MARCUS F. KEEP, M.D., PAUL A. DEMARE, M.D., AND LYNN S. ASHBY, M.D.

Transcription

1 J Neurosurg (Suppl) 102: , 2005 Gamma knife surgery for refractory postherpetic trigeminal neuralgia: targeting in one session both the retrogasserian trigeminal nerve and the centromedian nucleus of the thalamus MARCUS F. KEEP, M.D., PAUL A. DEMARE, M.D., AND LYNN S. ASHBY, M.D. The Gamma Knife Center of the Pacific, Honolulu, Hawaii; Division of Neurosurgery, University of New Mexico, Albuquerque, New Mexico; Pacific Radiation Oncology, The Queen s Medical Center, Honolulu, Hawaii; and The Gamma Knife Center at the Barrow Neurological Institute, Phoenix, Arizona Object. The authors tested the hypothesis that two targets are needed to treat postherpetic trigeminal neuralgia (TN): one in the trigeminal nerve for the direct sharp pain and one in the thalamus for the diffuse burning pain. Methods. Three patients with refractory postherpetic TN were treated with gamma knife surgery (GKS) through a novel two-target approach. In a single treatment session, both the trigeminal nerve and centromedian nucleus were targeted. First, the trigeminal nerve, ipsilateral to the facial pain, was treated with 60 to 80 Gy. Second, the centromedian nucleus was localized using standard coordinates and by comparing magnetic resonance images with a stereotactic atlas. A single dose of 120 to 140 Gy was delivered to the target point with a single 4-mm isocenter. Patients were followed clinically and with neuroimaging studies. Pain relief was scored as excellent (75 100%), good (50 75%), poor (25 50%), or none (0 25%). Follow up ranged from 6 to 53 months. There were no GKS-related complications. Two patients died of unrelated medical illnesses but had good or excellent pain relief until death. One patient continues to survive with 44 months follow up and no decrease in pain intensity, but with a decreased area of pain. Conclusions. Combined GKS of the centromedian nucleus and trigeminal nerve in a single treatment session is feasible and safe, and the effect was promising. A larger study is required to confirm and expand these results. KEY WORDS gamma knife surgery postherpetic neuralgia trigeminal nerve centromedian nucleus thalamus G amma knife surgery may be an effective and safe noninvasive treatment for functional indications. It can produce ablative lesions with excellent precision for small anatomical targets in the brain. This is particularly true for patients suffering from refractory cancer pain and those for whom the side effects of analgesic agents are intolerable. Postherpetic TN is a condition for which radiosurgery could play a part, given the discreet targets involved in the complex mechanisms of pain perception and the generally aged or infirm population affected most commonly by this disorder. Varicella Zoster virus can result in two clinical conditions. Primary varicella infection is well known as chicken pox. It is hypothesized that during the period of primary infection the virus is transmitted via the cutaneous nerves to the sensory ganglia where it remains latent but can be reac- Abbreviations used in this paper: CMN = centromedian nucleus; DBS = deep brain stimulation; GKS = gamma knife surgery; MR = magnetic resonance; PHN = postherpetic neuralgia; PHTN = postherpetic trigeminal neuralgia; TN = trigeminal neuralgia. 276 tivated years later. The reactivation results in a dermatomal vesicular eruption known as herpes zoster. The incidence of herpes zoster increases with age and is highest (5 10:1000 population) after the age of 50 years. The dermatomes most frequently affected by this condition include the trigeminal nerve, especially the V 1 or ophthalmic division, and the T3 L3 nerve roots. 26 The most common complication of herpes zoster is the development of a dermatomal pain syndrome called zosterassociated pain. Viral reactivation and vesicular eruptions result in damaged primary sensory neurons. Pain related to the reactivated disease may precede the eruption of vesicles and may continue during the acute infection as dermatomespecific acute neuritis. 1 For the majority of persons this severe pain resolves with the healing of the vesicles. Pain that persists beyond 1 month becomes a chronic condition known as PHN and is more commonly seen in persons of advanced age, with 50% of patients being older than 50 years, and 75% of affected patients being older than 65 years. 1 For a minority of patients the pain becomes permanent and highly resistant to treatment. 5 In the case of facial pain, or PHTN, one or more of the three branches of the

