PV QA 3 - Poster Viewing Q&A 3
TU_14_3255 - Predicting dose-volume-histograms for intracavitary high-dose-rate brachytherapy for gynecologic applications
Tuesday, October 23
1:00 PM - 2:30 PM
Location: Innovation Hub, Exhibit Hall 3
Predicting dose-volume-histograms for intracavitary high-dose-rate brachytherapy for gynecologic applications
C. V. Guthier1, M. King2, L. J. Lee2, and R. A. Cormack1; 1Department of Radiation Oncology, Brigham and Women’s Hospital and Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, 2Department of Radiation Oncology, Dana-Farber Cancer Institute/Brigham and Women's Hospital, Boston, MA
Purpose/Objective(s): Intracavitary high-dose-rate brachytherapy (HDR-BT) utilizing a tandem and ring applicators (TR) is an essential component of cervical cancer treatment. Due to the fixed TR geometry and patient specific anatomy, treatment planning is not intuitive and the planner must compromise between under dosing the planning target volume (PTV) or over overdosing surrounding organs at risks (OARs). The goal of this work is to provide a tool that predicts achievable PTV and OAR dosimetric metrics (DM) to illuminate the tradeoffs as the planner develops a patient specific treatment plan.
Materials/Methods: Due to the TR HDR-BT specific setup limitations, i.e. the applicator is most likely not in the center and the resulting dose distribution does not necessarily follow the shape of the PTV, existing approaches for external beam DVH prediction methods are not applicable to HDR-BT. Inspired by EBRT methods, we propose the following: (1) Based on a standardized treatment plan, the resulting isodose lines divide the actual treatment plan into individual shells. (2) For each shell and structures, the differential DVHs are fitted using a generalized extreme value distribution. (3) The sum of the individual differential DVHs form the final DVH. The hypothesis of using standardized plans is that patient specific variations considered in the individual plans are correlated with the volume overlap of the structures with the isodose shells of the standardized plan. The standardized plan is optimized with equally weighted dwell-times delivering the prescribed dose to point A and one third to point B. The IRB- approved retrospective study contains 40 patients from which 25 were randomly chosen for training and the remaining for validation of the model. To predict DVHs for PTV, rectum, bladder, and sigmoid an average model is used. The predictive power of the model is evaluated using residual analysis, statistical tests and the outcome of a potential replanning
Results: The range residual analysis for PTV was found to be between SRPTV=-0.014 to SRRectum=0.006 for the training cohort and between SRPTV=-0.013 to SRRectum=0.008 for the validation set. A two sample Kolmogorov-Smirnov test showed no statistical significant difference between the training and validation cohort (p>0.05). Using the predicted DMs for OARs as constraints for re-planning, the difference in PTV coverage was (0.02±0.03) and based on a Wilcoxon signed-rank test (p>0.05), was not statistically significant
Conclusion: With the presented methodology it was demonstrated that DVH prediction methods are feasible in HDR-BT. The standardized model provides a means to efficiently capture patient specific variation and allow DVHs prediction with a limited number of patients. Using achievable DMs as constraints for treatment planning algorithms, the presented framework supports the user in finding a patient specific treatment plan.
Author Disclosure: C.V. Guthier: None. M. King: None. L.J. Lee: Employee; Brigham and Women's Physician Organization. Research Grant; Bridge Expansion Grant, Astra Zeneca, Joint Center of Radiation Therapy, Dana-Farber Cancer Institute. R.A. Cormack: Honoraria; ASTRO, American Brachytherapy Society, American Association of Physicists in Medicine. Travel Expenses; American Brachytherapy Society, American Association of Physicists in Medicine. Board Member; New England Chapter of AAPM.