AUTHOR(S): Klun, Boris, M.D., Ph.D. Department of Neurosurgery, University Medical Center, Ljubljana, Slovenia Neurosurgery 30; 49-52, 1992 ABSTRACT: The results of the treatment of trigeminal neuralgia by neurovascular decompression or partial sensory rhizotomy in a personal series of 220 patients are presented. Microvascular decompression was performed in 178 patients and partial sensory rhizotomy in 42. The mean follow-up was 5.2 years. Immediate pain relief was achieved in 94% of all patients, but the rate dropped to 84% during the follow-up period. The recurrence rate in the microvascular decompression group was 6% and in the PSR 49%. Permanent sequelae occurred in 4 patients (loss of hearing, 1; loss of corneal reflex, 1; lesion of the portio minor, 2), but transitory complications (impaired hearing caused by hematotympanum and diplopia) were more frequent, especially in the beginning of the series. Elderly patients tolerated the procedure very well and the percentage of complications was evenly distributed in all age groups. Three patients died. No patient developed painful dysesthesias or anesthesia dolorosa. There were no differences in the outcome, considering sex and age. The duration of symptoms did not influence the prognosis. Patients with severe compression did better than those with a mild one, and patients with an arterial compression did better than those with a venous one. Trigeminal neuralgia in multiple sclerosis is seldom relieved by microvascular decompression. The experience of the surgeon reduces the number of negative findings considerably. KEY WORDS: Microvascular decompression; Partial sensory rhizotomy; Trigeminal neuralgia INTRODUCTION A milestone in the treatment of trigeminal neuralgia was the understanding of vascular compression of the nerve, described as early as 1932 by Dandy (5), rediscovered by Gardner (6), and fully recognized by Jannetta (7). In spite of some controversy, microvascular decompression (MVD) represents today the best and the most widely used surgical treatment for neuralgia of the Vth nerve. Vascular compression of other cranial nerves and the understanding of specific clinical pictures represent today a solid basis for a new and challenging field of functional surgery of the lower cranial nerves. PATIENTS AND METHODS Two hundred twenty consecutive patients were
examined, operated on, and followed up by the author. This figure includes 42 patients operated on elsewhere using different methods, including avulsions or neurotomies of singular branches. None had a previous MVD. Included also are 11 patients with tumors, if the only complaint was facial pain, regardless of whether the tumor was diagnosed during the diagnostic workup or found accidentally at operation. The male/female ratio was 94 to 126. The age distribution is shown in Table 1. The oldest patient was 84 years, and the youngest, 22 years. There were 124 patients with right-sided pain, 92 with left-sided pain, and 4 with bilateral pain. Excluded were patients with atypical pain. Nevertheless, 40 patients (8%) showed some symptoms, including permanent pain of different intensity, unusual triggers, or pronounced vegetative signs. Three patients had typical trigeminal pain and also typical cluster headaches, a dominant complaint in 2 of them. Three patients had multiple sclerosis. A total of 227 operations were performed including three that were repeated operations during the first hospital stay. Four patients with bilateral pain were operated on in two stages. Operations for late recurrences are not included in these figures. Late recurrences were classified as all patients with recurrent lasting pain after the hospital discharge. The diagnostic work-up consisted of an accurate history, and, since 1980, a computed tomographic scan. Magnetic resonance imaging was not used for technical reasons. Angiography was employed only in cases of suspected vascular pathology and, in the beginning of the series, in 2 patients with cluster headache. All patients but 2 were operated on in the lateral decubitus position. These were operated on in the sitting and supine positions, respectively. A small craniectomy, about 2 cm in diameter, just large enough to allow the introduction of a small retractor, was sufficient in almost all cases. The petrous vein was coagulated and cut, if in the field of view, and the arachnoid covering of the Vth nerve was opened. The nerve was inspected for vascular compression and arteries, if found, were separated from the nerve by a Teflon felt prosthesis. The muscle was used only once. In 3 patients the artery was suspended to the tentorium, and in 1, a sclerotic vertebral artery was separated from the nerve by an aneurysmal cuff clip. The veins, if the cause of the compression, were coagulated and cut. Glue was not used to fix the Teflon. In patients having no clear evidence of vascular compression, partial sensory rhizotomy (PSR) was performed. This consisted of sectioning one-third or less of the portio major, taking into consideration the expected branch distribution. The section was performed as close as possible to the brain stem. Patients were asked to give their consent for PSR in case no compression was found. Corticosteroids and antibiotics were never used before or during the operation, except in tumors; corticosteroids were occasionally administered postoperatively. Intravenous fluids, from 2500 to 3000 ml, were
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Neurosurgery 1992-98 January 1992, Volume 30, Number 1 49 Microvascular Decompression and Partial Sensory Rhizotomy in the Treatment of Trigeminal Neuralgia: Personal Experience with 220 Patients Experimental and Clinical Study
