Role of adjuvant or salvage radiosurgery in the management of unresected residual or progressive glioblastoma multiforme in the pre-bevacizumab era

OBJECT After initial standard of care management of glioblastoma multiforme (GBM), relatively few proven options remain for patients with unresected progressive tumor. Numerous reports describe the value of radiosurgery, yet this modality appears to remain underutilized. The authors analyzed the outcomes of early adjuvant stereotactic radiosurgery (SRS) for unresected tumor or later salvage SRS for progressive GBM. Radiosurgery was performed as part of the multimodality management and was combined with other therapies. Patients continued to receive additional chemotherapy after SRS and prior to progression being documented. In this retrospective analysis, the authors evaluated factors that affected patient overall survival (OS) and progression-free survival. METHODS Between 1987 and 2008 the authors performed Gamma Knife SRS in 297 patients with histologically proven GBMs. All patients had received prior fractionated radiation therapy, and 66% had undergone one or more chemotherapy regimens. Ninety-six patients with deep-seated unresectable GBMs underwent biopsy only. Of those in whom excision had been possible, resection was considered to be gross total in 68 and subtotal in 133. The median patient age was 58 years (range 23-89 years) and the median tumor volume was 14 cm(3) (range 0.26-84.2 cm(3)). The median prescription dose delivered to the imaging-defined tumor margin was 15 Gy (range 9-25 Gy). The median follow-up duration was 8.6 months (range 1.1-173 months). Cox regression models were used to analyze survival outcomes. Variables examined included age, residual versus recurrent tumor, prior chemotherapy, time to first recurrence, SRS dose, and gross tumor volume. RESULTS The median survival times after radiosurgery and after diagnosis were 9.03 and 18.1 months, respectively. The 1-year and 2-year OS after SRS were 37.9% and 16.7%, respectively. The 1-year and 2-year OS after diagnosis were 76.2% and 30.8%, respectively. Using multivariate analysis, factors associated with improved OS after diagnosis were younger age (< 60 years) at diagnosis (p < 0.0001), tumor volume < 14 cm(3) (p < 0.001), use of prior chemotherapy (p = 0.001), and radiosurgery at the time of recurrence (p < 0.0001). Multivariate analysis showed that younger age (p < 0.0001) and smaller tumor volume (< 14 cm(3)) (p = 0.001) were significantly associated with increased OS after SRS. Adverse radiation effects were seen in 69 patients (23%). Fifty-eight patients (19.5%) underwent additional resection after SRS. The median survivals after diagnosis for recursive partitioning analysis Classes III, IV and V+VI were 31.6, 20.8, and 16.7 months, respectively. CONCLUSIONS In this analysis 30% of a heterogeneous cohort of GBM patients eligible for SRS had an OS of 2 years. Radiosurgery at the time of tumor progression was associated with a median survival of 21.8 months. The role of radiosurgery for GBMs remains controversial. The findings in this study support the need for a funded and appropriately designed clinical trial that will provide a higher level of evidence regarding the future role of SRS for glioblastoma patients in whom disease has progressed despite standard management.