Thirty patients with primary cerebral tumors WHO III and IV previously treated, undergoing evaluation for tumoral recurrence, they underwent (18)FDG-PET study, MRI and PMRI. PET uptake was determined by visual inspection and was quantified by use of standard uptake values, the ratio of tumor uptake to normal tissue and were z scored using automated voxel-based comparison. PMRI was quantified by use of ratios of cerebral blood volume (rCBV). The accuracies were determined by comparing imaging data with histologic findings and clinical follow up of up to 21 mo.
Results: Sensitivity (Se), specificity (Sp), positive predictive value (PPV), negative predictive value (NPV) and accuracy were 100 %, 82 %, 90 %, 100 % and 93 % respectively for the PET/MRI fusion and 68 %, 82 %, 87 %, 60 % and 73 % respectively for PMRI. There were two false positive cases for PET/MRI fusion that were confirmed by biopsy: chronic inflammation; and foreign body granulomas. The receiver operating characteristic (ROC) curve analysis showed statistically significant difference (p = 0.0225).
Conclusions: (18)FDG SUVs, glucose uptake ratios and 3D stereotactic surface projections in brain tumors were not a reliable measure for evaluating recurrent tumors. PET/MRI fusion was more sensitive and accurate than PMRI for imaging recurrent primary brain tumors. The region of interest can be visually analyzed on the PET/MRI fusion images and described as recurrent tumor when any activity (lower, equal or greater than the contralateral cortex) is presented in the zone of hyperintensity seen on the post-gadolinium T1-weighted MRI.