cover and back 23-4 for website - barrowneuro.org

TREATMENT OPTIONS FOR THORACOLUMBAR BURST FRACTURES

BARROWBARROW

VOLUME 23 ◊ NUMBER 4 ◊ 2007

BARROW NEUROLOGICAL INSTITUTE OF ST. JOSEPH’S HOSPITAL AND MEDICAL CENTER • PHOENIX, ARIZONA

Q U A R T E R L Y

© Barrow, 2009

THE MISSION OF THE BARROW NEUROLOGICALINSTITUTE IS TO ADVANCE THE KNOWLEDGEAND PRACTICE OF MEDICINE IN THE NEURO-SCIENCES THROUGH BASIC AND CLINICALRESEARCH, THE EDUCATION OF MEDICALPROFESSIONALS, AND INNOVATION INCLINICAL TECHNIQUES AND TECHNOLOGY.

www.thebarrow.org

This issue’s cover illustrates a burst fracture of the L1vertebra and the methods of treatment for thoracolum-bar burst fractures. The article by Horn et al. on page4 discusses the efficacy of nonoperative treatment andcompares the outcomes of anterior and posterior surgi-cal approaches. Bracing with a custom-fit orthosis sta-bilizes the fracture and restricts movement of the spineduring healing. Operative treatment involves either ananterior corpectomy and internal fixation or posteriorreduction and fixation with pedicle screws and rods.The illustration is by Kristen Larson.

C O M M E N T S

Thoracolumbar burst fractures present a treatment dilemma. Whether they should be treated con-servatively or by surgical stabilization has not been clearly defined. Some patients with this conditionwho are treated nonoperatively have excellent outcomes. Because of the potential for devastatingneurological deficits, however, some clinicians prefer surgical stabilization. In this issue, Horn et al.retrospectively reviewed Barrow’s experience with conservative management of thoracolumbar burstfractures in patients with no neurological deficits. The authors found that nonoperative managementof burst fractures appears to be a safe initial treatment option. They also examined the relative effi-cacy of posterior and anterior approaches for the treatment of burst fractures when surgical treatmentis necessary.

Lekovic et al. report three cases of delayed vasospasm after clipping of incidentally discovered aneu-rysms and discuss possible mechanisms for vasospasm in the absence of subarachnoid hemorrhage.Prompt attention to changes in patients’ neurological condition after aneurysm clipping can help max-imize the chances for a good outcome.

Three interesting case reports round out this issue. Thiex et al. report a woman with a 3-year his-tory of pulsating bruit and dizziness who was found to have a dissecting aneurysm of the superiorcerebellar artery. The authors discuss treatment options for dissecting aneurysms in this unusual loca-tion. Garrett et al. report a child whose orbit was penetrated by a small tree branch while sleddingand examine the difficulties of distinguishing wooden objects on computed tomographic and mag-netic resonance imaging. Finally, Zargarpour et al. discuss the unusual occurrence of a thoracic spi-nal cord injury without radiographic abnormality.

Such cases highlight the broad range of problems that can result from injury or other lesions of thebrain and spinal cord and underscore the many challenges confronted by clinicians caring for thesepatients. We hope that you find these articles educational. Please consider sending a tax-deductibledonation in the enclosed self-addressed stamped envelope to help us continue to share our experi-ences with the medical community throughout the world.

Robert F. Spetzler, MDEditor-in-Chief

Senior EditorShelley A. Kick, PhD

Assistant EditorDawn Mutchler

Production EditorJaime L. Hoffman

Editorial AssistantsDiantha L. LeavittAndrew L. Wachtel

Medical IllustratorsMark Schornak, MSKristen L. Larson, MS

Three-Dimensional ModelerMichael Hickman

Graphic DesignerJaime L. Hoffman

Section Editors

NeurologyWilliam R. Shapiro, MD

NeurosurgeryVolker K. H. Sonntag, MD

NeurobiologyRonald J. Lukas, PhD

NeuroanesthesiologyPeter A. Raudzens, MD

NeuroradiologyShahram Partovi, MD

NeuropathologyStephen W. Coons, MD

BNI News EditorAndrew G. Shetter, MD

Neuroscience NursingVirginia Prendergast RN NP-C CNRN

Robert F. Spetzler

THE OFFICIAL JOURNAL OF THE

BARROW NEUROLOGICAL INSTITUTE OF

ST. JOSEPH’S HOSPITAL AND MEDICAL CENTER

VOLUME 23 ◊ NUMBER 4

C O M M E N T S

O R I G I N A L A R T I C L E

4

12

Analysis of Conservative Management and Comparisonof Neurological Outcomes Between Anterior and Pos-terior Approaches for Thoracolumbar Burst FracturesEric M. Horn, Iman Feiz-Erfan, Nicholas C. Bambakidis, StephenM. Papadopoulos, Volker K. H. Sonntag, and Nicholas Theodore

Bracing is a reasonable first-line treatment for thoracolumbar burstfractures. If conservative treatment fails, both anterior and posteriorapproaches are acceptable surgical options associated with no dif-ferences in neurological outcomes.

Delayed Ischemic Neurological Deficit After ElectiveClipping of Incidentally Discovered Aneurysms:Three Case ReportsGregory P. Lekovic, David Jho, P. Roc Chen, Volker K. H. Sonntag,Joseph M. Zabramski, and Robert F. Spetzler

Even in the absence of subarachnoid blood, delayed vasospasmcan represent a significant complication after an otherwise uncom-plicated aneurysmal clipping.

B A R R O W

C A S E R E P O R T S

4Analysis of Conservative Manage-ment and Comparison of Neurolog-ical Outcomes Between Anterior andPosterior Approaches for Thoraco-lumbar Burst Fractures

INFORMATION FOR SUBSCRIBERS

Readers are invited to submit letters, comments, or questions to the Editor, Neuroscience Publications;Barrow Neurological Institute; 350 West Thomas Road; Phoenix, AZ 85013-4496.

Correspondence regarding the Barrow Quarterly should be directed to Neuroscience Publications; BarrowNeurological Institute; 350 West Thomas Road; Phoenix, AZ 85013-4496 (602) 406-3593, or e-mail at [email protected].

Subscriber Information: To request or delete a subscription, or to notify of an address change or receipt oftwo or more copies of the Barrow Quarterly, please phone or write; Barrow Quarterly, Neuroscience Pub-lications; Barrow Neurological Institute; 350 W. Thomas Road; Phoenix, AZ 85013; (602) 406-3593 Mon-day through Friday, 8–5 local time or e-mail at [email protected].

The Barrow Quarterly is published by the Barrow Neurological Institute as part of its medical educationprogram and may be received free of charge, upon request. (All those receiving a complimentary copy whowish to help in defraying the costs of publication and distribution are invited to do so.)

© Copyright, 2009. Barrow Neurological Institute. All rights reserved. ISSN 1539-1914.

Editor-in-ChiefRobert F. Spetzler, MD

Linda HuntPresidentSt. Joseph’s Hospitaland Medical Center

Q U A R T E R L Y

23 Dissecting Aneurysm of the Superior CerebellarArtery: Case ReportRuth Thiex, Iman Feiz-Erfan, Francisco Ponce, Mohammad AliAziz-Sultan, and Robert F. Spetzler

A rare dissecting aneurysm of the superior cerebellar artery in apatient with long-standing symptoms is wrapped successfully.

12Delayed Ischemic Neurological

Deficit After Elective Clippingof Incidentally Discovered

Aneurysms: Three Case Reports

27Transorbital Penetrating

Tree Branch During a Sledding Accident

31 Motor Vehicle Accident Causing Thoracic SpinalCord Injury Without Radiographic AbnormalityJasmine Zargarpour, Poya Hedayati, and Shahram Partovi

Pediatric patients with neurological deficits but no negative findingson computed tomography or plain radiography should undergo MRimaging to rule out a rare thoracic spinal cord injury that might oth-erwise go undiagnosed.

27 Transorbital Penetrating Tree Branch During aSledding AccidentMark Garrett, Eric M. Horn, and Robert F. Spetzler

Surgical removal of tree branch fragments that penetrated the leftorbit of a 10-year-old girl provides an excellent clinical outcome.

4 BARROW QUARTERLY • Vol. 23, No. 4 • 2007

Analysis of Conservative Manage-ment and Comparison of Neuro-logical Outcomes Between Ante-rior and Posterior Approaches forThoracolumbar Burst Fractures

Thoracolumbar burst fractures arecommon injuries with the poten-

tial for causing significant morbidity de-pending on the stability of the fracture.7Over the past 50 years, treatment forthese injuries has progressed from non-operative bed rest to sophisticated fixa-tion techniques.2,8,12,13,34,35,40 However,the definition of stability for these frac-tures and the need for surgical stabiliza-tion remain controversial. In a recentrandomized, prospective clinical trial,outcomes of operative and nonopera-tive treatment for thoracolumbar burstfractures without neurologic deficitwere equivalent.48 No patients treatednonoperatively in this study needed sur-gical fixation, but this finding may reflectthe relatively small sample and incom-plete follow-up. Consequently, ques-tions still remain about the efficacy ofnonoperative and operative treatmentof these fractures. A recent systematicreview of all pertinent literature con-cluded that the optimum treatment ofthese injuries is yet to be determined.However, there appears to be a trend fa-voring initial nonoperative treatmentfor patients with no neurologic defi-cits.43

The appropriate surgical approach forpatients with a neurologic deficit or forthose with a grossly unstable fracture isalso controversial. Over the past 30 years,treatment of these injuries has evolvedfrom primarily posterior approaches toanterior approaches and recently is be-ginning to return to posterior approach-es.2,6,10,11,14,16,22,23,25,28,39,40,44-47 These trendshave paralleled significant advances in in-strumentation that have allowed stabilityto be attained without combined anteri-or and posterior fixation. Several direct

We retrospectively reviewed our experience with 86 patients with thoracolum-bar (T11-L2) burst fractures (1) to determine the efficacy of nonoperative treat-ment, (2) to determine predictors of successful nonoperative treatment, and (3)to compare outcomes of patients treated with an anterior or posterior approach.Patients treated initially with bracing were analyzed for failures and those failingbracing were compared with those treated successfully. All patients treated sur-gically were analyzed and those treated with an anterior approach were com-pared to those treated with a posterior approach. Twenty-eight patients were ini-tially treated nonoperatively with a brace and 58 underwent surgery. Of the 23patients initially treated nonoperatively that were available for follow-up, 8 failedbracing and underwent surgical fixation. Including the nonoperative failures, 66patients underwent surgical fixation: 48 from an anterior approach and 18 froma posterior approach. Although the difference was not significant, patients whounderwent a posterior approach who were available for follow-up had more com-plications (7/17) than those undergoing an anterior approach (8/43). There wereno differences in neurologic outcomes between the two groups. Even with ahigher than expected failure rate, nonoperative treatment of thoracolumbar burstfractures remains a safe, initial management option. If surgery is necessary fortreatment of thoracolumbar burst fractures, either an anterior or posterior ap-proach may be acceptable.

Key Words: burst fractures, paraplegia, spinal cord injury, spine, surgicalapproach, surgical fixation, trauma

Abbreviations Used: AO ASIF, Association for the Study of Internal Fixa-tion; ARDS, acute respiratory distress syndrome; ASIA, American SpinalInjury Association; CSF, cerebrospinal fluid; CT, computed tomography;CVA, cerebrovascular accident; DVT, deep venous thrombosis; MR, mag-netic resonance

Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and MedicalCenter, Phoenix, Arizona

Current Address: †Department of Neurological Surgery, Indiana University School of Medicine, In-dianapolis, Indiana; ‡Department of Surgery, Division of Neurosurgery, Maricopa Medical Center,Phoenix, Arizona; §Department of Neurological Surgery, Case Western Reserve University Schoolof Medicine and Neurological Institute, University Hospitals Case Medical Center, Cleveland, Ohio

Eric M. Horn, MD, PhD†

Iman Feiz-Erfan, MD‡

Nicholas C. Bambakidis, MD§

Stephen M. Papadopoulos, MD

Volker K. H. Sonntag, MD

Nicholas Theodore, MD


Copyright © 2009, Barrow Neurological Institute

comparisons have found few differencesin the outcomes of patients treated via ananterior or posterior approach.6,11,20,49

Furthermore, the only randomized,prospective trials comparing the two ap-proaches included only patients withoutneurologic deficits.11,49 Whether there isa difference in neurologic outcomes be-tween the two approaches is still unclear.We therefore reviewed our experiencewith thoracolumbar burst fractures (1) todetermine the efficacy of nonoperativetreatment, (2) to determine predictors ofsuccessful nonoperative treatment, and (3)to compare outcomes of patients treatedwith an anterior or posterior approach.

MethodsThe cases of patients presenting at our

institution with a traumatic thoracolum-bar burst fracture were reviewed retro-spectively. All patients with an acute frac-ture (less than 14 days old) at a single levelfrom T11 to L2 were included in thestudy. Furthermore, only burst fracturesthat lacked significant distraction of theposterior element or complete subluxa-tion were included in the study. The frac-tures primarily corresponded to the As-sociation for the Study of InternalFixation (AO ASIF) fracture types A3,B2, and B3.21 The inpatient and outpa-tient charts of each patient were reviewed,and their presenting computed tomo-graphic scans were used to measure frac-ture parameters.

Between December 1999 and Sep-tember 2005, 86 patients satisfied the in-clusion criteria. Their mean age was 39years (range, 9-77 years). The most com-mon fracture level was L1 (62 patients)followed by T12 (21 patients), L2 (2 pa-tients), and T11 (1 patient).

There were two stratifications forcomparison. Twenty-eight patients (meanage, 43 years) initially treated with brac-ing alone were compared to all 58 patients(mean age, 37 years) treated with surgicalfixation. All 28 nonoperative patients un-derwent placement of a custom-fit mold-ed plastic thoracolumbosacral orthoticfrom the same manufacturer (Hanger Or-thotic, Bethesda, MD). The 58 opera-tive patients underwent either an anteri-

or corpectomy and internal fixation via athoracoabdominal approach or posteriorreduction and fixation using predomi-nantly pedicle screws and rods.

Forty-eight patients (mean age, 37years) treated via an anterior corpecto-my and internal fixation (anterior ap-proach) were compared to 18 patients(mean age, 37 years) treated via posteri-or reduction and fixation (posterior ap-proach). This comparison included pa-tients who failed nonoperative treatmentand subsequently underwent surgical fix-ation. A corpectomy of the fracturedvertebra was performed in patients un-dergoing anterior fixation. Then eitheran allograft, rigid titanium cage packedwith autograft, or expandable titaniumcage packed with autograft was used forfusion material. In all anterior patients,fixation was limited to the immediatelyadjacent vertebrae, and either a screw-plate or screw-rod system was used. Sev-enteen patients treated via a posterior ap-proach had pedicle screws and rodsplaced between one and two adjacentlevels above and below the fracture site.In some cases posterolateral fusion wasperformed using autograft with bone fu-sion extenders. In one patient, a hookand rod system was used for posterior fix-ation.

