Gamma Surgery for Pain

The lack of anatomic and pathophysiologic background knowledge of the mechanisms of pain makes management of pain by open or closed stereotactic techniques largely unsatisfactory. Early results using gamma surgery to produce thalamotomies for pain control were published by Steiner et al.¹ All 52 patients treated suffered from terminal cancer and were treated prior to the advent of computed tomography (CT) or magnetic resonance imaging (MRI). Pneumoencephalography was used to target the thalamic centrum medianum-parafasciculus (CM-Pf) complex. Good pain relief was obtained in 8 patients and moderate pain relief in 18. The patients had in general only temporary relief of pain. Of those with good pain relief, 5 died without recurrence of pain between 1 and 13 months after the procedure, and 3 had recurrence of pain at 3, 6, and 9 months. Doses between 100 and 250 Gy were tested. The collimators used were 3 × 5 and 3 × 7 mm. Observation of an actual lesion was only possible in 21 of 36 patients who had a postmortem examination. Not surprisingly, the presence of a lesion was associated with relief. Lesions were only reliably created with doses greater than 160 Gy. The most effective lesions were more medially located near the wall of the third ventricle, and the greatest relief was for face or arm pain.

These early results were not very encouraging. However, with improvements in neuroimaging and alternate target selection, it is possible that more effective lesions can be produced. Recent reports seem to support this expectation. Hayashi et al. reported significant pain reduction in patients with severe cancer pain and post-stroke thalamic pain after gamma knife lesioning of the hypophysis.^2,3 Using the 4-mm collimator and doses of 140 to 180 Gy, Young et al. have published effective pain relief in patients with chronic, intractable pain following medial thalamotomy with the gamma knife.^4,5 In a series of 15 patients followed for more than 3 months after a radiosurgical medial thalamotomy, 4 (27%) were pain free, and 5 others (33%) had greater than 50% pain relief.⁴ Additional investigation must be conducted before the role of the gamma knife for pain treatment can be fully defined.

Gamma Surgery for Trigeminal Neuralgia

The first time radiosurgery was used to treat trigeminal neuralgia, Leksell treated two patients with the stereotactic technique using orthovoltage x-rays.¹⁶ The patients treated with this method were followed up for 17 years, during which time both remained pain free. With the introduction of the gamma knife, a series of 46 patients were treated in Stockholm with less encouraging results.⁶ The target in these cases was the Gasserian ganglion, and targeting was by bony landmarks or cisternography. In the first 24 patients, where stereotactic cranial x-rays were used for targeting, 33% of patients were pain free at 6 months and 8.3% at a mean follow-up of 26 months. Transoral cisternography with tantalum dust suspended in glycerol was used in a group of 22 patients and 59% of the patients became pain free at 6 months and 18% at 26 months. A later report by Lindquist et al. stated that approximately 50% of the patients initially became pain free, but neuralgia recurred for most of them several years after radiosurgery.⁷ With advances in neuroimaging, most notably MRI, gamma surgery for trigeminal neuralgia was revisited. However, the focus of treatment shifted from the ganglion to the nerve root entry zone or the cisternal segment of trigeminal nerve.⁸

A number of centers have since shown the safety and at least short-term pain relief with gamma surgery for trigeminal neuralgia (Fig. 107-1). Maesawa et al. reported complete pain relief without medication in 47.7% of patients at the initial follow-up and in 40% of patients at the last follow-up in 220 patients with a median follow-up of 22 months (range 6–78 months).⁹ In a series of 117 consecutive patients who were followed up for an average of 26 months (range 1–48 months), Pollock et al. reported an actuarial rate of freedom from pain without medication in 57% and 55% of patients at 1 and 3 years, respectively.¹⁰ Tawk et al. in a series of 38 patients followed up for a median of 24 months observed pain relief without medication in 44% at the 3-month evaluation but in only 16% of patients at the 24-month follow-up visit.¹¹ Regis et al. reported a series of 100 patients with a minimum follow-up of 12 months. Fifty-eight patients were pain free without medication.⁸

FIGURE 107-1 A gamma knife treatment plan illustrating the targeting of posterior root of trigeminal nerve on T1-weighted images (A, axial view; B, sagittal view) and constructive interference steady-state images (C, axial view; D, sagittal view). A 4-mm collimator was used. The circle represents the 50% isodose line and encircles the entry zone of the posterior root of the right trigeminal nerve.