RESULTS The results are summarized in Table 2. The followup time was from 12 years to 6 months, with a mean of 5.2 years. The outcome was not influenced by sex, age, or the duration of symptoms, but a considerable difference existed between the group with vascular compression and the group who underwent partial rhizotomy. The immediate result was a complete relief of 94% in total, 96% in the MVD group, and 86% in the PSR group. After 5 years, however, this figure dropped to 84% in total, but the recurrence rate in both groups was quite different. In the MVD group, there were only 5 recurrences and 6 patients having some minor residual pain (6%), in contrast to the PSR group with 10 failures and 12 patients with partial relief, an almost 50% recurrence rate. Patients with previous destructive procedures showed poorer results. Twenty-seven of them (64%) experienced some lasting burning pain or paresthesias of different intensity in the affected branch. In the 3 patients with multiple sclerosis, no vascular compression was found, and a PSR was performed in each case. It was thought that an intrinsic factor was responsible for the pain, because all patients with multiple sclerosis and symptomatic trigeminal neuralgia were treated by glycerol chemoneurolysis. Two patients with MVD continued to have occasional pain, which resolved within 2 and 3 weeks, respectively. One patient developed a contralateral pain during the hospital stay, which she never experienced before. This pain progressed to a severe neuralgia, which necessitated a MVD 6 months later. Two patients experienced a recurrence of an epidermoid tumor, and one a recurrence of an arachnoid cyst. The tumors were operated on again and the cyst treated by a cystotransversal sinus shunt. Ten patients had a second operation; 7 are free of pain. Three have some residual pain that is controlled by medication. The experience of this surgeon seems to reduce the number of negative findings considerably. During the consecutive 5-year period, the corresponding figures for negative findings were: 1975 to 1979, 14%; 1980 to 1984, 7%; and 1985 to 1989, 3%. Complications Almost all patients experienced headache during the first 2 days. The majority complained of nausea or vomiting for about 12 hours. Elderly patients generally had less severe headache than younger ones. Labial herpes is not always mentioned in the records, but occurred in about one-half of the patients with MVD and in nearly all PSR patients. It seems that the appearance and severity of herpes correlates with the degree of the manipulation of the nerve. One patient developed short, transient weakness of
the VIIth nerve, 1 (with an epidermoid tumor) developed hearing loss, and 9 had temporary hearing impairment. The latter occurred in the early years, when mastoid cells were opened, resulting in hematotympanum with the resulting conductive hearing impairment lasting for some weeks. Six patients had transient diplopia. In 1 it lasted 3 months, whereas in the rest it resolved within 3 to 8 weeks. This complication occurred at the beginning of the series and was always caused by inadvertent touching of the trochlear nerve by the suction tip. There was no similar complication during the past 6 years. Sensory loss of various degrees was present in all patients who underwent PSR, but none of them lost the corneal reflex. This was, however, the case in one patient in the MVD group. Minor sensory deficits were seen occasionally in the MVD group. The degree of sensory loss in the PSR group varied widely, even in the cases in which it was thought that similar lesions were performed. Some patients showed a specific kind of dissociated sensory loss, similar to the one occurring after spinal chordotomy. Painful anesthesia or anesthesia dolorosa were never seen in either group. Two patients had a lesion of the Vth nerve motor branch, which in one was permanent (Table 3). There was not a single case of postoperative cerebrospinal fluid leak, otherwise not rare in the posterior fossa operations performed by the same surgeon. A small dural flap and slack dura at the time of closure could be an explanation. Three patients died. One had an extensive cerebellar angioma and developed a cerebellar hemorrhage, one had a cerebellar infarction, and the third had a massive intraoperative air embolism. This was the patient operated on in the sitting position. Treatment of recurrences With some exceptions, recurrences were first treated medically. A few patients refused to take medication and required immediate treatment. In the first years, a policy of repeated exploration was pursued. The rationale for this was that some vascular compressions may have been overlooked because of a lack of experience. Later, we found that a second operation seldom revealed unexpected findings and generally resulted in a PSR. Ten patients underwent a second operation and the rest were treated by thermocoagulation during the first 3 years, and after that by retro-Gasserian glycerol injection. In 3 patients operated on again, a vessel missed at the first procedure was found, and in one (the only patient in whom a piece of muscle was used) we found a resorbed prosthesis with severe adhesions. A slipped Teflon felt was never seen. All patients undergoing a second operation with no signs of vascular compression or expansive lesions underwent PSR. In the past 4 years, glycerol chemoneurolysis has been the treatment of choice in most recurrences; however, the increasing recurrence rate in these patients necessitates a review of this policy. DISCUSSION
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
administered the first day, and analgesics were given liberally during the first 2 days. Most patients were discharged by the 7th postoperative day.
