Validation of IBS-GEC ESTRO-ABS Computed Tomography-Based Contouring Recommendations Using Prebrachytherapy Magnetic Resonance Imaging in Patients With Cervical Cancer Treated With Definitive Radiation Therapy.

PURPOSE: Consensus guidelines from IBS-GEC ESTRO-ABS (Indian Brachytherapy Society- Groupe Européen de Curiethérapie - European Society for Radiotherapy and Oncology- American Brachytherapy Society) for computed tomography (CT)-based contouring have recently introduced possibilities of CT-based adaptive brachytherapy (BT) for cervical cancer. The present study attempts to validate the guidelines by using magnetic resonance imaging (MRI) at diagnosis and in a pre-BT environment.

METHODS AND MATERIALS: The study involved patients with stages IIB to IVA carcinoma of the cervix who had pretreatment MRI and were treated with definitive chemoradiation and hybrid-interstitial BT from April 1, 2022, to October 1, 2023. The target volumes-high-risk clinical target volume (HRCTV), intermediate-risk clinical target volume (IRCTV), and organs-at-risk that included bladder, rectum, and sigmoid-were prospectively delineated on BT planning CT using BT planning MRI for guidance, as per IBS-GEC ESTRO-ABS recommendations by a radiation oncologist. The dimensions, volumes, doses, and Dice similarity coefficient on CT were compared with those on MRI images.

RESULTS

Two hundred eight scans from 104 sets of 52 patients of CT and magnetic resonance scans were the basis for this study. With respect to dimensions, CT overestimated HRCTV maximum width, IRCTV width at point A, and maximum thickness of HRCTV and IRCTV compared with MRI (P values, <.001, <.001, .049, and .007, respectively). In analysis of organs-at-risk volumes, CT overestimated sigmoid and rectum volumes, with median difference of paired values of 7.2 (Interquartile range [IQR], 3.1-13.6) and 7.8 (IQR, 3.6-12.8) and P values of <.001 and .003, respectively. Bladder 2cc and sigmoid 2cc doses were underestimated with CT images, with median difference of paired values of 0.5 (IQR, 0.3-0.9) and 0.5 (IQR, 0.3-0.8) and P values of <.001 and .003, respectively. There was an insignificant difference in HRCTV and IRCTV doses between CT and MR images. The median Dice similarity coefficients of HRCTV, IRCTV, bladder, rectum, and sigmoid were 0.72, 0.78, 0.76, 0.68, and 0.46, respectively.

CONCLUSIONS

Despite the dimensional differences, there was no dosimetric difference in HRCTV and IRCTV. CT-based contouring using IBS-GEC ESTRO-ABS recommendations, with MRI at diagnosis and BT, similar to the MRI at diagnosis + pre-BT MRI + CT at brachytherapy environment, is safe, convenient, and addresses logistic issues in institutions, especially in low- and middle-income countries.