Gamma knife radiosurgery for metastatic melanoma: an analysis of survival, outcome, and complications.

OBJECTIVE: Although the mainstays for treatment of metastatic brain disease have been surgery and/or external beam radiation therapy, an increasing number of patients are being referred for stereotactic radiosurgery as the primary intervention for their intracranial pathological abnormalities. The lack of efficacy and cognitive and behavioral consequences of whole brain irradiation have prompted clinicians to select patients for alternative therapies. This study analyzes the effectiveness of Leksell gamma unit therapy for metastatic melanoma to the brain.

METHODS: We present our experience with 59 Leksell gamma unit treatment sessions in 45 consecutive patients who presented with metastatic melanoma to the brain. Five of these procedures were performed as salvage therapy for patients who needed second radiosurgical treatment for new lesions that were remote from the previous targets and were not included in the overall analyses.

RESULTS: The population included 78% male patients. The mean patient age was 53 years (age range, 24-80 yr). The mean time from diagnosis of primary melanoma to discovery of brain metastasis was 43 months (median, 27.5 mo; range, 1-180 mo). At the time of diagnosis of brain disease, 35.5% of the patients (16 of 45 patients) had neurological symptoms, 77.7% (35 of 45 patients) had known visceral metastases, and 11.1% (5 of 45 patients) had seizure disorders. Eighty-six percent of the lesions (80 of 93 lesions) were cortical, 12% (11 of 93 lesions) were cerebellar, 1% (1 of 93 lesions) were pontine, and 1% (1 of 93 lesions) were thalamic. Fifty-seven percent of the sessions (31 of 54 sessions) were performed for a single lesion, 24.1% (13 of 54 sessions) for two lesions, 9.2% (5 of 54 sessions) for three lesions, 7.4% (4 of 54 sessions) for four lesions, and 1.8% (1 of 54 sessions) for five lesions. The mean treatment volume was 5.6 cc, with a mean prescription of 21.6 Gy to the 56.0% mean isodose line. The median survival time of the patients in our population, using Kaplan-Meier curves, was 43 months from the time of diagnosis of primary melanoma (range, 3-180 mo) and 8 months (range, 1-20 mo) from the time of gamma knife treatment. Complications included seizures within 24 hours of the procedure in four patients, with transient nausea and vomiting in three patients, transient worsening of preprocedure paresis responsive to steroids in three patients, and increased confusion in one patient. All 45 patients were located for follow-up (mean follow-up duration, 1 yr). After gamma knife treatment, 78% of the patients (35 of 45 patients) experienced either improved or stable neurological symptomatology before death or at the time of the latest follow-up examination. There were 26 deaths (58%). The cause of death was determined to be neurological in only 2 of 45 patients (7.7%). Follow-up magnetic resonance images revealed a 97% local tumor control rate of gamma knife-treated lesions, with 28% radiographic disappearance (9 of 32 cases). Six patients developed new lesions remote from radiosurgical targets and underwent second procedures.

CONCLUSION: Although metastatic melanoma to the brain continues to have a foreboding prognosis for long-term survival, gamma knife radiosurgery seems to be a relatively safe, noninvasive, palliative therapy, halting or reversing neurological progression in 77.8% of treated patients (35 of 45 patients). The survival rate matches or exceeds those previously reported for surgery and other forms of radiotherapy. Only 7.7% of the patients in our study population who died as a result of metastatic melanoma (2 of 26 patients) died as a result of neurological disease. The routine use of therapeutic level antiseizure medication is emphasized, considering the findings of our review.