed testing of taste, which should allow adifferentiation, was not reported in thesecases. If a nerve compression by the expand-ing wall haematoma (pseudoaneurysm) is thepathogenetic mechanism, the IXth nerve ismore at risk. If, however, interruption of thenerve's blood supply is the working mecha-nism, the IXth and VIIth may be affectedequally. Finally, isolated tinnitus was notconsidered a sign of VIIIth nerve dysfunction.

Only cases in whom the diagnosis had beensupported by angiography showing a typicalpicture were considered.1 3 8 11 12

ResultsThirty-six patients reported under the head-ing of carotid dissection with signs of ipsilat-eral cranial nerve palsies were found in theliterature and analysed together with our twopatients (table 1).289 1012-31 In 35, the age andsex were given: the average age was 44 years,

1193

n:..

-.j.

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from

Sturzenegger, Huber

..a08L).:v. ....%.'9 ) .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

.d, :: .. ...SE ; '. i. - w.: ..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..<'~~~~~~~~~~~~~~~~~~~~~~~~~~Figure 4 (Patient 2) MR images at 1 5 tesla. The proton density weighted (TR = 2740ms, TE = 30 ms) contiguous axial sections at the level of the skuU base (al and a2)demonstrate narrow lumen (flow void) of left ICA (arrowhead in al) and a surroundingmass with mixed iso and hyperintense signal (arrows in al). A little more cranially ahyperintense signal adjacent to theflow void signal indicates mural haematoma (arrow ina2). These signal changes are even better visualised on Ti-weighted (TR = 800 ms,TE = 11 ms) axial sections with fat suppression technique (b1 and b2): The massasymmetrically surrounding the narrowed vessel lumen (arrows in bl) represents theaneurysm. The slit-like residual lumen (arrowhead in b2) is compressed by the muralhaematoma shown as hyperintense signal (arrow in b2).

ranging from 28 to 60 years. Thirty were men

(86%, mean age 47 years) and five women

(mean age 39 years). All but one patient28showed an ICA narrowing of variable degree:11 patients had a high grade stenosis with a

string sign, seven a long irregular slight nar-

rowing, 15 a segmental narrowing, and one a

tapering occlusion. Seventeen (49%) hadaneurysmal dilatation: nine had a fusiformpseudoaneurysm, two of them bilaterally;three had aneurysmal pouches (fig 1); threehad a saccular aneurysm (fig 5); one had a

double lumen; and in one patient no detailson the "aneurysm" were available. In threepatients detailed angiographic results werenot given. In 35 patients the site of the dissec-tion could be located: the prepetrosal seg-ment was involved in all, 11 showedextension of the dissection to the (lower) cer-vical segment, 11 to the intrapetrosal, andeight to the cavernous segment. In threepatients this information was not provided. In26 patients data about tortuosity of the ICAwere given: 10 patients (38%) had some

degree of tortuosity, four had kinking, threecoiling, and three moderate tortuosity. Fourpatients had documented findings suggestiveof fibromuscular dysplasia (FMD).Tortuosity as well as the signs of FMD maybe masked by the dissection itself. Only a

minority of the reported cases had docu-mented cerebral four vessel angiography andcomments concerning the presence of FMDand tortousity in other cranial arteries werefrequently lacking. Therefore, the relative fre-quency of these vascular alterations is cer-tainly underestimated.

In 16 patients MRI scans were performed,14 of which had a mural haematoma demon-strated. In two patients the MRI showed nohyperintense signal. In one of these the delaywas 8 weeks and the haematoma might haveresolved.28 In the other patient no details onMR technique or examination time weregiven in the presence of a typical angiographicresult.'0 In our two patients MRI demon-strated tortuosity (fig 2 bl) and also ananeurysm (fig 4 bI) in addition to the muralhaematoma.

