Journal of Clinical Neuroscience 20 (2013) 440–445

Contents lists available at SciVerse ScienceDirect

Journal of Clinical Neuroscience

journal homepage: www.elsevier .com/ locate/ jocn

Clinical Study

Long-term prognostic factors for microvascular decompressionfor trigeminal neuralgia

Kwang Wook Jo a, Doo-Sik Kong b, Ki-Sun Hong c, Jeong A. Lee b, Kwan Park b,⇑a Department of Neurosurgery, The Catholic University of Korea, Bucheon St. Mary’s Hospital, Bucheon, Republic of Koreab Department of Neurosurgery, Samsung Medical Center, Sungkyunkwan University, School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Republic of Koreac Department of Neurosurgery, Anam Hospital, Korea University of Medicine, Seoul, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history:Received 10 November 2011Accepted 10 March 2012

Keywords:Microvascular decompressionNeurovascular conflictPrognostic factorTrigeminal neuralgia

0967-5868/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.jocn.2012.03.037

⇑ Corresponding author. Tel.: +82 2 3410 3496; faxE-mail address: kwanpark@skku.edu (K. Park).

The purpose of this retrospective study was to identify preoperative imaging characteristics and surgicalfindings that predict pain relief after microvascular decompression (MVD) for trigeminal neuralgia (TN).This study included 141 patients with follow-up ranging from 6 months to 10 years (mean follow-up = 26.3 months). Preoperative images were assessed in 90 patients who were evaluated with construc-tive interference in steady-state (CISS) MRI in the last 6 years. These findings were compared with theseverity of neurovascular conflict (NVC) found at operation to identify imaging findings useful for prog-nosis. Using Kaplan–Meier analysis, we found that the success rate of MVD was 91.1 ± 2.5% at 1 year andwas 76.3 ± 7.5% after 5 years. A higher degree of NVC at operation (p = 0.000), no vein compression(p = 0.049) and single vessel compression (p = 0.000) were good prognostic factors for pain relief. Twomeaningful positive MRI findings, specifically, the ‘‘cerebrospinal fluid rim sign’’ and the ‘‘deviation sign’’were statistically significantly associated with the severity of NVC at operation and MVD success(p = 0.000). In this study, 34 patients (24.1%) complained of facial numbness postoperatively, and the oralherpes simplex virus was reactivated in 19 patients (13.4%). The involvement of a single arterial offenderin NVC is the most important prognostic factor for MVD in TN, and the positive MRI findings describedin this report may be helpful in selecting patients for MVD.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Trigeminal neuralgia (TN) is a facial pain syndrome character-ized by paroxysmal, shock-like pain attacks that are distributedalong the trigeminal nerve. Various surgical treatments have beenused to treat TN. Among these treatments, microvascular decom-pression (MVD) has been the most attractive treatment strategybecause it provides the highest rates of long-term patient satisfac-tion and the lowest rates of pain recurrence.1 However, not all pa-tients with TN are cured by MVD. Although reports of initial painrelief after MVD range from 86–98.2%, the recurrence rates rangefrom 15.1–30%.2–4 Furthermore, open surgery has a higher mortal-ity rate compared with other techniques used to treat TN includingpercutaneous techniques or stereotactic radiosurgery.2,3

Most cases of TN are caused by a neurovascular conflict (NVC) atthe root entry zone of the trigeminal nerve, and in several analyses,patients with more severe vascular compression of the trigeminalroot have reported greater relief of their symptoms after microvas-cular decompression.5–8 Therefore, prediction of NVC severity

ll rights reserved.

: +82 2 3410 0048.

using preoperative imaging may be useful for predicting long-termsuccess and could help guide patient selection for MVD treatment.

This retrospective study was performed to determine the imag-ing predictors of pain relief after MVD. In addition, the preopera-tive patient factors and surgical findings were analyzed. Thepreoperative imaging studies were compared with the surgicalfindings to identify imaging findings predictive of success.

2. Methods

In this study, 153 consecutive patients were treated with MVDfor typical TN by a single surgeon (K. Park) at the Samsung MedicalCenter in Seoul, Korea between January 1998 and March 2010 andwere evaluated. Within this group, 12 patients were excluded dueto a lack of recorded postoperative results. Therefore, only 141 pa-tients were included in this study.

