Microvascular Decompression for Trigeminal Neuralgia with Special Reference to Delayed Recurrence Tomokazu
GOYA,
Department
Shinichiro
WAKISAKA
of Neurosurgery,
Miyazaki
and
Kazuo
KINOSHITA
Medical College, Miyazaki
Abstract Thirty-five patients with trigeminal neuralgia underwent microvascular decompression . Complete remission was obtained in 33 patients , while one was fair and another unchanged postoperatively. The clinical and operative findings were reviewed, analyzing the direction of vascular compression of the trigeminal nerve and the distribution of pain in the peripheral regions . There were some weak cor relations between the direction of vascular compression and the distribution of pain . Neuralgia in the region of second branch of the trigeminal nerve (V2) or in the regions of V2 and third branch of the nerve (V3) was caused by compression from the ventral or ventro-rostral direction , in the region of first branch of the nerve from the ventro-caudal direction , and in the V3 region from the ventral, ros tral, and dorsal directions of the nerve in general . In two patients who had had complete remission after first operation, trigeminal neuralgia recurred . They had typical intermittent painful attacks with a background of continuous dull pain or painful dysesthesia caused by Ivalon® sponges inserted be tween the nerve and the offending vessel. Complete remission was again obtained after removal of these sponge pieces. We would like to stress continuous dull pain or painful dysesthesia in cases of delayed recurrence as indicators for re-exploration . Key words:
trigeminal
neuralgia,
microvascular
decompression
Introduction
pain and atypical neuralgia such as dull pain with pressure sensation are suspected to originate from vascular compression of the trigeminal nerve at its root entry zone.9,'°I In our service, 35 patients of trigeminal neuralgia underwent MVD (36 operations). We report the results of clinical analysis in these cases. We also pre sent unusual clinical courses in two recurrent cases, both of which were successfully treated by a second operation.
1989
November
28,
1988;
delayed
recurrence
Clinical
Microvascular decompression (MVD) through a suboccipital retromastoideal small craniectomy has recently become the usual treatment for idiopathic trigeminal neuralgia because of the elaborate op erative findings established by Jannetta.4,5> Typical trigeminal neuralgia such as electric or shooting
Received
,
Accepted
July
5,
Analysis
Thirty-five patients with trigeminal neuralgia under went MVD (36 operations) between 1982 and 1988. Twelve patients were male and 23 female, and their ages ranged from 31 to 75 years (mean, 57.8 years). The affected side was the right in 15 cases and the left in 20. Trigeminal neuralgia concomitant with hemi facial spasm was observed in two cases. The period from onset of the pain to operation ranged from 3 months to 25 years and averaged 6 years. I.
Distribution of trigeminal neuralgia Nine patients had trigeminal neuralgia in the sec ond division (V2) only, and four in the third division (V3) only. In eight cases, the neuralgia expanded from V2 toward V3, and in five, from V3 toward V2. In one case only, the neuralgia expanded from V3 to V2 and V 1, and in another, from V 1 toward V2 and V3. In three cases, the neuralgia expanded from V 1 to V2, and in three, from V2 to V l. In one case, the neuralgia was restricted to V1. Neuralgia in the
combination
of V 1 and
preoperative tients.
facial
II.
V3 was not
sensation
observed.
was normal
Offending vessels and direction
The
in all pa
of vascular com
pression The most frequently observed conflicting vessel was the superior cerebellar artery (SCA) (28 cases, 80%). Figure 1 shows four types of compression caused by the SCA. In type a, the trigeminal nerve was compressed from the superior aspect, and in type b, from the ventral aspect. In type c, one branch of the SCA looped around the nerve and compressed it at the inferior ventral portion. In type d, the offending vessel was located at the dor sal aspect of the nerve. The former two types were the most frequently observed, and the latter two were rare types. Another less frequent conflicting vessel was the anterior inferior cerebellar artery (AICA) (3 cases, 8.6%). In two cases, the rostral loop of the AICA simply compressed the nerve at the inferior aspect. In one peculiar case, the AICA penetrated the nerve
between the portio major and minor from ventral to dorsal. In three cases, the trigeminal nerve was impinged upon by both the SCA and AICA from both the rostral and caudal directions. In these cases, the vessel causing the neuralgia was difficult to identify. Therefore, all possible vessels were carefully sepa rated from the trigeminal nerve. Four cases had venous compression of the nerve, as illustrated in Fig. 1. The findings were similar for all these cases. The so-called pontine vein or satellite trigeminal vein directly entered the dura in the close vicinity of Meckel's cave and, on the way to the dura, compressed the nerve forming an obvious in dentation in the nerve. At surgery, these veins were also cauterized and cut. III. Relationships between direction of vascular compression and distribution of pain The direction of vascular compression confirmed during operations was compared with the distribu tion of pain in the 35 cases, but one in which the direction was not clearly determined (Fig. 2). In cases of V 1 pain, vascular compression was from the inferior-ventral direction and involved narrow region. Vascular compression in cases of V2 pain was from the superior-ventral direction, and for V3 pain from the general direction of the ventral, supe rior, and dorsal aspects of the nerve. In the most frequent cases with the combination of V2 and V3 pain, the direction of the vascular compression was from the superior-ventral aspect of the nerve, as in dicated by the superimposed range of semicircles of V2 and V3 in Fig. 2. From these correlations, it is possible that the distribution of pain suggests the direction of vascular compression and the offend ing artery. IV. Results and complications of MVD The follow-up periods ranged from 6 months to 6
Fig. 1
Schematic representation of four types of neurovascular compression due to the SCA. Curved arrows indicate the compressed por tion of the trigeminal nerve. Labeled struc tures include the petrosal vein (star), SCA (open arrowhead), pons (asterisk), trigeminal nerve (arrow), and pontine vein (closed ar ro whead).