In 27 patients an ipsilateral XIIth nervepalsy was reported, 10 times isolated, sixtimes in conjunction with IXth, Xth, andXIth nerve palsies, four times with Xth nervepalsy, five times with IXth and Xth nervepalsies, once with XIth nerve palsy, and oncewith IXth, Xth, VIIth, and Vth nerve palsies.Five patients had dysgeusia, three isolated,one in conjunction with Vth, and one withVIth nerve palsy. Among seven patients witha Vth nerve palsy, it was isolated in four. Allpatients with Vth, VIth, or IIIrd nerve palsyhad an extension of the dissection up to thecavernous segment. The compiled frequencyof the affected cranial nerves in the 38 caseswas as follows (table 1): XIIth: 27 times; IXth(including the five patients with dysgeusia):18 times; Xth: 17 times; XIth: 7 times; Vth: 7times; VIIth/VIth/IIIrd: once each. In 20patients (53%) an ipsilateral oculosympa-thetic paresis (Homer's syndrome) wasobserved.

In 13 patients information about outcomeof cranial nerve palsies was given: eightpatients had recovered after three weeks totwo years, usually within two to four months;five patients had persistent palsies after threeto 10 months and, hence, some might wellhave recovered later on.Ten patients (26%) had ischaemic symp-

toms. Six had a transient and two a persistenthemiparesis contralateral to the dissection,and two had an ipsilateral amaurosis fugax.

DiscussionPeripheral cranial nerve palsy caused by ICAdissection is only rarely observed. In the largereported series totalling 356 patients withICA dissection, no cranial nerve palsies weredescribed.2332A3 Most of these patients withangiographically confirmed ICA dissectionwere presenting with cerebral ischaemic signsand symptoms. On the other hand, only 8(21 %) of 38 patients presenting with cranialnerve palsies due to ICA dissection collectedfrom the literature (table 1) showed cerebralischaemic involvement.

This discrepancy may be due to not notic-ing ipsilateral cranial nerve palsies in the pres-ence of signs of contralateral stroke. It might

1194 on A

ugust 22, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 Novem

ber 1993. Dow

nloaded from

Cranial nerve palsy in carotid dissection

A2

BI

t_F ..._1.

r.sF 4 ... b.

t

A3

*t

.,ls&

.{NV_j -

.:. F

.....

B2

C D

Figure 5 (Patient 2) Bilateral common carotid arteriogram (al to a3: sequential lateralviews, bi and b2: sequential anteroposterior projections of the left side; c and d lateral andanteroposterior views respectively of the right side). Left side: Lateral views demonstratedelayed ICA filling (al) simulating "string sign" in the early images (arrows in al),saccular aneurysm attached to coiling (arrowhead in a2 and a3,) and poor contrast of thesegment distal to the aneurysm (open arrow in a2 and a3) because ofstenosis (see b);contrast material is trapped in the aneurysm (a3). Anteroposterior views demonstrateaneurysm attached to the coiling (arrowhead in bl and b2) and high grade stenosisimmediately distal to the coiling (open arrow in bl and b2). Right side: Lateral (c) andanteroposterior (d) view show prepetrosal ICA coiling (arrowhead). Collateral blood supplyfrom right ICA to left MCA across circle of WiUis (arrows in d).

also indicate that ICA dissection as a cause oflower cranial nerve palsies is not readilyrecognised. In a recent review of 32 patientswith XIIth nerve palsy,44 carotid dissectionwas considered to be the cause in only onepatient, and this was a traumatic dissection.Unfortunately, no details on the diagnosticexamination of these patients are given.Therefore, the question remains, whethersome of the reported patients with so calledtraumatic XIIth nerve palsies or some ofthose attributed to brainstem ischaemia25 inreality had a carotid dissection causing thenerve palsy. In our own series of 31 patientswith ICA dissection, the two patientsreported here (6%) presented with isolatedcranial nerve palsies. In retrospect, we recallanother two patients who presented with iso-lated XIIth nerve palsy and ipsilateral neckpain. CT scan did not support the suspecteddiagnosis of a cervical tumour or infection.This was several years ago when ICA dissec-tion was not considered as a possible causeand hence no further investigations were per-formed.

