Imaging advances for the evaluation of brain tumors and neuroinflammation
The measurement of metabolic and cell physiology parameters with improved MRI techniques provides more detailed information on tissue biochemical composition and blood perfusion, facilitating the distinction between tumors and necrosis. Proton magnetic resonance (MR) spectroscopic imaging is used to evaluate cell metabolism through the detection of proton metabolites and their distribution. Chemicals detected include compounds containing choline, creatine, lactate, lipid, and N-acetylaspartate. Choline, a component of cell membrane phospholipids, is particularly useful for differentiating tumors from non-tumoral tissues because its content is increased in highly proliferative cell populations. Overall, higher choline levels are associated with disease progression or recurrence, whereas low levels of choline are found in necrotic lesions. The analysis of increased ratios of choline content in relation to other chemicals can result in an accuracy of up to 97% in separating tumors from non-tumoral necrotic tissues [31, 42].
Conventional MRI, with the T1/T2 mismatch criterion, had a specificity of 75% and a sensitivity of only 44% in distinguishing between tumors and inflamed lesions. PET scan combined the best sensitivity and specificity, respectively of 92% and 69%. PET remained superior compared to NMR spectroscopy for choline/N-acetylaspartate and choline/creatin ratios across different thresholds [45]. A retrospective study of 57 GBM patients examined with T2*-weighted dynamic susceptibility-weighted contrast material–enhanced (DSC) MRI found mean, maximum, and minimum relative peak height and relative cerebral blood volume were significantly higher in GBM cases compared to radiation necrosis cases. In contrast, mean, maximum, and minimum relative percentage of signal intensity recovery values were significantly lower in recurrent GBM compared to radiation necrosis [46]. Proton MR spectroscopy showed a temporary elevation of choline in 4 of 9 cases of necrosis, creating a confounding factor that could result in false positive findings for tumor recurrence [47]. Another study using proton MR spectroscopy in 11 patients who received high-dose radiotherapy revealed that cases of radiation necrosis had either increased lactate/creatine and phosphocreatine (Cr) ratio and decreased choline-containing compounds/phosphocreatine ratio compared to recurrent GBM, or reductions in all major metabolites [48]. A meta-analysis of 397 patients in 13 studies examined roles of several metabolites. MR spectroscopy and MR perfusion using Cho/NAA and Cho/Cr ratios and rCBV may increase the accuracy of differentiating necrosis from recurrent tumor in patients with primary brain tumors or brain metastases [49].
A meta-analysis of 6 studies with 118 patients and 134 scans indicated11C-choline PET as an accurate diagnostic method for the differentiation of tumor relapse from radiation induced necrosis in gliomas [50]. A study with 55 patients, followed up for at least 11 months, with suspected brain tumor recurrence or necrosis after radiotherapy, examined MRI, (18)F-FDG, and11C-choline PET/CT, concluding for the superiority of11C-choline PET/CT [51]. In a F98 orthotopic rat model of GBM, PET using (18)F-FDG and (18)F-FET PET were effective in discriminating GBM from radiation necrosis, with (18)F-FDG delayed PET being particularly useful [52]. A study with 50 patients showed that (11)C-methionine-PET was superior to both (11)C-choline and (18)F-FDG -PET for distinguishing GBM recurrence from radiation necrosis [53]. The LAT1 tumor-specific PET tracer 2-[18F]FELP PET is able to differentiate glioblastoma from radiation necrosis, and 2-[18F]FELP uptake is less likely to be contaminated by the presence of inflammation than the [18F]FDG signal [35].