A 20-year review of percutaneous balloon compression of the trigeminal ganglion

Transcription

1 J Neurosurg 94: , 2001 A 20-year review of percutaneous balloon compression of the trigeminal ganglion DAVID J. SKIRVING, M.B.B.S., AND NOEL G. DAN, F.R.A.C.S. Department of Neurosurgery, Concord Repatriation General Hospital, University of Sydney, Australia Object. The aim of this study was to investigate outcomes and complication rates associated with percutaneous balloon compression (PBC) of the trigeminal ganglion over a long follow-up period. Methods. This retrospective review was conducted in 496 patients with typical symptoms of unilateral trigeminal neuralgia who underwent 531 PBCs of the trigeminal ganglion between 1980 and The mean length of follow up was 10.7 years. The treatment used was a modification of that first described by Mullan and Lichtor in There were nine technical failures. Of the 522 successful procedures, prompt pain relief ensued in all patients except one. Recurrence of pain was found in 95 patients (19.2%) within 5 years and in 158 patients (31.9%) over the entire study period. Symptomatic dysesthesias occurred in 19 patients (3.8%), but corneal anesthesia and anesthesia dolorosa did not. Conclusions. In this review the authors present data on the largest cohort of patients with the longest follow up for this procedure in the current literature. The balance between the recurrence rate and troublesome sensory complications achieved in this series is favorable when compared with previously published studies on outcomes of PBC and the two alternative percutaneous methods, radiofrequency thermocoagulation and glycerol rhizolysis. The PBC procedure has additional advantages in that it is relatively straightforward and quick, and can be performed during a brief period of general anesthesia with no discomfort to the patient. This makes it an attractive first choice in the treatment of trigeminal neuralgia. KEY WORDS trigeminal neuralgia trigeminal ganglion percutaneous surgery radiofrequency lesion glycerol rhizolysis T HERE are three well-established percutaneous procedures for the treatment of trigeminal neuralgia, but no consensus as to which is best. Each method results in a certain number of recurrences and complications. The aim is pain relief with a minimum of unwanted sensory deficits. The senior author (N.G.D.) first performed PBC of the trigeminal ganglion in June 1980, following a visit and introduction to the method by its developer, Dr. Sean Mullan, before its first published description. This paper presents the results of 531 procedures performed in 496 patients between that time and the end of This is the largest such review of this method, with the longest follow up, in the literature. Our aim was to investigate the outcome and complication rates associated with this treatment. Clinical Material and Methods Patient Population The records of 496 consecutive patients who had undergone PBC of the trigeminal ganglion for typical symptoms of unilateral trigeminal neuralgia were reviewed. These patients had been selected after medical therapy Abbreviation used in this paper: PBC = percutaneous balloon compression. had either proved ineffective or had produced complications. The duration of pain before treatment ranged from 9 weeks to 7 years. There were 279 men (56%) and 217 women (44%) whose ages ranged from 18 to 86 years. A computerized tomography or magnetic resonance imaging study was obtained before the procedure in all patients. Nineteen patients had underlying pathological conditions: 11 (2.2%) had multiple sclerosis; in one patient the pain was deemed to be posttraumatic; and in seven patients tumors were found, which were treated with open surgery. Thirteen patients had undergone previous radiofrequency rhizotomy and seven patients had undergone previous microvascular decompression. There were 146 patients with more than one division affected: in 139 patients two divisions were affected and in seven patients three divisions were affected. The third division was the most frequently affected in 482 cases, followed by the second division in 96, and the first division in 64. Description of the Procedure The method we used was a modification of that first described by Mullan and Lichtor. 19 All procedures were performed in the operating theater after a brief period of general anesthesia had been induced with laryngeal mask or endotracheal intubation. Initially, a large needle was used, such as a liver biopsy needle or a 14-gauge cannula. Later in the series, a specially made reusable needle and trocar 913

