Novel Use of 0.35T MRI Guidance for High Dose Rate Brachytherapy in the Treatment of Cervical Cancer

Abstract

Intracavitary brachytherapy planning for gynecologic malignancies has previously relied on clinical set up, plain films, computed tomography (CT) and 3T diagnostic magnetic resonance imaging (MRI) guidance. We report the use of 0.35T MRI-guided brachytherapy planning for gynecologic malignancies at our institution, which we hypothesize would allow for shorter time to treatment, and shorter imaging times with relatively high-degree of image quality. Patients undergoing intracavitary brachytherapy with tandem and ovoid +/- interstitial parametrial needles for cervical malignancies at our institution who received treatment planned using the 0.35T MRI were reviewed. The high-risk CTV (HR-CTV), GTV (if visualized), bladder, sigmoid, nearby small bowel, and rectum were contoured. Typically, patients receiving 5 fractions of intracavitary brachytherapy would undergo planning using diagnostic MRI for the first and third fractions, and using 0.35T MRI, or CT scan, for the second, fourth, and fifth fractions. In all, 29 patients fit inclusion criteria from April 2015 to January 2017, ranging from stage IB1 to IVA (IB1: 9 patients; IB2: 1 patients; IIA1: 2 patients; IIA2: 3 patients; IIB: 7 patients; IIIB: 6 patients; IVA: 1 patient). A 0.35T non-contrast MRI was obtained while the tandem and ovoid instruments +/- interstitial needles were in place. Most patients received 5.5Gy x 5 fractions to the HR-CTV using high dose rate Ir-192 following 45Gy of external beam radiotherapy to the pelvis +/- simultaneous integrated boost to imaging-abnormal lymph nodes to 55 Gy. For all patients, the median HR-CTV size was 27.78cc, and the median D90 to the HR-CTV was 6.1Gy. The median D2cc to the bladder, rectum, bowel, and sigmoid were 5Gy, 2.5Gy, 2.6Gy, and 3Gy, respectively. The image acquisition time was approximately 2.5 min for the 0.35T MRI, and 15min for the diagnostic MRI. The median time from imaging start to initiation of treatment using the 0.35T MRI scan was 65min, compared to 105min using diagnostic MRI scan, (t-test, p < 0.01). The contrast-to-noise ratio (CNR) of tumor to cervix was significantly higher in the diagnostic MRI compared to 0.35T MRI; both diagnostic and 0.35T MRI scans had significantly higher CNR values compared to CT, (ANOVA and Tukey pair-wise tests, p<0.01). We report the first clinical use of 0.35T MRI-guided brachytherapy. The treatment using this approach expedited the image acquisition compared to diagnostic MRI, and provided significant advantage in visualizing the tumor and cervix compared to CT. This presents a feasible and reliable manner to image and plan gynecologic brachytherapy.