Extent of surgical resection of malignant astrocytomas of the spinal cord: Outcome analysis of 35 patients

OBJECTIVE: The optimal management of malignant intramedullary spinal cord astrocytomas remains controversial. Although radiotherapy has become the standard of care, the relationship between extent of resection and survival remains unclear. We report the outcomes of the surgical management of 35 malignant spinal cord astrocytomas and assess the association of extent of resection with survival after aggressive resection of these tumors. METHODS: An institutional intramedullary spinal cord tumor database (1990-2002) was reviewed to identify all patients treated for malignant astrocytomas of the spinal cord (anaplastic astrocytoma [AA] or glioblastoma multiforme [GBM]). Length of survival from surgery was charted by Kaplan-Meier plots, and association of extent of resection with Survival was assessed via log rank analysis for stratified covariates and Cox proportional-hazards model for continuous covariates. RESULTS: Twenty-seven (77%) and eight (23%) patients underwent resection of AA and GBM, respectively. Mean age was 29 +/- 16 years (range, 2-61 yr). Tumor involved six four vertebral levels. For AA cases, radical resection (no residual postoperative magnetic resonance enhancement) was achieved in 12 (44%) patients and subtotal resection (residual postoperative magnetic resonance enhancement) was achieved in 15 (56%). No GBM patients underwent radical resection (mean estimated resection, 70%). After surgery, two (6%) patients improved neurologically by modified McCormick score 19 (54%) remained stable, and 14 (40%) declined. Median overall survival for AA patients was 72 months (85% at 1 yr; 59% at 5 yr). Median overall survival for GBM patients was 9 months (31% at 1 yr; 0% at 5 yr). Subtotal versus radical resection of AA was associated with decreased overall survival (38 versus 78% at 4 yr, P = 0.028). Postoperative tumor dissemination was associated with decreased survival (P = 0.004). When adjusting for multiple comparisons (P < 0.006 needed for significance), a trend of increased survival was observed with radical resection (P = 0.023). CONCLUSION: Neurological function can be preserved with aggressive resection of malignant intramedullary spinal astrocytomas; however, motor decline may be observed in many cases. Radical resection of AA was associated with a trend of increased overall survival in nondisseminated AA cases. Radical surgery and radiotherapy of GBM was associated with poor survival, similar to historical controls of diagnostic biopsy and radiotherapy. A markedly shorter survival may be expected in cases in which AA disseminates along the neuraxis. Biopsy alone may not provide the best outcomes for patients with malignant spinal cord tumors.