Three patients had a combination of cluster headache and paroxysmal pain, triggered by rather typical trigeminal triggers. Vascular compression was found in all of them. They were relieved of cluster headache as well as of paroxysmal pain. One patient developed a contralateral neuralgia of the intermediate nerve a few years after a successful operation for trigeminal neuralgia, again successfully treated by MVD. There is no explanation for the low incidence of bilateral pain in the series. The age of the patients seems to be an important factor in the decision for treatment, as the majority of the patients belong to the advanced age group. No strict rules were set; each patient was evaluated individually. With a few exceptions, the age of 80 was considered an arbitrary limit. Interestingly, elderly patients showed the same percentage of complications as the younger ones; in general, they tolerated the procedure very well, had fewer postoperative headaches, and walked as early as the younger patients. One of the three deceased, however, was over the age of 80. Most complications were transitory. The permanent loss of hearing, which occurred once, is probably the only complication that cannot be prevented in all cases. This complication may be minimized with the routine recording of auditory potentials. In our series, it occurred in one patient operated on for an epidermoid tumor. The statoacoustic nerve was neither manipulated nor touched during the procedure. Obviously, the extreme vulnerability of the tiny cochlear endings remains the main concern in surgery of the facial pain, as well as with hemispasm. Temporary hearing impairment caused by hematotympanum is avoidable. The pattern of the compressing arteries was similar to that in other series (10,12) (Table 4). By far the most common vessel was one branch, rarely two, and seldom the stem, of the superior cerebellar artery. A considerable number of vessels remained unidentified, some because of technical difficulties, or because it was thought that the procedure should not be extended just for this purpose. A certain number of these vessels could be attributed, however, to large pontine branches shown in the anatomical studies by Klun and Prestor (9). In three patients with tumors, pain was obviously caused by a vessel pushed toward the nerve. At one time, atypical pain was suspected to be caused by compression located distally, whereas compressions at the entry zone produced typical pain. Lack of precise records taking this point into consideration leaves this question unanswered. The striking difference in the recurrence rate between the MVD and PSR groups is most probably a result of an overly cautious division of the nerve. An ideal ratio between an acceptable sensory loss and the prevention of recurrence is not easily determined. Our results are in contrast to the report by Adams et al. (2), who found no increase in incidence of vascular abnormalities later in the series, suggesting that the low rate of vascular compressions in their series could not be contributed to the experience of the surgeon. Our figures show just the opposite. The
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
The key to efficient treatment of patients with trigeminal neuralgia is to establish whether or not the pain is indeed trigeminal (3). In our practice, less than one-half of the patients referred with trigeminal neuralgia suffered from this condition. Problems neither clinically nor anatomically related to the trigeminal nerve are occasionally treated as trigeminal neuralgia, some times for years. The cause of trigeminal neuralgia is still controversial. Although a large percentage of patients show a vascular compression, some lesions are probably intrinsic; multiple sclerosis may be an example. Some questions remain unanswered. Asymptomatic vascular compressions were found in autopsies. On the other hand, in a small percentage of patients suffering from trigeminal neuralgia, no vascular contact was seen at operation (8). The question was therefore raised as to whether the manipulation itself, and traumatization of the nerve, may explain the benefit of the decompressive procedure (1). This interpretation seems highly unlikely. Manipulation of the nerve, even if rough, will seldom yield long-term benefits. Conversely, an offending vessel can occasionally be dissected without touching the nerve, with the same good results as in instances in which the nerve was manipulated. It is difficult to explain why the operation for trigeminal neuralgia leads to immediate relief of pain, in contrast to surgery for hemifacial spasm, which has essentially the same cause and is treated in the same way and yet in which the spasm may persist for several weeks. The high percentage of early good results declined to 84% during the follow-up period. The majority of recurrences were recorded in the group undergoing partial rhizotomy, most of them in the first year, whereas in the MVD group recurrences started in the second year and were evenly distributed in the following years. The total recurrence rate in the MVD group is only 6% in contrast to the PSR group, in which it reaches almost 50%, obviously as a result of too cautious a section of the nerve. There were no differences in the outcome considering sex and age. In contrast with similar series, the duration of symptoms did not influence the prognosis (4). The severity of compression played a significant role. Thirty-six patients with severe arterial compression did better than those with a mild one, and patients with arterial compression did better than those with a venous one (11). There were no recurrences in this subgroup, in contrast to two recurrences in the group of arterial compressions of milder degree, and two in the group of venous compressions. Patients subjected to previous destructive procedures had a less satisfactory outcome. The paroxysmal pain was abolished, but the burning, dysesthetic, permanent type of pain persisted in almost one-half of the patients. This kind of pain is difficult to control by medication. While some patients learned to cope, for some it remained a problem.
CONCLUSIONS The miserable existence to which patients with trigeminal neuralgia were condemned some decades ago, either because of unsuccessful medical treatment or because of destructive operations, seems to belong to the past. MVD represents a reasonable method with low morbidity and mortality and, above all, the recurrence rate is lower than in any alternative method. Unlike other methods that treat symptoms by damaging the nerve, it is not destructive and in most cases, etiological. Anesthesia dolorosa, the horrible complication, which can be appreciated only by patients and those who have treated them, is completely eliminated, which is not the case with the available destructive procedures. MVD, however, seems to give satisfactory results only when performed by an experienced surgeon. The success rate increases in parallel with the experience accumulated. An occasional operator may therefore prefer to use an alternative, technically easier method. As in similar procedures, the best method still remains the one the surgeon is most familiar with. Received for publication, March 26, 1991; accepted, final form, August 2, 1991. Reprint requests: Prof. Dr. Boris Klun, Department of Neurosurgery, Zaloska 7, 61000 Ljubljana, Yugoslavia. REFERENCES: (1-12) 1. 2.
