Three-dimensional imaging for presentation of the causative vessels in patients with hemifacial spasm and trigeminal neuralgia

ELSEVIER

Imaging

THREE-DIMENSIONAL IMAGING FOR PRESENTATION OF THE CAUSATIVE VESSELS IN PATIENTS WITH HEMIFACIAL SPASM AND TRIGEMINAL NEURALGIA Yoshiaki Kumon, M.D. Ph.D., Saburo Sakaki, M.D. Ph.D., Kanehisa Kohno, M.D. Ph.D., Shinsuke Ohta, M.D. Ph.D., Shiroh Ohue, M.D. Ph.D., and Hitoshi Miki, M.D. Ph.D. Department of Neurological Surgery and Department of Radiology, Ehime University School of Medicine, Ehime, Japan

Kumon Y, Sakaki S, Kohno K, Ohta S, Ohue S, Miki H. Threedimensional imaging for presentation of the causative vessels in patients with hemifacial spasm and trigeminal neuralgia. Surg Neural 1997;47:178-84. BACKGROUND

In patients with hemifacial spasm and trigeminal neuralgia, preoperative detection of the relationship between the blood vessels and the cranial nerves involved is essential. METHODS

We studied the causative vessels in 20 patients with hemifacial spasm and six patients with trigeminal neuralgia by means of magnetic resonance (MR) imaging with spoiled gradient recalled acquisition in the steady state (SPGR), MR angiography, and three-dimensional (3-D) imaging reconstructed from the data of SPGR MR imaging by the surface rendering method at a workstation. RESULTS

In all patients, the preoperative SPGRMR images demonstrated that the causative vessels compressed or were in contact with the root exit or root entry zone (REZ) of the facial or trigeminal nerve. These causative vessels were identified by inspection of the MR angiographic and 3-D images. The 3-D images provided clear information as to the anatomic relationship between the causative vessels and the REZ of these nerves. These findings were corroborated by the intraoperative findings. The symptoms were completely relieved after surgery in 18 of the patients with hemifacial spasm and in all six patients with trigeminal neuralgia. In all patients, sufficient decompression was depicted on the postoperative SPGRMR images at the causative vessels and the REZ of the nerve. CONCLUSION

SPGR MR images, MR angiography, and 3-D images are useful for the identification of the causative vessels in patients with hemifacial spasm or trigeminal neuralgia. The 3-D images are particularly useful for the simulation

Address reprint requests to: Yoshiaki Neurological Surgery, Ehime University cho, Onsen-gun, 791-02 Ehime, Japan. Received February 14, 1996; accepted 0090-3019/97/$17.00 PI1 SOO90-3019(96)00364-3

Kumon M.D. Ph.D., Department of School of Medicine, ShigenobuJuly 30, 1996.

planning of the operative procedure.

0 1997 by Elsevier

Science Inc. KEY

WORDS

Microvascular decompression, hemifacial spasm, trigeminal neuralgia, threedimensional imaging, SPGRMRI, MRA. t is accepted that hemifacial spasm and trigemiI nal neuralgia are caused by vascular compression of the root exit or root entry zone (REZ) of the facial or trigeminal nerve as evidenced by direct visualization at surgery and the symptom response to vascular decompression [2,3,9]. Other cranial nerve hyperactivity syndromes have also been proposed to arise by a similar mechanism [lo]. In patients with these diseases, preoperative detection of the relationship between the blood vessels and the REZ of these nerves is essential for the clarification of the pathogenesis, performance of appropriate operative manipulation, and increasing the likelihood of good operative results. Magnetic resonance (MR) imaging has recently been used to evaluate hemifacial spasm or trigeminal neuralgia for these purposes. T,-weighted and T,-weighted MR images were previously reported to be capable of depicting the causative vessels as flow void areas in patients with facial spasm or trigeminal neuralgia [8,11,14-161. These images were unclear, however, and the relationship between the vessels and the REZ of the facial nerve was not well demonstrated in the case of small causative vessels such as the anterior inferior cerebellar artery (AICA) or posterior inferior cerebellar artery (PICA). For the purpose of demonstrating the vascular structure clearly, some attempts have been made to change the image acquisition to gradient-echo imaging and the image direction to 655 Avenue