2 D. J. Skirving and N. G. Dan TABLE 1 Outcome in 496 patients who underwent PBC for trigeminal neuralgia Outcome Percentage of Patients immediate relief of pain 99 recurrence rate 5 yrs 19.2 over series (mean 10.7 yrs) 31.9 dysesthesias 3.8 anesthesia dolorosa 0 loss of corneal reflex 0 cranial nerve palsy 1.6 meningitis 0 were used. Briefly, the procedure is performed as follows: the patient is positioned supine with mild neck extension and approximately 30 of neck rotation away from the affected side. The entry point is 2 cm lateral to the angle of the mouth. The needle is passed medial to the coronoid process and aimed toward the plane intersecting a point 20 mm anterior to the external auditory meatus along the zygomatic arch, and the medial aspect of the pupil when the eye is looking forward (the Härtl landmarks). A No. 4 Fogarty balloon catheter is filled with undiluted radiographic contrast material to check the patency of the balloon. The catheter is inverted to extract the air and the stylet is replaced. The catheter is marked at the point at which 1 cm protrudes from the needle. Under fluoroscopic control, by using the C-arm in the oblique projection to visualize the foramen directly, the needle is advanced until it engages in the foramen ovale. This is usually accompanied by a short period of significant bradycardia. The Fogarty catheter is then advanced into the cave of Meckel under direct fluoroscopy. The balloon is slowly inflated to 0.75 ml by adding undiluted contrast material under fluoroscopic control. It was originally noted that ideally the balloon should be inflated until it assumes a pear shape with its stem end facing posteriorly. 19 This indicates that the balloon is starting to project out of the back of the cave toward the posterior fossa and that adequate compression is occurring. We saw this shape in most instances, but the ideal pear shape is not always attained because of the varying size of the cave. Our compression time was between 2 and 7 minutes, and we did not measure intraluminal balloon pressure. After compression, the balloon was deflated, and it and the needle were withdrawn together. Firm pressure was applied to the puncture site to prevent hematoma. The patient was usually discharged the day after the procedure. All patients were followed up by the senior author. Results The foramen ovale could not be penetrated in nine instances (1.8%). Of the successful attempts, all patients except one reported prompt relief of their pain. The recurrence rate was 19.2% (95 patients) within 5 years and 31.9% (158 patients) over the entire study period, with a mean length of follow up of 10.7 years (Table 1). The time to recurrence is shown in Table 2. The majority of recurrences were controlled with antiepileptic medications. Of TABLE 2 Time to recurrence in 158 patients treated with PBC Time to Recurrence No. of Recurrences 3 mos mos 23 1 to 2 yrs 14 2 to 5 yrs 46 5 to 17 yrs 63 those recurrences that were not controlled, 18 patients required one additional PBC, 14 required two more, and three required three more. Early facial numbness occurred in 441 patients (89%), but persisted for longer than 3 months in only 23 (4.6%). Even in these patients the numbness was well tolerated and gradually diminished. Symptomatic dysesthesias occurred in 19 patients (3.8%), and corneal anesthesia, anesthesia dolorosa, and aseptic meningitis did not occur. Seventeen patients (3.4%) had symptomatic masseter weakness, and eight patients (1.6%) had diplopia secondary to cranial nerve dysfunction. All of them recovered within 4 months. In one case the balloon was dislodged from the end of the catheter, most likely catching and then tearing on the needle. We did not attempt to remove the balloon from the retromaxillary space, and there have been no untoward symptoms as a result. In five instances there was a cheek hematoma that did not require intervention. Discussion There are three well-established percutaneous procedures for the treatment of trigeminal neuralgia, but no consensus as to which is best; each method is associated with recurrences and complications. Comparisons of the three techniques are made difficult by the following circumstances: 1) imprecise use of descriptive terms such as hypesthesia or dysesthesia; 2) differences in the size of the studies and thus the experience of the surgeon; 3) variable lengths of follow up, which can easily bias the results, especially the recurrence rates; and 4) inclusion of patients who do not have typical trigeminal neuralgic pain. 14,20 However, the