For each stratification, Student’s t-tests were used to compare means, andχ2 analysis was used to test proportions.Significance was set at p < 0.05. Thisstudy was approved by our Institution-al Review Board.

ResultsPatients initially treated with bracing

alone had less severe injuries, both radio-graphically and neurologically, than op-erative patients (Table 1). The mean hos-pital stay for the 28 nonoperative patientswas 6.4 days (range, 1-21 days): 24 pa-tients were discharged to home and 2each were discharged to a rehabilitation orskilled nursing facility. Of these 28 pa-tients, 23 (82%) were available for long-term follow-up (mean, 10.1 months).The other five patients lived outside Ari-zona or were from Mexico and could notbe reached for follow-up. One monthafter bracing, one patient in the nonop-erative group died from associated trau-matic injuries.

Of the 23 patients with adequate fol-low-up, 8 (35%; mean age, 41 years) failednonoperative treatment in a mean of 20days (range, 3 to 79 days). The primaryreasons for their failure were signs of in-stability or increased kyphosis on lateraldynamic radiographs (upright and supine).Despite these structural failures, no patientexperienced neurologic injury from thefailed bracing. When the severity of frac-tures was compared between the 15 non-operative patients in whom bracing wassuccessful (mean age, 46 years; Fig. 1) andthose who failed (Fig. 2), the only differ-ence was a greater extent of spinal canalcompromise in the failed group (Table 2).

Including the patients who failedbracing and subsequently underwentsurgical fixation, 66 patients underwentanterior or posterior fixation. Of these,

5

Horn et al: Analysis of Conservative Management and Comparison of Neurological Outcomes Between Anterior and Posterior Approaches for Thoracolumbar Burst Fractures

BARROW QUARTERLY • Vol. 23, No. 4 • 2007

Variable Nonoperative Operative

Vertebral height loss 32.6 40.7*(mean %)

Spinal canal compromise 30.1 48.0*(mean %)

Kyphosis 7 12.5*(mean %)

Neurologically intact 96 64*(% ASIA E)

*p < 0.05.

Table 1. Presentation of 28 Patients Treated Nonopera-tively and 58 Patients Treated Operatively

48 anterior approaches were performed:44 initially and 4 after failing bracing.In contrast, 14 patients initially and 4patients who failed bracing underwenta posterior approach. Patients under-going an anterior approach had a greaterdegree of spinal canal compromise andkyphosis than the group undergoing theposterior approach (Table 3). These dataindicate that our institutional bias was totreat structurally severe injuries with an-terior corpectomy and fixation. It ap-pears that the decision was not based onthe presence of neurologic compromise,which was similar in the two groups.

There were no differences in meanblood loss between anterior and poste-rior approaches (989 ml vs. 1338 ml, re-spectively) or in the length of postoper-ative hospital stay (9.1 vs. 7.9 days). Inthe anterior group, 43 of the 48 patients(90%) had adequate follow-up (mean,10.3 months) and 17 of the 18 (94%) pa-tients in the posterior group were avail-able for a mean follow-up of 9.6 months.Patients undergoing an anterior ap-proach had more structurally severe frac-tures and a lower rate of complications(8/43, 19%) than patients undergoing aposterior approach (7/17, 41%), but thedifference in the rate of complicationsfailed to reach significance (p > 0.05).No differences were noted in the com-plication rates associated with the differ-ent types of hardware used for fixation inpatients undergoing either an anterioror posterior approach. Details concern-ing the types of complications in eachgroup are presented in Table 4. Finally,there were no differences in the neuro-logic recovery (Fig. 3) of patients un-dergoing an anterior (7/43, 16%; Fig. 4)or posterior approach (3/17, 18%; Figs.5 and 6).

DiscussionCompared to earlier reports,3-5,15,

32,36,37,41,47,48 our data show that a highrate of failure was associated with non-operative treatment. Although most au-thors have reported successful treatmentusing a nonoperative approach, there arenotable exceptions. In one cohort of22 patients, six patients failed nonoper-

6



Figure 1. (A) Initial reconstructed sagittal CT scan of a 51-year-old patient with a L1 burstfracture treated by bracing alone. (B) Lateral radiograph obtained 12 months after injuryshows that the fracture remained stable, despite some persistent loss of vertebral height.

Figure 2. (A) Initial sagittal T2-weighted MR image of a 39-year-old patient with a L1 burst frac-ture. (B) Supine and (C) upright lateral radiographs obtained 12 days after injury showed a9.3-degree increase in kyphosis while in the brace. After bracing failed, the fracture was fix-ated via an anterior approach. (D) Lateral radiograph obtained 18 months after injury confirmsspinal stablity.

A B

A B

C D

ative treatment and underwent surgicalstabilization.25 In another early study,17% of patients treated nonoperativelydeveloped neurologic symptoms thatnecessitated surgical fixation.8

Failure of Conservative TreatmentWhy so many patients in our series

failed nonoperative treatment is unclear.Overall, loss of vertebral height, degreeof spinal canal compromise, and degreeof kyphosis in the patients treated non-operatively were all less than in the pa-tients treated operatively (Table 1). Thisfact led to the choice of nonoperativetreatment in the first place. When wecompared patients who failed nonop-erative treatment to those in whom itwas successful, the only difference iden-tified was a greater degree of spinal canalcompromise at admission in the former(Table 2). Logically, the most impor-tant biomechanical indicators of insta-bility are degree of kyphosis and loss ofvertebral height. Therefore, the differ-ence in spinal canal compromise wasprobably not a significant factor under-lying the failure of bracing.

One explanation for the high rate offailure associated with nonoperative treat-ment in our patients could be the use ofa custom-fit plastic brace rather than acast, but it seems unlikely. Several stud-ies have used similar types of bracing suc-cessfully.36,41,48 In fact, in one study pa-tients treated nonoperatively without theuse of a brace were ambulated immedi-ately and showed excellent results.37

Another possibility is a low thresh-old for nonoperative failure and subse-quent surgical treatment. Because ourstudy was retrospective and there wereno preexisting criteria for failure ofnonoperative treatment, this possibilitycannot be discounted. However, all pa-tients who failed nonoperative treat-ment had radiographic demonstrationof increased kyphosis or overt instabili-ty when placed upright in the brace.Furthermore, all patients who failed hadsevere, persistent pain that greatly re-stricted ambulation in the brace. De-spite differences in the criteria used fordetermining failure of nonoperativebracing between our study and previ-

7



Variable Successful† Failed‡

Vertebral height loss 32.2 30.1(mean %)


Kyphosis 5.9 7.6(mean degrees)†15 patients, ‡8 patients, *p < 0.05.

Table 2. Outcome of Initial Bracing of NonoperativePatients

Variable Anterior Approach Posterior Approach

Vertebral height loss 41.6 33.5*(mean %)


Kyphosis 12.8 9.3*(mean degrees)

Neurologically intact 65 72(% ASIA E)*p < 0.05.

Table 3. Presentation of 48 Patients Treated via an An-terior Approach and 18 Patients Treated via a PosteriorApproach

Anterior Approach† No. Posterior Approach‡ No.

Surgical Complications

Instability 1 Instability 1

Infected hardware 1 Hardware failure 2

Incomplete 1 Incomplete 1decompression decompression

CSF leak 1 Retained drainage 1catheter

Superficial wound 1infection

Medical Complications

CVA 1 DVT 1

ARDS 1

Pneumonia 1

DVT 1

Total complications, †19% (8/43); ‡41% (7/17).

Table 4. Complications in Patients Undergoing an Anterior or Posterior Approach

ous reports, we believe that our criteriawere appropriate for the managementof these patients.

Despite the high rate of failure asso-ciated with nonoperative treatment inour patients, none developed a neuro-logic deficit, including the patients whofailed bracing and ultimately underwentsurgical fixation. The relative safety ofnonoperative treatment is well estab-lished.3-5,15,36,37,41,47,48 In fact, only twostudies have reported neurologic dete-rioration during nonoperative treat-ment.9,32 When the present report is in-

cluded, the neurologic complicationrate is less than 2% of the 360 reportedpatients treated nonoperatively. Thus,even if the failure rate associated withnonoperative treatment is as high as re-ported in the present cohort, such con-servative management still represents asafe and effective initial treatment forthoracolumbar burst fractures.

Posterior ApproachPosterior reduction and pedicle screw-

rod fixation are widely used when surgi-cal fixation is needed for unstable thora-

columbar burst fractures. Before the earlyreports of transpedicular fixation forthoracolumbar burst fractures, variousfixation devices, including the Luqueloop and Harrington rods,9,10,23,44,46,47

were used. Although posterior fixa-tion using these older techniques wasefficacious, the use of pedicle screw-rod techniques has allowed shorter fix-ation segments with equivalent fusionrates.1,2,11,17,19,33,36,39,42

In our series, the most commontechnique of posterior fixation (11 of 18patients) was a pedicle screw-rod con-struct that spanned two levels above andbelow the fracture site. In six patients,however, the construct spanned only theimmediately adjacent levels, and a hook-rod system was used in one patient.

We found no correlation betweentype of hardware used and complica-tions, although hardware failures werefound in one patient for each type ofhardware used. Because the number ofpatients receiving each type of hardwarewas small, statistical comparison of com-plication rates across types of hardwareis impossible. High failure rates havebeen associated with the use of smallsegment pedicle screw-rod systems (onelevel above and below).26,42 Thus, inmost cases where posterior fixation isused to treat thoracolumbar burst frac-tures, the longer (two levels above andbelow) pedicle screw-rod construct isrecommended.

One criticism of the posterior ap-proach for thoracolumbar burst fracturesis the relative inability to completely re-move bone fragments compressing thethecal sac in the anterior portion of thespinal canal. Although there is little dis-agreement that adequate decompressioncan be performed from an anterior ap-proach, it is becoming widely acceptedthat the posterior approach is equally ef-fective. In fact, excellent decompressionof the spinal canal has been associated withnewer techniques used to access the ver-tebral body from a posterior approach.17,38

These transpedicular or lateral extracavi-tary approaches are especially needed incases of neurological injury because dis-traction alone is inadequate for the re-duction of most bony fragments.31

8



Figure 3. ASIA scores at initial presentation and last follow-up for patients treated via an(A) anterior or (B) posterior approach.

A

B

Our data support the ability to achieveadequate decompression of the spinalcanal from a posterior approach. Afterundergoing a posterior approach for resid-ual compression, none of our patientsneeded anterior decompression. The onepatient who underwent a second opera-tion for inadequate decompression after aposterior approach needed only an ex-tension of a partial laminectomy. Fur-thermore, most patients with an initialneurologic deficit improved significantlyafter posterior decompression and fixa-tion. This finding correlates with reportsthat have found significant improvementof neurologic deficits in patients with tho-racolumbar burst fractures after posteriordecompression and fixation.2,6,20,39,40,45,47

Anterior ApproachOver the past 30 years, anterior ap-

proaches for surgical fixation of thora-columbar burst fractures have becomemore common.14,16,22,24,29,34,35,49 Becausethe impingement of bony fragments onthe thecal sac can be visualized directly,an anterior approach yields more com-plete decompression than a posterior ap-proach.11 In several reports, patientswith thoracolumbar burst fractures andneurologic deficits treated via an anteri-or approach recovered significant neu-rologic function.14,16,22,24,34,35 The ques-tion remains, however, whether anteriordecompression offers a significant ad-vantage in terms of neurologic outcomecompared to a posterior approach. Noprospective, randomized trials havespecifically analyzed this issue. Howev-er, several retrospective studies havefound no differences in neurologic out-come associated with these two ap-proaches.6,20,35

It is well established that the degree ofinitial spinal canal compromise corre-

9



Figure 4. (A) Sagittal T2-weighted MR image of a 46-year-old patient immediately afterpresentation shows an L1 burst fracture. (B) Reconstructed sagittal CT scan shows a solidfusion 12 months after anterior fixation and fusion.

Figure 5. (A) Initial reconstructed sagittalCT scan of a 19-year-old patient showsan L1 burst fracture. (B) The fracture wasreduced and fixated from a posterior ap-proach, and the patient’s outcome wasexcellent. (C) After 9 months, however,the hardware failed due to incomplete fu-sion. (D) An anterior approach was per-formed for definitive treatment.

A B

A B

C D

lates with the presence of a neurologicdeficit.18,27 There is no correlation,however, between the extent of spinalcanal compromise and the severity ofneurologic deficits.27,30 More impor-tantly, no correlation has been foundbetween the initial extent of spinal canalcompromise and eventual neurologicoutcome.18,30 Furthermore, the extentof decompression has no effect on neu-rologic outcome, regardless of the man-ner of decompression.18,30 Given thatan anterior approach typically yields bet-ter decompression than a posterior ap-proach, these data may partially accountfor the lack of differences in neurologicoutcomes associated with anterior andposterior approaches.

ConclusionsNonoperative management of tho-

racolumbar burst fractures is a reason-able first-line treatment: There were noneurologic injuries in patients initiallytreated with bracing. Nonetheless, a sig-nificant number of these patients mayeventually require operative treatment.Although a higher complication ratewas associated with the posterior ap-proach compared to the anterior ap-proach, there were no differences in

neurologic outcomes between the twoapproaches. We advocate nonoperativemanagement as the initial treatment forthoracolumbar burst fractures in patientswithout neurologic deficits. When pa-tients with a neurologic deficit need de-compression, either anterior or poste-rior decompression and fixation areacceptable options.