0 1997 by Elsevier Science Inc. of the Americas, New York, NY 10010

3-D Image in Microvascular Decompression

II

Surg Neurol 1997;47:178-84

Summary of the Clinical Features of the Patients HFS

Patients (n) Age Q Mean (SD) Range Sex Male Female Side Left Right Duration (yr) Mean (SD) Range Causative vessels Relief of symptoms Abbreviations: SCA = superior AICA = anterior

20 47 (10) 21-69

TN

6 53 G9 42-65

6 14

3 3

11 9

3 3

4 (3) 0.6-15 AICA: 10 PICA: 10 complete: 18 partial: 2

4 (3) 0.1-10 SCA: 5 PICA: 1 complete: 6

HFS = hemifacial spasm; TN = trigeminal neuralgia; cerebellar artery; PICA = posterior cerebellar artery; inferior cerebellar artery.

the oblique sagittal view [13], and the methods such as MR tomographic angiography have been developed [ 11. MR imaging with spoiled gradient recalled acquisition in the steady state (SPGR) is a high speed scan technique that can demonstrate vessels as high-intensity areas clearly, as does gradient-echo imaging. Although SPGR MR images can demonstrate the causative vessels in patients with hemifacial spasm and trigeminal neuralgia, it does not consistently visualize the courses of both the vessels and the nerves on the same plane. That anatomic relationship between these structures must be deduced. In order to demonstrate the threedimensional (3-D) relationship between the causative vessels and the REZ of the facial or trigeminal nerve as clearly as possible, we attempted to present the causative vessels in patients with hemifacial spasm and trigeminal neuralgia by means of 3-D imaging, with SPGR MR imaging, and MR angiography.

PATIENTSANDMETHODS Twenty patients with hemifacial spasm and six patients with trigeminal neuralgia were studied. The patient characteristics including sex, age, affected side, and duration of disease are summarized in Table 1. MR imaging was performed using a 1.5-Tesla MR imaging system (Signa, GE Medical Systems, Milwaukee, WI). SPGR MR imaging was performed with

179

a relaxation time of 40 msec, echo time of 6.9 msec, flip angle of 25 degrees, slice thickness of 0.8 mm, matrix of 256 X 192, field of view of 20 cm, and slab thickness of 48 mm. In the MR angiographic studies, we employed 3-D SPGR MR imaging with maximum intensity projection. The acquired data from SPGR MR imaging was transferred to a workstation (Allegro, ISG Corp., Mississauga, Canada) for the creation of 3-D imaging. The first step of the image processing was the segmentation of the structures. The brain surface and the cranial nerves were automatically segmented by a region growing method, and the segmentation of the arterial system was performed automatically by a threshold method after the setting of the threshold range. The segmented images of the structures were reconstructed for 3-D images by the surface rendering method. It usually took 30 minutes for this image processing. The presence of the causative vessels was evaluated using the axial SPGR MR images on the planes rostra1 to and coincident with the REZ of the facial nerve in the patients with hemifacial spasm, and on the plane coincident with the REZ of the trigeminal nerve in patients with trigeminal neuralgia. These causative vessels were identified using MR angiography, and the 3-D relationship between the causative vessels and the REZ of the facial or trigeminal nerve was evaluated using 3-D images.

RESULTS HEMIFACIAL SPASM In all 20 patients with hemifacial spasm, the preoperative SPGR MR images demonstrated that the causative vessel either compressed or was in contact with the REZ of the facial nerve on the affected side. On the contralateral side, neither compression of or contact with the REZ of the facial nerve was seen in any patient. The causative vessels were depicted as stippled or linear high-intensity areas on SPGR MR images in 10 patients each. These causative vessels were identified by inspection of the MR angiographic images; those shown as stippled areas were identified as the PICA in all 10 patients, and those shown as linear areas were identified as the AlCA in the other 10 patients. The 3-D images provided clear information as to the anatomic relationship between the causative vessels and the REZ of facial nerves; the causative vessels compressed or contacted the rostra1 portion of the facial nerve at the REZ. These findings were confirmed intraoperatively in all patients. The symptoms were completely relieved postoperatively in 18 patients, and partially