The prognostic factors were assessed by a review of the medicalrecords and were classified into two categories, specifically, thepreoperative patient factors and the surgical findings. The preoper-ative patient factors included demographic characteristics andindividual symptoms of trigeminal pain. The surgical findingsincluded the vessel type causing the compression and the NVC

Table 1Overall outcomes of microvascular decompression for patients with trigeminalneuralgia

Years after operation 0 1 2 3 4 5

Success (%) 93.6 81.0 74.4 74.4 64.8 58.3Acceptable success (%) 97.9 91.1 89.5 89.5 76.3 76.3

K.W. Jo et al. / Journal of Clinical Neuroscience 20 (2013) 440–445 441

severity, which were obtained from operative reports and videos.The preoperative images were reviewed for 90 patients who wereevaluated using constructive interference in steady-state (CISS)MRI in the last 6 years. These findings were then compared withthe degree of NVC severity found at operation to identify the pre-dictive imaging findings.

To avoid any confounders due to inter-observer variability, asingle neurosurgeon (K. Park) and a single clinical nurse specialist(J. A. Lee) evaluated all patients. The outcomes were assessed by aclinical follow-up evaluation and were divided into three catego-ries as follows: (1) excellent: complete pain relief without medica-tion; (2) good: substantial pain relief but with minimal persistentpain without medication or complete pain relief with well-toler-ated medication; and (3) poor: little or no pain relief or severe paindespite medication. Moreover, two analytical models were con-structed. In one model, an excellent outcome after MVD was con-sidered to be a success. In the second model, a good or anexcellent outcome was classified as an ‘‘acceptable success’’.

Fig. 1. Kaplan–Meier curves of patients’ pain-free survival rate after microvasculardecompression for trigeminal neuralgia for: (a) success (excellent outcome), and (b)acceptable success (excellent or good outcome).

All surgical procedures were performed via a retrosigmoid sub-occipital approach, which has been previously described.9–11

Briefly, during surgery, the brainstem-evoked potentials and thesomatosensory-evoked potentials were closely monitored. Thedura was closed with several pieces of muscle interposed betweenthe interrupted sutures to prevent cerebrospinal fluid (CSF)leakage.10

The data were analyzed using the Statistical Package for the So-cial Sciences software (version 18.0, SPSS, Chicago, IL, USA). The pa-tients without known failure were censored based on the date oftheir latest follow-up evaluation. The Kaplan–Meier survival anal-ysis was performed, and log-rank tests were used to compare thesurvival curves. The confidence level for statistical significancewas a probability value of < 0.05, and all statistical tests performedwere two-tailed.

3. Results

3.1. Overall outcome after microvascular decompression

The follow-up evaluation period ranged from 6 months to10 years with a mean of 26.3 months. The Kaplan–Meier curvesfor success (excellent outcome) and acceptable success (excellentor good outcome) are shown in Fig. 1a, b. An immediate, completepostoperative relief from pain was achieved in 120 patients(85.1%), with immediate good relief being attained in 16 patients(11.3%) and poor relief in five patients (3.6%). Using the Kaplan–Meier analysis, the estimated probability of being pain-free at1 year was 91.1 ± 2.5% and was estimated to be 76.3 ± 7.5% after5 years of follow-up evaluation (Table 1).

In addition, 34 patients (24.1%) complained of facial numbnesspostoperatively, and the oral herpes simplex virus was reactivatedin 19 patients (13.4%). The herpes simplex viral infections weretreated with an anti-viral ointment, and all affected patients werecured without further complication. Although one patient (0.7%)developed delayed facial nerve palsy (House-Brackmann GradeIII) 2 weeks after surgery, this patient recovered fully after 2 weeksof prednisolone treatment. Furthermore, one patient (0.7%) devel-oped ipsilateral mild sensory hearing impairment (pure-tone audi-ometry 40 dB), and one patient (0.7%) exhibited cerebellarhemispheric hemorrhage and obstructive hydrocephalus. Afterexternal ventricular drainage, the patient recovered without anyneurological sequelae. In addition, one patient (0.7%) had a woundinfection, and one patient (0.7%) had mild mastication impairment.

