Purpose: To investigate and compare calculated dose in cervical cancer patients using Varian’s model-based dose calculation algorithm Acuros 13.5, and standard Task Group 43 (TG-43) formalism.
Methods: A set of locally advanced cervical cancer patients treated with MR based HDR brachytherapy using tandem and ovoid applicators and free-hand interstitial needles were selected. These implants were stabilized using Alatus vaginal balloons with a 5% IsoVue to 95% saline fill ratio. The current clinical flow utilizes manual digitization of all applicators followed by planning and dose calculation using the TG-43 formalism. For this study, every patient fraction was replanned using the Fleture-Suit Delclos-style solid applicator (SA) library in Brachy Vision that matched the set used clinically. The Ti needles were copied from the clinical plan. The dose-to-medium-in-medium (D₍m,m₎) was calculated using Acuros and several key equi-effective dose (EQD2) dosimetric parameters were compared between the two formalisms. We report D₉₀, D₉₈ for CTV₍HR₎ and CTV₍IR₎, and D₍2cm³₎ for organs at risk (OAR) including bladder, rectum, sigmoid, bowel, and vagina. Mean and absolute differences from the two plans were investigated with linear regression, and Bland Altman analysis.
Results: In this subset of 15 patients each with 5 fractions, where prescription dose is 550 cGy, the Acuros plans resulted in an average dose reduction to CTV₍HR₎ by 4.5%±0.37 (D₉₈) and 4.4% ±0.22% (D₉₀), and to CTV₍IR₎ by 4.1%±0.19 (D₉₈) and 4.0%±0.17 (D₉₀). Normal structures also showed a dose reduction to D₍2cm³₎ regions of bladder as 5.0%±0.31, rectum 8.1%±0.39, sigmoid 5.1%±0.43, bowel 6.2%±0.30, and vagina 4.8%±0.28.
Conclusion: Acuros estimates lower doses to target and normal structures. However, considering that the cervical brachytherapy consensus guidelines are based on TG43 data, it is unclear if using MBDCA calculations will be a clinical improvement.