2 Gamma knife surgery for postherpetic neuralgia nerve on one side of the face can be affected, usually in the forehead. Typically, the triad components of PHTN are 1) chronic, unrelenting burning pain; 2) paroxysms of lancinating pain similar to that experienced in typical TN, but more severe; and 3) allodynia or dynamic mechanical hyperalgesia induced by any light cutaneous stimuli. 3,5 Paresthesias and decreased normal sensation are also common, as well as permanent cutaneous scarring. The risk of developing this debilitating pain syndrome may be reduced by aggressive treatment with antiviral therapy (famvir and valacyclovir) during the acute vesicular outbreak. 28 Medications can be helpful in reducing either the intensity or frequency of discomfort, but in general PHTN is refractory to both medical and surgical therapies. Carbamazepine compounds, gabapentin, and tricyclic antidepressants seem the most effective, but dose-limiting side effects are particularly problematic for elderly patients or those with associated illnesses. Opioid and nonopioid analgesics add little benefit for patients with the neuropathic pain of PHTN, with the reported exception of high-dose oxycodone. 1,5 Because of the limitations of medical management, patients with refractory pain seek other forms of relief. 29 Anesthetic blocks and ablative surgical procedures have been attempted but with little success except for radiofrequency ablation of the nucleus caudalis in the dorsal root entry zone, which can produce up to 71% good or excellent relief in PHTN. These procedures involve major neurosurgery with significant neurological morbidity in up to 90% of patients. 4 For this reason interest continues in the evaluation of less invasive surgical procedures such as radiosurgery for this condition. 6 Gamma knife surgery has become a commonly used procedure for the treatment of medically refractory classic TN. The treatment is effective with excellent results and minimal risk of complications. The treatment produces complete pain relief in more that 50% of patients and reduces pain by half in an additional 30%. The most frequently reported side effect is facial numbness and paresthesias occurring at a rate of approximately 6 to 10% ,19,33,35 For many it has obviated the need for surgically invasive microvascular decompression. Relapse is easily managed with retreatment. 7,21 Unfortunately, this success is not replicated for those with PHTN. 26,30,35 In a single study 44% good or better pain relief was demonstrated ( 40 60% residual pain) from PHTN treating only the trigeminal nerve with GKS. 26 Thalamotomy has been performed stereotactically with a variety of methods, including radiofrequency ablation, DBS, and GKS. Lesioning of the medial thalamic nuclei has been demonstrated to reduce pain especially of the face and upper torso. First, Leksell 15 and later Steiner and colleagues, 22 and Young, et al., 31 performed series of gamma thalamotomies targeting the CMN of the thalamus to treat chronic pain, with success for pain syndromes other than PHTN. Intentional modification of specific nuclei of the thalamus has been accomplished through the development of several neurosurgical techniques. Specifically, stereotactic procedures such as radiofrequency ablation, implantation of DBS electrodes, and gamma knife thalamotomy have been performed as adjuncts to the general management of intractable pain syndromes, including pain from cancer. 8,15,22,24,27,31,32,34 Lesions of the CMN and related medial thalamic nuclei block transmission to the cortex, resulting in pain relief to the face and upper areas of the body. Due to the organization of the thalamus, there is no loss of normal sensation or the protective fast component of pain; however, the perception of the suffering component of pain appears to be reduced. New approaches to managing PHTN are needed. We developed a hypothesis that when the trigeminal nerve and CMN are treated at the same time in the same patient the response and pain relief could be improved. Therefore, we offered GKS to patients with intractable PHTN, with treatment in a singe session of both the trigeminal nerve and the CMN of the thalamus for palliative management of severe refractory PHTN. Illustrative Cases Clinical Material and Methods Case 1. This 83-year-old man had a history of five-vessel coronary bypass surgery and Type II diabetes mellitus. He presented with an 18-month history of PHTN involving the left V 1 ophthalmic branch. He began experiencing severe pain