years. Complete remission was obtained in 33 (94%) of 35 cases. Pain reduced by half in one case and another was unchanged. In two cases, the pain re curred after complete remission by MVD and again completely disappeared after a second operation. These two cases will be discussed later. As a complication of MVD, hearing loss devel oped in the initial three cases. In one case, cerebro spinal fluid leakage into the middle ear cavity oc curred afterr insufficient packing of the opened mastoid air cells. No such complication has been ex perienced since this event. The petrosal sacrificed in some cases without problems.
vein was
V.
Fig. 2
Delayed recurrence In our two recurrent cases, interesting clinical courses and findings were observed. Case 1: A 65-year-old female presented with inter mittent pain in the right cheek, especially exacer bated by mouth opening, since August, 1984. This paroxysmal pain continued for 1 or 2 minutes only and there was no pain during the interphase. This pain gradually expanded to the right upper lip. In August, 1985, she had a nerve block at the infraor bital foramen and the pain was resolved until recur rence after 8 months. As severe intermittent pain occurred again, MVD was performed in March, 1987. Two branches of the SCA, together compress ing the trigeminal nerve from above, were sepa rated from the nerve by inserting a small piece of Ivalon R sponge between the brainstem and these branches (Fig. 3). Although this was sufficient for decompression, another piece of sponge was also placed between the proximal portion of the SCA and the nerve in the vicinity of the dura, as the possibility of compression of the nerve by the caudal loop of the SCA in the future was considered. The pain com pletely disappeared just after the operation. Seven months after the operation, however, parox ysmal pain suddenly recurred in the right upper lip while taking a meal. Four months later, paroxysmal sharp pain had begun to superimpose on dull throb bing pain persisting all day long. As carbamazepine was ineffective for the pain, she was readmitted to our service. At the second operation, performed just 1 year after the first MVD, one of the two sponges previously inserted was tightly conglutinated to the nerve and, furthermore, compressed and distorted it downward at the petrous ridge. This Ivalon sponge was carefully separated and removed from the nerve. The neuralgia completely disappeared just after the operation and her facial sensation remained normal. Case 2: A 61-year-old female noticed tingling
Schematic representation of direction of vas cular compression and distribution of pain in the axial section of the trigeminal nerve at its root entry zone. A large oval and three small, round areas with oblique lines indicate the portio major and minor. The left side of these areas corresponds to the ventral aspect of the nerve, the right side to the dorsal, and the up per side to the rostral. Labeled (V1, V2, and V3) converging lines on the nerve indicate the facial regions involved with neuralgia and the direction of the vascular compression. Outer thick semicircular lines indicate the peripheral regions involved.
Fig.
3
Illustrated loop
operative
of the SCA
findings
in Case
and the trigeminal
1. nerve
Two near
Ivalon
sponge
the petrous
pieces ridge .
are placed
between
the caudal
Fig. 4
Case 2. Plain skull x-ray film (left) and CT scan (right) show abnormal calcification (arrow) on the petrous ridge.
