Gamma knife radiosurgery for treatment of cerebral metastases from non-small-cell lung cancer.

Purpose: To evaluate clinical and physico-dosimetric variables affecting clinical outcome of patients treated with Gamma Knife radiosurgery (GKRS) for brain metastases from non-small cell lung cancer (NSCLC).

Methods And Materials: Between 2001 and 2006, 373 patients (298 men and 75 women, median age 65 years) with brain metastases from NSCLC underwent GKRS. All of them had KPS ≥ 60%, eight or fewer brain metastases, confirmed histopathological diagnosis and recent work-up (<3 months). Thirty-five patients belonged to recursive partitioning analysis (RPA) Class I, 307 patients were in RPA Class II, 7 patients were in RPA Class III. Median tumor volume was 3.6 cm(3). Median marginal dose was 22.5 Gy at 50% isodose.; median 10 Gy and 12 Gy isodose volumes were 30.8 cm(3) and 15.8 cm(3), respectively. Follow-up with MRI was performed every 3 months. Overall survival data were collected from internal database, telephone interviews, and identifying registries.

Results: Mean follow-up after GKRS was 51 months (range, 6 to 96 months); mean overall survival was 14.2 months. Of 373 patients, 29 were alive at time of writing, 104 had died of cerebral progression, and 176 had died of systemic progression. In 64 cases it was not possible to ascertain the cause. Univariate and multivariate analysis were adjusted for the following: RPA class, surgery, WBRT, age, gender, number of lesions, median tumor volume, median peripheral dose, and 10 Gy and 12 Gy volumes. Identified RPA class and overall tumor volume >5 cc were the only two covariates independently predictive of overall survival in patients who died of cerebral progression.

Conclusions: Global volume of brain disease should be the main parameter to consider for performing GKRS, which is a first-line therapy for patient in good general condition and controlled systemic disease.