Increased relative biological effectiveness and periventricular radiosensitivity in proton therapy of glioma patients.

PURPOSE: Currently, there is an intense debate on variations in intra-cerebral radiosensitivity and relative biological effectiveness (RBE) in proton therapy of primary brain tumours. Here, both effects were retrospectively investigated using late radiation-induced brain injuries (RIBI) observed in follow-up after proton therapy of patients with diagnosed glioma. METHODS: In total, 42 WHO grade 2-3 glioma patients out of a consecutive patient cohort having received (adjuvant) proton radio(chemo)therapy between 2014 and 2017 were eligible for analysis. RIBI lesions (symptomatic or clinically asymptomatic) were diagnosed and delineated on contrast-enhanced T1-weighted magnetic resonance imaging scans obtained in the first two years of follow-up. Correlation of RIBI location and occurrence with dose (D), proton dose-averaged linear energy transfer (LET) and variable RBE dose parameters were tested in voxel- and in patient-wise logistic regression analyses. Additionally, anatomical and clinical parameters were considered. Model performance was estimated through cross-validated area-under-the-curve (AUC) values. RESULTS: In total, 64 RIBI lesions were diagnosed in 21 patients. The median time between start of proton radio(chemo)therapy and RIBI appearance was 10.2 months. Median distances of the RIBI volume centres to the cerebral ventricles and to the clinical target volume border were 2.1 mm and 1.3 mm, respectively. In voxel-wise regression, the multivariable model with D, D × LET and periventricular region (PVR) revealed the highest AUC of 0.90 (95 % confidence interval: 0.89-0.91) while the corresponding model without D × LET revealed a value of 0.84 (0.83-0.86). In patient-level analysis, the equivalent uniform dose (EUD11, a = 11) in the PVR using a variable RBE was the most prominent predictor for RIBI with an AUC of 0.63 (0.32-0.90). CONCLUSIONS: In this glioma cohort, an increased radiosensitivity within the PVR was observed as well as a spatial correlation of RIBI with an increased RBE. Both need to be considered when delivering radio(chemo)therapy using proton beams.

Impact of range uncertainty on clinical distributions of linear energy transfer and biological effectiveness in proton therapy.

PURPOSE: Increased radiation response after proton irradiation, such as late radiation-induced toxicity, is determined by high dose and elevated linear energy transfer (LET). Steep dose-averaged LET (LETd ) gradients and elevated LETd occur at the end of proton range and might be particularly sensitive to uncertainties in range prediction. Therefore, this study quantified LETd distributions and the impact of range uncertainty in robust dose-optimized proton treatment plans and assessed the biological effect in normal tissues and tumors of patients. METHODS: For each of six cancer patients (two brain, head-and-neck, and prostate), two nominal treatment plans were robustly dose optimized using single- and multi-field optimization, respectively. For each plan, two additional scenarios with ±3.5% range deviation relative to the nominal plan were derived by global rescaling of stopping-power ratios. Dose and LETd distributions were calculated for each scenario using the beam parameters of the corresponding nominal plan. The variability in relative biological effectiveness (RBE) and probability of late radiation-induced brain toxicity (PIC ) was assessed. RESULTS: The optimization technique (single- vs multi-field) had a negligible impact on the LETd distributions in the clinical target volume (CTV) and in most organs at risk (OARs). LETd distributions in the CTV were rather homogeneous with arithmetic mean of LETd below 3.2 keV/µm and robust against range deviations. The RBE variability within the CTV induced by range uncertainty was small (≤0.05, 95% confidence interval). In OARs, LETd hotspots (>7 keV/µm) occurred and LETd distributions were inhomogeneous and sensitive to range deviations. LETd hotspots and the impact of range deviations were most prominent in OARs of brain tumor patients which translated in RBE values exceeding 1.1 in all brain OARs. The near-maximum predicted PIC in healthy brain tissue of brain tumor patients was smaller than 5% and occurred adjacent to the CTV. Range deviations induced absolute differences in PIC up to 1.2%. CONCLUSIONS: Robust dose optimization generates LETd distributions in the target volume robust against range deviations. The current findings support using a constant RBE within the CTV. The impact of range deviations on the considered probability of late radiation-induced toxicity in brain tissue was limited for robust dose-optimized treatment plans. Incorporation of LETd in robust optimization frameworks may further reduce uncertainty related to the RBE-weighted dose estimation in normal tissues.

Material assignment for proton range prediction in Monte Carlo patient simulations using stopping-power datasets.

Motivation and objective. For each institute, the selection and calibration of the most suitable approach to assign material properties for Monte Carlo (MC) patient simulation in proton therapy is a major challenge. Current conventional approaches based on computed tomography (CT) depend on CT acquisition and reconstruction settings. This study proposes a material assignment approach, referred to as MATA (MATerial Assignment), which is independent of CT scanner properties and, therefore, universally applicable by any institute. MATERIALS AND METHODS: The MATA approach assigns material properties to the physical quantity stopping-power ratio (SPR) using a set of 40 material compositions specified for human tissues and linearly determined mass density. The application of clinically available CT-number-to-SPR conversion avoids the need for any further calibration. The MATA approach was validated with homogeneous and heterogeneous SPR datasets by assessing the SPR accuracy after material assignment obtained either based on dose scoring or determination of water-equivalent thickness. Finally, MATA was applied on patient datasets to evaluate dose differences induced by different approaches for material assignment and SPR prediction. RESULTS: The deviation between the SPR after material assignment and the input SPR was close to zero in homogeneous datasets and below 0.002 (0.2% relative to water) in heterogeneous datasets, which was within the systematic uncertainty in SPR estimation. The comparison of different material assignment approaches revealed relevant differences in dose distribution and SPR. The comparison between two SPR prediction approaches, a standard look-up table and direct SPR determination from dual-energy CT, resulted in patient-specific mean proton range shifts between 1.3 mm and 4.8 mm. CONCLUSION: MATA eliminates the need for institution-specific adaptations of the material assignment. It allows for using any SPR dataset and thus facilitates the implementation of more accurate SPR prediction approaches. Hence, MATA provides a universal solution for patient modeling in MC-based proton treatment planning.