sharp pain periodically in the left cheek and oral cavity since February, 1985. The paroxysms became more frequent and severe. The pain had always been typical tic douloureux, including the presence of trigger area around the left nasal labial angle. Five months after the onset of pain, the first MVD was performed in another hospital. The pain com pletely disappeared after this operation. One year and 2 months later, the tingling pain recurred intermittently on the left side of the nose and cheek. Continuous dull pain gradually devel oped, and then, as in Case 1, paroxysmal sharp pain occurred with the background continuous dull pain. In February, 1988, 3 years after the first MVD, she was admitted to our service. The neurological ex amination was entirely normal. There was no evidence of sensory deficit in the distribution of the left trigeminal nerve. A skull x-ray film (Towne's projection) revealed a bone defect caused by the previous craniectomy and a faint calcified lesion on the petrous ridge (Fig. 4 left). This calcification was confirmed by a computed tomographic (CT) scan (Fig. 4 right). There were no symptoms suggesting the presence of a space-occupying lesion. At the sec ond operation, one small artery located on the dorsal surface of the trigeminal nerve had been well separated from the nerve by an Ivalon sponge placed between this artery and the nerve. Therefore, the entry zone of the trigeminal nerve could not be di rectly observed because of the relatively large mass of sponge. After careful separation from the nerve and complete removal of this sponge, the trigeminal nerve was observed to be depressed toward the prepontine
cistern and had become flat. Histological
ly, there
were fine calcified
flecks
within
this Ivalon
sponge, as indicated by x-ray. At the ventral aspect of the root entry zone of the trigeminal nerve, a branch of the SCA was compressing the nerve. Presumably, this arterial compression had been over looked at the first operation. This branch of SCA was pushed ventrally toward the prepontine cistern and a small Ivalon sponge was inserted between the brainstem and the artery. The neuralgia was com pletely relieved without neurological deficit. In these two recurrent cases, Ivalon sponges placed as separators had become tightly bound to the nerve and compressed it, presumably causing the recur rence of trigeminal neuralgia.
Discussion I.
Fiber arrangement in the cisternal portion of the trigeminal nerve There is some controversy concerning the fiber ar rangement in the cisternal portion of the trigeminal nerve. Afferent fibers from three peripheral branches of the trigeminal nerve are reported to run in specific constant portions of the nerve bundle at the pontine cistern.') Other reports suggest that, although there is some regular distribution, fibers from each periph eral branch intermingle within the nerve in this por tion.',') Previously, during trigeminal rhizotomy through the suboccipital retromastoideal route, the nerve root has been incised partially or totally from the inferior aspect.',') In cases of partial resection of the nerve, according
facial hypesthesia expands to the degree of incision.
from For
V3 to V2 neuralgia
in V l . the nerve is incised at the ventral side. Ac cording to these clinical observations, V 1 seems to be located in the ventral side of the nerve and V3 in the dorsal aspect at the pontine cistern.'' However, Gudmundsson et al.') stated that the cross section of the portio major between the petrous apex and the pons is elliptical, and the angle between the longest diameter of this elliptical cross section and the long axis of the body can vary from 10 to 80°. This means that VI, V2, and V3 occur in turn from rostral (superior) to caudal (inferior) when this an gle is near vertical, and VI, V2, and V3 in turn from ventral to dorsal when this angle is near horizontal. This may explain the variations in sensory loss oc curring after partial section of the trigeminal nerve in the posterior fossa. Pelletier et al." reported that the fibers from the peripheral branches intermingle, although some regular arrangement is seen, and the fibers also intermingle even in terms of functions and modalities of sensation. Kerr" also reported that at the entrance to the pons of the trigeminal nerve, V 1 is located in the ventral side and V3 in the dorsal side. Another complicating factor of this fiber ar rangement is that the bundle of the trigeminal nerve rotates somewhat between the ganglion and the pons.'''' Therefore, there about the fiber arrangement trigeminal nerve.
are differing opinions within the bundle of the
II.
Direction of vascular compression and distribu tion of pain As mentioned above, there is no consensus about the fiber arrangement in the cisternal portion of the trigeminal nerve. Even in simple vascular compres sion from the rostral direction, the distribution of the facial pain would be variable. Nevertheless, some constant patterns can be expected even in our small number of cases, in spite of the many varia tions. From this study, the following assumptions can be made. Neuralgia in the territory of V2 and V3 is caused by the SCA compression of the trigeminal nerve from the ventral direction. Pain in the V 1 ter ritory alone or in V 1 and V2 territories is caused by the AICA compression from the ventral and caudal (inferior) directions. Pain in the V3 territory alone is caused by the SCA compression from the rostral (superior) and dorsal directions. We experienced one unusual case in which pain in V2 and V3 was caused by the AICA penetrating be tween the portio major and minor. Because the cen tral myelin extends and tapers off distally in the cis ternal portion of the trigeminal nerve, it is better to dislocate the offending artery toward the periph eral side where the junctional area tapers off.',"'
Therefore, the AICA was dislocated distally toward Meckel's cave with a small sponge and muscle pieces. The neuralgia disappeared after this procedure.
