C

Ra

Ja

b

a

ARR2AA

KABROV

1

hbemnprrvwa

2

ndl

S2

0h

Clinical Neurology and Neurosurgery 115 (2013) 1520–1523

Contents lists available at SciVerse ScienceDirect

Clinical Neurology and Neurosurgery

journa l homepage: www.e lsev ier .com/ locate /c l ineuro

ase report

otational vertebral artery occlusion associated with occipitoatlantalssimilation, atlantoaxial subluxation, and basilar impression

oji Inamasua,b,∗, Masashi Nakatsukasab

Department of Neurosurgery, Fujita Health University School of Medicine, Toyoake, JapanDepartment of Neurosurgery, Saiseikai Utsunomiya Hospital, Utsunomiya, Japan

r t i c l e i n f o

rticle history:eceived 17 April 2011eceived in revised form8 November 2012ccepted 2 December 2012vailable online 1 January 2013

eywords:tlantoaxial subluxationasilar impressionotationalccipitoatlantal assimilationertebral artery

. Introduction

Rotational vertebral artery (VA) occlusion, also known as bowunter’s stroke, is a rare form of vertebrobasilar ischemia elicitedy head rotation, and it results from transient mechanical oblit-ration of the VA [1]. The VA at the C1–C2 junction is involvedost frequently, and hemodynamic compromise of the domi-

ant VA is usually responsible for ischemic episodes [1]. Variousathoanatomies of the craniovertebral junction (CVJ) have beeneported to be associated with rotational VA occlusion. Adult-onsetotational VA occlusion associated with a CVJ anomaly, however, isery rare. We describe a case of rotational VA occlusion associatedith occipitoatlantal assimilation, atlantoaxial subluxation (AAS),

nd basilar impression.

. Case report

This previously healthy 22-year-old male college student with

o history of a major head/neck trauma suddenly experiencedizziness and nausea when he abruptly turned his head to the

eft side. The episode was painless. His symptoms persisted for

∗ Corresponding author at: Department of Neurosurgery, Fujita Health Universitychool of Medicine, 1-98 Kutsukake, Toyoake 458-0831, Japan. Tel.: +81 562 93000; fax: +81 562 93 2000.

E-mail address: ginamasu@aol.com (J. Inamasu).

303-8467/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.clineuro.2012.12.005

several days, and he visited a local physician and was referredto our neurology department. He was alert and well oriented.Mild truncal ataxia and gait disturbance was noted on neurologicalexamination. Neurologic tests for hearing disturbance, includingthe Weber and Rinne’s, were negative. A mildly positive finger-to-nose test with negative Romberg’s sign suggested presence ofa small cerebellar lesion. He did not complain of neck pain afterneck flexion/extension, and dynamic X-rays of the cervical spineto evaluate translational instability were not performed. T2- anddiffusion-weighted images of the brain magnetic resonance (MR)imaging demonstrated a few small high-intensity spots in the cere-bellum (Fig. 1A). He was diagnosed with a cerebellar infarction andwas admitted. Based on clinical history, rotational vertebral arteryocclusion was suspected. A thin-slice coronal CT image indicatedthat the right VA might be squeezed between the right C1–C2 jointand a bony spur of the occipital bone (Fig. 1B). Two weeks afteradmission, digital subtraction angiography (DSA) of the cerebralarteries was performed. The left VA was hypoplastic and barelyperfused the left cerebellar hemisphere. The right VA, which wasdominant and perfusing both of the cerebellar hemispheres, wasmarkedly tortuous at the C1–C2 junction (Fig. 1C). When he wasasked to rotate his head to the left for 45◦, cessation of blood flowat the C1–C2 junction was noted (Fig. 1D), confirming the diagno-

sis of rotational VA occlusion. Precaution to ensure his safety hadbeen taken during the procedure: both an interventional neuro-radiologist and an anesthesiologist attended the angiography. Assoon as cessation of the flow was confirmed, he was asked to rotate

J. Inamasu, M. Nakatsukasa / Clinical Neurology and Neurosurgery 115 (2013) 1520–1523 1521

Fig. 1. (A) Diffusion-weighted MR imaging studies showing high intensity spots in the cerebellum, indicating acute cerebellar infarction. (B) A thin-slice coronal CT scanwith contrast suggesting that the right VA (arrow) might be squeezed between the right C1–C2 joint and a bony spur of the occipital bone (asterisk). (C) The right VA( misphw

tF(stbwCabibBiu

pOuneft

anteroposterior view), which was dominant and perfusing both of the cerebellar heas confirmed when the patient rotated his head to the left side for 45◦ .