180

Surg Neurol 1997:47:178-84

Kumon et al

SPGR MR image and MR angiography in a patient qSPGRwith right hemifacial spasm. (A): The preoperative MR image shows a high-intensity linear area that appears to be in contact with the REZ of the right facial nerve (arrow); (B): the preoperative MR angiographic image demonstrates that the right anterior inferior cerebellar artery (AICA) is the causative vessel (arrow).

relieved in two patients. Sufficient decompression at the facial nerve REZ was demonstrated by postoperative SPGR MR images in all patients. TRIGEMINAL NEURALGlA In all six patients with trigeminal neuralgia, the preoperative SPGR MR images demonstrated that the causative vessels compressed the REZ of the trigeminal nerve on the affected side, which was displaced laterally and bowed at the REZ. The causative vessels were visualized as spotty highintensity areas on SPGR MR images. On the contralateral side, contact of the vessel with the trigeminal nerve REZ was seen in three patients, but the nerve was not displaced laterally. The causative vessel was identified as the superior cerebellar artery (SCA) in five patients and the PICA in one patient by inspection of the MR angiographic and 3-D images. The 3-D images, again, clearly visualized the anatomic relationship between the causative vessel and the trigeminal nerve REZ, and compression by

Preoperative 3-D images (A and B) show that the first and third loops of the right AICA (arrowheads) are attached to the REZ of the right facial nerve (arrow).

the causative vessel of the medial portion of the trigeminal nerve at the REZ was observed in all patients. The operative findings were compatible with those demonstrated by SPGR MR imaging, MR angiography, and 3-D imaging, and all patients were free of neuralgia postoperatively. Sufficient decompression between the causative vessel and the trigeminal nerve REZ was noted on the postoperative SPGR MR images. ILLUSTRATIVE

CASES

woman presented with right hemifacial spasm. SPGR MR images demonstrated that the causative vessel was in contact with the REZ of the right facial nerve (Figure 1A). This vessel, depicted as a linear high-intensity area on SPGR MR images, was identified as the AfCA based on MR angiography (Figure 1B). The 3-D images provided clear information as to the anatomic relationship between the two structures (Pigure 2). CASE

1. This 39-year-old

3-D Image in Microvascular Decompression

Surg Neurol 1997;47:178-84

18 1

The postoperative SPGR El MR image shows decompression between the highintensity linear area and the right facial nerve REZ (arrow).

The first and third loops of the right AICA appeared to be attached to the rostra1 portion of the nerve at the REZ. However, no vessel was observed at the REZ of the left facial nerve. These findings were confirmed intraoperatively, and a Teflon prosthesis was inserted between the right facial nerve REZ and the AICA. The spasm was completely abolished postoperatively, and sufficient decompression between the causative vessel and the RRZ of the facial nerve was observed on the postoperative SPGR MR images (Figure 3). CASE 2. This 60-year-old woman presented with left trigeminal neuralgia. SPGR MR images demonstrated the causative vessel as a stippled highintensity area, which compressed the REZ on the left trigeminal nerve. The nerve was displaced laterally and bowed at the REZ. The contralateral trigeminal nerve was also contacted by a vessel at the REZ, but it was not displaced laterally (Figure 4A).

This causative

vessel was identified

as the superior

cerebellar artery (SCA) using MR angiography (Figure 4B). The 3-D images clearly depicted the anatomic relationship between the vessel and the REZ of the left trigeminal nerve, and the posterior branch of the SCA seemed to compress the REZ of the nerve (Figure 5). These findings were confirmed intraoperatively, and a Teflon prosthesis was inserted between the left trigeminal nerve REZ and the SCA. The neuralgia was completely relieved postoperatively, and adequate decompression

was observed

erative SPGR MR images (Figure 6).

on the postop-

1 SPGR MR image and MR angiography in a patient qSPGR with left trigeminal neuralgia. (A): The preoperative MR images discloses a high-intensity stippled area (small arrow) that appears to be compressing the RFZ of the left trigeminal nerve (large arrow) laterally; (J3): the preoperative MR angiographic image demonstrates that the left superior cerebellar artery (SCA) is the causative vessel (arrow).

