Dosimetric Effects of Adaptive Gynecological HDR Brachytherapy Planning on a per-Fraction Versus per-Implant Basis
Abstract
Purpose
HDR gynecologic brachytherapy can involve creating a new plan for each fraction of treatment or delivering a single plan for both fractions of an implant. This work retrospectively analyzes the differences in EQD2 doses to the CTVs and OARs between two planning paradigms: one plan delivered for two fractions and a new plan generated for each fraction.
Methods
Four patients treated with commercially available HDR hybrid applicators were analyzed. CTVs and OARs were contoured by a physician on CT scans taken for both fractions. These CTs and contours were sent to the Oncentra TPS. Applicator models were applied and catheter digitization was performed. To ensure consistency of the delivered plans, the dwell times from the first fraction were decayed to align with the source strength at the time of retrospective calculation of the second fraction. D90 (EQD2) to CTVs and D2cc (EQD2) to OARs were determined for both plan paradigms and the differences were analyzed using Bland-Altman analysis.
Results
The differences in D90 (EQD2) to the HRCTVs and IRCTVs ranged between -0.62 to 1.05 Gy and -0.43 to 0.32 Gy, respectively. D2cc (EQD2) for the bladder, bowel, and sigmoid differed within a range of -1.96 and 1.72 Gy across all patients, with an average bias of 0.12 Gy. Differences in rectal D2cc (EQD2) ranged from -0.40 to 6.09 Gy, with a bias of 2.41 Gy.
Conclusion
The Limits of Agreement for the CTVs and most OARs were narrow and biases were close to zero, indicating consistency in delivered dose across the implants between the two paradigms. The dose to the rectum varied the most, which can be attributed to high variability in filling and proximity to the target. Nevertheless, using one plan for the duration of an implant results in acceptable doses to the CTVs and OARs.