Verifying calculated dose to the breast in stereotactic radiotherapy: A practical phantom-based method

Reference dosimetry for the GammaPod breast radiosurgery device is determined in a polymethyl methacrylate (PMMA) phantom, while the clinical dose calculations assume the patient's breast has the atomic composition of water with a density (∼0.935 g/cm³). Direct measurement within this clinical breast-tissue equivalent material is desirable to validate the treatment planning system (TPS) with end-to-end testing and patient-specific quality assurance (PSQA).

This study aimed to validate the GammaPod TPS dose calculations by 3D printing a breast-equivalent phantom and to develop a practical and reproducible method for end-to-end testing and PSQA.

Multiple breast-shaped phantoms matching the L02 GammaPod cup geometry were fabricated; three using 3D printing and one using clear ice. The 3D-printed phantoms were constructed from polylactic acid (PLA) and premium PLA (PLApro) using optimized infill densities to achieve a physical density near 0.935 g/cm³, matching the TPS's breast-tissue model. These infills were determined by creating series of CT phantom plugs and choosing the ones with the closest Hounsfield units (HUs) and stopping power ratios (SPRs) to the TPS breast tissue model. A rotating insert design allowed multi-position measurements with an Exradin A1SL ionization chamber and thermoluminescent dosimeters (TLDs). CT imaging and dosimetry were performed following the clinical GammaPod workflow. An ice phantom, with a density of approximately 0.918 g/cm³, served as an independent validation reference. Measured doses were compared with TPS-calculated doses.

The breast-tissue mimicking phantoms yielded mean Hounsfield (HU) and Stopping Power Ratios (SPR) numbers of -81 HU ± 5 and 0.99 SPR for blue PLApro (92.5%), -69 HU ± 11 and 0.99 SPR for natural PLA (86.0%), -82 HU ± 5 and 1.00 SPR for the ice phantom corresponding to densities of 0.916 g/cm³, 0.928 g/cm³, and 0.915 g/cm³ respectively, comparable to the modeled breast tissue. Measured-to-calculated dose differences for the ice phantom were within ± 0.5%, while the 3D printed phantoms had an average difference of 1.2%, consistent with published dosimetric uncertainty limits of the GammaPod. The total estimated measurement uncertainty, including charge-to-dose conversion and material variability, was 1.7%. The 3D-printed phantoms demonstrated excellent reproducibility when printing parameters were held constant.

The GammaPod TPS accurately predicts dose in breast-equivalent materials. The 3D-printed breast phantom provides a reliable, water-composition breast-density material, and geometrically accurate tool for end-to-end testing and PSQA, eliminating the challenges associated with water measurements. This method supports routine clinical verification and offers a reproducible framework for ongoing GammaPod QA and commissioning.