182

Surg Neurol 1997;47:178-84

Kumon et al

SPGR MR image reveals decompresqthe. leftPostoperative sion between the high-intensity stippled area and trigeminal nerve REZ (arrow).

3-D images (A and B) show that the qpressingPreoperative posterior branch of the left SCA (arrowhead) is comthe REZ of the left trigeminal nerve (arrow).

DISCUSSION Using SPGR MR images, MR angiography, and 3-D images reconstructed from data of SPGR MR imaging by the surface rendering method using a workstation, we preoperatively demonstrated vascular compression on the affected side of the facial or trigeminal nerve, and confirmed this finding intraoperatively in patients with hemifacial spasm and with trigeminal neuralgia. It is important to detect the presence of the causative vessels in patients with hemifacial spasm or trigeminal neuralgia not only for the differential diagnosis of these diseases from other diseases, but also to enhance the efficacy of decompression surgery. For the purpose of evaluating the relationship between the blood vessels and cranial nerves, noninvasive neuroradiologic methods such as computed tomography (CT) or MR imaging have been used. CT scans with contrast medium demonstrated ipsilateral dolichoectasia of the vertebro-

basilar artery in 33 of 46 patients with hemifacial spasm [5]. CT scans, however, reveal only ectasia or malposition of the artery, and because of the poor posterior fossa resolution do not show the vascular interaction with the cerebral parenchyma or cranial nerves. For this purpose, the usefulness of MR imaging has been reported. Wong et al. reported that in a patient with trigeminal neuralgia, axial T,-weighted MR images demonstrated the causative vessel as a signal void area medial to the proximal portion of the trigeminal nerve [16]. Tash et al. [14] and Hutchins et al. [8] also noted that coronal T,weighted MR images demonstrated the causative vessels as signal void areas surrounding the ipsilatera1 trigeminal nerve in patients with this disease. The causative vessels were detected in three of 12 patients with hemifacial spasm examined by Keyaki et al. using axial and coronal T,-weighted MR images [ 111, although in these three patients the compressing vessel was the vertebral artery. As reported by Tash et al., in all 13 of their patients with hemifacial spasm, coronal T,-weighted MR images demonstrated abnormal vascular loops shown as signal void areas compressing the REZ of the facial nerve [ 151. However, the visualization of the vascular structure shown as a signal void area on the T,-weighted or T,-weighted images was not clear enough to consistently provide an accurate diagnosis. Nageseki et al. concluded that oblique sagittal gradient-echo MR images provide anatomic information demonstrating vascular compression as they consistently afford visualization of the course of the nerves and clearly depict the vessel as a high-intensity area [ 131. Adler et al. reported that MR tomographic angiography, in which tissue and

3-D Image in Microvascular Decompression

arterial structures are displayed on the same image, detected ipsilateral vascular compression of the facial nerve or the pons in 24 of 37 patients with hemifacial spasm [ 11. They could observe the causative vessels more clearly because of the presentation of the vessel as a high-intensity area, and they could observe smaller arteries, such as the AICA and PICA, than could be seen on T,-weighted or Tz-weighted MR images. The SPGR MR imaging used in our study is a high speed scan technique that demonstrates vessels as high-intensity areas clearly due to the flow-related enhancement effect. Furthermore, since the slice thickness can be reduced in this method, the relationship between the lesion and cerebral tissue such as vessels, cranial nerves, or brain parenchyma are demonstrated in full. Although SPGR MR images demonstrate the causative vessels in patients with hemifacial spasm or trigeminal neuralgia, it does not consistently visualize the courses of both the vessels and the nerves on the same plane. One must deduce the 3-D relationship between the blood vessels and the cranial nerves. Three-D images are currently reconstructed from MR images or CT scans, and used for the presentation of the relationship between the lesion and the blood vessels, cranial nerves, and cranium in patients with skull base lesions [4,6], and for the accurate depiction of the lesion location in patients with intracerebral tumors surrounding the central sulci [ 7,121. We accordingly attempted to study the causative vessels in our patients with hemifacial spasm and with trigeminal neuralgia by means of 3-D imaging from data of SPGR MR imaging using a workstation, in addition to SPGR MR images and MR angiography. The 3-D images consistently demonstrated not only the course of the cranial nerve but also the compressing blood vessel on the same plane. We were able to recognize the relationship between the causative vessel and the facial or trigeminal nerve from any direction, and the inspection from the same direction as the intended operative approach was very useful for planning prior to microvascular decompression. SPGR MR imaging, MR angiography, and 3-D imaging are very useful methods, not only because they directly and safely demonstrate the relationship between the vessels and the facial or trigeminal nerve, but also because they depict abnormalias tumors, cerebral aneurysms, ties such arteriovenous malformations, and to some degree other vascular anomalies. In particular, 3-D images are useful for simulation planning of the operative procedure. The application of cerebral angiography, however, in patients with hemifacial spasm or