3. 4.
5. 6. 7.
Adams CBT: Microvascular compression: An alternative view and hypothesis. J Neurosurg 57:1-12, 1989. Adams CBT, Kaye AH, Teddy PJ: The treatment of trigeminal neuralgia by posterior fossa microsurgery. J Neurol Neurosurg Psychiatry 45:1020-1026, 1982. Apfelbaum RI: Surgery for tic douloureux. Clin Neurosurg 31:351-368, 1983. Bederson JB, Wilson CB: Evaluation of microvascular decompression and partial sensory rhizotomy in 252 cases of trigeminal neuralgia. J Neurosurg 71:359-367, 1989. Dandy WE: Concerning the cause of trigeminal neuralgia. Am J Surg 24:447-455, 1934. Gardner WJ: Concerning the cause of trigeminal neuralgia and hemifacial spasm. J Neurosurg 19:947-958, 1962. Jannetta PJ: Microsurgical approach to the trigeminal nerve for tic douloureux. Prog Neurol Surg 7:180-200, 1976.
8.
9. 10.
11.
12.
Jannetta PJ: Trigeminal neuralgia: Treatment by microvascular decompression, in Wilkins RG, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, vol 3, pp 23572362. Klun B, Prestor B: Microvascular relations of the trigeminal nerve: An anatomical study. Neurosurgery 19:535-539, 1986. Piatt JH, Wilkins RH: Treatment of tic douloureux and hemifacial spasm by posterior fossa exploration: Therapeutic implications of various neurovascular relationships. Neurosurgery 14:462-471, 1984. Szapiro J Jr, Sindou M, Szapiro J: Prognostic factors in microvascular decompression for trigeminal neuralgia. Neurosurgery 17:920929, 1985. Zorman G, Wilson CB: Outcome following microsurgical vascular decompression or partial sensory rhizotomy in 125 cases of trigeminal neuralgia. Neurology 34:13621365, 1984.
COMMENTS The author deserves credit for carefully following and documenting a large series of patients who have been followed for a substantial period of time after their treatment for trigeminal neuralgia. I generally agree with the author's conclusions as to the efficacy of the procedures and the good long-term success. I would like to point out that, at autopsy, multiple sclerosis patients have plaques at the root entry zone of the nerve, the same site that is the source of compression in those who are not suffering from demyelinating disease. Thus, the site of pathology is the same, and the only difference is whether the etiology of this is extrinsic or intrinsic to the nerve. Obviously, the latter cannot be cured by an operation such as microvascular decompression. We therefore, as Dr. Klun has suggested, use percutaneous chemoneurolysis with glycerol to treat these patients and are pleased with the success rate of this procedure. It is not necessary, as Dr. Klun has said, to manipulate the trigeminal nerve, and indeed we have tried to do as "pure" an operation as possible, minimizing manipulation of the nerve as much as possible. In most cases, this means that the nerve itself does not need to be manipulated at all. As a result, we almost never see any sensory disturbances in the trigeminal nerve and yet have comparable success rates. Most of our patients have been operated in the sitting position with proper monitoring. This has proven to be quite safe and eliminates the need for suction, which can traumatize the trigeminal nerve or adjacent structures, as the author has noted occurred in one of his patients with IVth nerve palsy. Our initial success rate in over 400 patients treated with microvascular decompression over 15 years has been about 96% with long-term success at about 90%. An additional 10 to 12% of patients will have some recurrence of pain but at a much less severe
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
percentage of negative findings during the consecutive 5-year periods dropped from 14% to 3% within 15 years. The mortality was higher than in comparable series. One patient died after a technically demanding operation involving an extensive angioma, which represented a considerable risk per se; one died of air embolism before the era of standard protective measures. A recent patient developed a cerebellar infarction after an uneventful procedure, and thus the concern remains.
level than before, and it usually can be controlled by previously ineffective medical therapy. We therefore consider them improved but not cured. This is in agreement with longer follow-up on Jannetta's patients as well. The posterior rhizotomy at the brain stem level is indeed an effective alternative for treating the tic pain. As Dr. Klun has said, it does not alter the dysesthetic pain produced by more peripheral denervation procedures and it is difficult at times to be sure as to the extent of denervation necessary. In some of the few patients on whom we have performed this, therefore, we have had to supplement it with an additional percutaneous destructive procedure to get complete relief. Our experience with a small number of patients has been, however, that once relief is achieved with a posterior rhizotomy, they tend not to have frequent recurrences, but our experience is more limited than the author's.
