Objective: We performed a prospective cohort study in a homogeneous patient group with proven glioblastoma who received uniform therapy to determine the prognostic value and clinical correlates of [18F]fluorodeoxyglucose (FDG) PET at different stages of the disease.
Materials and methods: Fifteen newly diagnosed patients with glioblastoma, aged 52 +/- 9 years and with Karnofsky performance score (KPS) of at least 50, were treated with surgical resection, chemotherapy (nimustine), and radiotherapy. They were followed prospectively for 2 years. Clinical data (e.g., tumor progression, performance status, and survival) were recorded and glucose metabolism was measured with PET and FDG before and during radiochemotherapy.
Results: Median survival of all patients was 13 months and four patients were alive 24 months after surgery. All tumors were hypermetabolic compared with normal white matter. A metabolic index was calculated as the ratio of maximum residual tumor metabolism to contralateral normal brain metabolism. It was of prognostic value for patient survival and tumor recurrence already in the first postoperative and all following PET studies and was correlated with the KPS. Reduction of contralateral brain metabolism was more closely related to prognosis than the tumor metabolism proper.
Conclusion: These results underscore the prognostic relevance of metabolically active residual tumor tissue after surgical resection.