Surg Neurol 1997;47:178-84

183

trigeminal neuralgia seems limited only to the detection of the causative vessels and the diagnosis of cerebral aneurysm or arteriovenous malformation. Furthermore, this examination technique is not always safe, nor can it show directly the relationship between blood vessels and the REZ of the facial or trigeminal nerve. Therefore, cerebral angiography seems to be necessary for a definite diagnosis only when the MR imaging features suggest cerebrovascular disease. We emphasize that 3-D images reconstructed from data of SPGR MR imaging using a workstation are useful for the preoperative evaluation of the relationship between these cranial nerves and the causative vessels in symptomatic patients. We thank Mr. Kohsuke Ueda very much for his technical assistance.

REFERENCES 1. Adler CH, Zimmerman RA, Savino PJ, Bernardi B, Bosley TM, Sergott RC. Hemifacial spasm: evaluation by magnetic resonance imaging and magnetic resonance tomographic angiography. Ann Neurol1992;32:502-6. 2. Auger RG, Piepgras DC, Laws ER. Hemifacial spasm: results of microvascular decompression of the facial nerve in 54 patients. Mayo CIin Proc 1986;61:640-4. 3. Barker 11FG, Jannetta PJ, Bissonette DJ, Shields PT, Larkins MV, Jho HD. Microvascular decompression for hemifacial spasm. J Neurosurg 1995;82:201-10. 4. Chalif DJ, Dufresne CR, Ransohoff J, McCarthy JA. Three-dimensional computed tomographic reconstructions of intracranial meningiomas. Neurosurgery 1988;23:570-5. 5. Digre KB, Corbett JJ, Smoker WRK, McKusker S. CT and hemifacial spasm. Neurology 1988;38:111 l-3. 6. Gandhe AJ, Hill DLG, Studholme C, Hawkes DJ, Ruff CF, Cox TCS, Gleeson MJ, Strong AJ. Combined and three-dimensional rendered multimodal data for planning cranial base surgery: a prospective evaluation. Neurosurgery 1994;35:463-71. 7. Hu X, Tan KK, Levin DN, Galhotra S, Mullan JF, Hekmatpanah J, Spire JP. Three-dimensional magnetic resonance images of the brain: application to neurosurgical planning. J Neurosurg 1990;72:433-40. 8. Hutchins LG, Harnsberger HR, Jacobs JM, Apfelbaum RI. Trigeminal neuralgia (tic douloureux): MR imaging assessment’. Radiology 1990;175:837-41. 9. Jannetta PJ, Abbasy M, Maroon JC, Ramos FM, Albin MS. Etiology and definitive microsurgical treatment of hemifacial spasm: operative techniques and results in 47 patents. J Neurosurg 1977;47:321-8. 10. Jannetta PJ. Neurovascular compression in cranial nerve and systemic disease. Ann Surg 1990;192:51825. 11. Keyaki A, Makita Y, Nabeshima A, Tei T, Nioka H,

Iihara K, Matsuo M, Miki Y. Usefulness of magnetic resonance imaging in the patient with hemifacial spasm. Prog CT 1990;12:433-8 (in Japanese). 12. Levin DN, Hu X, Tan KK, Galhotra S. Surface of the

184

13.