However, the relatively frequent observa-tion of isolated cranial nerve palsies withoutcentral involvement (79%) as opposed to therare observation of cranial nerve palsies inICA dissection patients with cerebralischaemic symptoms could indicate that thesetwo groups have a different pathogenesis. Itis, for instance, conceivable that the formerhad subadventitial rather than subintimal dis-sections, since the subadventitial dissection isless likely to obstruct the vessel lumen andwould rather extend outwardly. Several factsfrom our analysis support such a hypothesis:(1) the higher prevalence of men in the ICAdissection patients with cranial nerve palsies(86%) as opposed to the general ICA dissec-tion population (51%),4 (2) like in traumaticICA dissection,'2 there is a high prevalence ofaneurysms in spontaneous ICA dissectionpatients with cranial nerve palsies (49%),while aneurysms in ICA dissection patientswith hemispheric signs are less frequent(10-33%)3 1133; (3) the higher prevalence ofextreme vessel tortuosity in the ICA dissec-tion patients with cranial nerve palsies (27%)compared with the general population(3-12%).45 Tortuosity may indicate a struc-ture wall anomaly making the vessel predis-posed to external mechanical injury.Tortuosity also does alter flow dynamicswhich may cause sheer injury to the vascularwall again making the affected vessel predis-posed to dissection. Could more frequentlyoccurring trivial and unreported traumas of aphysically more active male population befavouring subadventitial dissection in thepresence of a predisposing arteriopathy?

In the case of subadventitial dissection,muscular medial layer may impede narrowingof the lumen and the mural haematoma mayconsiderably expand the circumference of thevessels (figs 2b and 4b). The close topograph-ical relationship of the most frequentlyinvolved XIIth, IXth, and Xth nerves with thecarotid artery in the upper cervical parapha-

1195

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from

Sturzenegger, Huber

Table 1 Patients with spontaneous ICA dissection and cranial nerve palsies

Reference AgelSex

Kramer'4 41/F

Gros"5 38/M

Labauge"3 38/M

Cohen'6 31/F

38/M

Fisher" 45/M

53/M

Barbizet'7 42/M

Bradac"8 58/M

41/MGoodman'9 59/M

Side

left

left

left

left

left

left

left

right

left

leftright

Site ofICAdissection

prepetrosal

prepetrosal

prepetrosal

pre and intrapetrosal,cavernous (?)pre and intrapetrosalcavernous (?)cervical + prepetrosal

prepetrosal

cervical, prepetrosal,intrapetrosal, cavernousprepetrosal

prepetrosalprepetrosal

Angiography

segmental narrowing,aneurysm prepetrosalfusiform aneurysm

slight stenosis, pseudoaneurysm

segmental narrowingpseudoaneurysmsegmental narrowingpseudoaneurysmstring sign up topetrosal canal, pouchsegmental narrowing

string signemboli in MCA branchesstring sign, saccularaneurysmstring signstring sign

Tortuosity FMD

no yes

no no

yes

?> yes

no no

no no

no nokinking no

MRI (delay)

no

no

no

no

no

no

no

no

no

nono

Cranial nervepalsies

IX, X(?), XII left

XII left

XII left

V (1 + 2) left

V left

dysgeusiachorda tympani left ?dysgeusiachorda tympani left ?m and V right

IX, X, XI, XII left

XII leftXII right

Maitland937/F left pre- and intrapetrosal, string sign

cavenousHommel20 55/M left prepetrosal segmental narrowing

saccular aneurysm

no no no

no no no

VI left, dysgeusia(chorda tympani?)XII left

54/M right prepetrosal segmental narrowingfusiform aneurysm

yes no no

44/M

Gauthier'2 ?

Mokri2 ?

Mokri"2 47/M

Benrabah22 40/M

50/M

Vighetto23 48/MGoldberg24 49/MFrancis'" 46/F

Hess25 41/M

Lieschke26 42/M

Waespe27 41/M

Bradac28 28/M41/F

53/M

Dal Pozzo29 58/M

45/M

54/M

Vernay3 60/MPanisset3' 36/M

53/M

Sturzenegger 42/M

left

leftleft

bilateral

left

left

leftleftleft

right

right

bilateral

rightleft

left

left

right

left

leftright

left

right

cervical, prepetrosal

prepatrosal left

prepetrosal rightcervical, prepetrosal,cavernouscervical, prepetrosal,cavernousprepetrosalpre- and intrapetrosalhigh cervical, intra-petrosal, cavernouscervical, prepetrosal