he neck back to neutral position, and no adverse events occurred.urthermore, a sagittal CT image showed the atlantoaxial distanceAAD) of 9 mm (Fig. 2A), indicating the presence of atlantoaxialubluxation (AAS). The tip of the dens process was 6 mm abovehe McGregor‘s baseline (Fig. 2A), which indicated the presence ofasilar impression [2]. Concomitant occipitoatlantal assimilationas also noted (Fig. 2A, arrow). T2-weighted MR imaging of theVJ showed that basilar impression was mild, and there was onlyslight distortion of the brainstem (Fig. 2B). Combination of AAS,asilar impression and tortuous right VA seemed to have resulted

n rotation-induced obliteration of the artery and subsequent cere-ellar infarction. He was referred to us 30 days after admission.y that time, he was neurologically intact. Because of the instabil-

ty, a posterior fusion surgery was recommended, and he agreed tondergo the procedure.

An occipitocervical instrumented fusion down to C3 waslanned, since the C1 posterior element was absent. A Synthesccipital Cervical Fusion System (Synthes K.K., Tokyo, Japan) wassed. He was positioned prone under general anesthesia, with his

eck in neutral position. Pre- or intra-operative traction was notmployed, because there were no signs of brainstem or cord dys-unction. A midline occipital plate was fixated with three bicorticalitanium screws (Fig. 3A). For C2, polyaxial C2 screws were placed

eres, was markedly tortuous at the C1–C2 junction. (D) Cessation of the blood flow

bilaterally into the lamina via a translaminar trajectory (Fig. 3B) tominimize the risk of VA injury [3]. A shorter polyaxial screw wasplaced into each lateral mass of C3. Subsequently, the screws wereconnected with contoured titanium rods. An iliac crest autograftwas placed between the decorticated occiput and C2 spinous pro-cess. Postoperative course was uneventful, and a good alignment ofthe CVJ was noted. The AAD was reduced to 2 mm, which was withinnormal range (Fig. 3C). He was discharged 10 days after surgery.Sold bony fusion was achieved by 9 months after surgery, and he hasnot experienced any recurrences of ischemia. Mild inconveniencein head rotation is well tolerated.

3. Discussion

Rotational VA occlusion mostly occurs at the C1–C2 junction[1]. Although hemodynamic compromise of the dominant VA isusually responsible for the ischemic episodes, obliteration of thenon-dominant VA may cause symptomatic ischemia [4], and occa-sionally, ischemia may be of embolic rather than of hemodynamic

origin [5]. The presence of underlying pathoanatomies togetherwith disproportion in the VA size seem to be prerequisites for devel-opment of symptomatic ischemia [1,4,5]. Our patient was unique inthat he sustained a cerebellar infarction of the left posteroinferior

1522 J. Inamasu, M. Nakatsukasa / Clinical Neurology and Neurosurgery 115 (2013) 1520–1523

F nce beO imag

crvwmt

taurotVac

FA

ig. 2. (A) Sagittal CT scan showing atlantoaxial distance as 9 mm (star), and distaccipitoatlantal assimilation was also noted (arrow). (B) A sagittal T2-weighted MR

erebellar artery (PICA) territory after temporary obliteration of theight VA. Although it seems paradoxical, a closer look at the rightertebral angiogram revealed that most of the left PICA territoryas perfused by the right VA (Fig. 1C), and we suspect that theechanism of ischemia in our patient was hemodynamic rather

han embolic.Various pathoanatomies of the CVJ are associated with rota-

ional VA occlusion. Many of them are due to degenerative changes,nd they are mechanically stable [1,4,5]. Most reported cases ofnstable rotational VA occlusion either have os odontoideum,heumatoid arthritis, or ankylosing spondylitis [1,4,5]. Those withs odontoideum are usually children, whereas those with rheuma-

oid arthritis or ankylosing spondylitis are elderly [1,4,5]. RotationalA occlusion associated with an unstable CVJ anomaly is raremong young to middle-aged adults. In our patient, it is likely thatombination of AAS, basilar impression and tortuous right VA have

ig. 3. Postoperative CT scans. (A) An axial scan showing the trajectory of C2 translaminasagittal CT scan showing good alignment of the craniovertevral junction as well as a red

tween the McGregor‘s baseline (dotted line) and odontoid tip as 6 mm (asterisk).ing showing slight distortion of the brainstem due to basilar impression.

