Outcomes of 3D MRI Based HDR Brachytherapy with Hybrid Multichannel Vaginal Cylinder Applicator and Freehand Needles for Treatment of Vaginal Disease

Purpose/Objective(s) Multichannel vaginal cylinder (MCVC) is unable to cover residual vaginal cancer >7 mm thick or with supra-vaginal extension. Freehand needles inserted directly into areas of disease can extend isodose lines to allow for better coverage of disease if clinically indicated. We report outcomes of use of MCVC with freehand needles at our institution. Materials/Methods All patients with malignancies involving vagina treated from 2014-2021 with HDR BT at our institution with MCVC with freehand needles were identified. Clinical characteristics including primary malignancy, initial stage, date of diagnosis and recurrence (if applicable), prior RT, use of EBRT, and BT dose and fractionation were collected. Number of freehand needles used per fraction, 3D imaging used for planning, high-risk clinical target volume (HR-CTV) at each fraction, dose to 2cc (D2cc) of organs at risk, and HR-CTV D90 were recorded. Initial response based on examination and imaging and overall local control outcomes were recorded. Local failure was defined as time to recurrence either within HR-CTV or marginal failure, calculated from date of last BT fraction, with censorship at date of last follow-up. Local control was analyzed via Kaplan-Meier method. Results 33 patients were identified with median follow-up 1.9 years (IQR 1.0-2.9 years). 19 patients (57.6%) had primary endometrial cancer with vaginal recurrence/disease. Remaining patients had primary vaginal cancer or involvement secondary to other primaries. 8 patients had recurrent disease after previous course of RT. 24 patients received EBRT with median dose of 45 Gy in 25 fractions, with remaining patients treated with BT alone. BT prescription dose ranged from 20-36 Gy in 3-7 fractions, with a total of 168 BT fractions delivered. Median number of freehand needles utilized was 3, with freehand needles utilized in 157/168 total fractions. MRI-based plan was done for at least 1 fraction in 30 patients (90.9%). Median HR-CTV at time of BT was 19.7cc. Median D2cc to the bladder, rectum, and sigmoid were 56.4 Gy, 55.7 Gy, and 45.4 Gy, respectively. Median D90 to the HR-CTV for pts treated with EBRT plus BT was 76.7 Gy and for BT alone was 48.8 Gy. 29 patients (87.9%) had complete local response to BT on initial examination and/or follow-up imaging. 1 and 2-year local control rates for those who did not have prior RT was 89.5% and 89.5%, respectively. 1 and 2-year local control rates for those who had previous RT was 58.3% and 29.2%, respectively. 1 patient had vaginal laceration requiring surgical repair and 1 patient with disease progression developed small bowel obstruction 1 month after BT, with no additional acute grade 3+ toxicities identified. Conclusion Our approach with MCVC plus freehand needles with MRI-based planning was feasible and safe. Initial local response is excellent, with acceptable local control rates and low rate of serious acute toxicities. Multi-institutional experiences are needed to further validate this approach for treatment of vaginal disease.