Clinical Implementation and Evaluation of an MRI-Only Workflow for Interstitial High-Dose-Rate Brachytherapy In Gynaecologic Cancer
Abstract
Purpose
To clinically implement and evaluate an MRI-only treatment planning (MRTP) workflow for interstitial high-dose-rate (HDR) brachytherapy in gynecologic cancers. In contrast to the conventional CT/MRI-based planning (CTP) approach, MRTP eliminates CT imaging, thereby avoiding additional radiation exposure, reducing procedure time, and removing uncertainties associated with inter-modality image registration.
Methods
Twenty patients undergoing interstitial HDR brachytherapy were imaged with CT and 3T MRI. For MRTP, catheter reconstruction was performed using an optimized pointwise encoding time reduction with radial acquisition (PETRA) MRI sequence which enhanced catheter–tissue contrast, combined with a Syed-Neblett implant model in Oncentra Brachy (Elekta Brachytherapy). Reconstruction accuracy was assessed by measuring 3D dwell position displacements between MRTP and CTP plans. Dosimetric comparisons were performed using DVH- and EQD2-based metrics for clinical target volume (CTV) D90 and organ-at-risk (OAR) D2cc. Spatial dose agreement was quantified using Dice similarity coefficients (DSC) for volumes encompassed by the 65%–150% isodose levels. Statistical significance was evaluated using the Wilcoxon signed-rank test.
Results
Catheter reconstruction using the PETRA sequence and implant model was successful along the full catheter length in all patients. The mean 3D dwell position displacement between MRTP and CTP was 2.27±1.5mm, attributable to anatomical variation, applicator shifts, and CT-MRI registration uncertainty inherent to the standard workflow. MRTP demonstrated higher CTV D90 values compared to CTP, indicating improved target coverage, while maintaining comparable OAR doses. The mean EQD2 (DVH) difference for CTV D90 was 3.87±4.5Gy (9.26±8.1%, p<0.0001), with all OAR D2cc differences within ±0.6Gy (±2%). DSC values exceeded 0.70 across the 65%–125% isodose levels, confirming strong spatial agreement between dose distributions.
Conclusion
An MRI-only workflow for interstitial HDR gynecologic brachytherapy was successfully implemented and demonstrated dosimetric accuracy comparable to conventional CT/MRI-based planning while eliminating registration uncertainty, supporting a streamlined, patient-centered treatment planning approach suitable for routine clinical practice.