number of published series in the literature to date, with the addition of our own, does allow meaningful comparisons, even though a statistical analysis is not feasible. A review of the literature reveals that the results of PBC 1,3,5,9,12 15,17,23 match those of radiofrequency lesioning 2,4,8,11,12,18,21,24,26,28 and glycerol rhizolysis, 7,10,12,16,25,29 often with fewer adverse effects. Our approach has been to offer the entire range of options, with the preference given to the simplest technique. Our study of PBC confirms the favorable outcomes reported in previously published studies, justifying the increasing popularity of the procedure. We have not compared the percutaneous techniques to microvascular decompression because this has been done before. 2,8,11,17,27,28 Of the successfully treated patients, all but one experienced immediate relief of their pain. Balloon compression has been considered to have a lower success rate than ra- 914

3 Percutaneous balloon compression of the trigeminal ganglion TABLE 3 Literature review of PBC series* Complication (%) Loss of No. of Recurrence Length of Dyses- Anesthesia Corneal CN Menin- Authors & Year Procedures Rate (%) Follow Up thesia Dolorosa Reflex Palsy gitis Belber & Rak, / mos 7 yrs 6 0 Fraioli, et al., yrs 0.6 Frank & Fabrizi, <3 yrs Lichtor & Mullan, yrs (est) 10 yrs Lobato, et al., / mos Meglio, et al., yrs 6.7 Peragut, et al., mos Abdennebi, et al., mos Brown & Gouda, / mos 10 yrs Correa & Teixeira, / yrs? *CN = cranial nerve; est = estimated; = not given;? = result given but ambiguous. Numerals shown to the right of the virgule indicate numbers of patients. Value estimated as 30% (28% between 1 and 10 years) because not all patients were followed for 10 years. diofrequency thermocoagulation for immediate relief of pain, 28 but our results indicate just as good an initial response. The recurrence rate was low. In their 10-year follow-up study of 100 patients, Lichtor and Mullan 14 reported a recurrence rate of 20% for 5 years and 28% for patients followed up for 1 to 10 years. These authors concluded that a recurrence rate of 20% over 5 years was to be expected with the PBC method, and they estimated the recurrence rate over 10 years to be close to 30%. They believed these figures could be improved with a repeated procedure and supplemental use of carbamazepine to 10% at 5 years and 20% at 10 years. We have achieved a recurrence rate of 19.2% within 5 years and 31.9% over a mean follow-up period of 10.7 years. A percutaneous procedure cannot be judged by recurrence rates alone; it must be critically assessed for its rates of symptomatic dysesthesia, anesthesia dolorosa, and corneal anesthesia. A balance must be sought, but this balance should err on the side of less being better than more. 20 Recurrence does not equate with failure; it is inevitable if intolerable adverse effects are to be minimized. Nugent 20 stated that the number of their patients who required a second radiofrequency procedure would have been lower if it had been their intent in the initial procedure to make more dense lesions. Eighty percent of their patients who underwent a second procedure were happy with the results. Our complication rate was low; the incidence of annoying dysesthesia was 3.8%. None of our patients suffered from anesthesia dolorosa or corneal anesthesia. Lichtor and Mullan 14 reported that 4% of their patients suffered from dysesthesia. However, their method changed midway through their study; in the first 60 patients the balloon compression times were 5 to 7 minutes (range minutes), whereas in the final 40 patients a standard 1-minute compression was performed because this compression time seemed to be as effective as any greater duration. All four of their patients in whom dysesthetic paresthesia developed had been treated with the longer compression times. In the last 40 patients of their series plus an additional 25 unreported cases with 1-minute compression times, there was only one case of disturbing dysesthesia. They concluded that this rate of less than 2% was a more appropriate figure. We have not correlated the rate of adverse sensory deficits with compression time. The success of subsequent procedures is another important consideration in the effectiveness of a technique. 