become a mainstay in the management of this painful condition.

Transcription

1 J Neurosurg 96: , 2002 Predictors of outcome in surgically managed patients with typical and atypical trigeminal neuralgia: comparison of results following microvascular decompression ELIZABETH C. TYLER-KABARA, M.D., PH.D., AMIN B. KASSAM, M.D., MICHAEL H. HOROWITZ, M.D., LOUISE URGO, P.A.C., CONSTANTINOS HADJIPANAYIS, M.D., ELAD I. LEVY, M.D., AND YUE-FANG CHANG, PH.D. Departments of Neurosurgery, Radiology, and Otolaryngology, Center for Cranial Nerve Disorders, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania T Object. Microvascular decompression (MVD) has become one of the primary treatments for typical trigeminal neuralgia (TN). Not all patients with facial pain, however, suffer from the typical form of this disease; many patients who present for surgical intervention actually have atypical TN. The authors compare the results of MVD performed for typical and atypical TN at their institution. Methods. The results of 2675 MVDs in 2264 patients were reviewed using information obtained from the department database. The authors examined immediate postoperative relief in 2003 patients with typical and 672 with atypical TN, and long-term follow-up results in patients for whom more than 5 years of follow-up data were available (969 with typical and 219 with atypical TN). Outcomes were divided into three categories: excellent, pain relief without medication; good, mild or intermittent pain controlled with low-dose medication; and poor, no or poor pain relief with large amounts of medication. The results for typical and atypical TN were compared and patient history and pain characteristics were evaluated for possible predictive factors. Conclusions. In this study, MVD for typical TN resulted in complete postoperative pain relief in 80% of patients, compared with 47% with complete relief in those with atypical TN. Significant pain relief was achieved after 97% of MVDs in patients with typical TN and after 87% of these procedures for atypical TN. When patients were followed for more than 5 years, the long-term pain relief after MVD for those with typical TN was excellent in 73% and good in an additional 7%, for an overall significant pain relief in 80% of patients. In contrast, following MVD for atypical TN, the long-term results were excellent in only 35% of cases and good in an additional 16%, for overall significant pain relief in only 51%. Memorable onset and trigger points were predictive of better postoperative pain relief in both atypical and typical TN. Preoperative sensory loss was a negative predictor for good long-term results following MVD for atypical TN. KEY WORDS trigeminal neuralgia microvascular decompression outcome RIGEMINAL neuralgia was first described in the late seventeenth century by Fehr, who wrote a eulogy that was discussed in a study by Schmidt. 9 As related by Stookey and Ransohoff, 10 John Locke later described TN that afflicted the Countess of Northumberland, wife of the English Ambassador to France. Brown, et al., 2 report that Nicholas Andre coined the term tic douloureux and in 1756 discussed surgical treatments in five patients who suffered from a cruel and obscure illness, which causes... in the face some violent motions, some hideous grimaces, which are an insurmountable obstacle to the reception of food, which put off sleep.... In 1773, John Fothergill 4 described 16 patients with TN. In 1966, Jannetta and Rand 6 proposed that the origin of TN rests in the compression of the trigeminal nerve by small blood vessels near the brainstem. Since that time, MVD has Abbreviations used in this paper: MVD = microvascular decompression; TN = trigeminal neuralgia. become a mainstay in the management of this painful condition. In its classic or typical form, TN involves a unilateral, lancinating, electrical pain in one or more of the trigeminal nerve distributions. Patients usually describe typical trigger points on the face and triggering stimuli or activities. The pain generally has a memorable onset, may be positional in nature, and has variable periods of remission. Most patients describe some modicum of pain relief from medications that include carbamazepine, phenytoin, or baclofen. Atypical TN denotes a syndrome in which patients describe unilateral pain which, although it is in a trigeminal distribution, is more burning or aching in nature. This discomfort may be continuous or nearly continuous and rarely responds to any of the aforementioned medications. Some patients have previously suffered from typical TN, but over the years it has evolved into an atypical condition. These patients may have moved through a stage called transitional TN, which has characteristics of both the typical and atypical disease. The trigeminal pain syndromes were re- 527

