To integrate 3D MR spectroscopy imaging (MRSI) in the treatment planning system (TPS) for glioblastoma dose painting to guide simultaneous integrated boost (SIB) in intensity-modulated radiation therapy (IMRT).

Glioblastoma, a high-grade glial infiltrating tumor, is the most frequent malignant brain tumor in adults and carries a dismal prognosis. External beam radiotherapy (EBRT) increases overall survival but this is still low due to local relapses, mostly occurring in the irradiation field. As the ratio of spectra of choline/N acetyl aspartate> 2 (CNR2) on MR spectroscopic imaging has been described as predictive for the site of local relapse, we hypothesized that dose escalation on these regions would increase local control and hence global survival. In this multicenter prospective phase III trial for newly diagnosed glioblastoma, 220 patients having undergone biopsy or surgery are planned for randomization to two arms. Arm A is the Stupp protocol (EBRT 60 Gy on contrast enhancement + 2 cm margin with concomitant temozolomide (TMZ) and 6 months of TMZ maintenance); Arm B is the same treatment with an additional simultaneous integrated boost of intensity-modulated radiotherapy (IMRT) of 72Gy/2.4Gy delivered on the MR spectroscopic imaging metabolic volumes of CHO/NAA > 2 and contrast-enhancing lesions or resection cavity. Stratification is performed on surgical and MGMT status. This is a dose-painting trial, i.e. delivery of heterogeneous dose guided by metabolic imaging. The principal endpoint is overall survival. An online prospective quality control of volumes and dose is performed in the experimental arm. The study will yield a large amount of longitudinal multimodal MR imaging data including planning CT, radiotherapy dosimetry, MR spectroscopic, diffusion and perfusion imaging. NCT01507506 , registration date December 20, 2011.

In this study, a radiomics analysis was conducted to provide insights into the differentiation of radionecrosis and tumor progression in multiparametric MRI in the context of a multicentric clinical trial. First, the sensitivity of radiomic features to the unwanted variability caused by different protocol settings was assessed for each modality. Then, the ability of image normalization and ComBat-based harmonization to reduce the scanner-related variability was evaluated. Finally, the performances of several radiomic models dedicated to the classification of MRI examinations were measured. Our results showed that using radiomic models trained on harmonized data achieved better predictive performance for the investigated clinical outcome (balanced accuracy of 0.61 with the model based on raw data and 0.72 with ComBat harmonization). A comparison of several models based on information extracted from different MR modalities showed that the best classification accuracy was achieved with a model based on MR perfusion features in conjunction with clinical observation (balanced accuracy of 0.76 using LASSO feature selection and a Random Forest classifier). Although multimodality did not provide additional benefit in predictive power, the model based on T1-weighted MRI before injection provided an accuracy close to the performance achieved with perfusion.

Intensity-modulated radiation therapy (IMRT) for total body irradiation (TBI) is practiced in several centers using the TomoTherapy System. In this context the planning target volume (PTV) is the entire body including the skin. A safety margin in the air surrounding the body should be added to take into account setup errors. But using inverse planning, over-fluence peak could be generated in the skin region to insure dose homogeneity. This work proposes to study the performance of the use of a virtual bolus (VB). A VB is a material placed on the skin surface during planning, but absent for the real treatment. The optimal VB that compensates large setup errors without introducing a high-dose increase or hot spots for small setup errors was determined. For two cylindrical phantoms, 20 VBs with different densities, thicknesses or designs were tested. Dose coverage of the PTV (V95%) in the presence of simulated setup errors was computed to assess the VB performance. A measure of the dose increase in the phantom center due to the absence of the VB during treatment was also achieved. Finally, the fluence peak at the phantom edge was measured in complete buildup conditions using a large phantom and a detector matrix. Using these VBs, simulated setup errors were compensated to a minimum value of 2.6 and 2.1 cm for small and large phantom, respectively (and only 1.2 and 1.7 cm with no VB). An optimal double-layer VB was found with a density of 0.4?\text{kg}.{\mathrm{m}}^{?3} and a total thickness of 8 mm; an inner layer of 5 mm was declared as the target for the treatment planning system and an additional layer of 3 mm was added to avoid the over-fluence peak. Using this VB, setup errors were compensated up to 2.9 cm. The dose increase was measured to be only +1.5\% at the phantom center and over-fluence peak was strongly decreased. PACS numbers: 87.53 Bn, 87.55 D-, 87.55 de, 87.55 dk