14 It has been noted previously 14,20 and is apparent to surgeons treating this problem that a recalcitrant group exists. This group is hard to cure, no matter which method is used. In our series 18 patients required one additional balloon compression, 14 required two more, and three required three more. The latter three have achieved intermittent relief, but their pain continues to recur, and we consider their cases to be recalcitrant. This small number may be due to the fact that only 20 patients in our series had undergone unsuccessful treatment with other methods. The results of the published series for PBC are shown in Table 3. Abdennebi, et al., 1 published a series of 200 patients who had a recurrence rate of 32.5% after a mean follow-up period of 51 months. There was a 3% rate of severe dysesthesia and a 2.5% rate of diminished corneal reflex without keratitis. These authors altered their compression time partway through their series, from between 7 and 9 minutes to 6 minutes. Severe dysesthesia and hypesthesia had occurred with the initial times, but dysesthesias became exceptional with the shorter time. Brown and Gouda 5 reviewed 141 consecutive patients who underwent 182 PBCs with a mean follow up of 22 months and a recurrence rate of 26%. The rate of severe numbness or annoying dysesthesia was 3.5%. No anesthesia dolorosa or loss of the corneal reflex was reported. Correa and Teixeira 9 report a recurrence rate of 20% over 4 years. They reported no anesthesia dolorosa or corneal anesthesia despite the fact that 11% of the patients were being treated for first-division pain. One patient had permanent sixth cranial nerve paresis. These authors used balloon compression for 1 minute, as Lichtor and Mullan have advocated. Frank and Fabrizi 13 used both the radiofrequency and 915

4 D. J. Skirving and N. G. Dan the PBC method, and advocated 2- to 3-minute compressions. They reported results in 700 patients (radiofrequency) and 212 patients (PBC). The recurrence rate was 25% for both methods, within a follow-up period of 3 years. The radiofrequency group had a 0.6% incidence of anesthesia dolorosa and a 1% incidence of corneal anesthesia, whereas those treated with PBC had neither. These authors concluded that balloon compression was their method of choice for first- and second-division pain because of the absence of ocular dysfunction. They considered that a satisfactory result with the balloon method for thirddivision pain required a massive compression of more than 5 minutes... causing excessive damage to the first and second divisions. Most of our compressions were 5 minutes or longer and our results show that this did not cause excessive damage to the other nerve divisions. Lobato, et al., 15 reported a recurrence rate of 9.7%, with a follow-up period of 6 to 54 months. Marked hypesthesia occurred in six patients, and in four of them the symptoms were seriously annoying. The authors attributed their higher rate of sensory deficits to their use of more vigorous compression. However, they still had no cases of anesthesia dolorosa or corneal anesthesia. They measured balloon intraluminal pressures and achieved a strict pear shape, correlating their success in obtaining this shape with the recurrence rate. They again used a 1-minute compression time. In their discussion, they pointed out that Lichtor and Mullan had advocated inflating the balloon until it began to take on a pear shape rather than to its full size, to minimize annoying numbness. Lobato, et al., reported that when they tried this, they were only able to control the neuralgia transiently. According to these authors, inflating the balloon to a fully developed pear shape resulted in long-term control. However, the degree of compression (which they estimated from the intraluminal balloon pressure and the shape and size of the fluoroscopic image) did not correlate with sensory deficits, which they believed was due in part to individual anatomical variation. We did not monitor intraluminal balloon pressure or record the exact numbers that conformed to the pear shape. We achieved this shape in most instances, but the ideal pear shape is not always possible because of the varying size of the cave. The variable times of balloon compression among the different series demonstrates that the optimal compression time is not firmly established, but that favorable results can be achieved with a range of compression times. Meglio, et al., 17 noted that their recurrence rate depended on compression time: it was 42.8% after 2 years in those patients who had been treated with 4- to 10-minute compression times; however, for those who had undergone only 1- to 3-minute compressions the recurrence rate was 84%. These results do not agree with Lichtor and Mullan s 1-minute compression results, and the high overall recurrence rate differs from the rest of the