during the 1st week of this outbreak, with a 3/10 intensity and superimposed attacks of lancinating (burning and stabbing) pain of 7/10 intensity that lasted approximately 2 minutes and occurred at least six times daily. He also suffered from marked allodynia. Additionally, he described a dysesthetic sensation of jangling across his left cheek. Multiple medication schedules were tried with only minimal relief and coupled with side effects of confusion, nausea, and lethargy. He considered his condition to be agonizing and suggested that he could not live any longer without relief. Physical examination revealed mottled depigmentation over the left forehead. Reduced sensation over the V 1 and V 2 dermatomes and an absent corneal response on the left were the only relevant neurological findings. Invasive procedures, such as dorsal root entry zone radiofrequency ablation and DBS implantation were discussed, but given the patient s advanced age and cardiac condition neither he nor his physicians believed he was a good candidate for these procedures. Case 2. This 61-year-old man had undergone liver transplantation because of hepatitis and alcoholic liver disease. He had also been found to have a malignant hepatoma and was being treated with chemotherapy and immunosuppressive transplant therapy. He developed severe herpes zoster of the left side of his face affecting all three distributions of the trigeminal nerve with pain and numbness involving his face and tongue that persisted following resolution of his vesicular rash. Despite intensive medical treatment his facial pain remained unbearable with the best relief on maximum therapy reducing it to only 5/10 pain intensity. Additional medication options were limited by his compromised hepatic function, and he was not considered a candidate for open surgical intervention. Case 3. This 56-year-old man had no significant other illnesses. A few weeks after the resolution of the rash of rightsided V 1 herpes zoster, he developed PHTN of V 1 and V 2. This had persisted for 21 months despite aggressive antiviral treatment with famvir during his acute phase. He also complained of allodynia. Intensive medical treatment had not helped and he ranked his pain as 5 6/10 at rest, but when touched on his forehead or cheek, even with a puff of 277

3 M. F. Keep, P. A. DeMare, and L. S. Ashby FIG. 1. Representative treatment plan of T 1 -weighted MR images on GammaPlan. Axial, coronal, and sagittal reconstructions are seen. Planned GKS targets are outlined as green spheres on each axis. Upper: Axial image with first target (green/yellow sphere) at the left CMN. Lower: Axial image with second target (green sphere) at the right trigeminal nerve. air, the pain would escalate to 10/10. He described the quality of his pain as sharp and stabbing, lasting only a second, and then a slow throbbing pain lasting 5 seconds. Physical examination showed restricted facial expression, but intact sensation to light touch. The remainder of his cranial nerves were normal. 278

4 Gamma knife surgery for postherpetic neuralgia Following a multidisciplinary evaluation, it was believed that all three patients might benefit from GKS targeting both the trigeminal nerve and the CMN of the thalamus. Treatment Planning and GKS Technique In one radiosurgical session, two targets (trigeminal nerve and CMN) were identified and treated using the following procedure. The Leksell stereotactic frame was fixed to the patient s skull with the guidance of ear bars to place the upper midbrain at the frame s cartesian center where targeting is most accurate and to minimize pitch or yaw during the frame application. Magnetic resonance images (Gyroscan ACS-NT 1.5 tesla; Philips) were obtained for localization of the retrogasserian trigeminal nerve root and for localization of the CMN of the thalamus. Axial 1-mmthick T 1 -weighted MR images were obtained utilizing both three-dimensional SPGR and three-dimensional inversionrecovery techniques. Proton density T 2 -weighted spin echo MR images were also obtained. Gray and white matter distinction of the brain parenchyma was readily differentiated with these sequences. Axial computerized tomography (Tomoscan AV; Philips) scans were also acquired through the brain to confirm the absence of magnetic distortion on MR images and for enhanced precision of localization of anatomical structures. Images were loaded into the Gamma Plan workstation (Elekta Instrument AB, Stockholm, Sweden) registered, and anatomical landmarks of the anterior commissure, posterior commissure, and midline floor of the third ventricle were identified. Using a computer-assisted algorithm program, 17 it was possible to identify the CMN 0 to 5 mm (average 3.5 mm) superior to the intercommissural line, between 8 and 12 mm (average 10 mm) posterior to the midcommisssural line, and between 8 12 mm (average 10 mm) lateral to the midline. 