prepetrosal

prepetrosal bilateral

cervical, prepetrosalcervical, prepetrosal

prepetrosal

prepetrosal

prepetrosal

cervical, pre- and intrapetrosalprepetrosalpre- and intrapetrosal

pre- and intrapetrosal,cavernous?cervical, prepetrosal

45/M left pre- and intrapetrosal

string sign

stenosis; no further details

stenosis, no further detailsprovidedsegmental stenosispseudoaneurysmpseudoaneurysmstring sign

string sign

string sign

long irregular narrowingtapered narrowing

tapering occlusion

string sign

pseudoaneurysm bilateral,irregular slight narrowinglong irregular narrowinglong irregular slightnarrowing, pseudoaneurysmpseudoaneurysm, no stenosis

slight segmental narrowing

slight segmental narrowing

slight irregular segmentalnarrowing, small pouchsegmental stenosisslight stenosis

stenosis, double lumen

long distance irregular wall,slight stenosis, small pouchsegmental severe stenosissaccuar aneurysm withincoiling

no no no

?? no?? no

?? no

kinking no

kinkingno

?> no

yes ?no nono yes

no no

kinking no

no no

no nocoiling no

kinking ? no

no no

no no

yes ?

coiling nobilateral

coiling nobilateral

no

IX, X, XII left

IX, X, XII leftIX, X, XII left

dysgeusia

IX,X,XI,XII left

none rghtno V left (hypereresthesia)

no VI left (hyperesthesia)

no XII leftyes, delay: 16 days IX,X,XI,XII leftyes, delay: ? V left, dysgeusia

yes, delay: 13 days X,XII right

yes, delay: about X,XII right10 daysyes; delay: about IX,X,XI,XII left10 days none rightyes, delay: 10 days X,XII rightyes, delay: 15 days XII left

yes, delay: 8 weeks XII leftno hyperintensesignalyes, delay: 19 days IX,X,XI,XII left

yes, delay: 40 days XI,XII right

yes, delay: 27 days IX,X,XII left

yes delay: 8 weeks XII leftyes, delay:? IX,X,XI,XII right

yes, delay?: V,VII,IX,X,XII left

yes, delay: 8 days IX,X right

yes, delay: 2 days XII left

Year indicates year of publication; ICA, intemal carotid artery; MCA, middle cerebral artery; FMD, fibromuscular dysplasia; TIA, transient ischaemic attack; ?,means that the respective criterion could not be analysed because of lack of data.

ryngeal space (fig 3) are a strong argumentfor a nerve compression by the expandedcarotid wall as the underlying mechanism. Inmost cases the dissection begins 2 to 4 cm

distal to the carotid bulb extending craniallyand showing a maximal distension at the Clor C2 vertebral level. At this point the IXthcranial nerve crosses the ICA laterally, theXth lies posteriorly and medially, and the

XIIth crosses the ICA a little more caudally,in general at the C3 level. The fact that inpatients with involvement of upper cranialnerves, dissection usually extended into thepetrosal and cavernous segment (table 1) isanother argument for a compression to be thecausative mechanism. Nerve ischaemia due tocompromised blood supply is a second possi-ble mechanism.4647 The IXth to XIIth nerves

X, XII right

1196

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from

Cranial nerve palsy in carotid dissection

Homer's Pain TIlAstroke Special remarkssyndrome

..- 1W;- -- r - _1_I _ :@1:__yes len ear, tace, neck

no

no

yes left

yes left

yes

yes

yes

yes

yesyes

no

no

lefttemple left

face left

eye, cheek left

temple left

face, eye left

face, eye left

neck, temple,head rightneck left

temple leftneck right

face, neck,eye lefteye, temple left

no temple, occipi-tal right

no yaw, ear, faceleft

yes face leftno face, neck left

yes left occipital left

no right noyes face, neck, eye

leftyes face, eye left

yes left neck, temple leftyes face leftyes face, neck left

no orbital right

no temple, eyeright

no neck, face left

no neck rightyes head

no head, neck left

no neck

yes neck right

yes eye, front left

yes occipital leftno occipital, eye

rightno occipital, neck

leftno yaw, ear, face

right

no nuchalandarteriography)

no

right arm + facetransientparesis right armtransientno

no

right arm + facetransientno

hemiplegia leftpersistentno

nohemiparesis lefttransientno

no

no

hemiplegia rightaphasia, persist

noright arm,transient

no

no

no

no

nono

hemiparesis lefttransientno

no

noamaurosis fugax

no

no

no

no

nono

no

amaurosis fugax

no

aissecuon very ilkelydissection probable

persistent XII nerve palsy

"Raeder's syndrome"