References1. Andress HJ, Braun H, Helmberger T, et al: Long-

term results after posterior fixation of thoraco-lumbar burst fractures. Injury 33:357-365, 2002

2. Been HD, Bouma GJ: Comparison of two typesof surgery for thoraco-lumbar burst fractures:combined anterior and posterior stabilisation vs.posterior instrumentation only. Acta Neurochir(Wien) 141:349-357, 1999

3. Butler JS, Walsh A, O’Byrne J: Functional out-come of burst fractures of the first lumbar verte-bra managed surgically and conservatively. IntOrthop 29:51-54, 2005

4. Cantor JB, Lebwohl NH, Garvey T, et al: Nonop-erative management of stable thoracolumbarburst fractures with early ambulation and brac-ing. Spine 18:971-976, 1993

5. Chow GH, Nelson BJ, Gebhard JS, et al: Func-tional outcome of thoracolumbar burst fracturesmanaged with hyperextension casting or brac-ing and early mobilization. Spine 21:2170-2175,1996

6. Danisa OA, Shaffrey CI, Jane JA, et al: Surgicalapproaches for the correction of unstable thora-columbar burst fractures: a retrospective analy-sis of treatment outcomes. J Neurosurg 83:977-983, 1995

7. Denis F: The three column spine and its signifi-cance in the classification of acute thoracolumbarspinal injuries. Spine 8:817-831, 1983

8. Denis F: Spinal instability as defined by the three-column spine concept in acute spinal trauma.Clin Orthop Relat Res (189):65-76, 1984

9. Denis F, Armstrong GW, Searls K, et al: Acute tho-racolumbar burst fractures in the absence of neu-rologic deficit. A comparison between operativeand nonoperative treatment. Clin Orthop RelatRes (189): 142-149, 1984

10. Dick W, Kluger P, Magerl F, et al: A new device forinternal fixation of thoracolumbar and lumbarspine fractures: The ‘fixateur interne’. Paraplegia23:225-232, 1985

11. Esses SI, Botsford DJ, Kostuik JP: Evaluation ofsurgical treatment for burst fractures. Spine15:667-673, 1990

12. Frankel HL, Hancock DO, Hyslop G, et al: Thevalue of postural reduction in the initial manage-ment of closed injuries of the spine with para-plegia and tetraplegia. I. Paraplegia 7:179-192,1969

13. Guttman L: Surgical aspects of the treatment oftraumatic paraplegia. J Bone Joint Surg31B:399-403, 1949

14. Haas N, Blauth M, Tscherne H: Anterior plating inthoracolumbar spine injuries. Indication, tech-nique, and results. Spine 16:S100-S111, 1991

15. Hitchon PW, Torner JC, Haddad SF, et al: Man-agement options in thoracolumbar burst frac-tures. Surg Neurol 49:619-626, 1998

16. Kaneda K, Abumi K, Fujiya M: Burst fractures withneurologic deficits of the thoracolumbar-lumbarspine. Results of anterior decompression andstabilization with anterior instrumentation. Spine9:788-795, 1984

17. Kaya RA, Aydin Y: Modified transpedicular ap-proach for the surgical treatment of severe tho-racolumbar or lumbar burst fractures. Spine J4:208-217, 2004

18. Kim NH, Lee HM, Chun IM: Neurologic injuryand recovery in patients with burst fracture ofthe thoracolumbar spine. Spine 24:290-293,1999

19. Knop C, Fabian HF, Bastian L, et al: Late resultsof thoracolumbar fractures after posterior instru-mentation and transpedicular bone grafting.Spine 26:88-99, 2001

20. Korovessis P, Piperos G, Sidiropoulos P, et al: Spi-nal canal restoration by posterior distraction oranterior decompression in thoracolumbar spinalfractures and its influence on neurological out-come. Eur Spine J 3:318-324, 1994

21. Magerl F, Aebi M, Gertzbein SD, et al: A compre-hensive classification of thoracic and lumbar in-juries. Eur Spine J 3:184-201, 1994

22. McAfee PC, Bohlman HH, Yuan HA: Anterior de-compression of traumatic thoracolumbar frac-tures with incomplete neurological deficit using aretroperitoneal approach. J Bone Joint SurgAm 67:89-104, 1985

23. McAfee PC, Yuan HA, Lasda NA: The unstableburst fracture. Spine 7:365-373, 1982

10



Figure 6. (A) Initial reconstructed sagittal CT scan of a 46-year-old patient with a T12 burstfracture. Two years after posterior fixation, (B) a lateral radiograph showed a solid construct.

A B

24. McDonough PW, Davis R, Tribus C, et al: Themanagement of acute thoracolumbar burst frac-tures with anterior corpectomy and Z-plate fixa-tion. Spine 29:1901-1908, 2004

25. McEvoy RD, Bradford DS: The management ofburst fractures of the thoracic and lumbar spine.Experience in 53 patients. Spine 10:631-637,1985

26. McLain RF, Sparling E, Benson DR: Early failureof short-segment pedicle instrumentation for tho-racolumbar fractures. A preliminary report. JBone Joint Surg Am 75:162-167, 1993

27. Meves R, Avanzi O: Correlation between neuro-logical deficit and spinal canal compromise in198 patients with thoracolumbar and lumbar frac-tures. Spine 30:787-791, 2005

28. Mikles MR, Stchur RP, Graziano GP: Posterior in-strumentation for thoracolumbar fractures. J AmAcad Orthop Surg 12:424-435, 2004

29. Mirbaha MM: Anterior approach to the thoraco-lumbar junction of the spine by a retroperitoneal-extrapleural technic. Clin Orthop Relat Res(91):41-47, 1973

30. Mohanty SP, Venkatram N: Does neurological re-covery in thoracolumbar and lumbar burst frac-tures depend on the extent of canal compro-mise? Spinal Cord 40:295-299, 2002

31. Mueller LA, Mueller LP, Schmidt R, et al: The phe-nomenon and efficiency of ligamentotaxis afterdorsal stabilization of thoracolumbar burst frac-tures. Arch Orthop Trauma Surg 126:364-368,2006

32. Mumford J, Weinstein JN, Spratt KF, et al: Tho-racolumbar burst fractures. The clinical efficacyand outcome of nonoperative management.Spine 18:955-970, 1993

33. Parker JW, Lane JR, Karaikovic EE, et al: Suc-cessful short-segment instrumentation and fu-sion for thoracolumbar spine fractures: A con-secutive 41/2-year series. Spine 25:1157-1170,2000

34. Sasso RC, Best NM, Reilly TM, et al: Anterior-onlystabilization of three-column thoracolumbar in-juries. J Spinal Disord Tech 18 Suppl:S7-14,2005

35. Schnee CL, Ansell LV: Selection criteria and out-come of operative approaches for thoracolum-bar burst fractures with and without neurologicaldeficit. J Neurosurg 86:48-55, 1997

36. Shen WJ, Liu TJ, Shen YS: Nonoperative treat-ment versus posterior fixation for thoracolumbarjunction burst fractures without neurologic defi-cit. Spine 26:1038-1045, 2001

37. Shen WJ, Shen YS: Nonsurgical treatment ofthree-column thoracolumbar junction burst frac-tures without neurologic deficit. Spine 24:412-415, 1999

38. Snell BE, Nasr FF, Wolfla CE: Single-stage tho-racolumbar vertebrectomy with circumferentialreconstruction and arthrodesis: surgical tech-nique and results in 15 patients. Neurosurgery58:ONS-8, 2006

39. Stromsoe K, Hem ES, Aunan E: Unstable verte-bral fractures in the lower third of the spine treat-ed with closed reduction and transpedicular pos-terior fixation: a retrospective analysis of 82fractures in 78 patients. Eur Spine J 6:239-244,1997

40. Tasdemiroglu E, Tibbs PA: Long-term follow-upresults of thoracolumbar fractures after posteriorinstrumentation. Spine 20:1704-1708, 1995

41. Tezer M, Erturer RE, Ozturk C, et al: Conserva-tive treatment of fractures of the thoracolumbarspine. Int Orthop 29:78-82, 2005

42. Tezeren G, Kuru I: Posterior fixation of thora-columbar burst fracture: short-segment pediclefixation versus long-segment instrumentation. JSpinal Disord Tech 18:485-488, 2005

43. Thomas KC, Bailey CS, Dvorak MF, et al: Com-parison of operative and nonoperative treatmentfor thoracolumbar burst fractures in patients with-out neurological deficit: a systematic review. JNeurosurg Spine 4:351-358, 2006

44. Wang GJ, Whitehill R, Stamp WG, et al: The treat-ment of fracture dislocations of the thoracolum-bar spine with halofemoral traction and Harring-ton rod instrumentation. Clin Orthop Relat Res(142):168-175, 1979

45. Weyns F, Rommens PM, Van Calenbergh F, et al:Neurological outcome after surgery for thora-columbar fractures. A retrospective study of 93consecutive cases, treated with dorsal instru-mentation. Eur Spine J 3:276-281, 1994

46. Willen J, Lindahl S, Irstam L, et al: Unstable tho-racolumbar fractures. A study by CT and con-ventional roentgenology of the reduction effectof Harrington instrumentation. Spine 9:214-219,1984

47. Willen J, Lindahl S, Nordwall A: Unstable thora-columbar fractures. A comparative clinical studyof conservative treatment and Harrington instru-mentation. Spine 10:111-122, 1985

48. Wood K, Buttermann G, Mehbod A, et al: Oper-ative compared with nonoperative treatment ofa thoracolumbar burst fracture without neuro-logical deficit. A prospective, randomized study.J Bone Joint Surg Am 85-A:773-781, 2003

49. Wood KB, Bohn D, Mehbod A: Anterior versusposterior treatment of stable thoracolumbar burstfractures without neurologic deficit: A prospec-tive, randomized study. J Spinal Disord Tech18 Suppl:S15-S23, 2005

11



12 BARROW QUARTERLY • Vol. 23, No. 4 • 2007

Delayed Ischemic NeurologicalDeficit After Elective Clipping ofIncidentally Discovered Aneu-rysms: Three Case Reports

Cerebral vasospasm (also known asintracranial arterial spasm) is char-

acterized by intense localized contrac-tion of arterial smooth muscle and canbe defined by both angiographic andclinical components.61 Angiographicvasospasm is the radiographic evidenceof arterial narrowing, classically usingcontrast angiography, but alternativelydocumented by CT or MR angiogra-phy. Clinical vasospasm is the onset ofneurological deficits corresponding tothe regions of arterial narrowing thoughtto be related to cerebral ischemia or in-farction.

The timing of vasospasm in relationto the causative stimulus defines the typeof mechanism likely to be involved.Chemical vasospasm associated withSAH or neurotrauma occurs in a delayedfashion, persists for days to weeks, and isusually synonymous with the develop-ment of delayed ischemic neurologicaldeficits.15,20 In contrast, mechanical va-sospasm, which follows manipulationof intracranial arteries, occurs immedi-ately and only lasts for hours.57 Despitethe increasing amount of research andexperience with cerebral vasospasm, ourunderstanding of its pathophysiology isstill incomplete, and prophylaxis or ther-apies are less than ideal.

Of special interest is the phenomenonof symptomatic vasospasm after an un-complicated craniotomy for incidentalaneurysms, because this entity does notfit neatly into existing theories of vaso-spasm.2,4,13,40,46,59 Vasospasm associatedwith unruptured aneurysms has been re-ported.6,9,17,28,38,41,46 Nonetheless, severalfactors require that these reports be inter-preted with caution when compared tothe type of cases represented by our pa-

Delayed vasospasm after subarachnoid hemorrhage (SAH) is a well-recognizedphenomenon commonly attributed to the chemical irritation of intracranial ves-sels by blood or its breakdown products. However, delayed postoperative va-sospasm in the absence of SAH is an unusual event that does not easily lend it-self to popular theories about the origin of vasospasm. We present three casesof severe vasospasm that produced delayed ischemic neurological deficits afterthe patients underwent uncomplicated surgical clipping of incidentally discov-ered, unruptured aneurysms. In all three cases, the vasospasm failed to respondto hyperdynamic therapy. After treatment, however, the last two patients signifi-cantly improved both clinically and angiographically. One patient was treated byplacement of cisternal catheters and continuous papaverine infusion; the otherpatient responded to endovascular treatment with intra-arterial nicardipine. Theseobservations suggest that delayed vasospasm, although rare, is a significant po-tential complication after uncomplicated elective craniotomy for aneurysmal clip-ping, even in the absence of subarachnoid blood. Moreover, this phenomenonmay have implications relevant to our understanding of the pathophysiology ofvasospasm itself.

Key Words: aneurysm, cerebral vasospasm, ischemia, neurological def-icit

Abbreviations Used: ACoA, anterior communicating artery; CSF, cere-brospinal fluid; CT, computed tomography; ICA, internal carotid artery;MCA, middle cerebral artery; MR, magnetic resonance; PCoA, posteriorcommunicating artery; SAH, subarachnoid hemorrhage


Current Address: †House Ear Clinic, Los Angeles, California; ‡Department of Neurosurgery, Har-vard Medical School, Boston, Massachusetts; §Department of Neurosurgery, the University of TexasMedical School at Houston, Houston, Texas

Gregory P. Lekovic, MD, PhD, JD†

David Jho, MD‡

P. Roc Chen, MD§

Volker K. H. Sonntag, MD

Joseph M. Zabramski, MD

Robert F. Spetzler, MD



tients. First, many of these reports predatethe era of MR imaging and CT. Second,most reported cases involve patients withsymptomatic aneurysms (i.e., manifest-ing with cranial neuropathy, presumablyafter increasing in size, or with headache).These aneurysms did not manifest withovert SAH. Nonetheless, the aneurysmswere dynamic lesions that could haveleaked subarachnoid blood—just not inquantities sufficient to be detected by CTor lumbar puncture. This issue has gainedimport as the widespread availability ofnoninvasive screening studies such as CTangiography and MR angiography hasincreased the detection rate of incidentalaneurysms. Whereas vasospasm afteraneurysmal SAH is routinely attributedto spasmogens contained in subarachnoidblood or to its breakdown products, theoccurrence of vasospasm in patients withunruptured, incidental aneurysms requiresalternative explanations.

Illustrative CasesWe describe two cases of delayed va-

sospasm after uncomplicated electiveclipping of incidental aneurysms and in-clude a patient reported previously(Table 1).6

Case 1In March 2003 a 58-year-old woman

with a long-standing history of multiple

sclerosis presented with four incidentalsaccular aneurysms involving the ACoA,left PCoA, left ICA terminus, and leftMCA (Fig. 1A and B). She had no fam-ily history of intracranial aneurysms. Hermedical history was significant for mul-tiple sclerosis, and her only medicationwas conjugated estrogen hormone re-placement therapy. Neurologically, shewas intact except for long-standingweakness of the proximal lower extrem-ities, which was attributed to multiplesclerosis.

The patient underwent an elective leftmodified orbitozygomatic craniotomy formicrosurgical clipping of the four aneu-rysms and wrapping of an aneurysmal di-latation of the left MCA (JMZ). FourYasargil titanium aneurysm clips wereused. At surgery no evidence of previousSAH was found. Postoperative angiogra-phy confirmed patency of the parent ves-sels and well-placed clips with no appar-ent residual aneurysm or spasm (Fig. 1Cand D). The procedure was otherwiseuneventful, and the patient was admittedto the intensive care unit.