21 This location was confirmed in the axial, coronal, and sagittal planes. Small modifications of target coordinates were made based on the patient s brain anatomy and comparison with the Schaltenbrand and Wahren stereotactic brain atlas. 19 The second target was then selected as the retrogasserian portion of the trigeminal nerve root. Radiosurgery was performed using a model B Leksell Gamma Knife unit (Elekta Instrument AB). See Fig. 1 for representative GammaPlan target placement. All three patients were treated while in the supine position at a gamma angle of 110. The CMN was treated with a single 4-mm isocenter with a maximum dose of 120 to 140 Gy. The retrogasserian portion of the trigeminal nerve root was treated with a single 4-mm isocenter with a maximum dose of 60 to 80 Gy placed so that the 27% isodose line just touched the brainstem. The first patient was treated by targeting the CMN and trigeminal nerve root on the same side as the pain. Ipsilateral GKS lesioning of the CMN has been shown to be equally effective to the contralateral side, likely due to bilateral innervation. Because of early concerns about imaging distortion it was decided to place lesions on the same side of both the CMN and trigeminal nerve and to place them as close to the frame center as possible. The dose to the CMN was 120 Gy and to the trigeminal nerve 60 Gy. The lower doses were used to reduce dose interaction related complications. Following validation of imaging reliability, the doses in the other two patients were directed at the ipsilateral trigeminal nerve root and the contralateral CMN, and the dose to each target was increased by 20 to 80 Gy and 140 Gy, respectively. Following GKS, the patients were discharged the same day. Pain medications were continued according to the patients preoperative dose schedules. Patients were followed up clinically and pain relief was scored as excellent (75 100%), good (50 75%), poor (25 50%), or none (0 25%). Results All patients tolerated the procedures well and no complications due to GKS were observed during the follow-up period. Patients were evaluated at routine scheduled intervals by the neurosurgeon and radiation oncologist. The patients primary care was then returned to the referring physicians with ongoing communications with the Gamma Knife team. See Fig. 2 for representative 3-month post-gks MR imaging. The patient in Case 1 was evaluated 2 weeks following GKS, at which time he reported no change in his pain. At 1 month, he was no longer having the severe attacks of pain (7/10) that were previously occurring every 3 to 4 hours. He reported the same basal level of pain 3/10 as before treatment. This was more frequent in the morning or when he would touch the affected area on his left forehead. He also continued to describe a jangling sensation in the V 2 distribution as before treatment; however, at 1 month it was less intense. At that time he continued to take gabapentin but had been able to reduce his nortriptyline dose in half. Through the rest of the year, he reported reduced pain and was returned to the care of his primary physician. Until the time of his death nearly 4.5 years later, from unrelated pneumonia, he continued to enjoy a marked reduction in facial pain (75%), which was considered to be an excellent and durable (53 months) response to treatment. The patient in Case 2 was evaluated 1 week after GKS, at which time he reported a considerable decrease in pain with a reduction of intensity from 5/10 with hydromorphone every 3 hours, to 2/10 without the need for hydromorphone. He ranked his pain relief as greater than 50%. He was evaluated again at 1 month when he reported that his facial pain had markedly improved. His only complaint was persistent numbness and pain along the left side of his tongue, but otherwise he felt well, had gained weight, and had been able to discontinue his gabapentin and reduce his daily dose of oxycontin. He continued to take numerous transplant-related medications. At his 2-month visit, the patient reported that his reduced facial pain remained stable while undergoing chemotherapy for his metastatic cancer. At his 3-month postprocedure visit, he was suffering from severe bone pain related to his primary malignancy, but he continued to report consistently good ( 50%) relief of facial pain until his cancer-related death 6 months after GKS. The patient in Case 3 remains alive at the time of this report, more than 4 years after GKS. Pretreatment he ranked his pain as 5 6/10 at rest, increasing to 10/10 if the affected area of his right forehead or cheek was touched. Six months after GKS the size of the affected area was reduced but the pain intensity was unchanged. At that point, he was unable 279