3.2. Characteristics of the series and prognostic factors

3.2.1. Preoperative patient factorsThe patient demographics and individual symptoms of trigem-

inal pain are shown in Table 2. The age at surgery ranged from17 years to 79 years (average: 53.1 years), and the duration ofthe symptoms before surgery ranged from 1 month to 360 months(average: 61.4 months). Forty patients (28.4%) had undergone aprior trigeminal procedure for neuralgia. The sex, symptom dura-tion before MVD, history of previous trigeminal procedure, topog-raphy of trigeminal pain, preoperative hypesthesia and the numberof involved trigeminal divisions were not significant prognostic

Table 2Preoperative prognostic factors of patients who had microvascular decompression fortrigeminal neuralgia

Variable Number (%)(N = 141)

P

Success Acceptablesuccess

Sex 0.231 0.159Male 47 (33.3)Female 94 (66.7)

Age (years) 0.248 *0.013< 50 49 (34.8)50–70 79 (56.0)> 70 13 (9.2)

Symptom duration beforeMVD

0.805 0.834

< 2 years 36 (25.5)2–6 years 66 (46.8)> 6 years 39 (27.7)

Previous trigeminalprocedures

0.457 0.397

No 101 (71.6)Yes 40 (28.4)

Rhizotomya 13 (9.2)Peripheral nerveprocedureb

21 (14.9)

Radiosurgeryc 6 (4.3)

Side 0.306 0.275Right 96 (68.1)Left 45 (31.9)

Preoperative hypesthesia 18 (12.8) 0.094 0.156Topography 0.132 0.076

V1 7 (5.0)V2 50 (35.5)V3 36 (25.5)V1 + V2 20 (14.2)V2 + V3 21 (14.9)V1 + V2 + V3 7 (5.0)

Extent (No. of divisions) 0.496 0.7691 93 (65.9)2 41 (29.1)3 7 (5.0)

* Indicates a statistical significance.a Rhizotomy involved radiofrequency or glycerol procedures.b Peripheral nerve procedure involved alcohol injection or radiofrequency

lesioning.c Radiosurgery involved gamma knife or cyberknife.

MVD = microvascular decompression, V1 = ophthalmic branch, V2 = maxillarybranch, V3 = mandibular branch.

Table 3Operative prognostic factors of patients who had microvascular decompression fortrigeminal neuralgia

Patient characteristic Number (%)(N = 141)

p

Success Acceptablesuccess

Compressing vesselSCA 106 (75.2)Vein 35 (24.8)AICA 16 (11.3)Others 6 (4.3)

Vein compression 0.083 *0.018Yes 35 (24.8)No 106 (75.2)

Neurovascular conflictseverity

*0.000 *0.000

Grade 0 3 (2.1)Grade I 43 (30.5)Grade II 67 (47.5)Grade III 28 (19.9)

No. of compressing vessels *0.000 *0.0000 3 (2.1)1 113 (80.1)2 22 (15.6)3 3 (2.1)

* Indicates statistical significance. AICA = anterior inferior cerebellar artery,SCA = superior cerebellar artery

442 K.W. Jo et al. / Journal of Clinical Neuroscience 20 (2013) 440–445

factors. However, age < 70 years (log rank, p = 0.013) was associ-ated with the prognosis for acceptable success (Table 2).

3.2.2. Operative findingsThe operative findings are summarized in Table 3. The superior

cerebellar artery (SCA) was the most commonly compressing ves-sel, and was found in 106 patients. The SCA alone was involved in84 patients (59.6%). Vein compression was found in 19 patients(13.5%), and compression by a vein alone or a combined venouscompression was a poor prognostic factor for success (log rank,success: p = 0.083 and acceptable success: p = 0.018).

The severity of the NVC was assessed in all patients. When sev-eral compressing vessels were found, the major conflict alone wasassessed. The severity was rated on a scale of 0 to III. Three patients(2.1%) had a grade 0 severity: no vessel was compressing. Forty-three patients (30.5%) had a grade I severity: vessel compressionof the nerve without indentation. Sixty-seven patients (47.5%)had a grade II severity: mild or moderate indentation of the nerveby the vessel. Twenty-eight patients (19.9%) had a grade III sever-ity: severe indentation or displacement in the nerve by the vessel.