The patient’s initial postoperativecourse was unremarkable. As she neareddischarge on postoperative Day 4, shesuddenly developed difficulties with find-ing words. CT of the head showed un-remarkable postoperative changes that in-cluded an extra-axial fluid collection inthe left frontal lobe. The CT was nega-

tive for SAH. Angiography, however,showed severe vasospasm of the left M1(Fig. 1E and F), which improved aftertreatment with hyperdynamic therapy(i.e., hypertensive hypervolemic hemo-dilution) (Fig. 1G and H). On postoper-ative Day 4, the patient’s craniotomy wasre-opened to remove the muslin wrapand to place cisternal catheters throughwhich papaverine was irrigated continu-ously for 4 days (Fig. 1I). She made acomplete clinical and angiographic re-covery and was discharged home on post-operative Day 16. At the patient’s 2- and7-month follow-up visits, her only com-plaint was positional headache. Neuro-logically, she was at her baseline.

Case 2In April 2004 a 53-year-old woman

sought treatment after experiencingfloaters and flashes of light through herleft eye and occasional left retro-orbitalpain for 10 months. Her medical histo-ry was negative for aneurysm or for afamily history of SAH. She was takingconjugated estrogens for postmenopausalhormone replacement therapy. At pre-sentation she was neurologically intact.

Four months after the patient’s initialpresentation, she underwent elective ce-rebral angiography, which showed a 12-mm x 8-mm aneurysm involving theparaclinoid segment of the ICA just dis-tal to the ophthalmic artery (Fig. 2A, B,

13

Lekovic et al: Delayed Ischemic Neurological Deficit After Elective Clipping of Incidentally Discovered Aneurysms: Three Case Reports


Patient Age/Sex Location No. Clips Delayed deficits Angiographic findings Treatment

1 58/F ACoA 1 Aphasia Severe vasospasm HHH, papaverineL PCoA 1 left M1

L ICA 1L MCA 1

2 53/F L paraclinoid 5 Aphasia Severe vasospasm left HHH, IA nicardipineICA A1/M1, ICA

3 54/F R MCA 1 Hemiparesis Diffuse severe vasospasm HHHbifurcation of right ICA, ACA, MCA

territories

HHH =hyperdynamic therapy (hypertensive, hypervolemic hemodilution); IA = intra-arterial; ACoA = anterior communicatingartery; L = left; PCoA =posterior communicating artery; ICA = internal carotid artery; MCA = middle cerebral artery; R = right;ACA = anterior cerebral artery

Table 1. Clinical Summary of Patients

14



D E

G H

F

I

Figure 1. Case 1. Preoperative (A) oblique and (B) lateral digital subtraction angiograms show four saccular aneurysms—one on theACoA, one on the left PCoA, one on the left ICA, and one on the left MCA. Postoperative (C) anteroposterior and (D) lateral angiogramsafter initial clipping of the aneurysms confirm satisfactory placement of the clips and the absence of vasospasm. On postoperative Day4, (E) anteroposterior and (F) lateral views of left ICA angiographic injections show severe vasospasm. On postoperative Day 6, (G) pos-teroanterior and (H) lateral angiographic views show that the vasospasm has improved. (I) CT of the head after the muslin wrappedaround the mild proximal aneurysmal dilatation of the left MCA has been removed and after the cisternal catheters through which pa-paverine was irrigated continuously for 4 days have been placed.

CA B

15



Figure 2. Case 2. Initial preoperative (A) anteroposterior and (B) lateral angiographic views show the irregular wide-neck aneurysmon the left ophthalmic artery. (C) Preoperative lateral unsubtracted left ICA angiogram shows the location and size of the aneurysmon the left ophthalmic segment of the ICA. On postoperative Day 4 before intra-arterial nicardipine therapy was instituted, (D) poste-rior-anterior and (E) lateral angiographic views of left ICA injections show severe vasospasm. On postoperative Day 6, (F) posterior-anterior and (G) lateral angiographic views show radiographic improvement of the vasospasm. On postoperative Day 8, (H) antero-posterior and (I) lateral angiograms show continued improvement of vasospasm.

A B C

D E F

G H I

and C). The irregular aneurysm en-compassed a 6-mm segment of the ICAand projected laterally. Due to thebroad-based nature of the aneurysm andto its proximity to the ophthalmic ar-tery, endovascular treatment was not at-tempted. The patient was recom-mended for a craniotomy and aneurysmclipping.

Subsequently, the patient underwentan elective left modified orbitozygomat-ic craniotomy. Treatment included tem-porary occlusion of the left ICA andcommon carotid artery and a left anteri-or clinoidectomy to expose the cavern-ous portion of the ICA. At surgery therewas no evidence of prior SAH. Five an-eurysm clips (RFS) were used to clip-reconstruct the ICA, and the craniotomywas closed in the usual manner. Therewere no complications.

The patient’s initial postoperativecourse was unremarkable. On postop-erative Day 4, however, she suddenlydeveloped intermittent difficulties withfinding words. Hyperdynamic thera-py was instituted. Angiography showedsevere vasospasm of the distal left ICA(Fig. 2D and E), and intra-arterial ni-cardipine (9 mg) was administered.Her problems with speech improvedrapidly.

Angiography performed on postop-erative Day 6 showed significant im-provement of vasospasm involving thedistal left ICA, A1, and M1 (Fig. 2F andG). Intra-arterial nicardipine was againadministered.

On postoperative Day 8, transcrani-al Doppler ultrasonography showed el-evated values for the left MCA. Thepatient developed mild confusion andoccasionally required reorientation toplace and time. She also exhibited someinappropriate speech and subtle rightpronator drift. Repeat angiographyshowed that mild residual vasospasm ofthe left distal ICA, A1, and M1 was con-tinuing to improve (Fig. 2H and I). Hy-perdynamic therapy was continuedwithout further endovascular interven-tion. By postoperative Day 10, the pa-tient’s neurological status had improvedto her baseline. She was discharged onpostoperative Day 16.

Case 3Previously, we reported the case of a

54-year-old woman with an asymp-tomatic aneurysm involving the rightMCA bifurcation. The aneurysm wasdetected incidentally on CT after shesustained a concussion during an as-sault.6 The 2-month gap between theassault and elective clipping of the an-eurysm effectively ruled out neurotrau-ma as a cause of vasospasm. The un-ruptured aneurysm was clipped (VKHS).During surgery transient focal spasm ofthe MCA was treated successfully withpapaverine. After a brief uneventfulhospital course, the patient was dis-charged.

While at home on postoperative Day9, the patient developed left hemipare-sis. Angiography confirmed diffuse va-sospasm of the right supraclinoid ICA,anterior cerebral artery, and MCA. De-spite aggressive treatment with volumeexpansion and dexamethasone, the pa-tient eventually developed an infarctionin the distribution of the right MCA.When she recovered, mild weakness ofher left hand persisted. Mechanical,chemical, and hypothalamic causes ofthe vasospasm were ruled out. The pro-posed pathogenesis was local release ofchemical factors from the wall of the an-eurysm or local autonomic dysfunction.

DiscussionWe reviewed all cases of vasospasm

that followed clipping of unruptured an-eurysms treated at our institution (Table1) and reported in the literature (Table2) between 1975 and 2005 (i.e., sincethe advent of CT, Table 1). Includingour patients, there are 19 such cases.These cases can be further divided intotwo groups: early vasospasm (i.e., onsetwithin 24 hours of surgery) and delayedvasospasm (onset thereafter).

Immediate postoperative mechani-cal vasospasm cannot be considered thesame phenomenon as delayed ischemicneurological deficits resulting from truedelayed vasospasm. Mechanical vaso-spasm induced by the manipulation orelectrostimulation of cerebral arteries ischaracterized by an early onset (within

hours) and usually runs a transient self-resolving course that lasts no more than1 or 2 days.19,22,43,57 True delayed vaso-spasm occurs as early as 3 days after ini-tial SAH. Its occurrence peaks 5 to 10days after SAH. It can persist 2 to 3weeks and, if left untreated, can causepermanent neurological deficits ordeath.12,15,33,36,55,58,65

In a previous analysis of the literaturepublished before the advent of CT (be-tween 1970 and 1980), Peerless foundeight cases of vasospasm that followeduncomplicated clipping of unrupturedaneurysms. Forty-five similar cases wereeliminated due to identifiable compli-cating factors such as SAH from neuro-trauma or aneurysmal leakage, priorSAH from aneurysmal rupture, me-chanical trauma to vessels from clips orretractors, and thromboembolic events.40

However, these eight angiographicallyconfirmed cases with symptomatic va-sospasm occurring within hours of sur-gery are consistent with a mechanicalcause. Of three cases described bySimeone and Peerless,46 two patients(Cases A and B) developed vasospasmimmediately after surgery. These pa-tients could also be categorized as in-stances of mechanical vasospasm. Sime-one and Peerless reported three otherpatients (Cases C,D,and E ) who devel-oped delayed vasospasm after aneurysmclipping. One patient (Case C) had pre-sented with headache, and two (CasesD and E) had presented with third cra-nial nerve palsy.46 Another patient (CaseF ) presented with both headache andvasospasm at surgery, but no frank evi-dence of previous SAH was observed.

Between 1991 and 1998, Kitizawaand coworkers clipped 30 unrupturedparaclinoid ICA aneurysms, of whichnine (33%) developed postoperative va-sospasm.28 Of these nine cases, four pa-tients had moderate postoperative an-giographic vasospasm and had neverbeen symptomatic. Only three patientswith severe angiographic vasospasm (de-fined as luminal narrowing greater thanor equal to 70%) were thought to havesymptomatic vasospasm (delayed is-chemic neurological deficits). Thesethree patients were treated successfully

16



with hyperdynamic therapy plus hyper-baric oxygen or intra-arterial papaver-ine.28

Kitizawa et al.28 retrospectively ana-lyzed 11 factors involved in the devel-opment of vasospasm. They found thattwo factors, temporary occlusion of theICA and the presence of multiple clips,were associated with delayed vasospasm.The authors suggested that intraopera-tive bleeding around the dural ring andthe aforementioned factors, which in-creased mechanical stimulation of thevascular wall, may amplify the risk ofdelayed postoperative vasospasm associ-ated with unruptured aneurysms in-volving the paraclinoid ICA. Othervariables such as age, gender, history ofprior SAH, aneurysm size or number,left or right side, operation time, expo-sure of the cervical ICA, and extent ofdural ring incision failed to correlatewith the development of delayed vaso-spasm. Although the small sample re-quires these statistical correlations to beinterpreted with caution, such a highincidence of delayed vasospasm associ-ated with uncomplicated clipping ofunruptured aneurysms in the anatomicregion of the paraclinoid ICA is still re-markable.

Paolini et al. described a patient whodeveloped delayed vasospasm after un-complicated aneurysmal clipping onpostoperative Day 28.38 This interval isthe longest reported delay between elec-tive clipping and onset of symptomaticvasospasm. On postoperative Day 1,their patient had postretraction edemaand a small hemorrhage in the frontallobe. She remained asymptomatic untilpostoperative Day 28 when she devel-oped left lower facial droop and leftupper extremity weakness. Treated suc-cessfully with volume expansion and anti-platelet therapy, she returned to herbaseline neurological status within 12hours.

Interestingly, all 10 patients who de-veloped delayed ischemic neurologicaldeficits (Table 2) were women. Onepatient in our series had a history ofmultiple sclerosis, and immunologicalprocesses may contribute to the patho-genesis of vasospasm.37 As in two of our

patients, two or more clips have beenplaced in many of the reported patientsduring surgical treatment of their un-ruptured aneurysm (Table 2). Kitazawaet al. reported that multiple clips mayincrease the risk of developing cerebralvasospasm.28

Implications for Pathophysiology ofVasospasm

There is some research on delayedvasospasm after skull base surgery andneurotrauma,26,34,39,56 but the mecha-nisms involved in cerebral vasospasm arebest studied relative to SAH.15,33,55,61,65

Delayed vasospasm after SAH are mostpopularly attributed to chemoreceptor-mediated irritation of arteries by bloodor its breakdown products such as oxy-hemoglobin, especially since the amountand distribution of blood in aneurysmalSAH are proportional to the occurrenceof vasospasm.12,14,15,33,36,55,58,65

The concept of endothelial damagealone serving as a stimulus for vasospasmis based on a primate model. The in-tracranial portion of the ICA of mon-keys was cannulated with a 30-gaugeneedle using a technique that avoidsSAH or occlusion of the lumen.46 Ofnine monkeys undergoing this proce-dure, six developed severe local arterialconstriction that persisted about 3 days.Within 1 day one monkey developedrapidly fatal vasospasm of all ipsilateralcerebral arteries.46

Catheter-induced vasospasm is awell-recognized complication of angi-ography, even in the absence of vesselrupture or dissection.11,30 These obser-vations indicate that endothelial injuryin the absence of significant subarach-noid blood is sufficient to produce severevasospasm. The cause may be a spas-mogenic agent released from endothe-lial cells or direct endothelial dysfunc-tion itself. It has been suggested thatdynamic aneurysms or surgical injurycan produce endothelial damage. Theproduction of prostacyclin is thereby re-duced, rendering damaged intracranialarteries susceptible to unopposed vaso-constriction by thromboxanes andprostaglandin endoperoxides in the cir-culating blood or CSF.7,17 Another pos-

sible cause, membrane-bound tissue fac-tor in the CSF of patients with aneu-rysmal SAH, is predictive of vasospasm,brain injury, and overall outcome.23

Other proposed causes of chemicalvasospasm include various cellular agentsderived from injured vessels or cerebraltissue such as endothelin-1, free radicals,lipid peroxidation agents, and phos-pholipase C pathway components suchas protein kinase C, diacylglycerol, andintracellular calcium.32,62 Oxyhemo-globin or endothelin-1 can directly pro-long smooth muscle contraction. Freeradicals can inhibit smooth muscle re-laxation by inactivating nitric oxide.Lipid peroxidases can increase proteinkinase C and intracellular calcium forsmooth muscle contraction.32,50

In addition to local vasoactive medi-ators, generalized circulating pharmaco-logic agents can produce chemical vaso-spasm or sensitize vessels to spasmogens,especially in regions of damaged en-dothelia. Certain chemical mediatorssuch as methysergide, cocaine, and fe-male hormones can exacerbate or pro-duce forms of pharmacologically in-duced vasospasm in the intracranialvasculature.8,31,51 Females tend to de-velop vasospastic phenomena in vascu-lar beds throughout the body moreoften than males,16,27,42 and it may bemore than coincidence that two of ourcases were postmenopausal women onhormone-replacement therapy.

