Objective: This study used quantitative radiological imaging to determine the effect of surgical resection on postoperative survival of patients with malignant astrocytomas. Previous studies relied on the surgeons' impressions of the amount of tumor removed, which is a less reliable measure of the extent of resection.
Methods: Information concerning possible prognostic factors was collected for 75 patients undergoing magnetic resonance imaging or computed tomography preoperatively and within 10 days postoperatively. Image analysis of the neuroradiological studies was conducted to quantify pre- and postoperative total tumor volumes and enhancing volumes. Univariate and multivariate proportional hazards models were used to analyze the regression of survival regarding 22 covariates that might affect survival. The covariates that were entered included age, gender, tumor grade, cumulative radiation dose, chemotherapy, seizures as a first symptom, Karnofsky performance status at presentation, pre- and postoperative total and enhancing tumor volumes, ratio of pre- to postoperative total and enhancing tumor volumes, tumor location, surgeon's impression of the degree of resection, and subsequent surgery.
Results: There were 23 patients with anaplastic astrocytomas and 52 with glioblastomas multiforme. The estimated mean survival time was 27 months for patients undergoing gross total resection, 33 months for subtotal resection, and 13 months for open or stereotactic biopsy. Five factors that were significant predictors of survival in multivariate analysis were tumor grade, age, Karnofsky performance status, radiation dose, and postoperative complications (P < 0.05). In univariate analysis, tumor grade, radiation dose, age, Karnofsky status, complications, presence of enhancing tumor in postoperative imaging, and postoperative volume of enhancing tumor were significantly associated with survival (P < 0.05).
Conclusion: We conclude that the most important prognostic factors affecting survival of patients with anaplastic astrocytomas and glioblastomas multiforme are tumor grade, age, preoperative performance status, and radiation therapy. Postoperative complications adversely affect survival. Aggressive surgical resection did not impart a significant increase in survival time. Surgical resection may improve survival, but its importance is less than that of other factors and may be demonstrable only by larger studies.