The severity of the NVC was significant with respect to the out-come. The more severe the NVC found at surgery, the greater thelikelihood of success or acceptable success (log rank, p = 0.000and p = 0.000, respectively; Fig. 2a, b). Using the Kaplan–Meieranalysis, the likelihood of success at 5 years was 80.3% for patientswith a grade II NVC and 96.4% for patients with a grade III NVC.However, for patients with a grade 0 or I NVC severity, the likeli-hood of success was 33.3% and 0%, respectively. The estimatedmean pain-free times were 2.1, 13.4, 23.3 and 23.3 months forNVC grade 0, I, II and III patients, respectively. The percentage ofpatients with acceptable success at 5 years was 83.4% for grade IIpatients and 96.4% for grade III patients.

In addition, 113 (80.1%) patients had only one compressing ves-sel found at surgery, and the surgical results were best in these pa-tients. Three patients (2.1%), who did not have a compressingvessel exhibited the lowest success and acceptable success rate.Moreover, 25 patients (17.7%) who had double or triple offendingvessels showed intermediate outcomes (Fig. 3a, b). Double or triplecompressing vessels were a prognostic factor (log rank, success:p = 0.000 and acceptable success: p = 0.000).

3.2.2.1. Positive imaging findings. The preoperative MRI were re-viewed for 90 patients who were evaluated with a CISS sequencein the final 6 years of the study and were compared with theNVC severity found at operation (Table 4). Two meaningful positiveMRI findings, the ‘‘CSF rim’’ and the ‘‘deviation sign,’’ were ob-served in 43 patients (47.8%). The most common positive imagingfinding was a CSF rim sign; defined as a narrow CSF signal betweenthe nerve and the vessel on axial imaging. The vessel is surroundedby a CSF rim, and that rim is again by the nerve (Fig 4). This signrepresents vertical or oblique compression that is sufficient tocause indentation. The CSF rim sign was observed in 28 (31.1%) pa-tients, and all was correlated with grade II or III NVC severity atoperation (Figs 4, Supplementary Fig. 1). The second most commonand easily identified sign was the deviation sign, which was ob-served in 15 patients (16.7%). Trigeminal nerve deviation causedby horizontal compression was easily identified on axial CISSimages (Fig. 5). A simple nerve bending or penetration by theoffending vessel was also observed (Supplementary Fig. 2). Of the

Fig. 2. Kaplan–Meier curves of pain-free survival rate after microvascular decom-pression for trigeminal neuralgia according to the degree of neurovascularcompression severity: (a) success, and (b) acceptable success.

Fig. 3. Kaplan–Meier curves of pain-free survival rate after microvascular decom-pression for trigeminal neuralgia according to the number of offending vessels: (a)success; and (b) acceptable success.

Table 4Preoperative image findings of patients who had microvascular decompression fortrigeminal neuralgia

Image findings NVC grade Total (%)

0 I II III (N = 90)

Positive 0 2 18 23 43 (47.8)

K.W. Jo et al. / Journal of Clinical Neuroscience 20 (2013) 440–445 443

43 patients with positive MRI findings, 41 patients (95.3%) had agreater than grade II NVC severity. The relationship between theMRI findings and NVC severity was statistically significant(p = 0.000). The sensitivity and specificity of the MRI findings forgrade II or III NVC were 62.1% and 91.7%, respectively.

CSF rim sign 0 0 14 14 28 (31.1)Deviation sign 0 2 4 9 15 (16.7)

Negativea 3 19 23 2 47 (52.2)

CSF = cerebrospinal fluid, NVC = neurovascular conflict.a A negative finding means that there was no CSF rim sign or deviation sign.