III. Venous compression Idiopathic trigeminal neuralgia is thought to be caused by arterial compression, but indentation of the trigeminal nerve by simultaneous venous com pression associated with arterial compression can be observed. In 1981, Jannetta4' reported that there were 57 venous abnormalities and 96 complex ab normalities of both arteries and veins, out of a total 411 patients undergoing MVD for trigeminal neuralgia. Apfelbaum" also confirmed venous com pression alone in 12 of 55 patients undergone MVD. This suggests that the occurrence of venous com pression may be quite high. The lateral pontine vein and the transverse pontine vein, or so-called aberrant trigeminal vein, run from the pons to the dura in the vicinity of Meckel's cave, and join the petrosal vein or enter the dura directly. Before en trance, these veins sometimes compress the nerve and form an indentation mostly at the peripheral side near Meckel's cave. In our experience, however, importance of venous compression for the neuralgia and its characteristics, compared to arterial compres sion, was difficult to assess: Continuous dull pain in the territory of V2 was induced by venous com pression alone and this pain was not a typical tri geminal neuralgia. IV. Recurrence of pain after MVD If pain relief is not obtained immediately after the operation, the cause may be overlooking of offend ing vessels or inadequate implants. A common cause of recurrence within 6 months after the op eration is said to be recollateralization of intrinsic pontine veins. Late recurrence appears to be most commonly due to the continuous elongation of ar teries which enter the root entry zone of the nerve or displacement of the implant from the nerve.') In our Case 1, the neuralgia recurred because one of two Ivalon sponges, inserted between the nerve and the offending vessels, had adhered to the nerve and compressed it. Continuous painful dysesthesia or dull pain, with a different nature from the pain caused by arterial compression, developed due to the adhesion. In addition to this continuous compres sion, the so-called "secondary missile phenome non" 5' occurred. Therefore, typical trigeminal neural gia was superimposed on the background continu ous dull pain. Interestingly, there was no sensory disturbance on the face, although compression of the nerve by an Ivalon sponge was apparent.
In Case 2, the offending artery located on the dor sal surface of the trigeminal nerve was well decom
trigeminal
3)
pressed by insertion of an Ivalon sponge between this vessel and the nerve. There was no pain for more than 1 year. However, the Ivalon sponge gradually compressed the nerve and displaced it ventrally. Fi nally, this nerve contacted the caudal loop of the SCA
located
pulsatile trigeminal
ventrally
to
the
nerve
and
suffered
5)
compression. It was supposed that typical neuralgia recurred due to this arterial com
pression, in addition to the continuous compression caused by the adhering Ivalon sponge. In these two cases, a common characteristic of the pain on trigeminal
4)
recurrence neuralgia
was that typical was superimposed
intermittent on the back
ground continuous dull pain. Therefore, the follow ing precautions are important to prevent recurrence: The size of the separator should be the minimum essential. The separator should not be introduced between the nerve and the offending vessel and should not contact the nerve. If late recurrence of trigeminal neuralgia occurs in cases who had com plete remission for a while after MVD, we should consider such causes as occurred in our cases.
References Apfelbaum RI: A comparison of percutaneous radiofrequency trigeminal neurolysis and micro vascular decompression of the trigeminal nerve for the treatment of tic douloureux. Neurosurgery 1: 16 -21, 1977 2) Gudmundsson K, Rhoton AL Jr, Rushton JG: De tailed anatomy of the intracranial portion of the
6)
7)
8)
9)
10)
11)
nerve.
J Neurosurg
35: 592-600,
1971
Hussein M, Wilson LA, Illingworth R: Patterns of sensory loss following fractional posterior fossa Vth nerve section for trigeminal neuralgia J Neurol Neurosurg Psychiatry 45: 786-790, 1982 Jannetta PJ: Vascular decompression in trigeminal neuralgia, in Samii M, Jannetta PJ (eds): The Cranial Nerves. Berlin, Springer, 1981, pp 331-340 Jannetta PJ, Bissonette DJ: Management of the fail ed patient with trigeminal neuralgia. Clin Neurosurg 32: 334-347. 1985 Keller JT, Van Loveren H: Pathophysioiogy of the pain of trigeminal neuralgia and atypical facial pain: A neuroanatomical perspective. Clin Neurosurg 32: 275-293, 1985 Kerr FWL: The divisional organization of afferent fibers of the trigeminal nerve. Brain 86: 721-732, 1963 Pelletier VA, Poulos DA, Lende RA: Functional localization in the trigeminal root. J Neurosurg 40: 504-513, 1974 Roski RA, Horwitz SJ, Spetzler RF: Atypical trigeminal neuralgia in a 6-year-old boy. Case report. J Neurosurg 56: 424-425, 1982 Szapiro J Jr, Sindou M, Szapiro J: Prognostic fac tors in microvascular decompression for trigeminal neuralgia. Neurosurgery 17: 920-929, 1985 Tarlov IM: Structure of the nerve root. Arch Neurol Psychiat 37: 555-583, 1937
1)
Address of
reprint
requests
Neurosurgery,
Kihara,
Kiyotake-cho,
16, Japan.