2 A. B. Kassam, et al. TABLE 1 Characteristics of 2264 patients in whom 2675 MVDs were performed for TN (2003 for typical and 672 for atypical TN) TN Diagnosis (no. of patients) Family No. of No. of History Transi- MVDs* Patients of TN (%) Typical Atypical tional Ratio (4.0) NA (4.1) (7.5) total (4.0) *Of 2264 patients treated, 346 had multiple MVDs. Abbreviation: NA = not applicable. The ratio of the number of patients with typical to the number with atypical TN. cently described by Burchiel and Slavin 3 as a continuum rather than discrete diagnoses, which is consistent with our description of transitional TN. In this paper we look at the results of 2003 MVDs for typical TN and 672 of these procedures for atypical TN that were performed at the University of Pittsburgh Medical Center before This helped us to determine the patient characteristics predictive for outcome following MVD and also to assess surgical complication rates in these two populations. In an earlier paper from our institution, investigators reviewed the results of MVD for typical TN. 1 In that paper the authors reviewed 1185 patients who underwent MVD for typical TN between 1972 and 1991 and looked at more than 1 year of follow up in 1155 patients. In our paper we have reviewed an additional 1079 patients who underwent MVD for either typical or atypical TN, and we looked at long-term outcomes in patients for whom 5 or more years of follow-up data were available. Clinical Material and Methods The TN database at the University of Pittsburgh was reviewed to identify all patients who underwent MVD for TN before Previously collected data relating to complications and outcomes were combined with data collected through phone calls to patients to yield the information in this manuscript. Follow up ended in December All data were analyzed by a statistician (Y.F.C.), who then made graphs of the results. Outcomes were divided into three categories as follows: 1) excellent denoted patients who were pain free off medications; 2) good denoted patients with infrequent pain controlled with low-dose medications; and 3) poor described patients with no postsurgical pain relief on or off medications. Population and Demographics Between February 1972 and December 1995, 2675 MVDs were performed in 2264 patients for the treatment of TN at the University of Pittsburgh Medical Center. Of these, 2003 procedures were for typical and 672 were for atypical TN (Table 1). Sixty-one percent of the patients were female. The median age of all patients was 55 years, with typical TN occurring at a median age of 57 (range 5 TABLE 2 Age and family history among the 2264 patients with TN who were treated with MVD 87 years) and atypical TN at a median age of 46.5 (range years). Of patients with typical TN, 4.8% had a family history of the syndrome, whereas only 1.5% with atypical TN had a family history of the disease (Table 2). A total of 2003 MVDs for typical TN were performed in 1739 patients, with 240 of them undergoing multiple MVDs. Sixty percent of patients had right-sided pain and 8% had bilateral symptoms. Of the 1795 patients who were treated preoperatively with carbamazepine, medical treatment ultimately failed in 92%. Other aspects of the patients clinical histories are noted in Table 3. Five years or more of follow-up data were available in 974 patients (range years, mean 11.3 years, median 10.5 years). A total of 672 MVDs for atypical TN were performed prior to 1996 in 537 patients, of whom 114 had multiple MVDs. Fifty percent of patients had right-sided pain and 12.5% had bilateral symptoms. Of the 441 patients who were treated preoperatively with carbamazepine, medical treatment ultimately failed in 96%. Other aspects of the patients clinical histories are noted in Table 3. Five or more years of follow-up data were available in 220 patients (range years, mean 9.8 years, median 7.9 years). When evaluating the 346 patients who had multiple MVDs at our institution we found that the ratio of typical to atypical TN decreased as the number of MVDs increased. The ratio for single procedures was significantly greater than for multiple procedures (Table 1). Additionally, we identified eight patients who were initially treated for typical TN who were subsequently treated for atypical TN. Postoperative Complications Results TN Diagnosis Factor Typical Atypical Transitional no. of patients median age (yrs)* mean (SD) 55.9 (12.0) 47.7 (12.6) 44.5 (14.2) family history of TN (%) *Patients age at MVD; if patients had undergone multiple MVDs, their age at first MVD was considered. Abbreviation: SD = standard deviation. Probability value less than when typical was compared with atypical TN. Results of MVD mean nothing if the surgical risks associated with the procedure are excessive. Complications for MVD in patients with typical and atypical TN are shown in Table 4. The most notable difference between the complications for typical and atypical TN surgery was the much higher rate of significant meningismus. Immediate Postoperative Pain Relief Typical TN. As shown in Table 5, excellent immediate postoperative pain relief was achieved in 80.3% of patients who underwent MVD for typical TN, 16.5% received good (partial) relief, and 3.2% had no response: a 96.8% signifi- 528

