Multimodality correlative study of canine brain tumors: Proton magnetic resonance spectroscopy, positron emission tomography, and histology

RATIONALE AND OBJECTIVES. Structural/functional relationships in an induced canine brain tumor were studied using protonTmagnetic resonance spectroscopy ('H-MRS), positron emission tomography (PET), and histology. METHODS. Proton-MRS and PET data of implanted canine brain tumors were correlated with quantitative analysis of the tissue composition within the MRS and PET regions of interest (ROIs). Linear regression analysis was employed to correlate the 'H-MRS and PET data with the percent tumor and the percent total lesion (comprising tumor plus associated pathology ie, edema, cysts, hemorrhage, inflammation) within the ROI.RESULTS. Using 'H-MRS, N-acetyl aspartate concentrations were indirectly correlated with the amount of tumor (P = .058), as well as the amount of tumor plus associated pathology (P = .032) within the ROI. Total creatine concentrations were indirectly correlated with the amount of tumor and the amount of tumor plus associated pathology within the ROI (P < .05). Lactate concentrations were directly correlated with the amount of tumor (P == .053) and the amount of tumor plus associated pathology (P = .058) within the ROI. Using PET, Oxygen metabolic rates were indirectly correlated with the amount of tumor and with the amount of tumor plus associated pathology within the ROI (P < .05). Glucose metabolic rates were directly correlated with both the amount of tumor and with the amount of tumor plus associated pathology at P < .05. Proton-MRS measured concentrations of choline and PET measured values for blood flow, and oxygen extraction showed correlations with the amount of tumor and with the amount of tumor plus associated pathology at P≥ .080. CONCLUSIONS. The PET and MRS data were complementary with respect to suggesting anaerobic glucose metabolism for the tumor. Unlike other tumors, no increase in choline was noted in the canine tumor .