14. 15. 16.

Surg Neurol 1997:47:178-84

Kumon et al

brain: three-dimensional MR images created with volume rendering. Radiology 1989;171:277-80. Nagaseki Y, Horikoshi T, Omata T, Ueno T, Uchida M. Nukui H, Tsuji R, Sasaki H. Oblique sagittal magnetic resonance imaging visualizing vascular compression of the trigeminal or facial nerve. J Neurosurg 1992;77: 379-86. Tash RR, Sze G, Leslie DR. Trigeminal neuralgia: MR imaging features. Radiology 1989;172:767-70. Tash R, DeMerritt J, Sze G, Leslie D. Hemifacial spasm: MR imaging features. Am J Neuroradiol 1991;12:83942. Wong BY, Steinberg GK, Rosen L. Magnetic resonance imaging of vascular compression in trigeminal neuralgia. J Neurosurg 1989;70:132-4.

COMMENTARY

I think Dr. Kumon and his coinvestigators present a good study of 20 patients with hemifacial spasm and six patients with trigeminal neuralgia, using a 3-D MR imaging technique. There is always an intimate

neurovascular

rela-

tionship in the CP angle, such as the trigeminal nerve with the SCA or AICA; VII and VIII cranial nerves with the AICA; and lower cranial nerves with PICA loops. I have observed many cases, with endoscopy and through operating microscope, and noted almost every time that the lower cranial nerves were associated with or in contact with the PICA; VII and VIII always with the AICA meatal loop. My previous investigation regarding prediction of neurovascular conflict through an enhanced CT scan, high resolution thin slice MR scan, or even angiography failed to demonstrate any accurate, fully operative diagnosis of neurovascular compression. In many cases, an enhanced CT scan will show a hypertrophied vertebral artery or tortuous basilar artery, or sometimes the AICA or SCA looping around the nerve. However, if one compares 100 patients within a specified age group with or without neurovascular compression symptoms and without any known diagnosis, one cannot determine neurovascular

compression

disease with neu-

roradiologic examinations. Although these authors have performed a good study, in order to predict neurovascular compression disease through neuroradiologic study, they will need at least 50-100 patients with comparative studies in both normal and patient groups. In conclusion, I would assume that 3-D MRA images may be useful in visualizing the posterior fossa vessels; however, I would disagree that 3-D MR im-

ages are helpful in diagnosing pression syndromes. Takanori

neurovascular

Fukushima,

com-

M.D., DMSc.

Pittsburgh,Pennsylvania This is a very interesting rovascular

impingement

paper demonstrating in cases considered

neufor mi-

crovascular decompression. It also confirms the postoperative decompression of the cranial nerve root entry zones in cases treated with successful results. The paper represents an up-to-date and optimal use of the MR imaging technique available presenting in many centers. It is, however, rather timeconsuming and may not be available for a majority of patients for a considerable period of time. In cases with typical trigeminal neuralgia and hemifacial spasm, there is presently no difference of opinion about the indication for microvascular decompression and no further neuroradiologic evaluation is usually necessary. The 3-D demonstration of the causative vessels preoperatively may, however, facilitate a rapid perioperative identification of the vessel responsible for the compression and therefore, minimize the risk of complications from further exploration and stretching of the nervous structures. After excluding intracranial tumors and multiple sclerosis by means of MR investigation, it is still the clinical picture, the age of the patient, and an appropriate description of the various options for treatment to the patient that decides the treatment. There are always a considerable number of patients however, whose symptoms are atypical in one or more respects and who require further evaluation. In these cases, the MR imaging with 3-D reconstruction will be most helpful and could definitely be of decisive value in determining whether a surgical exploration should be undertaken or not. In this way, the technique will contribute to the costeffectiveness of the management of suspected neurovascular compression conditions and also save some patients from operative procedures that will be of no benefit to them. This may also be valid for cases in which an unsuccessful microvascular decompression has been performed or in patients with recurring pain. Sten HGkansson,

M.D., Ph.D.

Stockholm,Sweden