3 Outcome predictors in typical and atypical trigeminal neuralgia TABLE 3 Clinical history in typical and atypical TN cant pain reduction with surgery. When patients available for 5 years or more of follow up were studied separately for their immediate postoperative results, 84.1% had experienced excellent results, 14.1% had a good result, and 1.8% had no surgical response: a 98.2% significant pain reduction with surgery. Atypical TN. Excellent immediate postoperative pain relief was achieved in 46.9% of patients who underwent MVD for atypical TN, 39.7% received good relief, and 13.4% had no response: an 86.6% significant pain reduction with surgery. When patients available for 5 years or more of follow up were studied separately for their immediate postoperative results, 58.2% had experienced excellent results, 33.2% had a good result, and 8.6% had no surgical response: a 91.4% significant pain reduction with surgery. Long-Term Postoperative Pain Relief Typical TN. As shown in Table 5, long-term excellent pain relief in the patients available for follow up was achieved in 73.7%, good pain relief was attained in 6.8%, and 19.5% had no long-term pain relief. Therefore, significant pain relief in this population declined from 98.2% immediately postoperatively to 80.5% at 5 years or more after MVD. Atypical TN. Long-term excellent pain relief in patients available for follow up was achieved in 34.7%, good pain relief was attained in 16.4%, and 48.9% had no long-term pain relief. Therefore, significant pain relief in this population declined from 91.4% immediately postoperatively to 51.1% at 5 years or more after MVD. In addition to analyzing overall outcomes, both immediate and long-term, in patients with typical and atypical TN, we investigated whether outcomes in our patient population differed based on presenting components of the clinical history. These included memorable onset of pain, presence of a trigger point, history of MVD, previous destructive procedure for management of TN, presence of a preoperative trigeminal distribution sensory loss, family history, and presence of bilateral pain. Findings for typical TN are summarized in Table 6 and for atypical TN in Table 7. The majority of patients had been treated with carbamazepine before undergoing MVD. We found that treatment TN Diagnosis (%) Typical Atypical Clinical History* (2003 ops) (672 ops) memorable onset of pain sensory loss trigger points bilat pain prior MVD prior destructive procedure carbamazepine use no response at op in carbamazepine users *All items except prior MVD were significantly different (p 0.001) between patients with typical and atypical TN. TABLE 4 Complications associated with 2003 MVDs for typical and 672 for atypical TN* No. W/ Diagnosis (%) Complication Typical Atypical hematoma 3 (0.1) 3 (0.4) edema 9 (0.4) 6 (0.9) infarction* 2 (0.1) 6 (0.9) hydrocephalus 3 (0.1) 0 palsy CN VII 34 (1.7) 6 (0.9) CN VIII 29 (1.4) 7 (1.0) other CN 93 (4.6) 16 (2.4) CSF leak 33 (1.6) 24 (3.6) pseudomeningocele 5 (0.2) 3 (0.4) meningitis 5 (0.2) 5 (0.7) meningismus 257 (12.8) 146 (21.7) death 4 (0.2) 1 (0.1) * CN = cranial nerve; CSF = cerebrospinal fluid. Probability value less than 0.01 for comparison between typical and atypical cases. in 92% of patients who had taken carbamazepine for typical TN had failed before MVD, whereas carbamazepine failed in 96% of patients with atypical TN before MVD. Trends were similar when immediate 30-day postoperative results were compared with long-term results at 5 years or more in all patients who underwent MVD and those in the subpopulation in whom 5 years or more of follow-up data were available. Along with these findings, a review of the data demonstrated that bilateral tic was a negative predictor for longterm pain control following MVD for typical TN (p = 0.001), as was preoperative sensory loss for atypical TN (p 0.05). Previous MVD or destructive lesions were negative predictors for immediate postoperative pain relief following surgery for typical TN (p 0.05), but not for long-term outcomes. For typical TN, trigger points were positive predictors for better long-term outcome (p 0.05). Memorable onset of pain and trigger points were positive TABLE 5 Results following 2003 MVDs for typical and 672 for atypical TN* TN Diagnosis (%) Result Typical Atypical immediate postop pain relief (all patients) cure partial no relief postop pain relief ( 5 yrs) cure partial no relief long-term outcome ( 5 yrs) cure partial relief failure * Five years or more of follow up was available in 969 of the typical cases and in 219 of the atypical ones. Probability value less than

4 A. B. Kassam, et al. TABLE 6 Outcomes according to clinical history in patients with typical TN Follow Up 5 Years (%) Postop Relief, All Patients (%) Postop Relief Long-Term Results Clinical Finding Cure Partial None Cure Partial None Cure Partial Relief Failure memorable onset of pain* yes no preop sensory loss* yes no bilat pain yes no prior MVD* yes no prior destructive procedure* yes no trigger points* yes no * Probability value less than 0.05 in postoperative relief. Probability value less than 0.05 in long-term results. predictors for better immediate postoperative pain relief for both typical and atypical TN (p 0.05). Discussion As has been previously discussed, MVD is a safe and effective treatment for typical TN. 1 Our results for postoperative pain relief after MVD for typical TN are similar to those previously reported from our institution. 