Addressing the dose perturbation of metallic implant in spinal Stereotactic Body Radiotherapy (SBRT)

This study compares methods of addressing the dose perturbation occurred in planning and delivery stage of spinal adjuvant Stereotactic Body Radiotherapy (SBRT) caused by the presence of titanium implants. Dose prediction by TPS was conducted while the CT number of titanium was (a) default, calculated by AAA algorithm ("No Overridden"); (b) relative electron density, mass density matched, calculated by AAA algorithm ("AAA"); (c) relative electron density, mass density matched, calculated by AXB algorithm ("AXB"). In (i) phantom study, dose predictions were compared with measurement conducted with 2 mm, 6 mm and 10 mm thickness of titanium alloy Grade 5 (Ti6Al4V) using field sizes of 1 × 1 cm², 2 × 2 cm², and 4 × 4 cm². In (ii) planning study, retrospective dose predictions on patient plans were carried out to evaluate the impact on clinical outcome. The mean discrepancies (%) between measurement and "No Overridden", "AAA", "AXB" at tissue - titanium interface (titanium - tissue interface) were respectively -16.86 (10.93), -14.05 (11.24), 0.71 (1.54) for 2 mm and 6 mm thickness of titanium; and -18.42 (11.20), -18.29 (11.04), -10.14 (0.01) for 10 mm thickness of titanium, respectively. The patient study results by "AXB" yielded a significant deficit in tumour volume coverage at prescribed dose compared to "AAA" and "No Overridden" by 5.23 % and 9.06 %, respectively. "No Overridden" and "AAA" can potentially generate acceptable prediction in specific scenarios where the depth of target remains approximately unchanged while the gantry rotates. The AXB algorithm is recommended to be used in routine practice involving titanium implants. (paper)