"Raeder's syndrome"

recovery of VIth nerve palsywithin 2 monthsrecovery of XIIth nerve palsywithin 2 monthscontrol angiography after 2 months:persistent aneurysm, no stenosisrecovery ofXth nerve palsy within3 days and of XIIth within 6 monthscontrol angiography after 8 months;persistent aneurysm, no stenosisXIIth nerve palsy persistentcontrol angiography after 3 months:normal ICA left, MCA branch occlusionno details on angiographyno details on angiography

no further details provided

asymptomatic after 2 1/2 y

MRI = normal (?) delay not given

bilateral dissection,right side asymptomaticXII paresis persistent after 4 monthsXII paresis persistent after 6 months

XII paresis persistent after 1 year,dissection not proved

CT also positive. Follow up MR. after 13months; haematoma resolved.CT also positive. Follow up MRI after 14months: haematoma resolvedCT negative. Follow up MRI after 6weeks: smaller still persistent haematoma.

recovery within 2 months

recovery within 6 months

bilateral coiling at MRI. Follow up MRIafter 7 weeks: haematoma resolved.clinical recovery within 10 weeks.bilateral coiling (at MRI

Partial resolution of stenosis after 7 days(Ultrasound). Clinical recovery within 3weeks.

are supplied by the ascending pharyngealartery and the Vth and VIIth by the middlemeningeal artery.4647 Both arteries arise fromthe external carotid artery which may bedirectly affected by dissection or compressedby the expanding dissected ICA. There arefew data supporting this pathogenesis. Thedistinction of these two mechanisms is, how-ever, irrelevant for the diagnostic and thera-

peutic management of the patients.The proof of carotid artery dissection as

the cause of cranial nerve palsies is a problemnot yet solved. Angiography, considered asthe standard by most authors, can detect acompromise of the vessel lumen only above acertain degree. The recognition of a lowgrade narrowing of the lumen is difficult, withconsiderable interobserver disagreement.48 Inthe case of a subintimally located dissectionthe mural haematoma will in the first placeprotrude the intima into the vessel lumen andproduce a variable degree of stenosis up toocclusion. These cases are easily depicted byangiography.381' 12 As the majority of ICAdissections seem to occur in the subintimallayer' angiography is the diagnostic method ofchoice especially if there are clinical or ultra-sound signs of cerebrovascular circulationcompromise.6 Yet, even in these cases,angiography is by no means always pathogno-monic. The advent of magnetic reso-nance angiography allows evaluation ofluminal caliper non-invasively as well.

However, if the dissection occurs in thesubadventitial layer, considerable wall expan-sion into the carotid space and displacementof the adjacent parapharyngeal space mayoccur (figs 2 b and 4 b) The resultinghaematoma can compress adjacent structuressuch as the lower cranial nerves (fig 3) with-out causing relevant vessel narrowing. Thesecases may not be detected by angiography.Not only can no vessel anomaly be seen, butalso the clinical picture with isolated cranialnerve palsies is not even considered an indi-cation for angiography. The question there-fore arises whether the reported prevalence ofsubintimal dissection"34 iS merely a conse-quence of patient selection and the diagnosticmethod used, since angiography seems aninadequate method to detect subadventitialdissection.The reported cases raise questions about

the significance of aneurysms and tortuosityfound on angiography both for establishingthe diagnosis and for looking for an underly-ing arteriopathy.Aneurysms are reported on angiography

in up to 33% of patients with ICA dissec-tion." We have observed three patientswith an aneurysm among 27 patients withICA dissection (1 %): two had small pouches(one of them patient 1) and one (patient 2)had a saccular aneurysm. It is often impossi-ble to determine whether the (saccular)aneurysms were present as congenital malfor-mations before dissection occurred orwhether they are the result of a persistentcommunication between the dissection cavityand the regular lumen of the artery.'5 27 28Aneurysms of the extracranial ICA seg-

ment are sometimes associated with FMD.50FMD in ICA dissection has been reported inas many as 22%.' However, FMD cannot beconsidered the main aetiology of ICA dissec-tion. The sex ratios observed in FMD andICA dissection strongly argue against that:FMD affects up to 80% women,50 and in thepresent review series 86% of patients with

1197

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from

Sturzenegger, Huber

ICA dissection were men. In addition,intracranial aneurysms occur frequently withFMD,50 but only exceptionally with ICA dis-section.