Wilkins suggested the concept ofhypothalamic or neurogenic vaso-spasm.56,60 He believed that damage tothe anterior hypothalamus can causesympathetic hyperactivity and increasecirculating catecholamines, with sub-sequent dysfunction of autoregulationand autonomic nervous control of ce-rebrovascular tone. Because the hypo-thalamus is involved in the autonomicregulation of vascular tone, hypotha-lamic dysfunction may be the essentialstep in the development of vaso-spasm.56,60

This argument is based on four ideas.(1) Unlike vasospasm in animal models,which is based solely on subarachnoidblood, vasospasm in humans has othercharacteristics. (2) Not all patients with

17



18



Authors Age/Sex History Aneurysm No. Clips Notable Factors Location Size (mm)

Simeone & 40/M NA L MCA NA NA Dynamic w/evidence of Peerless 197546 growth over 2 y

30/F NA L ICA NA NA Rickham reservoir used PCoA junct due to intraop spasm

of L ICA + MCA

32/F NA L ICA NA 0 Aneu mainly intracavernous, Ophth junct filled w/ copper wires to

promote clot formation

Raynor & 25/F NA L ICA NA NA Preop (4 d) spasm of L ICA Messer 198041 PCoA junct next to aneu neck, postop

incidental R ophth aneu

Friedman et al. 30/F NA L PCoA 10 1 Preop (1 mo) spasm of L ICA 198317 and basilar a. Dynamic w/

evidence of growth over 2 wks

Bloomfield & 54/F* HTN, postconcussion R MCA near 7 2 Neurotrauma 2 mo before aneu Sonntag 19856 syndrome, pyelonephritis bifurcation surgery, transient intraop focal

R MCA vasospasm treatedw/papaverine

Day & Raskin 42/F MS (mild paraparesis, R supraclin 10 NA Bloodless CSF x 219869 L optic atrophy) ICA/ fetal origin

PCA junct

Kitazawa et al. 29/M NA R paraclin ICA 7 3 None200528 (3 aneu)

39/F NA R paraclin ICA 6 2 None

52/F NA R paraclin ICA 5 1 None

57/M NA L paraclin ICA 5 1 None

59/F NA L paraclin ICA 5 1 None

62/F NA L paraclin ICA 10 3 Temp occlusion of ICA, Mild EDH

21/F* NA L paraclin ICA 4 1 Temp occlusion of ICA

53/F* NA L paraclin ICA 4 3 SAH(3 aneu)

63/F* NA L paraclin ICA 5 2 Temp occlusion of ICA,Mild EDH

Paolini et al. 47/F* Signif smoking R MCA “large” 2 Small R frontal hemorrhage w/ 200538 bifurcation postretraction edema on POD 1

Lekovic et al. 58/F* Postmenopausal HRT, ACoA, L PCoA, NA 4 Left MCA muslin wrap(current study) MS L ICA, L A1, L M1

53/F* Postmenopausal HRT, Ophth segment 8 5 Preop narrow caliber of L ICA diverticulosis, cholecys- of L ICAtectomy, appendectomy,hysterectomy, L shoulder+ bilateral knee surg

Table 2. Cases of Vasospasm Associated with Unruptured Aneurysms in the Literature

†word finding difficulty, ‡ophthalmoplegia due to incidental R ophthalmic aneurysm, §mild residual weakness of L hand, *cases oftrue delayed symptomatic vasospasm after incidental aneurysm clipping. aneu =aneurysm, ACA = anterior cerebral artery,ACoA = anterior communicating artery, angio = angiography, asympt = asymptomatic, CN =cranial nerve, CSF =cerebrospinalfluid, EDH =epidural hematoma, GOS =Glasgow Outcome Scale, GR =good recovery, HHH =hyperdynamic therapy (hemodilu-

19



Postoperative Symptoms Treatment GOSNeurological Vasospasm onsetDeficits & location

Aphasia, R hemiparesis Early (POD 0) clinical NA MDvasospasm, (POD 2) angiovasospasm of L ICA + MCA

Aphasia, RUE weakness Early (POD 0) clinical Papaverine + angio phenoxy- MDvasospasm, (POD 1) angio benzamine through Rickhamvasospasm of L ICA + MCA reservoir, mannitol + glycerol

Aphasia, somnolence, Delayed (POD 5) Intraarterial phenoxybenzamine, GRR hemiparesis L ICA,L ACA,L MCA steroids, + mannitol

Somnolence, R hemiparesis Preop early (POD 0) L ICA NA GR‡(Preop L CN 3 palsy from (increased spasmmass effect) since 4 d preop)

Asympt vasospasm Preop (1 mo) L ICA, NA GR(Preop L CN 3 palsy from basilar arterymass effect)

L hemiparesis Delayed (POD 9) R ICA – HHH + dexamethasone GR§supraclinoid segment, 4 mg/6hr/5dR ACA, R MCA

Headache Preop (1 wk) NA GRdiffuse vasospasm

None Delayed (POD 14) HHH GRmod vasospasm

Convulsions Delayed (POD 5) HHH GRmod vasospasm



Aphasia, R hemiparesis Delayed (POD 3) HHH GRmod vasospasm

Aphasia Delayed (POD 16) HHH GRmod vasospasm

Aphasia, Gertzmann Delayed (POD 12) HHH + intraarterial GRsyndrome severe vasospasm papaverine

Aphasia Delayed (POD 9) HHH + hyperbaric O2 GRsevere vasospasm

Aphasia, R hemiparesis Delayed (POD 5) HHH + hyperbaric O2 GRsevere vasospasm

L lower facial droop, Delayed (POD 28) HHH + antiplatelet therapy GRL hemiparesis severe vasospasm

Aphasia† Delayed (POD 4) HHH +intradural papaverine GRL M1

Aphasia† Delayed (POD 4) HHH + intraarterial nimodipine GRL distal ICA, L A1, L M1

tional hypertensive hypervolemic), HRT =hormone replacement therapy, HTN =hypertension, ICA = internal carotid artery,MCA = middle cerebral artery, MD = moderate disability, mod =moderate, MS = multiple sclerosis, ophth =ophthalmic, paraclinoid= paraclin, PCoA =posterior communicating artery, POD =postoperative day, RUE = right upper extremity, signif = significant, su-praclinoid = supraclin, and temp = temporary.

subarachnoid blood develop vasospasm,but it most often occurs in those withaneurysms located adjacent to the hy-pothalamus or involving vessels that sup-ply the hypothalamus. (3) Postmortemevidence has shown destructive lesionsin the anterior hypothalamus of patientswith aneurysmal SAH consistent withsympathetic dysfunction. (4) Finally, thestudy of postoperative and posttraumat-ic cerebral vasospasm provides similaranatomic support.56,60

Selective stimulation of nerves organglia has restored autonomic balancein animal models of SAH-induced va-sospasm.5,64 Futhermore, when appliedeither prophylactically53,54 or as treat-ment, reversible functional sympathec-tomy via cervical spinal cord stimula-tion with epidural electrodes hasimproved cerebral vasospasm in animalmodels.10,25 In a similar manner, spinalcord stimulation has been used success-fully to treat cerebral vasospasm in se-lect patients.45,49 That nitric oxide is in-volved as a mediator in this response18,52

suggests overlap with the mediators in-volved in chemical vasospasm. Thus,the two ideas of chemical and neuro-genic vasospasm can be combined: It ispossible that a spasmogenic substance isreleased from the hypothalamus in re-sponse to stimulation or damage. Wil-son and Field showed that porcine hy-pothalamic extract, but not porcinecerebral cortical tissue extract, couldproduce vasospasm when injected intothe cisterna magna of dogs; their findingindicates the presence of a spasmogenicsubstance in the hypothalamus.63

Management of Delayed IschemicNeurological Deficits After ElectiveClipping of Aneurysms

In our cases, as in the other reportedcases of delayed vasospasm, hyperdy-namic therapy used in conjunction withdexamethasone, antiplatelet drugs, hy-perbaric oxygen, and intradural or intra-arterial local vasodilators resulted ingood overall outcomes.6,28,38 Althoughthe potential benefits of these treatmentoptions seem to outweigh their risks,the small samples and lack of controlsfor comparison make it difficult to de-

termine which therapies are actually ef-fective. The only common treatmentfactor among all of these cases was hy-perdynamic therapy. Similarly, all pa-tients with moderate angiographic va-sospasm in the series reported byKitazawa et al. were treated effectivelywith hyperdynamic therapy alone; nonedeveloped symptomatic vasospasm.28

Although hyperdynamic therapyusing a combination of crystalloids andcolloids can exacerbate cerebral edema inpatients with acute infarcts,21 it is usefulin the prophylaxis and therapy of vaso-spasm in aneurysmal SAH.24,47 Despitethe use of invasive and noninvasive mon-itoring to guide treatment, hyperdy-namic therapy can also produce occa-sional cardiopulmonary complicationssuch as congestive heart failure and pul-monary edema related to volume over-load.44 Calcium antagonists, balloon an-gioplasty, or injection of intra-arterial orintradural vasodilators can be helpful ifvasospasm is severe or when aggressivemedical therapy has failed.1 Nonethe-less, outcomes have been excellent in re-sponse to a combination of the afore-mentioned therapies. Consequently,current experimental treatments such asspinal cord stimulation for producing re-versible functional sympathectomy andendothelin antagonists probably add novalue at this time for the treatment of de-layed vasospasm associated with unrup-tured incidental aneurysms.

Role of Routine AngiographyAfter aneurysms are clipped, the rou-

tine use of intraoperative (or immediatepostoperative) angiography is benefi-cial.29 However, whether the addition-al cost and risk of complications associ-ated with routine delayed angiographycan be justified for use in asymptomat-ic patients is unclear. Kitazawa et al.routinely obtained postoperative an-giograms in their series of unrupturedaneurysms a mean of 12 days after sur-gery.28 Although these authors discov-ered a surprisingly high rate of radio-graphic vasospasm, severe vasospasmonly occurred in the symptomaticgroup. Kitazawa et al. did not assess thecost-effectiveness of this strategy.28

Rather, a high-index of suspicionand prompt evaluation with angiogra-phy are warranted for patients withsymptoms consistent with delayed is-chemic neurological deficits after elec-tive clipping of aneurysms. Therefore,patients should be made aware of therare possibility that delayed vasospasmmay occur within the first two weeks oftreatment so that they are prepared toreturn for the management of vaso-spasm if necessary. As experience withthis population grows, it may becomepossible to use CT angiography to assessfor delayed postoperative vasospasm in asimpler and less invasive manner thanconventional angiography.3,35,48

ConclusionBased on our observations, we sug-

gest that vasospasm may be a rare butsignificant complication of elective cra-niotomy for aneurysmal clipping.Moreover, the incidence of delayed va-sospasm after elective clipping may beunderestimated because postoperativeangiograms are not routinely obtained.Therefore, a high index of clinical sus-picion for delayed ischemic neurologi-cal deficits is warranted for patients un-dergoing elective clipping of aneurysms,particularly in the paraclinoid regionand in women, or when multiple an-eurysm clips are placed. These patientsshould be counseled about the possibil-ity of delayed ischemic neurological def-icits. After aneurysm clipping, patientswith complaints consistent with delayedischemic neurological deficits shouldundergo angiography to rule out vaso-spasm as the cause. After uncomplicat-ed clipping of unruptured aneurysms,early recognition of vasospasm as a pos-sible cause of changes in patients’ neu-rological condition may allow appro-priate interventions to be instituted tomaximize the chances of a favorableoutcome.

20



21



References1. Anonymous: Mechanical and pharmacologic

treatment of vasospasm. AJNR Am J Neurora-diol 22:S26-S27, 2001

2. Adams CBT: Vasospasm: Occurence in condi-tions other than subarachnoid hemorrhagefrom a bleeding aneurysm, in Wilkins RH (ed):Cerebral Arterial Spasm: Proceedings of theSecond International Workshop on Cerebral Va-sospasm. Baltimore: Williams & Wilkins, 1980,p171

3. Anderson GB, Ashforth R, Steinke DE, et al: CTangiography for the detection of cerebral vaso-spasm in patients with acute subarachnoid hem-orrhage. AJNR Am J Neuroradiol 21:1011-1015, 2000

4. Aoki N, Origitano TC, al Mefty O: Vasospasm afterresection of skull base tumors. Acta Neurochir(Wien) 132:53-58, 1995

5. Atalay B, Bolay H, Dalkara T, et al: Transcornealstimulation of trigeminal nerve afferents to in-crease cerebral blood flow in rats with cerebralvasospasm: A noninvasive method to activatethe trigeminovascular reflex. J Neurosurg97:1179-1183, 2002

6. Bloomfield SM, Sonntag VK: Delayed cerebralvasospasm after uncomplicated operation on anunruptured aneurysm: Case report. Neurosur-gery 17:792-796, 1985

7. Boullin DJ, Bunting S, Blaso WP, et al: Respons-es of human and baboon arteries to prostaglandinendoperoxides and biologically generated andsynthetic prostacyclin: Their relevance to cere-bral arterial spasm in man. Br J Clin Pharmacol7:139-147, 1979

8. Buttner A, Mall G, Penning R, et al: The neu-ropathology of cocaine abuse. Leg Med (Tokyo)5 Suppl 1:S240-S242, 2003

9. Day JW, Raskin NH: Thunderclap headache:Symptom of unruptured cerebral aneurysm.Lancet 2:1247-1248, 1986

10. Ebel H, Schomacker K, Balogh A, et al: High cer-vical spinal cord stimulation (CSCS) increasesregional cerebral blood flow after induced sub-arachnoid haemorrhage in rats. Minim InvasiveNeurosurg 44:167-171, 2001

11. Eckard DA, Purdy PD, Girson MS, et al: Intraar-terial papaverine for relief of catheter-induced in-tracranial vasospasm. AJR Am J Roentgenol158:883-884, 1992

12. Endo S, Suzuki J: Experimental cerebral vaso-spasm after subarachnoid hemorrhage: Devel-opment and degree of vasospasm. Stroke8:702-707, 1977

13. Fein JM: Unruptured aneurysms and cerebralvasospasm, in Wilkins RH (ed): Cerebral Arte-rial Spasm: Proceedings of the Second Inter-national Workshop on Cerebral Vasospasm.Baltimore: Williams & Wilkins, 1980, pp 499-504

14. Fisher CM, Kistler JP, Davis JM: Relation of cere-bral vasospasm to subarachnoid hemorrhage vi-sualized by computerized tomographic scan-ning. Neurosurgery 6:1-9, 1980

15. Fisher CM, Roberson GH, Ojemann RG: Cere-bral vasospasm with ruptured saccular aneu-rysm—the clinical manifestations. Neurosur-gery 1:245-248, 1977

