Background: Adjuvant chemotherapy benefits a minority of patients with recurrent glioblastoma; a critical question therefore is how to select such responders. Metabolic reaction of the tumor, studied by positron emission tomography (PET) with [18F]-fluoro-2-deoxy-D-glucose (FDG), may be a suitable test of tumor responsiveness. In an effort to evaluate their prognostic value, the early effects of carmustine on brain and recurrent glioblastoma glucose metabolism were studied.
Methods: Positron emission tomography with [18F]fluoro-2-deoxy-D-glucose was used to study 10 patients with recurrent glioblastoma before and 24 hours after a first dose of carmustine (120 mg/m2). Absolute quantification of glucose metabolism was obtained, and metabolic changes were analyzed and confronted with survival duration.
Results: The mean ratios of pre- to postchemotherapy values were 0.98 +/- 0.06 (range, 0.90-1.06) for the cortex and 0.97 +/- 0.06 (range, 0.89-1.09) for the white matter. There was no statistical difference between pre- and post-chemotherapy metabolism. In the glioblastoma, individual ratios of pre- to postchemotherapy values were considerably more scattered than in the nonneoplastic tissues, ranging from 0.76 to 1.51. A significant positive correlation (P < 0.002) was found between post- to pre-chemotherapy ratios of glucose metabolism and survival length.
Conclusions: In patients with recurrent glioblastoma, the first administration of carmustine produces no acute change of glucose metabolism in the noninvolved cerebral tissues. Acute changes of glucose metabolism in tumor tissue in response to this treatment varied, and a hypermetabolic reaction possibly predicts longer survival.