The author has reviewed the operative findings and results of 220 consecutive patients treated surgically for trigeminal neuralgia. Nineteen percent of the patients had a negative exploration. Although the initial pain relief was excellent in both groups, the recurrence rate was much higher in those patients treated with partial rhizotomies. They had a 5% incidence of cranial nerve morbidity and 1% mortality. Although their incidence of negative explorations is somewhat higher than has been found in Jannetta's series (1%), their results do support the concept that vascular compression is the cause of trigeminal neuralgia in most cases, and that microvascular decompression will result in long-term relief of the patient's facial pain. The 50% incidence of labial herpes after microvascular decompression and 100% incidence of labial herpes after partial rhizotomy is of interest. Pazin et al. (1) have reported a 38% incidence of labial herpes after microvascular decompression. Paul B. Nelson Pittsburgh, Pennsylvania REFERENCES: (1) 1.
Pazin GJ, Ho M, Jannetta, PJ: Reactivation of herpes simplex virus after decompression of the trigeminal nerve root. J Infect Dis 138:405409, 1978.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Ronald I. Apfelbaum Salt Lake City, Utah
Table 1. Age at Operation
Table 3. Complications
Table 4. Compression Patterns
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Table 2. Operative Results
examined, operated on, and followed up by the author. This figure includes 42 patients operated on elsewhere using different methods, including avulsions or neurotomies of singular branches. None had a previous MVD. Included also are 11 patients with tumors, if the only complaint was facial pain, regardless of whether the tumor was diagnosed during the diagnostic workup or found accidentally at operation. The male/female ratio was 94 to 126. The age distribution is shown in Table 1. The oldest patient was 84 years, and the youngest, 22 years. There were 124 patients with right-sided pain, 92 with left-sided pain, and 4 with bilateral pain. Excluded were patients with atypical pain. Nevertheless, 40 patients (8%) showed some symptoms, including permanent pain of different intensity, unusual triggers, or pronounced vegetative signs. Three patients had typical trigeminal pain and also typical cluster headaches, a dominant complaint in 2 of them. Three patients had multiple sclerosis. A total of 227 operations were performed including three that were repeated operations during the first hospital stay. Four patients with bilateral pain were operated on in two stages. Operations for late recurrences are not included in these figures. Late recurrences were classified as all patients with recurrent lasting pain after the hospital discharge. The diagnostic work-up consisted of an accurate history, and, since 1980, a computed tomographic scan. Magnetic resonance imaging was not used for technical reasons. Angiography was employed only in cases of suspected vascular pathology and, in the beginning of the series, in 2 patients with cluster headache. All patients but 2 were operated on in the lateral decubitus position. These were operated on in the sitting and supine positions, respectively. A small craniectomy, about 2 cm in diameter, just large enough to allow the introduction of a small retractor, was sufficient in almost all cases. The petrous vein was coagulated and cut, if in the field of view, and the arachnoid covering of the Vth nerve was opened. The nerve was inspected for vascular compression and arteries, if found, were separated from the nerve by a Teflon felt prosthesis. The muscle was used only once. In 3 patients the artery was suspended to the tentorium, and in 1, a sclerotic vertebral artery was separated from the nerve by an aneurysmal cuff clip. The veins, if the cause of the compression, were coagulated and cut. Glue was not used to fix the Teflon. In patients having no clear evidence of vascular compression, partial sensory rhizotomy (PSR) was performed. This consisted of sectioning one-third or less of the portio major, taking into consideration the expected branch distribution. The section was performed as close as possible to the brain stem. Patients were asked to give their consent for PSR in case no compression was found. Corticosteroids and antibiotics were never used before or during the operation, except in tumors; corticosteroids were occasionally administered postoperatively. Intravenous fluids, from 2500 to 3000 ml, were
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Neurosurgery 1992-98 January 1992, Volume 30, Number 1 49 Microvascular Decompression and Partial Sensory Rhizotomy in the Treatment of Trigeminal Neuralgia: Personal Experience with 220 Patients Experimental and Clinical Study
RESULTS The results are summarized in Table 2. The followup time was from 12 years to 6 months, with a mean of 5.2 years. The outcome was not influenced by sex, age, or the duration of symptoms, but a considerable difference existed between the group with vascular compression and the group who underwent partial rhizotomy. The immediate result was a complete relief of 94% in total, 96% in the MVD group, and 86% in the PSR group. After 5 years, however, this figure dropped to 84% in total, but the recurrence rate in both groups was quite different. In the MVD group, there were only 5 recurrences and 6 patients having some minor residual pain (6%), in contrast to the PSR group with 10 failures and 12 patients with partial relief, an almost 50% recurrence rate. Patients with previous destructive procedures showed poorer results. Twenty-seven of them (64%) experienced some lasting burning pain or paresthesias of different intensity in the affected branch. In the 3 patients with multiple sclerosis, no vascular compression was found, and a PSR was performed in each case. It was thought that an intrinsic factor was responsible for the pain, because all patients with multiple sclerosis and symptomatic trigeminal neuralgia were treated by glycerol chemoneurolysis. Two patients with MVD continued to have occasional pain, which resolved within 2 and 3 weeks, respectively. One patient developed a contralateral pain during the hospital stay, which she never experienced