Tortuosity, next to FMD, is the secondmost common arterial wall anomaly cited tomake a vessel predisposed for dissection.34251In our review (table 1) 10 (38%) of 26patients whose angiograms could be judgedhad some degree of tortuosity. Seven (27%)had a kinking or coiling. According to largeseries of consecutive, selected carotid arteri-ograms slight tortuosity is present in 11% to31%, but extreme tortuosity (kinking andcoiling) in only 3% to 12%.45 Tortuosityseems to be more frequent in those patientswith dissection who have aneurysm forma-tion. Among the 10 patients with tortuositywe found seven with aneurysms (70%),whereas the 16 patients without tortuosityhad only five aneurysms (31%). Whetherextreme tortuosity alone, not associated withatherosclerosis or dissection, can cause cere-bral ischaemia or lower cranial nerve com-pression is a matter of debate.45

Application of MRI and a large index ofsuspicion may clarify several of the unsettledissues. MRI is the only non-invasive tech-nique to demonstrate mural haematoma andto show its longitudinal extension (fig 2 a)and the degree of wall expansion (fig 2 b and4 b).24 263149 Newer techniques such as chemi-cal shift imaging (fat suppression) (fig 4 bl,b2) make MRI more sensitive to even smallarterial dissections.49 MRI additionally givesinformation about surrounding tissues. Theinformation provided by MRI at present iscomplementary to that of angiography.

In conclusion, carotid dissection should beconsidered in isolated or combined lower cra-nial nerve palsies, especially if associated withipsilateral neck pain, headache, and Homer'ssyndrome. Descriptive terms like "carotidy-nia", Raeder's syndrome, "migraine withmiosis", and all the cranial nerve syndromeswith famous eponyms5 should be avoided infavour of a causal diagnosis. Today, MRIprovides a sensitive non-invasive method forthe diagnosis of ICA dissection if performedwithin the first weeks.

1 Hart RG, Easton JD. Dissection and trauma of cervico-cerebral arteries. In: Barnett HJM, ed. Stroke pathophysi-ology, diagnosis and management. New York: ChurchillLivingstone, 1986:775-88.

2 Mokri B, Sundt TM, Houser OW, Piepgras DG.Spontaneous dissection of the cervical internal carotidartery. Ann Neurol 1986;19:126-38.

3 Sellier N, Chiras J, Benhamou M, Bories J. Dissectionsspontanees de la carotide interne. Aspects cliniques,radiologiques etevolutifs. A propos de 46 cas. JNeuroradiol 1983;10:243-59.

4 Hart RG, Easton JD. Dissection of cervical and cerebralarteries. Neurol Clin 1983;1: 155-82.

5 Tanaka M, Isaka K, Morimatsu M, Hirai S.Jugular fora-_men syndrome. Neurology 1983,32.1 19-20.6 Sturzenegger M. Ultrasound findings in spontaneous

carotid artery dissection. The value of duplex sonogra-phy. Arch Neurol 1991;48:1057-63.

7 Silcox LE, Updegrove RA. Extracranial aneurysm of theinternal carotid artery. Arch Otolaryngol 1959;69:329-33.

8 Fisher CM, Ojemann RG, Roberson GH. Spontaneousdissection of cervicocerebral arteries. Can J Neurol Sci1978;5:9-19.

9 Maitland CH, Black J, Smith W. Abducens nerve palsydue to spontaneous dissection of the internal carotidartery. Arch Neurol 1983;40:448-9.

10 Francis KR, Williams DP, Troost BT. Facial numbnessand dysesthesia, new features of carotid artery dissec-tion. Arch Neurol 1987;44:345-6.