The reported rates of recurrence following radiosurgery for trigeminal neuralgia have ranged from 5% to 42% and probably are related to incomplete radiation effects on the targeted tissues (Table 107-1).

Pain free outcomes were usually achieved within several weeks after the gamma procedure. Fountas et al. reported that most patients with no previous surgeries responded within 4 weeks after treatment.¹² Pollock et al. reported that complete pain relief occurred within a median of 3 weeks (range 1–20 weeks).¹⁰ According to Regis et al., initial pain relief occurred after a median delay of 10 days (range 2–5 weeks) in 94% of patients.⁸

The incidence of new trigeminal dysfunction varies in the literature from 6% to 66%. Pollock et al. reported new trigeminal dysfunction in 43 patients (37%) and demonstrated an association between greater radiation doses and the risk of trigeminal nerve dysfunction—90 Gy as a maximum dose caused numbness in 50% of patients, whereas in only 15% of patients who received 70 Gy.¹³ They also reported a strong correlation between the development of new facial sensory loss and achievement and maintenance of pain relief—excellent pain-free outcome 4.5 times more likely in patients with new trigeminal deficits compared with those who had normal postradiosurgical trigeminal nerve function. Regis et al. reported that six patients in their series of 100 cases presented with facial paresthesia and four with facial hyperesthesia.⁸ In Young’s series, facial numbness occurred in 12%, 20%, and 29% of cases when a maximum dose of 70, 80, and 90 Gy, respectively, was used.¹⁴

At the University of Virginia, we have recently reviewed our treatment of 170 cases of typical trigeminal neuralgia with gamma surgery.¹⁵ There were 67 males and 93 females with a mean age of 63.8 years (range 23–95 years). Prior to gamma surgery, 24 patients had prior microvascular decompression, 44 had one or more glycerol injections, radio-frequency thermocoagulation, balloon compression or neurectomy prior to radiosurgery. In each case, the radiosurgical target chosen was 2 to 4 mm anterior to the entry zone of the trigeminal nerve into the pons. A prescription dose of between 30 and 45 Gy to the 50% isodose line was used. The maximum dose ranged from 60 to 90 Gy (median 80 Gy). Excellent outcome was defined as complete pain-free condition without medication. Pain relief of a different degree with or without medication was considered palliative outcome. Follow-up ranged from 6 months to 12 years (mean 4.2 years) (Fig. 107-2).

FIGURE 107-2 A gamma knife treatment plan illustrating the target for mesial temporal lobe epilepsy on T1-weighted images. The amygdala, anterior hippocampus, and parahippocampal gyrus are included in the 50% isodose line.

Thirty-four patients (20%) never had any improvement of their pain following gamma surgery. Fifty-six percent of patients became pain free without medication at some point after treatment. At 1, 2, 3, and 5 years follow-up, 50%, 44%, 42%, and 30% of patients, respectively, were pain free without medication. Twenty patients remained pain free without medication more than 5 years after gamma surgery (range 62–111 months). Palliative results were obtained by 16%, 10%, 8%, and 6%, respectively, at 1, 2, 3, and 5 years follow-up. Patients with two or more percutaneous procedures or neurectomy prior to gamma surgery had less favorable outcome.

Thirty-eight patients (22%) developed postradiosurgical facial numbness. Two patients considered the numbness to be worse than the original facial pain. The fact that they did not have pain relief after gamma surgery presumably contributed to their dissatisfaction. We also observed the association of facial numbness with higher rate of pain free outcome as reported by Pollock et al.¹³ Forty-five patients (26.4%) including 32 who were pain free without medication and 13 who had palliative results after gamma surgery eventually had a recurrence of pain at some point during the follow-up. Twenty-three patients underwent a second gamma procedure. Of 17 patients who were completely pain free for two to 60 months following the initial procedure, 12 became pain free without medication again after retreatment. Four of six patients who had only palliative outcome after the initial procedure became pain free without medication following the second treatment.