to: T. Goya, Miyazaki
M.D.,
Medical
Miyazaki-gun,
Department College,
5200
Miyazaki
889
GOYA,
Department
Shinichiro
WAKISAKA
of Neurosurgery,
Miyazaki
and
Kazuo
KINOSHITA
Medical College, Miyazaki
Abstract Thirty-five patients with trigeminal neuralgia underwent microvascular decompression . Complete remission was obtained in 33 patients , while one was fair and another unchanged postoperatively. The clinical and operative findings were reviewed, analyzing the direction of vascular compression of the trigeminal nerve and the distribution of pain in the peripheral regions . There were some weak cor relations between the direction of vascular compression and the distribution of pain . Neuralgia in the region of second branch of the trigeminal nerve (V2) or in the regions of V2 and third branch of the nerve (V3) was caused by compression from the ventral or ventro-rostral direction , in the region of first branch of the nerve from the ventro-caudal direction , and in the V3 region from the ventral, ros tral, and dorsal directions of the nerve in general . In two patients who had had complete remission after first operation, trigeminal neuralgia recurred . They had typical intermittent painful attacks with a background of continuous dull pain or painful dysesthesia caused by Ivalon® sponges inserted be tween the nerve and the offending vessel. Complete remission was again obtained after removal of these sponge pieces. We would like to stress continuous dull pain or painful dysesthesia in cases of delayed recurrence as indicators for re-exploration . Key words:
trigeminal
neuralgia,
microvascular
decompression
Introduction
pain and atypical neuralgia such as dull pain with pressure sensation are suspected to originate from vascular compression of the trigeminal nerve at its root entry zone.9,'°I In our service, 35 patients of trigeminal neuralgia underwent MVD (36 operations). We report the results of clinical analysis in these cases. We also pre sent unusual clinical courses in two recurrent cases, both of which were successfully treated by a second operation.
1989
November
28,
1988;
delayed
recurrence
Clinical
Microvascular decompression (MVD) through a suboccipital retromastoideal small craniectomy has recently become the usual treatment for idiopathic trigeminal neuralgia because of the elaborate op erative findings established by Jannetta.4,5> Typical trigeminal neuralgia such as electric or shooting
Received
,
Accepted
July
5,
Analysis
Thirty-five patients with trigeminal neuralgia under went MVD (36 operations) between 1982 and 1988. Twelve patients were male and 23 female, and their ages ranged from 31 to 75 years (mean, 57.8 years). The affected side was the right in 15 cases and the left in 20. Trigeminal neuralgia concomitant with hemi facial spasm was observed in two cases. The period from onset of the pain to operation ranged from 3 months to 25 years and averaged 6 years. I.
Distribution of trigeminal neuralgia Nine patients had trigeminal neuralgia in the sec ond division (V2) only, and four in the third division (V3) only. In eight cases, the neuralgia expanded from V2 toward V3, and in five, from V3 toward V2. In one case only, the neuralgia expanded from V3 to V2 and V 1, and in another, from V 1 toward V2 and V3. In three cases, the neuralgia expanded from V 1 to V2, and in three, from V2 to V l. In one case, the neuralgia was restricted to V1. Neuralgia in the
combination
of V 1 and
preoperative tients.
facial
II.
V3 was not
sensation
observed.
was normal
Offending vessels and direction
The
in all pa
of vascular com
pression The most frequently observed conflicting vessel was the superior cerebellar artery (SCA) (28 cases, 80%). Figure 1 shows four types of compression caused by the SCA. In type a, the trigeminal nerve was compressed from the superior aspect, and in type b, from the ventral aspect. In type c, one branch of the SCA looped around the nerve and compressed it at the inferior ventral portion. In type d, the offending vessel was located at the dor sal aspect of the nerve. The former two types were the most frequently observed, and the latter two were rare types. Another less frequent conflicting vessel was the anterior inferior cerebellar artery (AICA) (3 cases, 8.6%). In two cases, the rostral loop of the AICA simply compressed the nerve at the inferior aspect. In one peculiar case, the AICA penetrated the nerve
between the portio major and minor from ventral to dorsal. In three cases, the trigeminal nerve was impinged upon by both the SCA and AICA from both the rostral and caudal directions. In these cases, the vessel causing the neuralgia was difficult to identify. Therefore, all possible vessels were carefully sepa rated from the trigeminal nerve. Four cases had venous compression of the nerve, as illustrated in Fig. 1. The findings were similar for all these cases. The so-called pontine vein or satellite trigeminal vein directly entered the dura in the close vicinity of Meckel's cave and, on the way to the dura, compressed the nerve forming an obvious in dentation in the nerve. At surgery, these veins were also cauterized and cut. III. Relationships between direction of vascular compression and distribution of pain The direction of vascular compression confirmed during operations was compared with the distribu tion of pain in the 35 cases, but one in which the direction was not clearly determined (Fig. 2). In cases of V 1 pain, vascular compression was from the inferior-ventral direction and involved narrow region. Vascular compression in cases of V2 pain was from the superior-ventral direction, and for V3 pain from the general direction of the ventral, supe rior, and dorsal aspects of the nerve. In the most frequent cases with the combination of V2 and V3 pain, the direction of the vascular compression was from the superior-ventral aspect of the nerve, as in dicated by the superimposed range of semicircles of V2 and V3 in Fig. 2. From these correlations, it is possible that the distribution of pain suggests the direction of vascular compression and the offend ing artery. IV. Results and complications of MVD The follow-up periods ranged from 6 months to 6
Fig. 1
Schematic representation of four types of neurovascular compression due to the SCA. Curved arrows indicate the compressed por tion of the trigeminal nerve. Labeled struc tures include the petrosal vein (star), SCA (open arrowhead), pons (asterisk), trigeminal nerve (arrow), and pontine vein (closed ar ro whead).