16. Freedman RR, Mayes MD: Familial aggregationof primary Raynaud’s disease. Arthritis Rheum39:1189-1191, 1996

17. Friedman P, Gass HH, Magidson M: Vasospasmwith an unruptured and unoperated aneurysm.Surg Neurol 19:21-25, 1983

18. Goksel HM, Karadag O, Turaclar U, et al: Nitricoxide synthase inhibition attenuates vasoactiveresponse to spinal cord stimulation in an exper-imental cerebral vasospasm model. Acta Neu-rochir (Wien) 143:383-390, 2001

19. Gomez CR, Backer RJ, Bucholz RD: Transcrani-al Doppler ultrasound following closed head in-jury: Vasospasm or vasoparalysis? Surg Neu-rol 35:30-35, 1991

20. Grolimund P, Weber M, Seiler RW, et al: Timecourse of cerebral vasospasm after severe headinjury. Lancet 1:1173, 1988

21. Gupta D, Sharma BS, Gupta SK, et al: Postoper-ative hypertensive-hypervolaemic-haemodilution(Triple H) therapy in the treatment of vasospasmfollowing aneurysmal subarachnoid haemor-rhage. Neurol India 48:126-131, 2000

22. Heilbrun MP: The relationship of neurological sta-tus and the angiographical evidence of spasmto prognosis in patients with ruptured intracrani-al saccular aneurysms. Stroke 4:973-979, 1973

23. Hirashima Y, Nakamura S, Suzuki M, et al: Cere-brospinal fluid tissue factor and thrombin-antithrombin III complex as indicators of tissueinjury after subarachnoid hemorrhage. Stroke28:1666-1670, 1997

24. Hoh BL, Carter BS, Ogilvy CS: Risk of hemor-rhage from unsecured, unruptured aneurysmsduring and after hypertensive hypervolemic ther-apy. Neurosurgery 50:1207-1211, 2002

25. Karadag O, Eroglu E, Gurelik M, et al: Cervical spi-nal cord stimulation increases cerebral corticalblood flow in an experimental cerebral vasospasmmodel. Acta Neurochir (Wien) 147:79-84, 2005

26. Kasamo S, Asakura T, Yamamoto Y: [Posttrau-matic cerebrovascular narrowing. A case reportand review of the literature]. No Shinkei Geka12:1411-1416, 1984

27. Keller KB, Lemberg L: Prinzmetal’s angina. AmJ Crit Care 13:350-354, 2004

28. Kitazawa K, Hongo K, Tanaka Y, et al: Postoper-ative vasospasm of unruptured paraclinoid ca-rotid aneurysms: Analysis of 30 cases. J ClinNeurosci 12:150-155, 2005

29. Klopfenstein JD, Spetzler RF, Kim LJ, et al: Com-parison of routine and selective use of intraoper-ative angiography during aneurysm surgery: Aprospective assessment. J Neurosurg 100:230-235, 2004

30. Le Roux PD, Elliott JP, Eskridge JM, et al: Risksand benefits of diagnostic angiography after an-eurysm surgery: A retrospective analysis of 597studies. Neurosurgery 42:1248-1254, 1998

31. MacGregor EA: Female sex hormones and mi-graine. Rev Neurol (Paris) 161:677-678, 2005

32. Mayberg MR: Cerebral vasospasm. NeurosurgClin N Am 9:615-627, 1998

33. Mayberg MR, Batjer HH, Dacey R, et al: Guide-lines for the management of aneurysmal sub-arachnoid hemorrhage. A statement for health-care professionals from a special writing group ofthe Stroke Council, American Heart Association.Circulation 90:2592-2605, 1994

34. Mohr JP, Kase CS: Cerebral vasospasm. Part I. Incerebral vascular malformations. Rev Neurol(Paris) 139:99-113, 1983

35. Ochi RP, Vieco PT, Gross CE: CT angiography ofcerebral vasospasm with conventional angio-graphic comparison. AJNR Am J Neuroradiol18:265-269, 1997

36. Osaka K: Prolonged vasospasm produced bythe breakdown products of erythrocytes. J Neu-rosurg 47:403-411, 1977

37. Ostergaard JR: Risk factors in intracranial sac-cular aneurysms. Aspects on the formation andrupture of aneurysms, and development of ce-rebral vasospasm. Acta Neurol Scand 80:81-98, 1989

38. Paolini S, Kanaan Y, Wagenbach A, et al: Cere-bral vasospasm in patients with unruptured intra-cranial aneurysms. Acta Neurochir (Wien)147:1181-1188, 2005

39. Parkinson D, Bachers G: Arteriovenous mal-formations. Summary of 100 consecutive su-pratentorial cases. J Neurosurg 53:285-299,1980

40. Peerless SJ: Postoperative cerebral vasospasmwithout subarachnoid hemorrhage, in Wilkins RH(ed): Cerebral Arterial Spasm: Proceedings of theSecond International Workshop on Cerebral Va-sospasm. Baltimore: Williams & Wilkins, 1980,pp 496-498

41. Raynor RB, Messer HD: Severe vasospasm withan unruptured aneurysm: Case report. Neuro-surgery 6:92-95, 1980

42. Recober A, Geweke LO: Menstrual migraine.Curr Neurol Neurosci Rep 5:93-98, 2005

43. Schaller C, Zentner J: Vasospastic reactions inresponse to the transsylvian approach. SurgNeurol 49:170-175, 1998

44. Segal E, Greenlee JD, Hata SJ, et al: Monitoringintravascular volumes to direct hypertensive,hypervolemic therapy in a patient with vaso-spasm. J Neurosurg Anesthesiol 16:296-298,2004

45. Shinonaga M, Takanashi Y: Vasodilating effect ofspinal cord stimulation for cerebral vasospasm.Acta Neurochir Suppl 77:229-230, 2001

46. Simeone FA, Peerless SJ: Prolonged cerebral va-sospasm without subarachnoid hemorrhage, inSmith RR, Robertson JT (eds): SubarachnoidHemorrhage and Cerebrovascular Spasm. Spring-field, IL: Charles C Thomas, 1975, pp 206-225

47. Swift DM, Solomon RA: Unruptured aneurysmsand postoperative volume expansion. J Neuro-surg 77:908-910, 1992

48. Takagi R, Hayashi H, Kobayashi H, et al: Three-dimensional CT angiography of intracranial va-sospasm following subarachnoid haemorrhage.Neuroradiology 40:631-635, 1998

49. Takanashi Y, Shinonaga M: Spinal cord stimula-tion for cerebral vasospasm as prophylaxis. Neu-rol Med Chir (Tokyo) 40:352-356, 2000

50. Tanaka Y, Shigenobu K, Nakayama K: Effects ofendothelin-1 on the myogenic contraction of ca-nine cerebral artery in response to quick stretch.Res Commun Mol Pathol Pharmacol 109:95-101, 2001

51. Tanner JR: St. Anthony’s fire, then and now: Acase report and historical review. Can J Surg30:291-293, 1987

22



52. Toda N, Tanaka T, Ayajiki K, et al: Cerebral va-sodilatation induced by stimulation of the ptery-gopalatine ganglion and greater petrosal nerve inanesthetized monkeys. Neuroscience 96:393-398, 2000

53. Visocchi M, Argiolas L, Meglio M, et al: Spinal cordstimulation and early experimental cerebral spasm:The “functional monitoring” and the “preventingeffect”. Acta Neurochir (Wien) 143:177-185, 2001

54. Visocchi M, Di Rocco F, Meglio M: Protective ef-fect of spinal cord stimulation on experimentalearly cerebral vasospasm. Conclusive results.Stereotact Funct Neurosurg 76:269-275, 2001

55. Weir B, Grace M, Hansen J, et al: Time course ofvasospasm in man. J Neurosurg 48:173-178,1978

56. Wilkins RH: Hypothalamic dysfunction and in-tracranial arterial spasms. Surg Neurol 4:472-480, 1975

57. Wilkins RH: Intracranial vascular spasm in head in-juries, in Vinken PJ, Bruyn GW (eds): Handbook ofClinical Neurology. Injuries of the Brain and Skull,Part I. Amsterdam: North Holland, 1975, pp 163-198

58. Wilkins RH: The role of intracranial arterial spasmin the timing of operations for aneurysm. ClinNeurosurg 24:185-207, 1977

59. Wilkins RH: Intracranial arterial spasm after pro-cedures other than operations for intracranial an-eurysm, in Wilkins RH (ed): Cerebral ArterialSpasm: Proceedings of the Second InternationalWorkshop on Cerebral Vasospasm. Baltimore:Williams & Wilkins, 1980, pp 505-509

60. Wilkins RH: Possible role of the hypothalamus inthe development of intracranial arterial spasm,in Wilkins RH (ed): Cerebral Arterial Spasm: Pro-ceedings of the Second International Workshopon Cerebral Vasospasm. Baltimore: Williams &Wilkins, 1980, pp 266-269

61. Wilkins RH: Cerebral vasospasm. Crit Rev Neu-robiol 6:51-77, 1990

62. Wilkins RH: Cerebral vasospasm in conditionsother than subarachnoid hemorrhage. Neuro-surg Clin N Am 1:329-334, 1990

63. Wilson JL, Feild JR: The production of intracranialvascular spasm by hypothalamic extract. J Neu-rosurg 40:473-479, 1974

64. Yarnitsky D, Lorian A, Shalev A, et al: Reversal ofcerebral vasospasm by sphenopalatine ganglionstimulation in a dog model of subarachnoid hem-orrhage. Surg Neurol 64:5-11, 2005

65. Zubkov AY, Nanda A, Zhang JH: Signal trans-duction pathways in cerebral vasospasm. Patho-physiology 9:47-61, 2003

23BARROW QUARTERLY • Vol. 23, No. 4 • 2007

Dissecting Aneurysm of theSuperior Cerebellar Artery:Case Report

We report the rare case of a dis-secting aneurysm of the SCA

that presented without SAH and thatwas treated by wrapping.

Case ReportA 37-year-old woman had a 3-year

history of a pulsating whooshing soundin her left ear, recurrent episodes of dizzi-ness, light-headedness, gait ataxia, andtremor of the right hand. Her medicalhistory was unremarkable for systemicillnesses or trauma as was her family his-tory. She smoked two packs of cigarettesdaily. Six months before she was referredto our hospital, she had undergone MRimaging (Fig. 1A) after experiencingsyncope. The study showed flow voidswithin the left superior peripontine cis-tern consistent with a vascular nidus fedby the left SCA. Three months later,DSA showed that about the first cen-timeter of the left SCA appeared nor-mal. Thereafter, it followed an unusual-ly tortuous corkscrew-shaped course forabout 6 mm. It then straightened to itsnormal curvilinear configuration (Fig.1B-D). No projection showed a distinctanatomical relationship between the an-eurysm and the parent artery. Given theirregularity of the lesion, the interven-tional neuroradiologists refrained fromexploration and surgical inspection wasoffered.

A left-sided full orbitozygomatic cra-niotomy was performed. The superficialtemporal artery was dissected in antici-pation of microvascular bypass. The an-terior choroidal artery was followedposteriorly to the basilar artery andLiliequist’s membrane (Fig. 2A). Inci-sion of the tentorial edge further exposed

A 37-year-old woman had a 3-year history of a pulsating bruit in her left ear anddizziness. MR angiography suggested a vascular abnormality related to the leftSCA. DSA showed that the abnormality involved the anterior pontomesencephalicsegment of the SCA 1 cm distal to its origin from the basilar artery where its max-imum diameter was 6 mm. The patient underwent a left-sided full orbitozygo-matic craniotomy. A dissecting aneurysm of the left SCA was wrapped becausethe length of the aneurysm rendered revascularization impossible in case theSCA had to be sacrificed. The patient’s long-standing history of symptoms andthe location of the lesion make this dissecting aneurysm exceptional.

Key Words: dissecting aneurysm, orbitozygomatic approach, superiorcerebellar artery

Abbreviations Used: CN3, oculomotor nerve; DSA, digital subtractionangiography; MR, magnetic resonance; PCA, posterior cerebral artery;SAH, subarachnoid hemorrhage; SCA, superior cerebellar artery


Current Address: †Department of Surgery, Division of Neurosurgery, Maricopa Medical Center,Phoenix, Arizona; ‡Department of Neurosurgery, University of Miami Miller School of Medicine,Miami, Florida

Ruth Thiex, MD

Iman Feiz-Erfan, MD†

Francisco Ponce, MD

Mohammad Ali Aziz-Sultan, MD‡



the lesion, providing a good view of thetortuous portion of the dissecting aneu-rysm and the proximal side of the parentartery (Fig. 2B-C). A large segment ofdissection is extended distally. Conse-quently, it was impossible to revascu-larize this vessel in case it needed to besacrificed. Therefore, the dissecting an-eurysm was wrapped with cotton (Fig.2D). The patient tolerated the proce-dure well. She had no postoperative def-icits and was discharged 2 days after sur-gery.

DiscussionAneurysms on the main trunk of the

SCA are rare. Before 1990 only 13 cases,

all of which were saccular aneurysms,had been reported.6,15,19,21,22,25,30,35,38

Most arose from the origin of the SCA.In 1990 Hirose et al.10 first described afusiform aneurysm of the SCA in a 25-year-old patient with SAH. Theyclipped the SCA at a site distal to theorigin of a perforating artery, and thepatient recovered with mild cerebellarataxia. Since then both endovascular andsurgical techniques have been used totreat SCA aneurysms.5,20,26

Location of Dissecting AneurysmsDissecting aneurysms of the SCA are

extremely rare; only four such cases havebeen reported to date.10-12,29 The ori-gin of dissecting aneurysms remains un-

clear but seems to be related to athero-sclerosis and degenerative changes in thevessel wall. Defects of the internal elas-tic lamina may be the cause of dissec-tion in the wall of intracranial arteries,particularly in patients with congenitalor acquired medical abnormalities.1,3,8,18

As a dissection progresses, other con-tributing factors also may be involved,including hypertension,1,8,28 delivery,33

various arterial diseases,2,12,14 and ana-tomical characteristics of the artery.28,36

Most dissecting aneurysms occur inthe intracranial vertebrobasilar trunk.Thus, the vertebral artery undergoesmorphological changes after it pene-trates the dura. For example, the re-duction in the thickness of the adventi-tial and medial layers and the significantreduction or loss of elastic fibers in themedial and external laminae36 may con-tribute to the development of dissect-ing aneurysms.28

Dissecting aneurysms involving theintracranial peripheral arteries predomi-nantly occur in relatively young patientswho usually lack underlying abnormal-ities to explain the appearance of a dis-section. These vascular abnormalitieshave also been related to migraine,17 fi-bromuscular dysplasia,12 mixed connec-tive-tissue disease, angiitis, and trauma.28

The origin of the dissecting aneurysm ofthe SCA might have been associated withour patient’s history of heavy smoking.