resulted in the rotational VA occlusion: the right VA seemed tobe squeezed between the right C1–C2 joint and a bony spur ofthe occipital bone (Fig. 1B). Despite the fact that occipitoatlantalassimilation is a congenital anomaly, it is not clear whether AAS isalso congenital or not. The C2–C3 autofusion, which often coexistswith occipitoatlantal assimilation, may be involved in the devel-opment and progression of AAS [2], and that may explain why hedeveloped symptoms in his early 20s. Another possibility is thatboth AAS and occipitoatlantal assimilation were congenital. More-over, anomalous VAs may often be associated with occipitoatlantalassimilation.

Treatment of rotational VA occlusion that occurs at C1–C2

junction relies on underlying pathoanatomy. The use of thin-slice reconstructed CT images and 3D-CTA in addition to DSAhas made it possible to identify a lesion where the rotationalVA occlusion occurs (Fig. 1B) and to delineate treatment strat-

r screws. (B) A 3-D reconstructed scan of the occipitocervical fixation construct. (C)uced atlantoaxial distance (star).

ology

emrotoRhbotmlf[

fPdi(prs

[

[

[

[

[

J. Inamasu, M. Nakatsukasa / Clinical Neur

gy in individual patient. Although simple decompression surgeryay suffice in cases without CVJ instability [2], fusion may be

equired when there is a mechanical instability [3]. Regardingur patient, we had not been confident preoperatively whetherranslational instability was present, since he did not complainedf neck pain or neurologic deficits after neck flexion/extension.eduction of the AAD seen in the postoperative CT scan (Fig. 3C),owever, suggests that translational in addition to rotational insta-ility had been present. Conservative management consistingf administration of anticoagulants and use of a cervical collaro limit head rotation might have been another option, and it

ay be warranted for those who refuse surgery [6,7]. However,ong-term compliance of conservative management is unsatis-actory and most patients eventually choose to undergo surgery6,7].

An occipitocervical fusion instead of C1–C2 fusion was per-ormed in our patient, because of the absence of C1 posterior arch.lacement of a screw into an occipitalized C1 would be technicallyifficult, although it might not be impossible [3]. For C2 anchor-

ng, translaminar screws were placed bilaterally into the lamina

Fig. 3B), because we were concerned that placement of either C2edicle or C1–C2 transarticular screws might put him at significantisk of the right VA injury. The C2 translaminar screw con-truct has a biomechanical profile comparable with the other two

[

[

and Neurosurgery 115 (2013) 1520–1523 1523

standard posterior C1–C2 fixation techniques, and may particularlybe useful in circumstances where even a single VA injury may leadto disastrous consequence [3].

Conflict of interest

None of the authors have any financial relationships with otherpeople or organizations that could inappropriately influence theirwork.

References

1] Abdullah J, Caemaert J, Calliauw L, Verstraete K, Achten E. Bow hunter’s strokein a child: a case report. Acta Neurologica Belgica 1992;92:180.

2] Smoker WR. Craniovertebral junction: normal anatomy, craniometry, and con-genital anomalies. Radiographics 1994;14:255–77.

3] Wright NM. Posterior C2 fixation using bilateral, crossing C2 laminar screws:case series and technical note. Journal of Spinal Disorders and Techniques2004;17:158–62.

4] Morimoto T, Kaido T, Uchiyama Y, Tokunaga H, Sakaki T, Iwasaki S. Rotationalobstruction of nondominant vertebral artery and ischemia: case report. Journalof Neurosurgery 1996;85:507–9.

5] Tominaga T, Takahashi T, Shimizu H, Yoshimoto T. Rotational vertebral arteryocclusion from occipital bone anomaly: a rare cause of embolic stroke. Casereport. Journal of Neurosurgery 2002;97:1456–9.

6] Barton JW, Margolis MT. Rotational obstructions of the vertebral artery at theatlantoaxial joint. Neuroradiology 1975;9:117–20.

7] Lemole Jr GM, Henn JS, Spetzler RF, Zabramski JM. Bow hunter’s stroke.Barrow Q; 2001. p. 17 (http://www.thebarrow.org/Education And Resources/Barrow Quarterly/205248).