Optimizing Outcomes With Maximal Surgical Resection of Malignant Gliomas

Stephen J. Hentschel, MD, Raymond Sawaya, MD


Cancer Control. 2003;10(2) 

In This Article

Benefits of Maximal Resection

A recent study at The University of Texas M. D. Anderson Cancer Center found that a resection of 98% or more of the tumor volume was an independent variable associated with longer survival in patients with glioblastoma multiforme (GBM) ( Table 1 ).[2] The median survival for these patients was 13.4 months compared with 8.8 months for patients who had lesser resections (P<.0001). The pre-and post-operative magnetic resonance imaging (MRI) scans of these 416 patients with GBM were analyzed prospectively and volumetric data were collected to determine the extent of resection as accurately as possible.[3] This eliminated a drawback encountered in many studies in which nonquantitative and subjective descriptions, such as "gross total,""subtotal," and "partial," were used to describe the extent of surgical resection.[4,5,6] A multivariate analysis identified five independent predictors of survival, three of which were chosen for the outcome scale: tumor necrosis, patient age, and Karnofsky score ( Table 2 ). Using this analysis, groups were created to allow survival estimation based on the scale according to the extent of resection ( Table 3 ). For patients under 45 years of age with a Karnofsky score above 80 and no necrosis on their MRI scan, it is likely that these three factors outweigh the influence of the extent of resection; thus, a statistical difference was not found between the group with 98% or greater resection and the group with lesser resection. Also, the numbers for this younger group were too small to allow meaningful statistical analyses. The median survival in this group was nearly 3 years (35.3 months). A similar argument can also be applied to patients ≥65 years of age with a Karnofsky score of less than 80 and with necrosis on their MRI. Unfortunately, the median survival in this group of patients was less than 1 year (8.6 months).

This was not a randomized, controlled study, and thus the patients may have been partially "preselected" for better survival. That is, patients may have been selected for surgery if the treating surgeon believed that a maximal resection of the tumor was possible, whereas patients who were believed to have tumors that were not maximally resectable may not have been selected for surgery. Therefore, it is difficult to determine from a retrospective study whether the characteristics of the tumor or the extent of resection imparts a better prognosis in these patients. However, these two variables are closely related, and differentiating between them may be irrelevant from a practical perspective.

Malignant gliomas are rapidly growing tumors that can reach large sizes prior to detection. In the M. D. Anderson series,[2] the median preoperative tumor volume at presentation was 34 cm3. Compared with patients with lesser resections, those undergoing gross total resections have been shown to have better neurologic outcomes on follow-up without added perioperative morbidity or mortality.[1,7,8] In addition, with the significant reduction in mass effect, it is our experience that patients with maximal resections tolerate radiotherapy better and experience fewer side effects.

Fewer cells remaining after resection of a tumor should improve prognosis because fewer cells will need to be eliminated by either adjuvant therapies or the body's own defenses.[9] Despite the fact that a 99% resection of a tumor mass consisting of 109 cells leaves 107 cells alive, a reduction in the number of cells that may be resistant to therapy, that may undergo further malignant transformation, or that may release growth factors and immune inhibiting agents is potentially advantageous. The limits of adjuvant therapies in terms of cellular thresholds of effectiveness are not known. Also, an important consideration is the ability to provide tissue from tumor samples for research purposes. Samples from human brain tumor tissue are needed to develop new therapies since it is unlikely that significant advances toward a cure for glioblastoma will be achieved using animal models alone.

Although some authors contend that stereotactic biopsy can be used to guide the management of gliomas,[10,11] the discrepancy rate between the biopsy and the resected specimens is 38%, even with expert neuropathologic review.[12] These results are from a study of 81 consecutive patients at the M. D.Anderson Cancer Center, which found that the resulting discrepancy would have affected therapeutic regimens in 26% of cases and had prognostic implications in 38% of cases. Particularly in cases where a surgical resection of the lesion is possible, little is gained by performing a stereotactic biopsy prior to the open craniotomy unless an abscess, lymphoma, or other nonneoplastic lesion is likely.