years. Complete remission was obtained in 33 (94%) of 35 cases. Pain reduced by half in one case and another was unchanged. In two cases, the pain re curred after complete remission by MVD and again completely disappeared after a second operation. These two cases will be discussed later. As a complication of MVD, hearing loss devel oped in the initial three cases. In one case, cerebro spinal fluid leakage into the middle ear cavity oc curred afterr insufficient packing of the opened mastoid air cells. No such complication has been ex perienced since this event. The petrosal sacrificed in some cases without problems.
vein was
V.
Fig. 2
Delayed recurrence In our two recurrent cases, interesting clinical courses and findings were observed. Case 1: A 65-year-old female presented with inter mittent pain in the right cheek, especially exacer bated by mouth opening, since August, 1984. This paroxysmal pain continued for 1 or 2 minutes only and there was no pain during the interphase. This pain gradually expanded to the right upper lip. In August, 1985, she had a nerve block at the infraor bital foramen and the pain was resolved until recur rence after 8 months. As severe intermittent pain occurred again, MVD was performed in March, 1987. Two branches of the SCA, together compress ing the trigeminal nerve from above, were sepa rated from the nerve by inserting a small piece of Ivalon R sponge between the brainstem and these branches (Fig. 3). Although this was sufficient for decompression, another piece of sponge was also placed between the proximal portion of the SCA and the nerve in the vicinity of the dura, as the possibility of compression of the nerve by the caudal loop of the SCA in the future was considered. The pain com pletely disappeared just after the operation. Seven months after the operation, however, parox ysmal pain suddenly recurred in the right upper lip while taking a meal. Four months later, paroxysmal sharp pain had begun to superimpose on dull throb bing pain persisting all day long. As carbamazepine was ineffective for the pain, she was readmitted to our service. At the second operation, performed just 1 year after the first MVD, one of the two sponges previously inserted was tightly conglutinated to the nerve and, furthermore, compressed and distorted it downward at the petrous ridge. This Ivalon sponge was carefully separated and removed from the nerve. The neuralgia completely disappeared just after the operation and her facial sensation remained normal. Case 2: A 61-year-old female noticed tingling
Schematic representation of direction of vas cular compression and distribution of pain in the axial section of the trigeminal nerve at its root entry zone. A large oval and three small, round areas with oblique lines indicate the portio major and minor. The left side of these areas corresponds to the ventral aspect of the nerve, the right side to the dorsal, and the up per side to the rostral. Labeled (V1, V2, and V3) converging lines on the nerve indicate the facial regions involved with neuralgia and the direction of the vascular compression. Outer thick semicircular lines indicate the peripheral regions involved.
Fig.
3
Illustrated loop
operative
of the SCA
findings
in Case
and the trigeminal
1. nerve
Two near
Ivalon
sponge
the petrous
pieces ridge .
are placed
between
the caudal
Fig. 4
Case 2. Plain skull x-ray film (left) and CT scan (right) show abnormal calcification (arrow) on the petrous ridge.
