Evaluation of Carbon Fiber and Titanium Surgical Implants for Proton and Photon Therapy.

PURPOSE: Impending and actual pathologic fractures secondary to metastatic bone disease, lymphoma, or multiple myeloma often require intramedullary fixation followed by radiotherapy. Due to carbon's low atomic number, there is reduced CT-imaging artifacts and dose perturbation when planning postoperative radiation for carbon fiber (CF) rods. Herein, we characterize the dosimetric properties of CF implants compared to titanium-alloy (TA) for proton and photon.

METHODS AND MATERIALS: TA and CF samples were acquired from an implant manufacturer. Material characteristics were evaluated by CT scans with and without metal artifact reduction (MAR). Relative stopping power (RSP) was determined from the range pull-back of each sample in a 20 cm range proton beam. Photon transmission measurements were made in a solid water phantom and compared to the modeled dosimetry from the RayStation planning system.

RESULTS: CF caused no visible CT artifacts and MAR was not necessary for Hounsfield Unit (HU) determination (median 364 HU) or contouring, while TA (median 3071 HU) caused substantial artifacts, which were improved, but not eliminated by MAR. The proton RSP was measured as 3.204 for TA, and 1.414 for CF. For 6 MV photons, the measured transmission was 89.3% for TA and 98% for CF. CF RSP calculation and transmission from CT HU showed a physical density overestimate compared to measurements, which would cause a slight, but acceptable dose uncertainty (<10% proton range or 1% photon transmission).

CONCLUSION: With a density similar to bone, CF implants did not cause imaging artifacts and minimal dose perturbation compared to TA. Although the CF proton RSP is underestimated and the photon attenuation is overestimated by the HU, both effects are relatively small and may be most easily accounted for by planning with a 2 mm expansion around organs-at-risk beyond or in close proximity to the implant.