Comparison of TG-43 and advanced collapsed cone engine dose calculations in cervical HDR brachytherapy with plastic and metallic applicators

Abstract

Abstract Purpose: Conventional brachytherapy treatment planning based on the TG-43 formalism assumes a homogeneous water medium and neglects the effects of applicator materials and tissue heterogeneities. This study aimed to investigate the clinical impact of Model-Based Dose Calculation Algorithms (MBDCAs) by comparing the Advanced Collapsed Cone Engine (ACE) with the TG-43 formalism in cervical cancer brachytherapy using both plastic and metallic applicators. Methods: A total of 60 high-dose-rate (HDR) cervical cancer brachytherapy plans (30 using plastic applicators and 30 using metallic applicators) were retrospectively analyzed. Plans originally calculated using the TG-43 formalism were recalculated with the ACE algorithm. Dosimetric parameters for target volumes (HR-CTV and IR-CTV) included D90 and D98, while organs at risk (rectum, bladder, and sigmoid) were evaluated. Point A dose was also assessed. Statistical comparisons were performed with significance defined as p < 0.05. In addition, point-dose measurements using an in-house phantom were conducted to verify the calculated dose differences. Results: ACE consistently calculated lower doses than TG-43 across all evaluated parameters (p < 0.001). A significant effect of applicator material was observed. For HR-CTV D90, the mean difference between ACE and TG-43 was −2.03% for metallic applicators compared with only −0.88% for plastic applicators. The largest discrepancy was observed for bladder D0.1cc in the metallic applicator group (−2.34%). Phantom measurements demonstrated improved agreement between calculated and measured doses when ACE was used (%diff = 1.78-2.24) compared with TG-43 (%diff = 12.50-13.22). Conclusion: Applicator material, particularly metallic components, significantly influences dose calculation accuracy in cervical cancer brachytherapy. The ACE algorithm accounts for material heterogeneity and provides dose estimations that are more consistent with phantom measurements than those obtained using TG-43. These findings suggest that ACE may improve dose estimation accuracy, particularly in the presence of high-density applicators.