sharp pain periodically in the left cheek and oral cavity since February, 1985. The paroxysms became more frequent and severe. The pain had always been typical tic douloureux, including the presence of trigger area around the left nasal labial angle. Five months after the onset of pain, the first MVD was performed in another hospital. The pain com pletely disappeared after this operation. One year and 2 months later, the tingling pain recurred intermittently on the left side of the nose and cheek. Continuous dull pain gradually devel oped, and then, as in Case 1, paroxysmal sharp pain occurred with the background continuous dull pain. In February, 1988, 3 years after the first MVD, she was admitted to our service. The neurological ex amination was entirely normal. There was no evidence of sensory deficit in the distribution of the left trigeminal nerve. A skull x-ray film (Towne's projection) revealed a bone defect caused by the previous craniectomy and a faint calcified lesion on the petrous ridge (Fig. 4 left). This calcification was confirmed by a computed tomographic (CT) scan (Fig. 4 right). There were no symptoms suggesting the presence of a space-occupying lesion. At the sec ond operation, one small artery located on the dorsal surface of the trigeminal nerve had been well separated from the nerve by an Ivalon sponge placed between this artery and the nerve. Therefore, the entry zone of the trigeminal nerve could not be di rectly observed because of the relatively large mass of sponge. After careful separation from the nerve and complete removal of this sponge, the trigeminal nerve was observed to be depressed toward the prepontine
cistern and had become flat. Histological
ly, there
were fine calcified
flecks
within
this Ivalon
sponge, as indicated by x-ray. At the ventral aspect of the root entry zone of the trigeminal nerve, a branch of the SCA was compressing the nerve. Presumably, this arterial compression had been over looked at the first operation. This branch of SCA was pushed ventrally toward the prepontine cistern and a small Ivalon sponge was inserted between the brainstem and the artery. The neuralgia was com pletely relieved without neurological deficit. In these two recurrent cases, Ivalon sponges placed as separators had become tightly bound to the nerve and compressed it, presumably causing the recur rence of trigeminal neuralgia.
Discussion I.
Fiber arrangement in the cisternal portion of the trigeminal nerve There is some controversy concerning the fiber ar rangement in the cisternal portion of the trigeminal nerve. Afferent fibers from three peripheral branches of the trigeminal nerve are reported to run in specific constant portions of the nerve bundle at the pontine cistern.') Other reports suggest that, although there is some regular distribution, fibers from each periph eral branch intermingle within the nerve in this por tion.',') Previously, during trigeminal rhizotomy through the suboccipital retromastoideal route, the nerve root has been incised partially or totally from the inferior aspect.',') In cases of partial resection of the nerve, according
facial hypesthesia expands to the degree of incision.
from For
V3 to V2 neuralgia
in V l . the nerve is incised at the ventral side. Ac cording to these clinical observations, V 1 seems to be located in the ventral side of the nerve and V3 in the dorsal aspect at the pontine cistern.'' However, Gudmundsson et al.') stated that the cross section of the portio major between the petrous apex and the pons is elliptical, and the angle between the longest diameter of this elliptical cross section and the long axis of the body can vary from 10 to 80°. This means that VI, V2, and V3 occur in turn from rostral (superior) to caudal (inferior) when this an gle is near vertical, and VI, V2, and V3 in turn from ventral to dorsal when this angle is near horizontal. This may explain the variations in sensory loss oc curring after partial section of the trigeminal nerve in the posterior fossa. Pelletier et al." reported that the fibers from the peripheral branches intermingle, although some regular arrangement is seen, and the fibers also intermingle even in terms of functions and modalities of sensation. Kerr" also reported that at the entrance to the pons of the trigeminal nerve, V 1 is located in the ventral side and V3 in the dorsal side. Another complicating factor of this fiber ar rangement is that the bundle of the trigeminal nerve rotates somewhat between the ganglion and the pons.'''' Therefore, there about the fiber arrangement trigeminal nerve.
are differing opinions within the bundle of the
II.
Direction of vascular compression and distribu tion of pain As mentioned above, there is no consensus about the fiber arrangement in the cisternal portion of the trigeminal nerve. Even in simple vascular compres sion from the rostral direction, the distribution of the facial pain would be variable. Nevertheless, some constant patterns can be expected even in our small number of cases, in spite of the many varia tions. From this study, the following assumptions can be made. Neuralgia in the territory of V2 and V3 is caused by the SCA compression of the trigeminal nerve from the ventral direction. Pain in the V 1 ter ritory alone or in V 1 and V2 territories is caused by the AICA compression from the ventral and caudal (inferior) directions. Pain in the V3 territory alone is caused by the SCA compression from the rostral (superior) and dorsal directions. We experienced one unusual case in which pain in V2 and V3 was caused by the AICA penetrating be tween the portio major and minor. Because the cen tral myelin extends and tapers off distally in the cis ternal portion of the trigeminal nerve, it is better to dislocate the offending artery toward the periph eral side where the junctional area tapers off.',"'
Therefore, the AICA was dislocated distally toward Meckel's cave with a small sponge and muscle pieces. The neuralgia disappeared after this procedure.