literature. Peragut, et al., 23 reported a low recurrence rate, at the expense of high rates of dysesthesia and loss of the corneal reflex, without keratitis. There were no cases of anesthesia dolorosa. These results are not consistent with the rest of the series. Because of its consistently low rate of corneal anesthesia 3,5,9,13 15 it has been recommended by some investigators that PBC be used for isolated first-division pain. 3,12,23 Taha and Tew 28 suggested that more studies will be necessary to confirm or refute this. They recommended microvascular decompression for patients with isolated first-division pain who can tolerate a major operation but radiofrequency thermocoagulation for those who cannot. This paper attests to the validity of the earlier recommendation, as do other series in which this method has been used since that time. In recent published series numerous patients with first-division pain have been treated with the PBC method and there has been a consistent absence of corneal anesthesia. In our series, 64 patients were treated for first-division pain with no cases of corneal anesthesia. We advocate the use of PBC as the preferred treatment for isolated firstdivision pain. We believe that the PBC modality has a number of additional advantages. 1) It is performed during a brief period of general anesthesia, causing no discomfort to the patient, whereas the radiofrequency method requires an alert, cooperative patient to localize the lesioning and monitor sensory loss. Sensations experienced by patients undergoing radiofrequency lesioning range from a tolerable burning feeling to severe, even intolerable pain. 20 2) The technique is simple to perform. It demands nothing more technically difficult than placing a needle in the foramen ovale under fluoroscopic control, a skill which can be acquired rather rapidly. 14 Nugent 20 suggested that it may be the best treatment for those who do not treat many patients with trigeminal neuralgia. 3) It does not require special equipment such as a radiofrequency lesion generator, which reduces the cost. Whereas sterile glycerol may be difficult to acquire, the Fogarty balloons and needles/trocars are readily available. 4) It is quick because functional localization is not required. It has been said that a disadvantage of the procedure is the significant bradycardia, at times associated with hypotension, that is experienced when the needle engages the foramen ovale or on inflation of the balloon. 6,20 This bradycardia was noted routinely in our series. As did Mullan and Lichtor, 19 we used it to our advantage, first as an indication that the foramen ovale has been entered, and second to show that a secure compression of the balloon has occurred. Atropine was only needed on a few occasions and was delivered without incident. However, we are aware of two unreported cases in which asystole occurred. In one case the asystole ceased on removal of the needle, whereas in the other case atropine was required. On occasion (1.8%) there was a failure to penetrate the foramen ovale. The size of the needle may have contributed; it is required to be large to allow a Fogarty catheter to pass through it. These unsuccessful attempts mostly occurred in patients who were undergoing a second or subsequent procedure, and scarring of the entrance to the foramen is probably partly responsible. Conclusions It was suggested 25 years ago that the large number of choices for the treatment of trigeminal neuralgia meant that none of them were totally satisfactory. 22 As more patients have been treated with the various percutaneous methods, it has become clear that these treatments are effective at relieving trigeminal pain with few adverse effects. There is still no general agreement on the best per- 916

5 Percutaneous balloon compression of the trigeminal ganglion cutaneous method. The surgeon must be happy with the balance between recurrence of pain and the causation of sensory deficits with the chosen procedure. The balance achieved in this series is favorable when compared with previously published studies of PBC and the two alternative percutaneous methods radiofrequency thermocoagulation and glycerol rhizolysis. We emphasize that there were no cases of anesthesia dolorosa or corneal anesthesia in our large series. References 1. Abdennebi B, Mahfouf L, Nedjahi T: Long-term results of percutaneous compression of the gasserian ganglion in trigeminal neuralgia (series of 200 patients). Stereotact Funct Neurosurg 68: , Apfelbaum RI: A comparison of percutaneous radiofrequency trigeminal neurolysis and