Dissecting aneurysms in the periph-eral arteries of the posterior circulationare primarily located in the proximalportion of arterial branches. Anatom-ically, the SCA can be divided into twosegments. The cisternal segment ex-tends from the anterior pontomesen-cephalic to the lateral pontomesen-cepahlic to the cerebellomesencephalicsegment. The cortical segment is equiv-alent to the hemispheric, vermian, andmarginal branches.9,27 Of the 14 SCAaneurysms reported before 1990, seveninvolved the anterior pontomesen-cephalic segment. Thirteen percent ofvertebrobasilar system aneurysms arefusiform.37 Of the cases reported before1990, however, all but one aneurysm,which involved the main trunk of theSCA, were saccular.10

24

Thiex et al: Dissecting Aneurysm of the Superior Cerebellar Artery: Case Report


Figure 1. (A) Three months before DSA, MR angiography showed a vascular anomaly fedby the SCA. (B) Anteroposterior, (C) lateral, and (D) oblique right vertebral angiograms showa vascular convolution in relation to the pontomesencephalic segment of the left SCA.

A B

C D

Surgical IndicationsThe indications for surgical treat-

ment of dissecting aneurysms remaincontroversial. Dissecting aneurysms ofthe anterior circulation tend to mani-fest with ischemic symptoms; those in-volving the posterior circulation aremore likely to lead to SAH.8,37 Unlikeall reported cases of dissecting aneu-rysms of the SCA,10,11,29 our patient didnot present with SAH. Instead she hada long-standing history of a bruit in herleft ear. Only after syncope, whichprompted MR imaging, was the aneu-rysm dicovered incidentally. Whetherirritation from the tortuosity of the dis-secting segment of the SCA caused thebruit remains questionable. Lacking ev-idence of previous hemorrhage renderssecondary pathological changes subse-quent to hemorrhage unlikely.

Urgent surgical treatment for dis-secting aneurysms of the posterior cir-culation is recommended because therate of recurrent hemorrhage is high andthese patients have extremely poor out-comes if their lesion ruptures.32,37 How-ever, there are some arguments againstsurgical treatment. Spontaneous reso-lution of a dissecting aneurysm of theposterior cerebral artery17 and similar re-sults in the vertebrobasilar system24 favorconservative management in the ab-sence of neurological deterioration.The decision to administer surgicaltreatment in neurologically stable pa-tients who present with ischemia or norecurrent hemorrhage is controversial.Such cases can heal spontaneously orafter anticoagulant therapy or both.17,23

The need for surgery is unequivocal inpatients with recurrent SAH or angio-graphic progression of their dissec-tion.8,24 We offered our patient surgicalexploration for several reasons. First,the patient’s angiographic findings wereindefinite about the presence of a lesion.Second, the patient’s tortuous vascularanatomy was a concern. Finally, the pa-tient was young.

Surgical TreatmentSurgical treatment of dissecting an-

eurysms includes excision, trapping,proximal ligation, and reinforcement

with or without an intracranial-to-ex-tracranial bypass.3,7 Still, the optimalsurgical procedure has not yet been es-tablished. Aneurysmal excision or trap-ping seems to be the only effectivemeans of managing a ruptured dissect-ing aneurysm. As in our case, revascu-larization procedures7,13 or preservationof the blood flow of the parent arterymay be necessary.

In our patient, the dissection en-compassed an extensive portion of theSCA beginning 1 cm distal to its origin.Consequently, revascularization afterexcision of the tortuous dissecting an-eurysm or trapping was infeasible. Wetherefore reinforced the wall of the dis-

secting aneurysm by wrapping it. Ac-cording to Rhoton,27 the SCA bifur-cates into a rostral and caudal trunk 0.6to 34 mm (mean, 19 mm) from the ori-gin of the vessel. The bifurcation is nearthe point of maximal caudal descent ofthe artery on the lateral side of the brain-stem.

The most appropriate surgical pro-cedures should be selected based on thelocation of the aneurysm, its anatomicalrelation to the parent arteries, the ex-tent of the dissection, and the collateralcirculation. Regardless of the approachused, the exposure is typically deep andconfined, and the vascular anatomy iscomplex and variable. The intimate re-

25



Figure 2. Intraoperative photographs taken through the operating microscope. Theposterior cerebral artery and oculomotor nerve ride over the SCA. (A) Following the pos-terior cerebral artery and oculomotor nerve leads to the origin of the SCA. (B) Approxi-mately 1 cm distal to the origin of the SCA, the vessel becomes tortuous and elongated.Its maximum diameter was 6 mm. (C) After the arachnoid adhesions were dissected, thevessel wall appeared thin, prompting the surgeon (D) to wrap the dissecting aneurysm.

A B

C D

CN 3

SCA

SCA

cottondissectinganeurysm

PCA

26



lationship between these aneurysms andcranial nerves and perforating vessels in-creases the potential for surgical com-plications.26

Aneurysms involving the cisternalsegment may be approached through afull OZ approach, as in our case, orthrough the pterional transsylvian,31 sub-temporal transtentorial,6,25 or occipitaltranstentorial approaches.21 Aneurysmsinvolving a marginal branch of the cor-tical segment are usually clipped throughthe suboccipital approach. Those in-volving the hemispheric branch aretreated through an infratentorial supra-cerebellar or occipital transtentorial ap-proach. Finally, lesions involving the ver-mian branch are treated through asuboccipital,16 infratentorial supracere-bellar,4 or occipital transtentorial ap-proach.34

References1. Adams HP, Jr., Aschenbrener CA, Kassell NF, et

al: Intracranial hemorrhage produced by spon-taneous dissecting intracranial aneurysm. ArchNeurol 39:773-776, 1982

2. Austin MG, Schaefer RF: Marfan’s syndrome,with unusual blood vessel manifestations. AMAArch Pathol 64:205-209, 1957

3. Berger MS, Wilson CB: Intracranial dissecting an-eurysm of the posterior circulation: Report of sixcases and review of the literature. J Neurosurg61:882-894, 1984

4. Bozboga M, Canbolat A, Savas A, et al: Aneu-rysm arising from the medial branch of the su-perior cerebellar artery. Acta Neurochir (Wien)138:1013-1014, 1996

5. Chaloupka JC, Putman CM, Awad IA: Endovas-cular therapeutic approach to peripheral aneu-rysms of the superior cerebellar artery. AJNRAm J Neuroradiol 17:1338-1342, 1996

6. Dimsdale H, Logue V: Ruptured posterior fossaaneurysms and their surgical treatment. J Neu-rol Neurosurg Psychiatry 22:202-217, 1959

7. Durward QJ: Treatment of vertebral artery dis-secting aneurysm by aneurysm trapping andposterior inferior cerebellar artery reimplantation.Case report. J Neurosurg 82:137-139, 1995

8. Friedman AH, Drake CG: Subarachnoid hemor-rhage from intracranial dissecting aneurysm. JNeurosurg 60:325-334, 1984

9. Hardy DG, Peace DA, Rhoton AL, Jr: Microsur-gical anatomy of the superior cerebellar artery.Neurosurgery 6:10-28, 1980

10. Hirose Y, Nakamura T, Takamiya Y, et al: Fusiformsuperior cerebellar artery aneurysm presentingwith contralateral abducens nerve paresis—casereport. Neurol Med Chir (Tokyo) 30:119-122,1990

11. Ikeda K, Shoin K, Taguchi H, et al: Postpartumdissecting aneurysm of the superior cerebellarartery—case report. Neurol Med Chir (Tokyo)39:852-857, 1999

12. Kalyan-Raman UP, Kowalski RV, Lee RH, et al:Dissecting aneurysm of superior cerebellar ar-tery. Its association with fibromuscular dysplasia.Arch Neurol 40:120-122, 1983

13. Kawaguchi S, Sakaki T, Kamada K, et al: Dis-secting aneurysm of the posterior inferior cere-bellar artery—case report. Neurol Med Chir(Tokyo) 33:634-637, 1993

14. Kulla L, Deymeer F, Smith TW, et al: Intracranialdissecting and saccular aneurysms in polycystickidney disease. Arch Neurol 39:776-778, 1982

15. Logue V: Posterior fossa aneurysm. Clin Neu-rosurg 6:183-219, 1964

16. Mabuchi S, Kamiyama H, Abe H: Distal aneu-rysms of the superior cerebellar artery and pos-terior inferior cerebellar artery feeding an asso-ciated arteriovenous malformation: Case report.Neurosurgery 30:284-287, 1992

17. Maillo A, Diaz P, Morales F: Dissecting aneurysmof the posterior cerebral artery: Spontaneous res-olution. Neurosurgery 29:291-294, 1991

18. Manz HJ, Luessenhop AJ: Dissecting aneurysmof intracranial vertebral artery: Case report andreview of literature. J Neurol 230:25-35, 1983

19. Matricali B, Seminara P: Aneurysm arising fromthe medial branch of the superior cerebellar ar-tery. Neurosurgery 18:350-352, 1986

20. Nagasawa S, Kobata H, Aoki J, et al: A largethrombosed superior cerebellar artery aneurysm:A case report. Surg Neurol 45:36-38, 1996

21. Papo I, Caruselli G, Salvolini U: Aneurysm of thesuperior cerebellar artery. Surg Neurol 7:15-17,1977

22. Poppen JL: Vascular surgery of the posteriorfossa. Clin Neurosurg 6:198-210, 1958

23. Pozzati E, Andreoli A, Padovani R, et al: Dissect-ing aneurysms of the basilar artery. Neurosur-gery 36:254-258, 1995

24. Pozzati E, Padovani R, Fabrizi A, et al: Benign ar-terial dissections of the posterior circulation. JNeurosurg 75:69-72, 1991

25. Pozzati E, Tognetti F, Padovani R, et al: Superiorcerebellar artery aneurysms. Case report. J Neu-rosurg Sci 24:85-88, 1980

26. Redekop GJ, Durity FA, Woodhurst WB: Man-agement-related morbidity in unselected aneu-rysms of the upper basilar artery. J Neurosurg87:836-842, 1997

27. Rhoton AL, Jr: The cerebellar arteries. Neuro-surgery 47:S29-S68, 2000

28. Sasaki O, Koizumi T, Ito Y, et al: Dissecting aneu-rysm of the posterior cerebral artery treated withproximal ligation. Surg Neurol 37:394-401, 1992

29. Sato M, Kodama N, Sasaki T, et al: Aneurysmsarising from the cortical segment of the superiorcerebellar artery—two case reports. Neurol MedChir (Tokyo) 39:858-862, 1999

30. Sherman DG, Salmon JH: Ocular bobbing withsuperior cerebellar artery aneurysm. Case report.J Neurosurg 47:596-598, 1977

31. Sugita K: Microneurosurgical Atlas. Berlin:Springer-Verlag, 1985

32. Tikkakoski T, Leinonen S, Siniluoto T, et al: Iso-lated dissecting aneurysm of the left posterior in-ferior cerebellar artery: Endovascular treatmentwith a Guglielmi detachable coil. AJNR Am JNeuroradiol 18:936-938, 1997

33. Van de Kelft E, Kunnen J, Truyen L, et al: Post-partum dissecting aneurysm of the basilar artery.Stroke 23:114-116, 1992

34. Vassilakis D, Sirmos C, Selviaridis P: Aneurysmof the vermian branch of the superior cerebellarartery. J Neurosurg Sci 37:243-245, 1993

35. Vincent FM, Zimmerman JE: Superior cerebellarartery aneurysm presenting as an oculomotornerve palsy in a child. Neurosurgery 6:661-664,1980

36. Wilkinson IM: The vertebral artery. Extracranialand intracranial structure. Arch Neurol 27:392-396, 1972

37. Yamaura A: Diagnosis and treatment of vertebralaneurysms. J Neurosurg 69:345-349, 1988

38. Yasargil MG, Smith RD, Young PH, et al: Mi-croneurosurgery. Stuttgart: Georg Thieme Ver-lag, 1984


Transorbital Penetrating TreeBranch During a Sledding Accident

The morbidity and mortality associ-ated with intracranial wooden for-

eign bodies have repeatedly been dem-onstrated in the literature. The orbit isa pathway of low resistance for pene-trating objects to enter the intracranialcompartment. However, intracranialwood can be challenging to detect withcurrent imaging modalities. We presentthe case of a 10-year-old girl who had atree branch penetrate her orbit during asledding accident. Despite the danger-ous resting place of two branch frag-ments, her intracranial structures wereuninjured. Surgical removal of the frag-ments was successful, and the patienthad an excellent clinical outcome.

Case ReportA 10-year-old girl was sledding when

a small tree branch penetrated the leftorbit near the medial canthus. She wastransferred to the trauma service at ourinstitution. Her initial examinationruled out coexisting injuries. On ad-mission her Glasgow Coma Scale scorewas 15. Her left pupil was reactive todirect light, but she complained of blur-ry vision. An ophthalmologic exami-nation of that eye documented her vi-sion as 20/200 and the presence ofmedial rectus palsy. The remainder ofher neurologic examination was nor-mal. The patient was placed on intra-venous antibiotics shortly after admis-sion (1 gm of vancomycin twice a day,500 mg of metronidazole four times aday, and 1.5 gm cefotaxime three timesa day).

CT angiography showed that thebranch had fragmented. One piece hadtraversed the sphenoid sinus and lodged

Complications related to intracranial wooden foreign bodies are well known. Ifwooden foreign bodies are suspected, efforts should be made to remove themsurgically to avoid infection-related complications or even death. The authors pre-sent the case of a 10-year-old girl whose left orbit near the medial canthus waspenetrated by a small tree branch during a sledding accident. Different imagingmodalities showed that the branch had fragmented. One fragment rested alongthe lateral border of the pons, and one penetrated the sphenoid sinus and carotidcanal. The patient underwent successful surgical removal of the wooden frag-ments and made an excellent clinical recovery.

Key Words: cerebral abscess, foreign body, intracranial, trauma, wood

Abbreviations Used: CT, computed tomography; MR, magnetic reso-nance

Division of Neurological Surgery, Barrow Neurological Institute, St. Joseph’s Hospital and MedicalCenter, Phoenix, Arizona†Current Address: Department of Neurological Surgery, Indiana University School of Medicine,

Indianapolis, Indiana

Mark Garrett, MD

Eric M. Horn, MD, PhD†



in the medial carotid canal (Fig. 1). MRangiography showed that the internal ca-rotid artery was indented, but there wasno obvious vascular injury (Fig. 2). An-other piece of the branch had traveledintracranially and rested along the mesialtemporal lobe. Its tip lay in the ambientcistern between the tentorial edge andpons (Fig. 3). Multiple reconstructionCT scans and MR images were obtainedto determine the exact path of thebranch fragments (Fig. 4).