4. Discussion

Since Gardner proposed the theory of NVC, a growing body ofevidence indicates that compression of the trigeminal nerve rootat or near the dorsal root entry zone by a vessel is the major causeof TN.2,7,12–15 Although autopsy studies have revealed some degreeof contact between the trigeminal nerve and a blood vessel in14–60% of asymptomatic individuals, it is uncommon for arterialcontact to produce a distortion or grooving of the nerve.16–19 Con-

versely, studies of the vascular relationships of trigeminal nervesthat were surgically exposed for treatment of trigeminal neuralgiahave revealed that 78–91% of nerves showed compression or

Fig. 4. Axial preoperative constructive interference in steady-state MRI showingcerebrospinal fluid (CSF) rim signs. (a) Right trigeminal nerve is compressed bysuperior cerebellar artery (SCA) in the rostral direction and a grade II neurovascularconflict (NVC) was observed during surgery. A rail-like CSF rim around vessel isobserved. (b) Left trigeminal nerve is compressed by the superior cerebellar arterywith an acute angle, and a grade III NVC was found at surgery. An arrowhead-likeCSF signal surrounding the vessel is observed. (c) Right trigeminal nerve in contactwith a branch of the SCA and no CSF rim is observed. The NVC was grade I atoperation with mild compression without indentation.

Fig. 5. Axial preoperative constructive interference in steady-state MRI showingdeviation signs. (a) Right trigeminal nerve is deviated to the side (laterally) by theoffending artery. Grade III neurovascular conflict (NVC) was observed at surgery. (b)Right trigeminal rootlets are divided and penetrated by the offending artery. GradeIII NVC was identified at surgery.

444 K.W. Jo et al. / Journal of Clinical Neuroscience 20 (2013) 440–445

grooving by adjacent arteries.16–19 In a recent anatomical study oftrigeminal nerves in individuals not suffering from TN with 3.0-Te-sla MRI, there was a contact point in 45% of trigeminal nerves, but amild deviation was observed in only 7%. However, there was nomoderate or severe deviation observed.20 Other MRI studies haverevealed trigeminal neurovascular contact in asymptomatic pa-tients but a more severe contact in patients with TN.21,22 Thesepreviously described results indicate that an innocent proximityof the trigeminal nerve and vessels is frequently found with MRI,but our experience shows that a trend toward greater NVC severity

is more frequent at the affected trigeminal nerve than at the con-tralateral and normal trigeminal nerves.

In addition to NVC as a cause of TN, the degree of NVC is a prog-nostic factor for the long-term outcome after MVD treatment. Pre-vious studies have reported that patients with more severevascular compression of the trigeminal root have more successfulrelief of symptoms after MVD,5,6,8 and in a recent study conductedby Sindou et al., the presence of a clear-cut and marked vascularcompression at surgery was the most powerful prognostic factor.7

In our study, the degree of NVC at operation was the most impor-tant prognostic factor. Among 141 patients, 95 (67.4%) had a NVCworse than grade II, and the percentages of patients with successand acceptable success at 5 years were 80.3% and 96.4%, respec-tively. The patients with grade 0 or I NVC showed 2.1% and 30.5%success and acceptable success ratios at 5 years, respectively.

The prediction of the severity of NVC with a preoperative imag-ing study may be useful in differentiating symptomatic NVC frominnocent neurovascular contact, which may be valuable for patientselection for MVD treatment. Recently, MRI techniques have beendeveloped that allow for preoperative delineation of neurovascularanatomy in patients with TN. The technique of three-dimensional(3D) CISS MRI offers high contrast such that tiny structures, includ-ing the cranial nerves and vasculature at the ventral surface of thebrainstem, can be visually differentiated from the bright appear-ance of the CSF compartment.23,24 The severity of the NVC relatesto the degree of nerve compression, and may result in morpholog-ical changes, including distortion and indentation.22 Axial distor-tion and indentation that results from medial to lateralcompression may be found on axial images, and in our study, 15patients (16.7%) showed such a deviation. In cases of a vertical oroblique compression, the NVC severity could be precisely delin-eated on the axial images. On the coronal images, the exact siteof the NVC was perfectly perpendicular, due to the difficulty offinding an imaging plane that was perfectly perpendicular to thecourse of the nerve.22