before. This pain progressed to a severe neuralgia, which necessitated a MVD 6 months later. Two patients experienced a recurrence of an epidermoid tumor, and one a recurrence of an arachnoid cyst. The tumors were operated on again and the cyst treated by a cystotransversal sinus shunt. Ten patients had a second operation; 7 are free of pain. Three have some residual pain that is controlled by medication. The experience of this surgeon seems to reduce the number of negative findings considerably. During the consecutive 5-year period, the corresponding figures for negative findings were: 1975 to 1979, 14%; 1980 to 1984, 7%; and 1985 to 1989, 3%. Complications Almost all patients experienced headache during the first 2 days. The majority complained of nausea or vomiting for about 12 hours. Elderly patients generally had less severe headache than younger ones. Labial herpes is not always mentioned in the records, but occurred in about one-half of the patients with MVD and in nearly all PSR patients. It seems that the appearance and severity of herpes correlates with the degree of the manipulation of the nerve. One patient developed short, transient weakness of
the VIIth nerve, 1 (with an epidermoid tumor) developed hearing loss, and 9 had temporary hearing impairment. The latter occurred in the early years, when mastoid cells were opened, resulting in hematotympanum with the resulting conductive hearing impairment lasting for some weeks. Six patients had transient diplopia. In 1 it lasted 3 months, whereas in the rest it resolved within 3 to 8 weeks. This complication occurred at the beginning of the series and was always caused by inadvertent touching of the trochlear nerve by the suction tip. There was no similar complication during the past 6 years. Sensory loss of various degrees was present in all patients who underwent PSR, but none of them lost the corneal reflex. This was, however, the case in one patient in the MVD group. Minor sensory deficits were seen occasionally in the MVD group. The degree of sensory loss in the PSR group varied widely, even in the cases in which it was thought that similar lesions were performed. Some patients showed a specific kind of dissociated sensory loss, similar to the one occurring after spinal chordotomy. Painful anesthesia or anesthesia dolorosa were never seen in either group. Two patients had a lesion of the Vth nerve motor branch, which in one was permanent (Table 3). There was not a single case of postoperative cerebrospinal fluid leak, otherwise not rare in the posterior fossa operations performed by the same surgeon. A small dural flap and slack dura at the time of closure could be an explanation. Three patients died. One had an extensive cerebellar angioma and developed a cerebellar hemorrhage, one had a cerebellar infarction, and the third had a massive intraoperative air embolism. This was the patient operated on in the sitting position. Treatment of recurrences With some exceptions, recurrences were first treated medically. A few patients refused to take medication and required immediate treatment. In the first years, a policy of repeated exploration was pursued. The rationale for this was that some vascular compressions may have been overlooked because of a lack of experience. Later, we found that a second operation seldom revealed unexpected findings and generally resulted in a PSR. Ten patients underwent a second operation and the rest were treated by thermocoagulation during the first 3 years, and after that by retro-Gasserian glycerol injection. In 3 patients operated on again, a vessel missed at the first procedure was found, and in one (the only patient in whom a piece of muscle was used) we found a resorbed prosthesis with severe adhesions. A slipped Teflon felt was never seen. All patients undergoing a second operation with no signs of vascular compression or expansive lesions underwent PSR. In the past 4 years, glycerol chemoneurolysis has been the treatment of choice in most recurrences; however, the increasing recurrence rate in these patients necessitates a review of this policy. DISCUSSION
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
administered the first day, and analgesics were given liberally during the first 2 days. Most patients were discharged by the 7th postoperative day.
Three patients had a combination of cluster headache and paroxysmal pain, triggered by rather typical trigeminal triggers. Vascular compression was found in all of them. They were relieved of cluster headache as well as of paroxysmal pain. One patient developed a contralateral neuralgia of the intermediate nerve a few years after a successful operation for trigeminal neuralgia, again successfully treated by MVD. There is no explanation for the low incidence of bilateral pain in the series. The age of the patients seems to be an important factor in the decision for treatment, as the majority of the patients belong to the advanced age group. No strict rules were set; each patient was evaluated individually. With a few exceptions, the age of 80 was considered an arbitrary limit. Interestingly, elderly patients showed the same percentage of complications as the younger ones; in general, they tolerated the procedure very well, had fewer postoperative headaches, and walked as early as the younger patients. One of the three deceased, however, was over the age of 80. Most complications were transitory. The permanent loss of hearing, which occurred once, is probably the only complication that cannot be prevented in all cases. This complication may be minimized with the routine recording of auditory potentials. In our series, it occurred in one patient operated on for an epidermoid tumor. The statoacoustic nerve was neither manipulated nor touched during the procedure. Obviously, the extreme vulnerability of the tiny cochlear endings remains the main concern in surgery of the facial pain, as well as with hemispasm. Temporary hearing impairment caused by hematotympanum is avoidable. The pattern of the compressing arteries was similar to that in other series (10,12) (Table 4). By far the most common vessel was one branch, rarely two, and seldom the stem, of the superior cerebellar artery. A considerable number of vessels remained unidentified, some because