5 M. F. Keep, P. A. DeMare, and L. S. Ashby FIG. 2. Magnetic resonance images obtained 3 months after simultaneous left CMN gamma thalamotomy and right trigeminal nerve GKS (see Fig. 1). Axial Gd-enhanced T 1 -weighted MR images revealing enhancement at the left CMN (top left) and the right trigeminal nerve (top center) including its root entry zone, and coronal and sagittal same-sequence images revealing an enhancing lesion in left CMN (top right and bottom left), and a corresponding axial FLAIR sequence revealing the effect of GKS in the left CMN (bottom right). to characterize the quality of his pain clearly. Because of his limited response to GKS, he sought another neurosurgical team who performed classic microvascular decompression 10 months later. He had no additional pain relief from the microvascular decompression; however, he continued to report a progressive decrease in the area of his face affected until only the right forehead was still involved. This reduction in the affected area stabilized 3 years and 4 months after GKS. He is under the ongoing care of a pain specialist who continues to treat him with various medications. Mechanisms of PHTN Pain Discussion The rationale for targeting the trigeminal nerve and CMN of the thalamus is based in the anatomy and pathophysiology of facial pain. 2,9,23,28 Facial pain involves multiple pathways. The gasserian ganglion, containing the cell bodies of the trigeminal nerve s sensory nerves, is the equivalent of spinal dorsal root ganglia and contains the cell bodies of both tactile and nociceptive neurons. The secondary sensory neurons in the caudalis nucleus project brief sharp sensation primarily though the mostly crossed trigeminal meniscus and quintothalamic tract to the contralateral ventral posterior medial thalamic nucleus while maintaining a small ipsilateral connection. The ventral posterior medial relay nuclei transmit normal facial sensation and quick sharp pain to the cortex, where it reaches consciousness. Slow burning pain such as is experienced in PHTN is conducted to the CMN in a more diffuse manner. Input from both primary and secondary pain sensory neurons project to the adjacent reticular formation in the brainstem. The trigeminal inflow to the spinoreticular tract transmits pain impulses through a multisynaptic, diffusely projecting system of both crossed and uncrossed fibers projecting through the medulla and pons, and ultimately to the medial thalamic CMN and the adjacent intralaminar and parafascicular nuclei. These diffuse projecting nuclei have widespread connections to other thalamic nuclei and through the cortex affect the level of brain arousal and the degree of suffering associated with the pain. Herpes zoster produces damaged and dead primary sensory and pain neurons originating in the face. Damage to the peripheral nerve and gasserian ganglion with loss of normal sensory neurons accounts for some of the well-observed sensory loss. Due to transsynaptic degeneration, secondary sensory neurons in the nucleus caudalis either die or develop denervation supersensitivity. The transmission of normal sensation and pain from peripheral to central centers is also adversely affected so that normal sensory input is misdirected and spills into the pain pathways, possibly via ephaptic transmission and sprouting. This may be responsible for the overlay of lancinating pain elicited by minimal facial stimuli. The few remaining primary pain fibers that project onto secondary sensory neurons may themselves sprout and innervate more than the usual number of secondary sensory neurons, creating an amplification effect. Thus, a small normally innocuous input triggers an abnormally amplified response, producing severe pain. Furthermore, central reorganization of thalamic pain systems in response to the loss of peripheral input likely accounts for the chronic unrelenting burn of deafferentation pain. Rationale for Dual-Target GKS The two locations, peripheral and central, responsible for the pain in PHTN may account for the