Among the treatment alternatives for trigeminal neuralgia, microvascular decompression providing both immediate and long-lasting complete pain relief as well as relatively low incidence of trigeminal nerve dysfunction appears to come closest to what one would expect of a good treatment for trigeminal neuralgia. Although less effective than microvascular decompression, radiosurgery remains a reasonable treatment option for those patients who are unwilling or unable to undergo more invasive surgical approaches and it offers a low risk of mortality or major morbidity.

Gamma Surgery for Movement Disorders

Thalamotomy for tremor in Parkinson’s disease remains one of the most gratifying procedures in functional neurosurgery and defends its place in the therapeutic armory for those common cases in which drugs fail to stop the tremor. To avoid the potential risks of open thalamotomies, the prototype of the gamma knife was used by Leksell¹⁶ to produce thalamic lesions in five cases of tremor between 1968 and 1970. At that time, the intended target could not be visualized but was indirectly determined by using derived coordinates relative to the anterior and posterior commisures visualized by pneumoencephalography. Verification that a lesion had been produced could not be obtained because neither CT nor MRI was available. The fixation of the head of the patient for the radiosurgical procedure was also unsatisfactory, because the stereotactic frame used for target localization was too large to fit into the collimator helmet. Instead, fixation devices were applied onto a plaster-of-Paris helmet previously molded on the patient’s head. It is therefore not surprising that beneficial results were lacking.

In 1986, MRI was introduced at Karolinska Hospital, and better anatomic visualization of the target volume became possible. A new stereotactic frame compatible with MRI that also served as the fixation device in the gamma knife was introduced.¹⁶ These improvements paved the way for new attempts to relieve Parkinsonian tremor by gammathalamotomy, and two cases were treated using this improved methodology.⁷ The first procedure was performed using an 8-mm collimator, and the volume of the resulting lesion was much larger than intended (1.5 cm³). The tremor began to dwindle after 2 months, but a transient hemiparesis and mild speech disturbance ensued secondary to edema. The eventual outcome was, however, satisfactory, and 4 years after the treatment the patient returned free of tremor contralateral to the side of the thalamic lesion, asking for a second procedure to stop the tremor that had developed on the other side. The second patient was treated using a 4-mm collimator, which gave a smaller volume to the thalamic lesion. In this case, the clinical result was not satisfactory. The patient was treated a second time without relief of tremor. It is not clear whether the lack of effect was due to the atypical clinical picture in this patient or to the lack of physiologic corroboration of the target. In spite of experience from centers active in this field indicating that modern imaging techniques, especially MRI, may obviate the need for physiologic target definition, this assertion remains controversial.

Pioneering work in neuroanatomy and neurophysiology by Hirai, Jones, and Ohye has shed much light on the position, anatomic organization, and physiologic significance of the thalamus as it pertains to tremor, rigidity, and dyskinesia.^17–19 MRI guidance for selective thalamotomy in the treatment of Parkinsonian or essential tremors has been well established. The correlations between neuroanatomic and electrophysiologic findings in the human ventrolateral thalamic nuclei (e.g., VLa, VLp, VPLa, VPLc) are better understood. For gamma surgery, the difficulty arises in identifying the VLp and VLa nuclei in the human thalamus purely by radiologic methods. As such, thin-slice MRI, Surgiplan, and a neuroanatomic atlas may be required to treat these nuclei with the gamma knife.

Rand treated 18 cases of movement disorder with radiosurgery (R. Rand, personal communication, 1994). Of the seven patients with resting tremor, four responded to a nucleus ventralis lateralis (NVL) lesion with marked improvement in the tremor, and in two patients, rigidity improved as well. Eight other patients underwent radiosurgical pallidotomy for rigidity, and four had significant improvement. Two of three patients treated with a NVL lesion for intention tremor showed dramatic improvement.

Ohye (C. Ohye, personal communications, 2004) has performed gammathalamotomies on 56 patients for Parkinson’s disease, 21 with essential tremor, and 6 with intention tremor. Thalamotomies were performed using a single 4-mm shot and 130-Gy dose. Follow-up MRI revealed two different types of thalamic changes. One type was a round, punched-out lesion with a volume of less than 100 mm³, and the other was an irregular high-signal zone (volume up to 800 mm³