The decision was made to removethe foreign bodies surgically through acombined approach. The entrancewound was widened, and the orbit wasexplored to identify the branch frag-ments. A modified orbitozygomaticcraniotomy was performed to visualizethe largest intracranial branch during re-moval and to gain vascular control in theevent of injury. However, the woodenfragment, which was entirely containedwithin the leaflets of the tentorium, wasnever visualized through the cranioto-my.

Immediately after the fragment wasremoved, a small amount of subduralbleeding occurred but stopped sponta-neously. Although we believe that thetip of the fragment caused this bleedingas it penetrated the tentorial dura, directvisual verification could not be ob-tained. The bleeding stopped com-pletely, and no further exploration wasperformed. After the branch that hadpenetrated the carotid canal was re-moved, intraoperative angiography con-firmed the absence of vascular injury.

Postoperative CT raised concernsthat pieces of wood were retained with-in the orbit. Because of the high riskassociated with retained foreign bodies,the patient was returned to the operat-ing room for re-exploration. No fur-ther fragments were found, and the CTfindings were attributed to posttrau-matic liposis.

After the first surgery, the patient hadan afferent pupillary defect but her per-ception of light and motion was intact.After the second procedure, she lost allvision in that eye. Decadron was start-ed and continued (4 mg every 8 hours)until she was discharged. After consul-

tation with the infectious disease service,the antibiotics were discontinued at dis-charge with plans for close follow-up.

At her 1-month follow-up exami-nation, the patient had regained nearnormal vision in the left eye. Only slightmedial rectus palsy persisted. The pa-tient had no signs or symptoms of in-

fection and otherwise made an excel-lent recovery.

DiscussionOur patient’s case is of interest be-

cause of the dangerous path taken by thebranch fragments without injuring in-

28

Garrett et al: Transorbital Penetrating Tree Branch During a Sledding Accident


Figure 1. Sagittal CT angiogram shows the two major fragments of the branch. One frag-ment entered the intracranial compartment via the superior orbital fissure. The other trav-eled through the sphenoid sinus and its tip lodged in the carotid canal.

Figure 2. (A) Three-dimensional reconstruction MR angiogram shows a hypointensity (arrow)of the left internal carotid artery where the tip of the branch penetrated the carotid canal. (B)Axial MR angiogram shows the tip of the fragment that penetrated the carotid canal (arrow).

A B

tracranial vascular or parenchymal struc-tures. The fragments required surgicalremoval—the danger of intracranialwood is well known.3 Nishio et al., forexample, reported a 13-year-old girl

whose orbit was penetrated by a chop-stick.4 Initial CT failed to show awooden fragment, but 7 years later shedeveloped a brain abscess. The authorsalso reviewed 23 case reports of wood-

en foreign objects that penetrated thedura published since 1984. Sixteen caseswere complicated by central nervoussystem infections of which 12 were as-sociated with brain abscesses. The frag-ment in the carotid canal placed our pa-tient at risk not only for infection butalso for blood flow abnormalities.

Our patient’s surgical reexplorationhighlights the dilemma of diagnosingforeign wooden objects with currentimaging modalities. A thin-cut post-operative CT scan of the orbits showeda small, well-defined, iso- to hypodensearea that appeared consistent with a re-tained fragment. Both neurosurgeonsand neuroradiologists evaluated herpostoperative CT scans, and the deci-sion was made to reoperate. On CTwood is often hypointense and can bemistaken for air or even orbital fat.1,7

Consequently, some authors have ar-gued that MR imaging is superior toCT in this setting.

However, the appearance of woodon MR imaging can be quite variable.On MR imaging of the extremities, theappearance of wooden foreign bodiescan vary, but the signal intensity of woodon T1-weighted and T2-weighted MRimages is usually equal to or less thanthat of skeletal muscle.5 Smely andOrszagh confirmed this variability, re-porting cases that showed both hyper-and hypointensity on T1-weighted im-ages.6 These findings prompted an invitro study using CT and MR imagingto study the appearance of multipletypes of wood in both fatty and soft-tis-sue backgrounds that resembled anorbit.2 In that investigation, CT was su-perior to MR imaging for detectingwooden foreign bodies in the eye. Iffindings on CT or MR imaging raisesuspicion of a penetrating wooden for-eign body, surgical intervention seemswarranted to avoid the attendant risks.

ConclusionThis interesting clinical vignette

serves as an important reminder of is-sues regarding intracranial wood.Specifically, significant morbidity andmortality rates are associated with re-

29



Figure 3. Axial T1-weighted MR image shows the largest branch fragment, which pen-etrated the superior orbital fissure.

Figure 4. Three-dimensional reconstruction CT shows the intracranial fragment.



30

tained intracranial wooden objects. Al-though the radiographic diagnosis ofwooden objects remains difficult, highsuspicion based on either CT or MRimaging should prompt surgical removalof any foreign body.

References1. Hansen JE, Gudeman SK, Holgate RC, et al: Pen-

etrating intracranial wood wounds: Clinical limi-tations of computerized tomography. J Neuro-surg 68:752-756, 1988

2. McGuckin JF, Jr., Akhtar N, Ho VT, et al: CT andMR evaluation of a wooden foreign body in an invitro model of the orbit. AJNR Am J Neuroradi-ol 17:129-133, 1996

3. Miller CF, Brodkey JS, Colombi BJ: The dangerof intracranial wood. Surg Neurol 7:95-103,1977

4. Nishio Y, Hayashi N, Hamada H, et al: A case ofdelayed brain abscess due to a retained intra-cranial wooden foreign body: A case report andreview of the last 20 years. Acta Neurochir(Wien) 146:847-850, 2004

5. Peterson JJ, Bancroft LW, Kransdorf MJ: Wood-en foreign bodies: Imaging appearance. AJRAm J Roentgenol 178:557-562, 2002

6. Smely C, Orszagh M: Intracranial transorbital in-jury by a wooden foreign body: Re-evaluation ofCT and MRI findings. Br J Neurosurg 13:206-211, 1999

7. Specht CS, Varga JH, Jalali MM, et al: Orbitocra-nial wooden foreign body diagnosed by mag-netic resonance imaging. Dry wood can be iso-dense with air and orbital fat by computedtomography. Surv Ophthalmol 36:341-344,1992


Motor Vehicle Accident CausingThoracic Spinal Cord InjuryWithout Radiographic Abnormality

Spinal injuries are primarily sustainedby adults. However, SCIWORA,

defined as the occurrence of a spinal cordinjury without radiographic abnormali-ty, is predominantly found in pediatricpatients.5 Children with SCIWORAshow no evidence of an abnormality onplain film radiography or CT but mayhave positive neurological findings andcorresponding findings on MRI imag-ing.2 SCIWORA represents fewer than1% of all spinal cord injuries,3 but its di-agnosis has increased with advances inMR imaging technology and with in-creased awareness about the phenome-non. Of the approximately 1300 newcases of spinal cord injury in the pediatricpopulation each year, 19 to 34% are re-ported to be SCIWORA.4,5 This articledescribes a relatively rare case of SCI-WORA involving the thoracic spinalcord.

Case ReportA 22-month-old boy, who was an

unrestrained passenger, was ejected in arollover motor vehicle accident. At ad-mission he was intubated and sedated.A cursory examination showed no ob-vious injuries of the extremities. He wasresting in a stable position in bed. Hiscranial nerve function was intact. Helocalized purposely with both upper ex-tremities but demonstrated flaccid para-paresis and lack of strength in both lowerextremities. Examination showed norectal tone. Reflexes in his upper ex-tremities were 2+. Reflexes were absentin both lower extremities, but his pulsewas palpable bilaterally.

Initial plain film radiographs of thecervical, thoracic, and lumbar spine were

SCIWORA is most commonly caused by traumatic injury sustained in a motor ve-hicle accident or fall. The immature pediatric spine confers some protectionagainst direct osseous spinal injury but makes children susceptible to ligamen-tous and spinal cord injury. Trauma-induced spinal cord injury usually involvesthe cervical spine; we report a rare case of SCIWORA involving the thoracic spi-nal cord. Despite negative radiographic or CT findings, pediatric patients withneurologic deficits should undergo MR imaging to rule out spinal cord injury.

Key Words: pediatric, SCIWORA, spinal cord injury, trauma

Jasmine Zargarpour, BA†

Poya Hedayati, MD

Shahram Partovi, MD

Abbreviations Used: CT, computed tomography, MR, magnetic reso-nance, SCIWORA, spinal cord injury without radiographic abnormality,STIR, short tau inversion recovery

Division of Neuroradiology, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Cen-ter, Phoenix, Arizona

†Current address: Department of Pediatrics, University of Southern California, Los Angeles, Cali-fornia


negative for acute findings (Fig. 1). Sub-sequent CT of the cervical, thoracic (Fig.2), and lumbar spine showed no fracturesor subluxation. CT of the head was alsonegative.

Based on the patient’s neurologic find-ings, MR imaging was ordered (Fig. 3).The study demonstrated abnormal cordsignal intensity from T10 to T12. STIRsequences demonstrated no evidence ofbone marrow edema.

DiscussionIn terms of tautness, the pediatric

spine reaches adult maturation at 8 to 9years of age. SCIWORA is caused byligamentous flexibility and elasticity ofthe immature vertebral column. Notsurprisingly, two-thirds of the cases occurin patients 8 years or younger before thespine has matured. The malleability ofthe spine places the spinal cord under se-vere distortion, causing it to contact itsbony counterpart, potentially leading to

neurological problems.6 MR imagingmay show compressive, treatable lesionsthat were not apparent on plain radio-graphs. Spinal cord injury in pediatric pa-tients can include ligamentous instabili-ty, ischemia, infarction, or an evolvinglesion.1,8

In the uncommon case of a thoracicspinal cord injury, the mechanism of in-jury is often related to motor vehicle ac-cidents, crush injuries, or falls from aheight.9 Damage to the thoracic spine is

32

Zargarpour et al: Motor Vehicle Accident Causing Thoracic Spinal Cord Injury Without Radiographic Abnormality


Figure 1. Lateral radiograph of the thora-columbar spine appears normal.

Figure 2. Sagittal CTs of the (A) cervical and (B) thoracic spine also appear normal.

Figure 3. (A) Axial T2-weighted, (B) sagittal STIR, and (C) sagittal T2-weighted MR images of the infant’s thoracolumbar spine. The ab-normal cord signal extending from T10 to T12 is consistent with edema and hemorrhage. On STIR sequences (B), there was no evi-dence of bone marrow edema.

A B

A B C

33

Zargarpour et al: Motor Vehicle Accident Causing Thoracic Spinal Cord Injury Without Radiographic Abnormality


less common than damage to the cervi-cal spine because the former is protectedby the rib cage, abdomen, and thorax,which resist excessive flexion and exten-sion.5,9 Lesions involving the lower tho-racic segments, as in our patient, are usu-ally caused by a crushing effect more thanby flexion or extension.6,7 In contrast,injury to the cervical spine is common-ly caused by extreme flexion or exten-sion of the spinal cord.

When Pang coined the term SCI-WORA in 1982,8 MR imaging had onlyrecently been introduced. Thus, the de-finition was based purely on CT and plainx-rays. As the use of MR imaging hasbecome more widespread, such lesionsare increasingly detected. Therefore, theterm SCIWORA is now a misnomer be-cause most patients have an abnormality

that is detectable on MR imaging.2 Thisfact underlines the importance of MRimaging when a child has clinical neuro-logical deficits but no findings on CT orplain films.

References1. Baker C, Kadish H, Schunk JE: Evaluation of pe-

diatric cervical spine injuries. Am J Emerg Med17:230-234, 1999

2. Cirak B, Ziegfeld S, Knight VM, et al: Spinal in-juries in children. J Pediatr Surg 39:607-612,2004

3. Hendey GW, Wolfson AB, Mower WR, et al: Spi-nal cord injury without radiographic abnormality:Results of the National Emergency X-Radiogra-phy Utilization Study in blunt cervical trauma. JTrauma 53:1-4, 2002

4. Kokoska ER, Keller MS, Rallo MC, et al: Charac-teristics of pediatric cervical spine injuries. J Pe-diatr Surg 36:100-105, 2001

5. Launay F, Leet AI, Sponseller PD: Pediatric spinalcord injury without radiographic abnormality: Ameta-analysis. Clin Orthop Relat Res 433: 166-170, 2005

6. Pang D: Spinal cord injury without radiographicabnormality in children, two decades later. Neu-rosurgery 55:1325-1342, 2004

7. Pang D, Pollack IF: Spinal cord injury without ra-diographic abnormality in children—the SCI-WORA syndrome. J Trauma 29:654-664, 1989

8. Pang D, Wilberger JE, Jr.: Spinal cord injury with-out radiographic abnormalities in children. JNeurosurg 57:114-129, 1982

9. Yamaguchi S, Hida K, Akino M, et al: A case of pe-diatric thoracic SCIWORA following minor trau-ma. Childs Nerv Syst 18:241-243, 2002

Barrow Neurological InstituteSt. Joseph’s Hospital and Medical CenterCHW

® 350 West Thomas RoadPhoenix, Arizona 85013-44961-800-BARROW-1 (1-800-227-7691)or 602-406-6281www.thebarrow.org

C O N T E N T S

Robert F. Spetzler

C O M M E N T S


12 Delayed Ischemic Neurological Deficit After ElectiveClipping of Incidentally Discovered Aneurysms:Three Case ReportsGregory P. Lekovic, David Jho, P. Roc Chen, Volker K. H. Sonntag,Joseph M. Zabramski, and Robert F. Spetzler

Dissecting Aneurysm of the Superior CerebellarArtery: Case ReportRuth Thiex, Iman Feiz-Erfan, Francisco Ponce, Mohammad AliAziz-Sultan, and Robert F. Spetzler

23

Analysis of Conservative Management and Comparisonof Neurological Outcomes Between Anterior and Poste-rior Approaches for Thoracolumbar Burst FracturesEric M. Horn, Iman Feiz-Erfan, Nicholas C. Bambakidis, StephenM. Papadopoulos, Volker K. H. Sonntag, and Nicholas Theodore

4

Transorbital Penetrating Tree Branch Duringa Sledding AccidentMark Garrett, Eric M. Horn, and Robert F. Spetzler

27

Motor Vehicle Accident Causing Thoracic SpinalCord Injury Without Radiographic AbnormalityJasmine Zargarpour, Poya Hedayati, and Shahram Partovi

31


cover and back 23-4 for website - barrowneuro.org

Documents