2006 - Dose Optimization Methods for HDR Prostate Brachytherapy with Boost to Intraprostatic Dominant Lesions Using a TRUS-MRI Contour-Based Deformable Fusion Workflow

Purpose/Objective(s):

Real-time HDR prostate brachytherapy with dominant intraprostatic lesions (DIL) contouring has been performed in our clinic for many years. In April 2023, a new contour-based deformable TRUS-MRI registration system was acquired, allowing a MRI-guided DIL boosting.

Deformable fusions are performed in a first contouring step (before needle insertion) to place at least one catheter near each DIL location. However, prostate deformations during the implant affect the DIL shape, location and hence the final dosimetry. Without a second registration, this uncertainty is not taken into account.

The aim of this study is to evaluate if an implant class solution could be used instead of a DIL-driven needle placement. In such case, information of DIL position would not be needed before insertion. This would allow to perform only one fusion after needle insertion without loss of dosimetric quality.

Materials/Methods:

Our real-time workflow with MRI guided DIL boost consists of the following steps: 1) Patient setup and first TRUS acquisition, 2) Prostate and OAR contouring, 3) Deformable MRI fusion with DIL contoured, 4) Implant simulation and evaluation, 5) Live TRUS-guided catheter insertion according to simulation, 6) Second TRUS acquisition, contour adaption and needle reconstruction, 7) Dosimetry optimization and final evaluation, 8) Treatment delivery.

MRI scans are acquired in a 3T magnet without endorectal coil. DIL are delineated by an experienced radiologist on T2- and diffusion-weighted sequences in advance of the BT procedure.

For this work, 20 consecutive patients (24 DIL) treated with 15 Gy in a single fraction from June 2024 to January 2025 were replanned. Needle distributions were defined with the planning system’s automatic option selecting a 3 mm urethra protection. DIL locations were not taken into account to simulate the case in which they are contoured after catheter insertion.

Optimizations were considered clinically valid when the minimum dose to the DIL equaled the originally accepted by the oncologist within 1%, the loss in DIL D90 did not exceed 5%, and our institution’s PTV coverage dose objectives and OAR constraints kept the original fulfillment.

Results:

The table shows the differences for target coverage and OAR doses obtained with the simulated implant minus the original treatments. DIL Dmin and D90 matched the originally approved within ±1% and ±5%, respectively, in all patients except two. The maximum difference reported was +10.4% in DIL D90 coverage. Average difference in needle number was -1.2.

Conclusion:

In the light of the results, clinically acceptable dosimetries based on a class solution can be achieved and allow the possibility to perform MRI fusions after needle insertion, without two registration steps.

Average differences ± SD (%) for DIL, Prostate and OAR relevant dosimetric parameters

DIL Boost		Prostate				Urethra		Rectum
Dmin	D90	Dmean	D90	V100	V150	Dmax	Dmean	Dmax	Dmean
0.0±0.7	-1.0±4.0	-7±11	0.1±3.2	0.3±1.5	-5.1±7.3	2.3±2.8	3.8±7.0	-1.8±6.0	-0.9±5.1