1 Our longterm outcomes are slightly better. This in part reflects the fact that we included 27 patients who underwent more than two operations for recurrent or residual pain. In our review of factors predictive of immediate postoperative relief after surgery for typical TN we found that trigger points and memorable onset of pain were positive predictors of excellent outcomes, whereas preoperative sensory loss, previous MVD, and previous destructive le- TABLE 7 Outcomes according to clinical history in patients with atypical TN Follow Up 5 Years (%) Postop Relief, All Patients (%) Postop Relief Long-Term Results Clinical Finding Cure Partial None Cure Partial None Cure Partial Relief Failure memorable onset of pain* yes no preop sensory loss yes no bilat pain yes no prior MVD yes no prior destructive procedure yes no trigger points* yes no * Probability value less than 0.05 in postoperative relief. Probability value less than 0.05 in long-term results. 530

5 Outcome predictors in typical and atypical trigeminal neuralgia sioning were negative predictors. Only trigger points were positive predictors for good long-term outcomes, whereas bilateral pain was a negative predictor of excellent longterm outcomes. This is consistent with a previous review of bilateral tic pain at our institution, 8 from which we inferred that venous pathology is more common in these cases and that when the offending vessel is venous, recurrence is more likely. 1,5,7,11 The most striking finding in this study was the extremely poor results for atypical TN. Despite postoperative pain relief in 86% of patients, only 51% had long-term relief and only 35% had excellent long-term outcomes. Similarly to typical TN, memorable onset and trigger points were positive predictors for excellent postoperative relief after MVD for atypical TN. Preoperative sensory loss was a negative predictor for long-term pain relief. These results are consistent with our belief that there is a group of patients along the continuum of typical to atypical TN who may benefit from MVD. We refer to these as transitional patients. They are characterized as having the constant burning pain of atypical TN but also some components of typical TN, such as trigger points for sudden memorable onsets of their pain. Sensory loss would be more consistent with patients who have progressed farther along the continuum. 3 We found that the majority of patients who underwent MVD had taken carbamazepine and medical management had failed in these patients prior to surgery. Unfortunately, we do not have good records regarding which patients had had a prior response to carbamazepine. In the future we intend to analyze the role of prior carbamazepine responsiveness as a predictive factor. Conclusions As we have previously asserted, MVD is a safe and effective treatment for typical TN. As a rule, MVD for atypical TN has a poor outcome, with 50% of patients experiencing recurrent pain that is poorly controlled. We infer from the data that there is a subset of patients with atypical TN who have better outcomes. We plan to advance this work by reviewing the group of patients whom we refer to as having transitional TN to determine if they have better long-term outcomes. Acknowledgment We thank Peter J. Jannetta, M.D., who is the surgeon of record for the majority of the cases included in this study. References 1. Barker FG II, Jannetta PJ, Bissonette DJ, et al: The long-term outcome of microvascular decompression for trigeminal neuralgia. N Engl J Med 334: , Brown JA, Coursaget C, Preul MC, et al: Mercury water and cauterizing stones: Nicolas Andre and tic douloureux. J Neurosurg 90: , Burchiel KJ, Slavin KV: On the natural history of trigeminal neuralgia. Neurosurgery 46: , Fothergill J: Of a painful affection of the face, in Society of Physicians in London: Medical Observations and Inquires. London: T Cadell, 1773, Vol 5, pp Hamlyn PJ, King TT: Neurovascular compression in trigeminal neuralgia: a clinical and anatomical study. J Neurosurg 76: , Jannetta PJ, Rand RW: Transtentorial retrogasserian rhizotomy in trigeminal neuralgia by microneurosurgical technique. Bull LA Neurol Soc 31:93 99, Klun B: Microvascular decompression and partial sensory rhizotomy in the treatment of trigeminal neuralgia: personal experience with 220 patients. Neurosurgery 30:49 52, Pollack IF, Jannetta PJ, Bissonette DJ: Bilateral trigeminal neuralgia: a 14-year experience with microvascular decompression. J Neurosurg 68: , Schmidt JE: Medical Discoveries. Springfield, IL: Charles C Thomas, Stookey B, Ransohoff J: Trigeminal Neuralgia: Its History and Treatment. Springfield, IL: Charles C Thomas, Sun T, Saito S, Nakai O, et al: Long-term results of microvascular decompression for trigeminal neuralgia with reference to probability of recurrence. Acta Neurochir 126: , 1994 Manuscript received May 25, Accepted in final form October 24, Address reprint requests to: A. B. Kassam, M.D., Center for Cranial Nerve Disorders, University of Pittsburgh Medical Center, Suite B-400, 200 Lothrop Street, Pittsburgh, Pennsylvania [email protected]. 531