In this study, the CSF rim between the nerve and the vesselproved pathognomonic for distortion and indentation. All of theCSF rim signs were correlated with a grade II or III NVC severity,although they displayed a low level of sensitivity. Where vesselsshowed a vertical or oblique compression and with an indentationor deviation on the axial images, the offending vessel was fre-quently surrounded by a CSF rim because there was a gap betweenthe offending vessel and the groove in the nerve (SupplementaryFig. 1). In a previous study, NVC was defined as having no CSF be-tween the trigeminal nerve and the vessel, and 82% of asymptom-atic individuals showed vessel-nerve contact.25 In a recent study ofvascular compression of the trigeminal nerve in asymptomaticindividuals with 3-Tesla MRI 3D CISS sequences, ‘‘compression’’was defined as having no CSF visible between the nerve and thevessel on MRI, and the authors concluded that vascular compres-sion of the trigeminal nerve was a frequent finding in asymptom-atic individuals.26 We also found that MRI evidence of a vesselcompressing the trigeminal nerve is not necessarily pathological.This prior report indirectly demonstrated that neurovascular con-tacts without indentation or distortion have no CSF rim and sup-port our opposing view (see also Supplementary Fig. 2).26

Although the positive preoperative MRI findings were highlyspecific for NVC grade II or III, the level of sensitivity was low.Among the 47 patients who did not exhibit any positive MRI find-ings, 25 (53.2%) had an NVC of grade of II or III at operation. Theabsence of positive MRI findings did not predict long-term failureof MVD for TN; however, long-term success may be anticipatedwith positive MRI findings. Furthermore, ongoing novel trials usingnew imaging techniques, including the inner view of the 3D mag-netic resonance cisternogram and angiogram fusion imaging, arebeing performed to assess NVC severity.2 Thus, patient selection

K.W. Jo et al. / Journal of Clinical Neuroscience 20 (2013) 440–445 445

combining clinical judgment with preoperative NVC analysis onimaging will improve the long-term surgical outcomes of MVDfor TN.

5. Conclusion

The presence of a definitive NVC with a single arterial offenderat surgery is the most important prognostic factor in MVD for TN,and the preoperative MRI findings of deviation and a CSF rim signare correlated with a more severe NVC. Due to the limitations ofthis retrospective study, a thorough prospective study is requiredto clarify the significance of these positive MRI findings.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jocn.2012.03.037.

References

1. Tatli M, Satici O, Kanpolat Y, et al. Various surgical modalities for trigeminalneuralgia: literature study of respective long-term outcomes. Acta Neurochir(Wien) 2008;150:243–55.

2. Barker 2nd FG, Jannetta PJ, Bissonette DJ, et al. The long-term outcome ofmicrovascular decompression for trigeminal neuralgia. N Engl J Med1996;334:1077–83.

3. Sindou M, Leston J, Howeidy T, et al. Micro-vascular decompression for primaryTrigeminal Neuralgia (typical or atypical). Long-term effectiveness on pain;prospective study with survival analysis in a consecutive series of 362 patients.Acta Neurochir (Wien) 2006;148:1235–45 [discussion 45].

4. Tyler-Kabara EC, Kassam AB, Horowitz MH, et al. Predictors of outcome insurgically managed patients with typical and atypical trigeminal neuralgia:comparison of results following microvascular decompression. J Neurosurg2002;96:527–31.

5. Mendoza N, Illingworth RD. Trigeminal neuralgia treated by microvasculardecompression: a long-term follow-up study. Br J Neurosurg 1995;9:13–9.

6. Piatt Jr JH, Wilkins RH. Treatment of tic douloureux and hemifacial spasm byposterior fossa exploration: therapeutic implications of various neurovascularrelationships. Neurosurgery 1984;14:462–71.

7. Sindou M, Leston J, Decullier E, et al. Microvascular decompression for primarytrigeminal neuralgia: long-term effectiveness and prognostic factors in a seriesof 362 consecutive patients with clear-cut neurovascular conflicts whounderwent pure decompression. J Neurosurg 2007;107:1144–53.

8. Szapiro Jr J, Sindou M, Szapiro J. Prognostic factors in microvasculardecompression for trigeminal neuralgia. Neurosurgery 1985;17:920–9.