11 Houser WO, Mokri B, Sundt TM, Baker HL, Reese DF.Spontaneous cervical cephalic arterial dissection and itsresiduum: Angiographic spectrum. AJNR 1984;5:27-34.

12 Guthier G, Rohr J, Wildi E, Megret M. L'hematome dis-sequant spontane de l'artere carotide interne. ArchSuisses Neurol Psychiatr 1985;136:53-74.

13 Labauge R, Thevenet A, Gros C, Vlahovitch B, PeguretC, Frerebeau C. Les anevrysmes due segmentexocrinien de 1' axe carotidien et leur traitement chirur-gical. A propos de 13 observations personnelles. RevNeurol 1971;124:512-25.

14 Kramer W. Hyperplasie fibromusculaire et anevrysmeextracranien de la carotide interne avec syndrome para-pharyngien typique. Rev Neurol 1969;120:239-44.

15 Gros C, Vlahovitch B, Labauge R, Thevenet A, KuhnerA, Frerebeau Ph. Les anevrysmes extra-craniens de lacarotide interne. Neuro-chirurgie 1970;16:367-82.

16 Cohen DN, Zakov ZN, Salanga VD, Dohn DF. Raeder'sparatrigeminal syndrome. Am J Ophthalmol 1975;79:1044-9.

17 Barbizet J, Degos J-D, Lamotte J, Baulac M. Dissectionspontanee de la carotide inteme. Ann Mid Interne1978;129:73-6.

18 Bradac GB, Kaernbach A, Bolk-Weischedel D, Finck GA.Spontaneous dissecting aneurysm of cervical cerebralarteries. Neuroradiology 1981;21:149-54.

19 Goodman JM, Zink WL, Cooper DF. Hemilingual paraly-sis caused by spontaneous carotid artery dissection. ArchNeurol 1983;40:653-4.

20 Hommel M, Pollak P, Gaio JM, Pellat J, Perret J, ChateauR. Paralysies du nerf grand hypoglosse par deuxanevrysmes et un anevrysme dissequant de l'arterecarotide interne. Rev Neurol 1984;140:415-21.

21 Mokri B, Piepgras DG, Wiebers DO, Houser OW.Familial occurrence of spontaneous dissection of theinternal carotid artery. Stroke 1987;18:246-51.

22 Benrabah R, Bousser MG, Cabanis EA, Lopez A, MoreauJ. Syndrome de Claude-Bernard-Horner douloureuxrevelateur d'une dissection spontan&e de l'arterecarotide interne. Interet du bilan ultrasonique cervical.A propos de deux observations. BuUl Soc Ophthalmol Fr1988;88:763-8.

23 Vighetto A, Lisovoski F, Revol A, Trillet M, Aimrad G.Internal carotid artery dissection and ipsilateralhypoglossal nerve palsy. Y Neurol Neurosurg Psychiatry1990;53:530-1.

24 Goldberg HI, Grossmann RI, Gomory JM, Ashbury AK,Bilaniuk LT, Zimmermann RA. Cervical internalcarotid artery dissecting hemorrhage: diagnosis usingMRI. Radiology 1986;158:157-61.

25 Hess DC, Sehti KD, Nichols FT. Carotid dissection: anew false localising sign. J Neurol Neurosurg Psychiatry1990;53:804-5.

26 Lieschke GJ, Davies S, Tress BM, Ebeling P.Spontaneous internal carotid artery dissection present-ing as hypoglossal nerve palsy. Stroke 1988;19:1151-5.

27 Waespe W, Niesper J, Imhof HG, Valavanis A. Lower cra-nial nerve palsies due to internal carotid artery dissec-tion. Stroke 1988;19:1561-4.

28 Bradac GB, Riva A, Stura G, Doriguzz C. Dissectionspontanee de la carotide interne avec paralysie du XII. JNeuroradiol 1989;16: 197-202.

29 Dal Pozzo G, Mascalchi M, Fonda C, Cadelo M, Ronchi0, Inzitari D. Lower cranial nerve palsy due to dissec-tion of the internal carotid artery: CT and MR imaging.JComput Ass Tomogr 1989;13:989-95.