difficulty in treating it with a single-target approach. The hypothesis was submitted that two locations would need to be treated to ameliorate this pain: one peripheral and one central. The peripheral site is the retrogasserian trigeminal nerve. The peripheral GKS lesion should reduce the abnormal peripheral input that is believed to be in turn overamplified by the rewired central centers. It should reduce the lancinating component of PHTN. The abnormal pain output from the nucleus caudalis is transmitted primarily through the CMN; thus, treating the CMN as well might interrupt this transmission and possibly dampen the sensation of chronic burning pain. Thalamotomy and CMN Gamma Thalamotomy Tasker 24 has provided the most extensive metaanalysis of the results of therapeutic thalamotomy, establishing the importance of the CMN for the treatment of pain. Surgical thalamotomy for pain began as ventrocaudal thalamotomy with 82% of 22 treated patients experiencing pain relief, but at the expense of dysesthesia or paresis in 32%, leading the same surgeons to abandon the ventrocaudal procedure in favor of medial thalamotomy, which was found to yield better pain relief with fewer complications and no loss of sensation. 16,18 From 175 cases reviewed, Tasker reported an overall success rate of 46 to 70% achieved by invasive medial thalamotomy. Hitchcock and Teixeria 8 reported the results of medial thalamotomy for pain performed in 43 patients (19 CMN and 23 basal thalamotomies), with the best results seen in patients who underwent bilateral CMN thalamotomy. Complications reported in these series included transient confusion with permanent deficits seen only in patients who underwent bilateral procedures result- 280

6 Gamma knife surgery for postherpetic neuralgia ing in large lesions. 8,24 These surgical series confirmed that targeting the CMN reduces risk to more lateral thalamic structures and does not result in any deficit in the perception of the contralateral primary sensory modalities of pinprick, temperature, touch, or position. 25 Microelectrode recordings in 45 patients undergoing radiofrequency medial thalamotomy confirmed a 67% response rate in pain reduction without any sensory deficits, unless the lesions were large and extended into lateral structures. 9 Gamma knife surgery provides the same results as surgical thalamotomy. Young, et al., 31 provide the largest modern series of CMN gamma thalamotomies for pain. In addition to confirming efficacy, they have established that intraoperative electrophysiological monitoring used for target localization in DBS and radiofrequency procedures is not necessary because of the high precision and visual recognition of the anatomical targets with MR and computerized tomography imaging and correlation with an anatomical neurosurgical atlas, 20 based on correlation with previously identified radiofrequency targets. In a report from 1996, 10 patients underwent CMN gamma thalamotomy, which included one case of anesthesia dolorosa, one case of thalamic pain, one brainstem stroke, two cases of PHN, and five with pain from spinal cord injury. 31 Seven of 10 had good-to-excellent results, and three were considered treatment failures. There were no radiosurgery-associated complications. In a followup series in 1996, Young and colleagues 32,34 reported the results of treating 34 patients with gamma thalamotomy for pain, treating the intralaminar nuclei with doses ranging from 120 to 180 Gy, achieving some degree of pain relief in 60 to 64.7% of patients with a complication rate of 6% for most patients. The complication rate rose to 17% if there were bilateral targets, multiple isocenters, or a dose exceeding 160 to 180 Gy. This work by Young s group also indicated that an optimal dosimetry was 140 Gy delivered through a single 4-mm isocenter Conclusions Two of three patients achieved useful pain relief with combined lesions in the trigeminal nerve and CMN. If a relief rate of over 60% is maintained in a larger series over a longer time the method will be worth serious consideration. Acknowledgments We would like to thank Drs. Maurice Nicholson, Steve Holmes, Leon Liem, Lois Mastrofrancesco, Daniel Erdman, and Mr. Brent Murphy, Mr. Norman Lehto, Mr. Hong Guo, and Ms. Lisa Capps of The Gamma Knife Center of the Pacific; and Mr. Jeffrey Fiedler of The Gamma Knife Center at the Barrow Neurological Institute for their invaluable and excellent assistance. References 1. Beydoun A: Postherpetic neuralgia: role of gabapentin and other treatment modalities. Epilepsia 