9. Hitotsumatsu T, Matsushima T, Inoue T. Microvascular decompression fortreatment of trigeminal neuralgia, hemifacial spasm, and glossopharyngealneuralgia: three surgical approach variations: technical note. Neurosurgery2003;53:1436–41 [discussion 42–3].

10. Park JS, Kong DS, Lee JA, et al. Intraoperative management to preventcerebrospinal fluid leakage after microvascular decompression: dural closurewith a ‘‘plugging muscle’’ method. Neurosurg Rev 2007;30:139–42 [discussion42].

11. Sindou M, Leston JM, Decullier E, et al. Microvascular decompression fortrigeminal neuralgia: the importance of a noncompressive technique–Kaplan-Meier analysis in a consecutive series of 330 patients. Neurosurgery2008;63:341–50 [discussion 50–1].

12. Gardner WJ. Concerning the mechanism of trigeminal neuralgia and hemifacialspasm. J Neurosurg 1962;19:947–58.

13. Jannetta PJ. Arterial compression of the trigeminal nerve at the pons in patientswith trigeminal neuralgia. J Neurosurg 1967;26(Suppl.):159–62.

14. Miller JP, Magill ST, Acar F, et al. Predictors of long-term success aftermicrovascular decompression for trigeminal neuralgia. J Neurosurg2009;110:620–6.

15. Sindou M, Howeidy T, Acevedo G. Anatomical observations duringmicrovascular decompression for idiopathic trigeminal neuralgia (withcorrelations between topography of pain and site of the neurovascularconflict). Prospective study in a series of 579 patients. Acta Neurochir (Wien)2002;144:1–12 [discussion -3].

16. Haines SJ, Jannetta PJ, Zorub DS. Microvascular relations of the trigeminalnerve. An anatomical study with clinical correlation. J Neurosurg1980;52:381–6.

17. Hamlyn PJ. Neurovascular relationships in the posterior cranial fossa, withspecial reference to trigeminal neuralgia. 2. Neurovascular compression of thetrigeminal nerve in cadaveric controls and patients with trigeminal neuralgia:quantification and influence of method. Clin Anat 1997;10:380–8.

18. Hamlyn PJ, King TT. Neurovascular compression in trigeminal neuralgia: aclinical and anatomical study. J Neurosurg 1992;76:948–54.

19. Hardy DG, Rhoton Jr AL. Microsurgical relationships of the superior cerebellarartery and the trigeminal nerve. J Neurosurg 1978;49:669–78.

20. Kakizawa Y, Seguchi T, Kodama K, et al. Anatomical study of the trigeminal andfacial cranial nerves with the aid of 3.0-tesla magnetic resonance imaging. JNeurosurg 2008;108:483–90.

21. Miller JP, Acar F, Hamilton BE, et al. Radiographic evaluation of trigeminalneurovascular compression in patients with and without trigeminal neuralgia. JNeurosurg 2009;110:627–32.

22. Satoh T, Omi M, Nabeshima M, et al. Severity analysis of neurovascular contactin patients with trigeminal neuralgia: assessment with the inner view of the 3DMR cisternogram and angiogram fusion imaging. AJNR Am J Neuroradiol2009;30:603–7.

23. Tanrikulu L, Hastreiter P, Richter G, et al. Virtual neuroendoscopy: MRI-basedthree-dimensional visualization of the cranial nerves in the posterior cranialfossa. Br J Neurosurg 2008;22:207–12.

24. Yoshino N, Akimoto H, Yamada I, et al. Trigeminal neuralgia: evaluation ofneuralgic manifestation and site of neurovascular compression with 3D CISSMR imaging and MR angiography. Radiology 2003;228:539–45.

25. Kress B, Schindler M, Rasche D, et al. MRI volumetry for the preoperativediagnosis of trigeminal neuralgia. Eur Radiol 2005;15:1344–8.

26. Peker S, Dincer A, Necmettin Pamir M. Vascular compression of the trigeminalnerve is a frequent finding in asymptomatic individuals: 3-T MR imaging of 200trigeminal nerves using 3D CISS sequences. Acta Neurochir (Wien) 2009;151:1081–8.