Evaluation of dosimetric consistency between pre- and postoperative CT-guided noncoplanar template-assisted ¹²⁵I seed implantation in recurrent and metastatic thoracic tumors: a retrospective study

Objective Recurrent and metastatic thoracic malignancies remain challenging to manage because of their proximity to vital structures and the impact of respiratory motion on dose accuracy. This retrospective study aimed to assess the dosimetric consistency between preoperative planning and postoperative verification in computed tomography (CT)-guided 3D-printed noncoplanar template (3DPNCT)-assisted ¹²⁵I seed implantation. Methods Clinical and dosimetric data from 32 patients who underwent 3DPNCT-assisted seed implantation were retrospectively reviewed. Preoperative treatment plans were compared with postoperative dose verification to evaluate deviations in key indices, including target coverage, dose conformity, and homogeneity. Statistical comparisons were performed using appropriate paired analyses, and agreement between planned and delivered doses was further evaluated through Bland–Altman analysis. Results The mean D90 decreased from 142.47 ± 28.90 Gy to 135.06 ± 31.66 Gy ( p = 0.022). The mean V100 decreased from 93.19 ± 14.74% to 91.48 ± 12.37% ( p = 0.441), and V150 from 73.81 ± 16.20% to 69.30 ± 17.61% ( p = 0.059). No significant differences were observed in V200, CI, or EI (all p > 0.05). The homogeneity index (HI) increased slightly from 0.222 ± 0.114 to 0.257 ± 0.130 ( p = 0.049). Conclusions CT-guided 3DPNCT-assisted ¹²⁵I seed implantation achieves robust dosimetric reproducibility and favorable plan accuracy in thoracic tumors. Despite minor deviations related to respiratory motion and intraoperative uncertainties, this approach provides a reliable and clinically feasible option for managing complex thoracic malignancies. Further prospective studies incorporating motion management and real-time adaptive planning are warranted to optimize its clinical application.