40:S51 56, Bowsher D: Trigeminal neuralgia: an anatomically oriented review. Clin Anat 10: , Fields LF, Rowbotham M, Baron R: Post-herpetic neuralgia: irritable nociceptors and deafferentation: review. Neurobiology Disease 5: , Gorecki JP, Nashold BS: The Duke experience with the nucleus caudalis DREZ operation. Acta Neurochir Suppl 64:128 31, Gouda JJ, Brown JA: Atypical facial pain and other pain syndromes. Neurosurgical perspectives on Trigeminal Neuralgia. Neurosurg Clin North Am 8:87 100, Green AL, Nandi D, Armstrong G, et al: Post-herpetic trigeminal neuralgia treated with deep brain stimulation. J Clin Neurosci 10: , Hasegawa T, Kondziolka D, Spiro R, et al: Radiosurgery for refractory trigeminal neuralgia. Neurosurgery 50: , Hitchcock ER, Teixeira MJ: A comparison of results from centermedian and basal thalamotomies for pain. Surg Neurol 15: , Jeanmonod D, Magnin M, Morel A: Thalamus and neurogenic pain: physiological, anatomical and clinical data. Neuroreport 4: , Kondziolka, D, Lunsford LD, Flickinger JC: Gamma knife radiosurgery for the treatment of trigeminal neuralgia. Clin J Pain 18: 42 47, Kondziolka D, Perez B, Flickinger JC, et al: Gamma knife radiosurgery for trigeminal neuralgia: results and expectations. Arch Neurol 55: , Leksell L: A stereotactic apparatus for intracerebral surgery. Acta Chir Scand 99: , Leksell L: Stereotactic radiosurgery in trigeminal neuralgia. Acta Chir Scand 137: , Leksell L: The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 102: , Leksell L, Meyerson BA, Forster DMC: Radiosurgical thalamotomy for intractable pain. Confin Neurol 34:264, Mark V, Ervin F, Hackett T: Clinical aspects of stereotactic thalamotomy in the human in the treatment of chronic severe pain. Arch Neurol 3: , Ott K: An algorithm for empirical determination of stereotactic targets. Stereotact Funct Neurosurg 71:29 35, Riechert T: Die Chirurgische behandlung der zentralen schmerzzustande: Eischiesslich der stereotaktischen operation am thalamus und mesencephalon. Acta Neurochir 8: , Rogers CL, Shetter AG, Fiedler JA, et al: Gamma knife radiosurgery for trigeminal neuralgia: the initial experience from the Barrow Neurological Institute. Int J Radiat Oncol Biol Phys 47: , Schaltenbrand G, Wahren W: Atlas for Stereotaxy of the Human Brain. New York: Thieme, 1977, Smith KA, Rogers CL: Stereotactic radiosurgery for refractory trigeminal neuralgia, in Pollock BE (ed): Contemporary Stereotactic Radiosurgery: Technique and Evaluation. Armonk, NY: Futura, Steiner L, Forster DMC, Leksell L, et al: Gammathalamotomy in intractable pain. Acta Neurochir 52: , Sweet WH: Central mechanisms of chronic pain (neuralgias and certain other neurogenic pain), in JJ Bonica (ed): Pain. New York: Raven Press, 1980, Tasker, RR: Thalamotomy, stereotactic neurosurgery. Neurosurg Clin North Am 1: , Tsubokawa T, Moriyasu N: Follow-up results of centre median thalamotomy for relief of intractable pain. Confin Neurol 37: , Urgosik D, Vymazai J, Vladyka V, et al: Treatment of postherpetic trigeminal neuralgia with the Leksell gamma knife. J Neurosurg (Suppl 3) 93: , White JC, Sweet WH: Pain and the Neurosurgeon: A Forty- Year Experience. Springfield, IL: Charles C Thomas, 1969, pp Whitely, RJ: Varicella-Zoster virus infections, in Fauci, Brauwald, Isselbacher, et al. (eds): Harrison s Principles of Internal 281

7 M. F. Keep, P. A. DeMare, and L. S. Ashby Medicine, ed 14. New York: McGraw-Hill, 1998, Vol 1, pp Young RF: Clinical experience with radiofrequency and DREZ lesions. J Neurosurg 72: , Young RF: Functional neurosurgery with the Leksell gamma knife. Stereotact Funct Neurosurg 66:19 23, Young RF, Jacques DS, Rand RW, et al: Medial thalamotomy with the Leksell gamma knife for treatment of chronic pain. Acta Neurochir 62: , Young RF, Jacques DS, Rand RW, et al: A. Technique of stereotactic medial thalamotomy with Leksell Gamma Knife for the treatment of chronic pain. Neurol Res 17:59 65, Young RF, Vermeulen S, Grimm P, et al: Gamma knife radiosurgery for the treatment of trigeminal neuralgia: idiopathic and tumor related. Neurology 48: , Young RF, Vermeulen S, Grimm P, et al: Gamma knife thalamotomy for the treatment of persistent pain. Stereotact Funct Neurosurg 64: Young RF, Vermeulen S, Posewitz A: Gamma knife radiosurgery for the treatment of trigeminal neuralgia. Stereotact Funct Neurosurg 70: , 1998 Address reprint requests to: Marcus F. Keep, M.D., Division of Neurosurgery, 2ACC, University of New Mexico Health Science Center, University of New Mexico, Albuquerque, New Mexico, mkeep@salud.unm.edu. 282