Background: Previously, we demonstrated the high accuracy of alpha-[(11)C]methyl-L-tryptophan (AMT) PET for differentiating recurrent gliomas from radiation injury. The present study evaluated the prognostic value of increased AMT uptake in patients with previously treated high-grade glioma.
Methods: AMT-PET was performed in 39 patients with suspected recurrence of World Health Organization grades III-IV glioma following surgical resection, radiation, and chemotherapy. Mean and maximum standardized uptake values (SUVs) and unidirectional AMT uptake (K) were measured in brain regions suspicious for tumor and compared with the contralateral cortex (ie, background). Optimal cutoff thresholds for 1-year survival prediction were determined for each AMT parameter and used for calculating the prognostic value of high (above threshold) versus low (below threshold) values for post-PET overall survival (OS).
Results: In univariate analyses, 1-year survival was strongly associated with 3 AMT parameters (SUVmax, SUVmean, and tumor-to-background K-ratio; odds ratios: 21.3-25.6; P ≤ .001) and with recent change in MRI contrast enhancement (odds ratio: 14.7; P = .02). Median OS was 876 days in the low- versus 177 days in the high-AMT groups (log-rank P < .001). In multivariate analyses, all 3 AMT parameters remained strong predictors of survival: high AMT values were associated with unfavorable 1-year survival (binary regression P ≤ .003) and shorter overall survival in the whole group (Cox regression hazard ratios: 5.3-10.0) and in patients with recent enhancement change on MRI as well (hazard ratios: 7.0-9.3; P ≤ .001).
Conclusion: Increased AMT uptake on PET is highly prognostic for 1-year and overall survival, independent of MRI contrast enhancement and other prognostic factors in patients with a previously treated high-grade glioma.
Keywords: MRI; amino acid PET; high-grade glioma; survival; tryptophan.
© The Author(s) 2014. Published by Oxford University Press on behalf of the Society for Neuro-Oncology. All rights reserved. For permissions, please e-mail: [email protected].