Carbon-Ion Therapy Patient-Specific QA: Preliminary Results Derived Using 3D Polymer Gel Dosimetry in Anthropomorphic Inhomogeneous Geometry

Abstract

<h3>Purpose/Objective(s)</h3> Improved conformity in carbon-ion radiation therapy (CIRT) with respect to photon RT comes at the cost of a potential reduced delivery robustness due to patient setup uncertainties, approximations in dose calculations and carbon-ion range uncertainties. This work presents preliminary results of a study aiming to address these challenges using 3D polymer gel dosimetry in anthropomorphic and inhomogeneous geometry. <h3>Materials/Methods</h3> All irradiations were performed at a synchrotron facility, in one of the treatment rooms equipped with a horizontal fixed beam line and pencil beam scanning (PBS) modality. For basic dosimetric characterization, carbon pristine Bragg peaks (BP) (ranges 30, 70, 90 mm in water) and a spread-out BP (SOBP) (range 115 mm and modulation 40 mm) were delivered on a cylindrical PMMA container filled with VIPAR polymer gel. An anthropomorphic phantom equipped with a cylindrical polymer gel insert was used for the irradiation of a pseudo-clinical CIRT plan, following all the steps of the standard clinical workflow, from simulation CT to irradiation. Two contralateral beams were optimized with a TPS (v.8b) to deliver 15 Gy (RBE) on a target located approximately at the center of the brain, in proximity of critical OARs such as brainstem and optical pathways. The phantom was then MRI scanned post-irradiation to retrieve the irradiated brain volume in 3D. The measured irradiated volume was compared against corresponding TPS derived data. <h3>Results</h3> Carbon-ion range was spatially verified with an accuracy of < 1 mm. For both the pristine BP and the SOBP irradiations. A significant deviation of the polymer gels' response on high-LET has been observed, as expected from previous studies with proton irradiations. For the pseudo-clinical treatment plan, a qualitative comparison between doses measured in polymer gels, against the corresponding TPS calculations, revealed a satisfying geometric accuracy of dose delivery in 3D, thus showing a high 3D geometrical accuracy of dose delivery in a typical standard clinical workflow. <h3>Conclusion</h3> In this work, we showed that patient-specific 3D polymer gel dosimetry is applicable to CIRT, as a valid tool to evaluate dosimetric and spatial plan accuracy in the clinical environment. As expected, the LET dependence of polymer gel's response, especially in the high-LET region, poses severe limitations for accurate dose measurements. Nevertheless, the proposed methodology is a promising efficient and safe clinical way for verifying such a complex RT technique.