OBJECTIVE: To determine whether histogram values of the normalized apparent diffusion coefficient (nADC) and normalized cerebral blood volume (nCBV) maps obtained in contrast-enhancing lesions detected on immediate post-operative MR imaging can be used to predict the patient response to concurrent chemoradiotherapy (CCRT) with temozolomide (TMZ). MATERIALS AND METHODS: Twenty-four patients with GBM who had shown measurable contrast enhancement on immediate post-operative MR imaging and had subsequently undergone CCRT with TMZ were retrospectively analyzed. The corresponding histogram parameters of nCBV and nADC maps for measurable contrast-enhancing lesions were calculated. Patient groups with progression (n = 11) and non-progression (n = 13) at one year after the operation were identified, and the histogram parameters were compared between the two groups. Receiver operating characteristic (ROC) analysis was used to determine the best cutoff value for predicting progression. Progression-free survival (PFS) was determined with the Kaplan-Meier method and the log-rank test. RESULTS: The 99th percentile of the cumulative nCBV histogram (nCBV C99) on immediate post-operative MR imaging was a significant predictor of one-year progression (p = 0.033). ROC analysis showed that the best cutoff value for predicting progression after CCRT was 5.537 (sensitivity and specificity were 72.7% and 76.9%, respectively). The patients with an nCBV C99 of < 5.537 had a significantly longer PFS than those with an nCBV C99 of ≥ 5.537 (p = 0.026). CONCLUSION: The nCBV C99 from the cumulative histogram analysis of the nCBV from immediate post-operative MR imaging may be feasible for predicting glioblastoma response to CCRT with TMZ.
Adult ; Aged ; Antineoplastic Agents, Alkylating/*therapeutic use ; Brain/pathology/radiography ; Brain Neoplasms/*drug therapy/mortality/radiography ; Chemoradiotherapy ; Dacarbazine/*analogs & derivatives/therapeutic use ; Diffusion Magnetic Resonance Imaging ; Disease Progression ; Disease-Free Survival ; Female ; Glioblastoma/*drug therapy/mortality/radiography ; Humans ; Kaplan-Meier Estimate ; Male ; Middle Aged ; Proportional Hazards Models ; ROC Curve ; Retrospective Studies