30 Vernay D, Deffond D, Durif F, Georget AM, Dordain G.Paralysie du XII au cours d'une dissection carotidiennecervicale haute.Presse Med 1989;18:985-6.

31 Panisset M, Eidelman BH. Multiple cranial neuropathy asa feature of internal carotid artery dissection. Stroke1990;21:141-7.

32 Bogousslavsky J, Despland P-A, Regli F. Spontaneouscarotid dissection with acute stroke. Arch Neurol1987;44:137-40.

33 Steinke W. Aulich A, Hennerici M. Diagnose und Verlaufvon Carotisdissektionen. DMW 1989;114: 1869-75.

34 De Bray JM, Leftheriotis G, Causeret H, Pasquier JL,SaumetJL. Les dissections carotidiennes. Etude de 28cas.Yournal des Maladies Vasculaires 1990;15:321-5.

35 Hennerici M, Steinke W, Rautenberg W. High-resistanceDoppler flow pattern in extracranial carotid dissection.Arch Neurol 1989;46:670-2.

36 De Bray JM, Dubas F, Joseph PA, Causeret H, PasquierJP, Emile E. Etude ultrasonique de 22 dissections caro-tidiennes. Rev Neurol 1989;145:702-9.

37 Biller J, Hingtgen WL, Adams HP, Smoker WRK,Godersky JC, Toffol GJ. Cervicocephalic arterial dissec-tions: a ten-year experience. Arch Neurol 1 986;43:1234-8.

38 Pozzati E, Giuliani G, Acciarri N, Nuzzo G. Long-termfollow-up of occlusive cervical carotid dissection. Stroke1990;21:528-3 1.

39 Rosat P, Deramond H, Grunewald P, Le Gars D,Galibert P. L'anevrysme dissequant non traumatique dela carotide interne cervicale. Conduite a tenir.Neurochirurgie 1981;27:133-7.

40 Miillges W, Ringelstein EB, Weiller C, Leibold M,

1198

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from

Cranial nerve palsy in carotid dissection

Bruckmann H. Dissektionen der A. carotis interna -neue diagnostische und pathogenetische Aspekte.Fortschr Neurol Psychiat 1991;59: 12-24.

41 Kaps M, Dorndorf W, Damian MS, Agnoli L.Intracranial hemodynamics in patients with spontaneouscarotid dissection. Eur Arch Psychiatr Neurol Sci1990;239:246-56.

42 Luken MG, Ascherl GF, orrell JW, Hilal SH.Spontaneous dissecting aneu rsms of the extracranialinternal carotid artery. Clin Nefrosurg 1979;26:353-75.

43 Marx A, Messing B, Storch B, Busse 0. SpontaneDissektionen hirnversorgender Arterien. Nervenarzt1987;58:8-18.

44 Tommasi-Davenas C, Vighetto A, Confavreux C, AimnardG. Causes des paralysies du nerf grand hypoglosse. Apropos de 32 cas. La Presse Medicale 1990;19:864-8.

45 Weibel J, Fields WS. Tortuosity, coiling, and kinking of

the internal carotid artery. Neurology 1965;15:7-18 and462-8.

46 Lapresle J, Lasjaunias P. Cranial nerve ischaemic arterialsyndromes. A review. Brain 1986;109:207-15.

47 Lasjaunias P, Doyon D. L'artere pharygienne ascendantedans la vascularisation des dernieres paires craniennes(IX, X, XI, XII). JNeuroradiol 1978;5:287-301.

48 Croft RJ, Ellam LD, Harrison MJG. Accuracy of carotidangiography in the assessment of atheroma of the inter-nal carotid artery. Lancet 1980;1:997-1000.

49 Pacini R, Simon J, Ketonen L, Kido D, Kieburtz K.Chemical-shift imaging of a spontaneous internalcarotid artery dissection: case report. AJ7NR1991;12:360-2.

50 Mettinger KL, Ericson K. Fibromuscular dysplasia andthe brain. Stroke 1982;13:46-58.

51 Hart RG, Easton JD. Dissections. Stroke 1985;16:925-7.

1199

on August 22, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.56.11.1191 on 1 N

ovember 1993. D

ownloaded from