of technical difficulties, or because it was thought that the procedure should not be extended just for this purpose. A certain number of these vessels could be attributed, however, to large pontine branches shown in the anatomical studies by Klun and Prestor (9). In three patients with tumors, pain was obviously caused by a vessel pushed toward the nerve. At one time, atypical pain was suspected to be caused by compression located distally, whereas compressions at the entry zone produced typical pain. Lack of precise records taking this point into consideration leaves this question unanswered. The striking difference in the recurrence rate between the MVD and PSR groups is most probably a result of an overly cautious division of the nerve. An ideal ratio between an acceptable sensory loss and the prevention of recurrence is not easily determined. Our results are in contrast to the report by Adams et al. (2), who found no increase in incidence of vascular abnormalities later in the series, suggesting that the low rate of vascular compressions in their series could not be contributed to the experience of the surgeon. Our figures show just the opposite. The
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
The key to efficient treatment of patients with trigeminal neuralgia is to establish whether or not the pain is indeed trigeminal (3). In our practice, less than one-half of the patients referred with trigeminal neuralgia suffered from this condition. Problems neither clinically nor anatomically related to the trigeminal nerve are occasionally treated as trigeminal neuralgia, some times for years. The cause of trigeminal neuralgia is still controversial. Although a large percentage of patients show a vascular compression, some lesions are probably intrinsic; multiple sclerosis may be an example. Some questions remain unanswered. Asymptomatic vascular compressions were found in autopsies. On the other hand, in a small percentage of patients suffering from trigeminal neuralgia, no vascular contact was seen at operation (8). The question was therefore raised as to whether the manipulation itself, and traumatization of the nerve, may explain the benefit of the decompressive procedure (1). This interpretation seems highly unlikely. Manipulation of the nerve, even if rough, will seldom yield long-term benefits. Conversely, an offending vessel can occasionally be dissected without touching the nerve, with the same good results as in instances in which the nerve was manipulated. It is difficult to explain why the operation for trigeminal neuralgia leads to immediate relief of pain, in contrast to surgery for hemifacial spasm, which has essentially the same cause and is treated in the same way and yet in which the spasm may persist for several weeks. The high percentage of early good results declined to 84% during the follow-up period. The majority of recurrences were recorded in the group undergoing partial rhizotomy, most of them in the first year, whereas in the MVD group recurrences started in the second year and were evenly distributed in the following years. The total recurrence rate in the MVD group is only 6% in contrast to the PSR group, in which it reaches almost 50%, obviously as a result of too cautious a section of the nerve. There were no differences in the outcome considering sex and age. In contrast with similar series, the duration of symptoms did not influence the prognosis (4). The severity of compression played a significant role. Thirty-six patients with severe arterial compression did better than those with a mild one, and patients with arterial compression did better than those with a venous one (11). There were no recurrences in this subgroup, in contrast to two recurrences in the group of arterial compressions of milder degree, and two in the group of venous compressions. Patients subjected to previous destructive procedures had a less satisfactory outcome. The paroxysmal pain was abolished, but the burning, dysesthetic, permanent type of pain persisted in almost one-half of the patients. This kind of pain is difficult to control by medication. While some patients learned to cope, for some it remained a problem.
CONCLUSIONS The miserable existence to which patients with trigeminal neuralgia were condemned some decades ago, either because of unsuccessful medical treatment or because of destructive operations, seems to belong to the past. MVD represents a reasonable method with low morbidity and mortality and, above all, the recurrence rate is lower than in any alternative method. Unlike other methods that treat symptoms by damaging the nerve, it is not destructive and in most cases, etiological. Anesthesia dolorosa, the horrible complication, which can be appreciated only by patients and those who have treated them, is completely eliminated, which is not the case with the available destructive procedures. MVD, however, seems to give satisfactory results only when performed by an experienced surgeon. The success rate increases in parallel with the experience accumulated. An occasional operator may therefore prefer to use an alternative, technically easier method. As in similar procedures, the best method still remains the one the surgeon is most familiar with. Received for publication, March 26, 1991; accepted, final form, August 2, 1991. Reprint requests: Prof. Dr. Boris Klun, Department of Neurosurgery, Zaloska 7, 61000 Ljubljana, Yugoslavia. REFERENCES: (1-12) 1. 2.
3. 4.
5. 6. 7.
Adams CBT: Microvascular compression: An alternative view and hypothesis. J Neurosurg 57:1-12, 1989. Adams CBT, Kaye AH, Teddy PJ: The treatment of trigeminal neuralgia by posterior fossa microsurgery. J Neurol Neurosurg Psychiatry 45:1020-1026, 1982. Apfelbaum RI: Surgery for tic douloureux. Clin Neurosurg 31:351-368, 1983. Bederson JB, Wilson CB: Evaluation of microvascular decompression and partial sensory rhizotomy in 252 cases of trigeminal neuralgia. J Neurosurg 71:359-367, 1989. Dandy WE: Concerning the cause of trigeminal neuralgia. Am J Surg 24:447-455, 1934. Gardner WJ: Concerning the cause of trigeminal neuralgia and hemifacial spasm. J Neurosurg 19:947-958, 1962. Jannetta PJ: Microsurgical approach to the trigeminal nerve for tic douloureux. Prog Neurol Surg 7:180-200, 1976.
