2216 - Novel Method for Robust Catheter Insertion in HDR Prostate Brachytherapy
Presenter(s)
J. B. Schulz1, B. Bednarz1, J. M. Floberg2, and J. M. Slagowski2; 1Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, 2Department of Human Oncology, University of Wisconsin-Madison, Madison, WI
Purpose/Objective(s): High-dose-rate (HDR) brachytherapy requires precise catheter placement, especially when escalating dose to focal lesions identified on PSMA-PET/MRI. We propose a pre-procedural planning method that accounts for placement uncertainties during treatment planning for GTV-directed dose escalation. Our hypothesis is that patient-specific visualization of catheter positioning sensitivity can improve treatment precision while simultaneously optimizing catheter trajectories and dwell times. This approach aims to reliably boost GTV doses without compromising OAR constraints, while potentially reducing toxicity, procedure time, and improving quality of life.
Materials/Methods: We implemented and validated a Python-based treatment planning system using a modified FISTA algorithm to simultaneously robustly optimize catheter placement positions and dwell times. Our framework incorporates target coverage with focal GTV dose escalation, OAR sparing, and catheter sparsity. Critically, we also introduce a novel term that explicitly accounts for catheter insertion uncertainties, ensuring the recommended catheter positions remain robust to potential deviations during placement. To support clinical decision-making, the system simulates catheter displacements in six cardinal directions, and identifies worst-case-catheter-miss-scenarios that would compromise GTV dose delivery or increase OAR doses. The system provides intuitive visual feedback to guide decision-making during pre-implant planning.
Results: Testing retrospectively on a single-patient dataset with an MRI-delineated GTV abutting the urethra, demonstrated that our approach maintained robust focal boosting while achieving PTV V100% of 95%, bladder and rectum V80 < 1cc (0.8cc, 0.6cc), and urethra V118 < 0.1 cc. Visualization revealed anterior-posterior displacements posed the greatest risk to GTV boost delivery (10% D90 reduction with 2mm displacement), while medial-lateral shifts were better tolerated. Our approach maintained GTV D90 >120% of prescription across all simulated perturbations versus conventional plans where GTV D90 dropped below 110% in worst-case scenarios.
Conclusion: We developed the first methodology for robust catheter position placement optimization for focal HDR prostate brachytherapy that shifts the paradigm from reactive to proactive planning. By visualizing worst-case catheter miss scenarios, our approach enhances the precision and robustness of catheter placement for GTV dose escalation while meeting PTV coverage and OAR constraints. This methodology has implications for clinical practice by standardizing catheter insertion, potentially being used to guide intraoperative catheter placement, reducing total procedural time and improving overall plan quality. Future work includes validation on a larger patient dataset and development of patient-specific visualization tools to further democratize high-precision brachytherapy techniques.