microvascular decompression of the trigeminal nerve for the treatment of tic douloureux. Neurosurgery 1:16 21, Belber CJ, Rak RA: Balloon compression rhizolysis in the surgical management of trigeminal neuralgia. Neurosurgery 20: , Broggi G, Franzini A, Lasio G, et al: Long-term results of percutaneous retrogasserian thermorhizotomy for essential trigeminal neuralgia: Considerations in 1000 consecutive patients. Neurosurgery 26: , Brown JA, Gouda JJ: Percutaneous balloon compression of the trigeminal nerve. Neurosurg Clin North Am 8:53 62, Brown JA, Preul MC: Percutaneous trigeminal ganglion compression for trigeminal neuralgia. Experience in 22 patients and review of the literature. J Neurosurg 70: , Burchiel KJ: Percutaneous retrogasserian glycerol rhizolysis in the management of trigeminal neuralgia. J Neurosurg 69: , Burchiel KJ, Steege TD, Howe JF, et al: Comparison of percutaneous radiofrequency gangliolysis and microvascular decompression for the surgical management of tic douloureux. Neurosurgery 9: , Correa CF, Teixeira MJ: Balloon compression of the Gasserian ganglion for the treatment of trigeminal neuralgia. Stereotact Funct Neurosurg 71:83 89, Dieckmann G, Bockermann V, Heyer C, et al: Five-and-ahalf years experience with percutaneous retrogasserian glycerol rhizotomy in treatment of trigeminal neuralgia. Appl Neurophysiol 50: , Ferguson GG, Peerless BSJ, Girvin JP: Trigeminal neuralgia: a comparison of the results of percutaneous rhizotomy and microvascular decompression. Can J Neurol Sci 8: , Fraioli B, Esposito V, Guidetti B, et al: Treatment of trigeminal neuralgia by thermocoagulation, glycerolization and percutaneous compression of the gasserian ganglion and/or retrogasserian rootlets: long-term results and therapeutic protocol. Neurosurgery 24: , Frank F, Fabrizi AP: Percutaneous surgical treatment of trigeminal neuralgia. Acta Neurochir 97: , Lichtor T, Mullan JF: A 10 year-follow-up review of percutaneous microcompression of the trigeminal ganglion. J Neurosurg 72:49 54, Lobato RD, Rivas JJ, Sarabia R, et al: Percutaneous microcompression of the gasserian ganglion for trigeminal neuralgia. J Neurosurg 72: , Lunsford LD: Percutaneous retrogasserian glycerol rhizotomy, in Rovit RL, Murali R, Jannetta PJ (eds): Trigeminal Neuralgia. Baltimore: Williams & Wilkins, 1990, pp Meglio M, Cioni B, Moles A, et al: Microvascular decompression versus percutaneous procedures for typical trigeminal neuralgia: personal experience. Stereotact Funct Neurosurg 54/ 55:76 79, Menzel J, Piotrowski W, Penzholz H: Long-term results of Gasserian ganglion electrocoagulation. J Neurosurg 42: , Mullan S, Lichtor T: Percutaneous microcompression of the trigeminal ganglion for trigeminal neuralgia. J Neurosurg 59: , Nugent GR: Percutaneous techniques for trigeminal neuralgia, in Kaye AH, Black P (eds): Operative Neurosurgery. London: Churchill Livingstone, 2000, pp Nugent GR: Surgical treatment: radiofrequency gangliolysis and rhizotomy, in Fromm GH, Sessle BJ (eds): Trigeminal Neuralgia: Current Concepts Regarding Pathogenesis and Treatment. London: Butterworth-Heinemann, 1991, pp Onofrio BM: Radiofrequency percutaneous Gasserian ganglion lesions. Results in 140 patients with trigeminal pain. J Neurosurg 42: , Peragut JC, Gondin-Oliveira J, Fabrizi A, et al: Micro-compression du ganglion de Gasser. Un traitment de la nevralgie faciale essentielle. A propos de 70 observations. Neurochirurgie 37: , Rovit RL: Percutaneous radiofrequency thermal coagulation of the gasserian ganglion, in Rovit RL, Murali R, Jannetta PJ (eds): Trigeminal Neuralgia. Baltimore: Williams & Wilkins, 1990, pp Saini SS: Reterogasserian anhydrous glycerol injection therapy in trigeminal neuralgia: observations in 552 patients. J Neurol Neurosurg Psychiatry 50: , Sweet WH: Complications of treating trigeminal neuralgia; an analysis of the literature and response to questionaire, in Rovit RL, Murali R, Jannetta PJ (eds): Trigeminal Neuralgia. Baltimore: Williams & Wilkins, 1990, pp Sweet WH: The treatment of trigeminal neuralgia (tic douloureux). N Engl J Med 315: , Taha JM, Tew JM Jr: Comparison of surgical treatments for trigeminal neuralgia: reevaluation of radiofrequency rhizotomy. Neurosurgery 38: , Waltz TA, Dalessio DJ, Copeland B, et al: Percutaneous injection of glycerol for the treatment of trigeminal neuralgia. Clin J Pain 5: , 1989 Manuscript received May 24, Accepted in final form January 3, Address reprint requests to: Noel Dan, F.R.A.C.S., Specialist Medical Centre, 235 New South Head Road, Edgecliff NSW 2027, Australia. davidskirving@hotmail.com. 917