PO27 Presentation Time: 7:40 AM: Library-Based MRI-Only Treatment Planning (MRTP) for a Venezia Applicator Using MRI for Catheter Detection

Abstract

<h3>Purpose</h3> Interstitial HDR brachytherapy currently requires CT, in addition to MRI, to complete the treatment planning workflow. CT is necessary to visualize and localize catheters and template, while MRI is needed for delineation of the clinical target volume (CTV) and the organs-at-risk (OARs). An MRI-only treatment plan/planning (MRTP) approach that can eliminate CT from the workflow using an optimized ‘Pointwise Encoding Time Reduction with Radial Acquisition (PETRA)' MRI, is proposed. Typically, applicators lack MR contrast, making MRI-based digitization difficult for catheters within the applicator. The feasibility of utilizing an applicator library to complete MRTP in gynecological patients was investigated for a Venezia applicator. The clinical translation feasibility of MRTP was evaluated by conducting dosimetric comparisons between the MRTP and the clinical treatment plan (CTP). <h3>Materials and Methods</h3> An optimized PETRA MRI (TR\TE 3.3\0.7ms) was acquired on a 3T MAGNETOM Vida (Siemens) in three gynecological cancer patients treated using a Venezia applicator (Elekta) and interstitial needles. A library-based applicator model tool available on Oncentra Brachy TPS was used to aid catheter digitization of applicators. The library of Venezia applicator was used to digitize two lunar ovoids and tandem. Interstitial needles were manually digitized based on negative contrast between needles and surrounding tissue. The MRTP was generated by activating and assigning the same dwell times as in the CTP. The catheter displacements between two approaches were assessed by comparing dwell positions following the CT-to-MR transformation, as well as the displacement in catheter tip depth. The DVH and EQD₂ metrics for the CTV D₉₀ and the OARs D_2cc (rectum, bladder and sigmoid) were compared between CTP and MRTP. <h3>Results</h3> MRTP workflow corresponded to the average tip depth difference of -1.58 ± 2.49 mm, 0.50 ± 1.55 mm, and -0.38 ± 1.17 mm for lunar ovoids, tandem, and interstitial needles, respectively, compared to the conventional CT-based approach. Differences in catheter displacements were 3.34 ± 1.38mm, 2.49 ± 0.23mm and 2.26 ± 0.48 mm for lunar ovoids, tandem and interstitial needles, respectively. These differences corresponded to absolute DVH differences of -2%, -1% and 3% for CTV D₉₀ in these three patients. The average absolute DVH differences were -3.44 ± 0.02%, 1.77 ± 0.02%, and -0.98 ± 0.02% for OARs (rectum, bladder and sigmoid, respectively). The average relative total EQD2 differences were 0.12 ± 0.02%, -2.62 ± 0.01%, 1.41 ± 0.01% and -0.71 ± 0.02% for CTV, rectum, bladder, and sigmoid, respectively. <h3>Conclusions</h3> Preliminary implementation feasibility of a library-based MRTP for a Venezia applicator was demonstrated in three gynecologic cancer patients.