III. Venous compression Idiopathic trigeminal neuralgia is thought to be caused by arterial compression, but indentation of the trigeminal nerve by simultaneous venous com pression associated with arterial compression can be observed. In 1981, Jannetta4' reported that there were 57 venous abnormalities and 96 complex ab normalities of both arteries and veins, out of a total 411 patients undergoing MVD for trigeminal neuralgia. Apfelbaum" also confirmed venous com pression alone in 12 of 55 patients undergone MVD. This suggests that the occurrence of venous com pression may be quite high. The lateral pontine vein and the transverse pontine vein, or so-called aberrant trigeminal vein, run from the pons to the dura in the vicinity of Meckel's cave, and join the petrosal vein or enter the dura directly. Before en trance, these veins sometimes compress the nerve and form an indentation mostly at the peripheral side near Meckel's cave. In our experience, however, importance of venous compression for the neuralgia and its characteristics, compared to arterial compres sion, was difficult to assess: Continuous dull pain in the territory of V2 was induced by venous com pression alone and this pain was not a typical tri geminal neuralgia. IV. Recurrence of pain after MVD If pain relief is not obtained immediately after the operation, the cause may be overlooking of offend ing vessels or inadequate implants. A common cause of recurrence within 6 months after the op eration is said to be recollateralization of intrinsic pontine veins. Late recurrence appears to be most commonly due to the continuous elongation of ar teries which enter the root entry zone of the nerve or displacement of the implant from the nerve.') In our Case 1, the neuralgia recurred because one of two Ivalon sponges, inserted between the nerve and the offending vessels, had adhered to the nerve and compressed it. Continuous painful dysesthesia or dull pain, with a different nature from the pain caused by arterial compression, developed due to the adhesion. In addition to this continuous compres sion, the so-called "secondary missile phenome non" 5' occurred. Therefore, typical trigeminal neural gia was superimposed on the background continu ous dull pain. Interestingly, there was no sensory disturbance on the face, although compression of the nerve by an Ivalon sponge was apparent.
In Case 2, the offending artery located on the dor sal surface of the trigeminal nerve was well decom
trigeminal
3)
pressed by insertion of an Ivalon sponge between this vessel and the nerve. There was no pain for more than 1 year. However, the Ivalon sponge gradually compressed the nerve and displaced it ventrally. Fi nally, this nerve contacted the caudal loop of the SCA
located
pulsatile trigeminal
ventrally
to
the
nerve
and
suffered
5)
compression. It was supposed that typical neuralgia recurred due to this arterial com
pression, in addition to the continuous compression caused by the adhering Ivalon sponge. In these two cases, a common characteristic of the pain on trigeminal
4)
recurrence neuralgia
was that typical was superimposed
intermittent on the back
ground continuous dull pain. Therefore, the follow ing precautions are important to prevent recurrence: The size of the separator should be the minimum essential. The separator should not be introduced between the nerve and the offending vessel and should not contact the nerve. If late recurrence of trigeminal neuralgia occurs in cases who had com plete remission for a while after MVD, we should consider such causes as occurred in our cases.
References Apfelbaum RI: A comparison of percutaneous radiofrequency trigeminal neurolysis and micro vascular decompression of the trigeminal nerve for the treatment of tic douloureux. Neurosurgery 1: 16 -21, 1977 2) Gudmundsson K, Rhoton AL Jr, Rushton JG: De tailed anatomy of the intracranial portion of the
6)
7)
8)
9)
10)
11)
nerve.
J Neurosurg
35: 592-600,
1971
Hussein M, Wilson LA, Illingworth R: Patterns of sensory loss following fractional posterior fossa Vth nerve section for trigeminal neuralgia J Neurol Neurosurg Psychiatry 45: 786-790, 1982 Jannetta PJ: Vascular decompression in trigeminal neuralgia, in Samii M, Jannetta PJ (eds): The Cranial Nerves. Berlin, Springer, 1981, pp 331-340 Jannetta PJ, Bissonette DJ: Management of the fail ed patient with trigeminal neuralgia. Clin Neurosurg 32: 334-347. 1985 Keller JT, Van Loveren H: Pathophysioiogy of the pain of trigeminal neuralgia and atypical facial pain: A neuroanatomical perspective. Clin Neurosurg 32: 275-293, 1985 Kerr FWL: The divisional organization of afferent fibers of the trigeminal nerve. Brain 86: 721-732, 1963 Pelletier VA, Poulos DA, Lende RA: Functional localization in the trigeminal root. J Neurosurg 40: 504-513, 1974 Roski RA, Horwitz SJ, Spetzler RF: Atypical trigeminal neuralgia in a 6-year-old boy. Case report. J Neurosurg 56: 424-425, 1982 Szapiro J Jr, Sindou M, Szapiro J: Prognostic fac tors in microvascular decompression for trigeminal neuralgia. Neurosurgery 17: 920-929, 1985 Tarlov IM: Structure of the nerve root. Arch Neurol Psychiat 37: 555-583, 1937
1)
Address of
reprint
requests
Neurosurgery,
Kihara,
Kiyotake-cho,
16, Japan.
to: T. Goya, Miyazaki
M.D.,
Medical
Miyazaki-gun,
Department College,
5200
Miyazaki
889