8.
9. 10.
11.
12.
Jannetta PJ: Trigeminal neuralgia: Treatment by microvascular decompression, in Wilkins RG, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, vol 3, pp 23572362. Klun B, Prestor B: Microvascular relations of the trigeminal nerve: An anatomical study. Neurosurgery 19:535-539, 1986. Piatt JH, Wilkins RH: Treatment of tic douloureux and hemifacial spasm by posterior fossa exploration: Therapeutic implications of various neurovascular relationships. Neurosurgery 14:462-471, 1984. Szapiro J Jr, Sindou M, Szapiro J: Prognostic factors in microvascular decompression for trigeminal neuralgia. Neurosurgery 17:920929, 1985. Zorman G, Wilson CB: Outcome following microsurgical vascular decompression or partial sensory rhizotomy in 125 cases of trigeminal neuralgia. Neurology 34:13621365, 1984.
COMMENTS The author deserves credit for carefully following and documenting a large series of patients who have been followed for a substantial period of time after their treatment for trigeminal neuralgia. I generally agree with the author's conclusions as to the efficacy of the procedures and the good long-term success. I would like to point out that, at autopsy, multiple sclerosis patients have plaques at the root entry zone of the nerve, the same site that is the source of compression in those who are not suffering from demyelinating disease. Thus, the site of pathology is the same, and the only difference is whether the etiology of this is extrinsic or intrinsic to the nerve. Obviously, the latter cannot be cured by an operation such as microvascular decompression. We therefore, as Dr. Klun has suggested, use percutaneous chemoneurolysis with glycerol to treat these patients and are pleased with the success rate of this procedure. It is not necessary, as Dr. Klun has said, to manipulate the trigeminal nerve, and indeed we have tried to do as "pure" an operation as possible, minimizing manipulation of the nerve as much as possible. In most cases, this means that the nerve itself does not need to be manipulated at all. As a result, we almost never see any sensory disturbances in the trigeminal nerve and yet have comparable success rates. Most of our patients have been operated in the sitting position with proper monitoring. This has proven to be quite safe and eliminates the need for suction, which can traumatize the trigeminal nerve or adjacent structures, as the author has noted occurred in one of his patients with IVth nerve palsy. Our initial success rate in over 400 patients treated with microvascular decompression over 15 years has been about 96% with long-term success at about 90%. An additional 10 to 12% of patients will have some recurrence of pain but at a much less severe
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
percentage of negative findings during the consecutive 5-year periods dropped from 14% to 3% within 15 years. The mortality was higher than in comparable series. One patient died after a technically demanding operation involving an extensive angioma, which represented a considerable risk per se; one died of air embolism before the era of standard protective measures. A recent patient developed a cerebellar infarction after an uneventful procedure, and thus the concern remains.
level than before, and it usually can be controlled by previously ineffective medical therapy. We therefore consider them improved but not cured. This is in agreement with longer follow-up on Jannetta's patients as well. The posterior rhizotomy at the brain stem level is indeed an effective alternative for treating the tic pain. As Dr. Klun has said, it does not alter the dysesthetic pain produced by more peripheral denervation procedures and it is difficult at times to be sure as to the extent of denervation necessary. In some of the few patients on whom we have performed this, therefore, we have had to supplement it with an additional percutaneous destructive procedure to get complete relief. Our experience with a small number of patients has been, however, that once relief is achieved with a posterior rhizotomy, they tend not to have frequent recurrences, but our experience is more limited than the author's.
The author has reviewed the operative findings and results of 220 consecutive patients treated surgically for trigeminal neuralgia. Nineteen percent of the patients had a negative exploration. Although the initial pain relief was excellent in both groups, the recurrence rate was much higher in those patients treated with partial rhizotomies. They had a 5% incidence of cranial nerve morbidity and 1% mortality. Although their incidence of negative explorations is somewhat higher than has been found in Jannetta's series (1%), their results do support the concept that vascular compression is the cause of trigeminal neuralgia in most cases, and that microvascular decompression will result in long-term relief of the patient's facial pain. The 50% incidence of labial herpes after microvascular decompression and 100% incidence of labial herpes after partial rhizotomy is of interest. Pazin et al. (1) have reported a 38% incidence of labial herpes after microvascular decompression. Paul B. Nelson Pittsburgh, Pennsylvania REFERENCES: (1) 1.
Pazin GJ, Ho M, Jannetta, PJ: Reactivation of herpes simplex virus after decompression of the trigeminal nerve root. J Infect Dis 138:405409, 1978.
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Ronald I. Apfelbaum Salt Lake City, Utah
Table 1. Age at Operation
Table 3. Complications
Table 4. Compression Patterns
Redistribution of this article permitted only in accordance with the publisher’s copyright provisions.
Table 2. Operative Results
