Single Institute Experience with MRI-Guided Adaptive Brachytherapy for Locally Advanced Cervix Cancer: Long Term Outcomes and Toxicity Analysis

Abstract

We report over 10 years' experience of MRI-guided adaptive brachytherapy (MRIGABT) in locally advanced cervix cancer (LACC). A total of 162 patients with LACC FIGO stage IB-IVB were treated at our institute between 2010 and 2020. Treatment consisted of chemoradiotherapy (weekly intravenous cisplatin 40 mg/m², 5 cycles, 1 day per cycle, 45-50.4 Gy external beam radiotherapy (EBRT) in 1·8-2 Gy fractions, followed by MRIGABT. Target volume definition and dose reporting for MRIGABT was according to GEC-ESTRO recommendations. MRIGABT dose prescription was according to our institutional practice. Overall survival (OS) and disease-free survival (DFS) were the primary endpoints. Kaplan-Meier estimates were calculated for OS and DFS at 2, 5 and 10-years. Organ-specific late toxicity grade ≥3 (≥G3) (CTCAEv5.0) was reported, alongside rates of bowel fistula, stricture, and perforation. Using logistic regression, we explored the relationship between EQD2 D2cc bladder and ≥G3 genitourinary (GU) toxicity. We compared patient proportions developing ≥ G3 GU toxicity in those who received EQD2 ≥85 Gy versus <85 Gy. We examined the following predictors of ≥G3 gastrointestinal (GI) toxicity: EQD2 dose (≥65 Gy versus <65 Gy), pre-existing bowel conditions, nodal boost, and extended field EBRT. Median follow up was 4.7 years (IQR 3.3-7.1 years). Median EBRT dose was 50.4 Gy (IQR 50.4-50.4 Gy); 91% received chemotherapy. Median high-risk clinical target volume (HRCTV) was 23.6 cm3 (IQR 16.6-31.3 cm3). Median doses were as follows; D90 HRCTV 88.9 Gy EQD210 (IQR 84.1-91.3 Gy), median D2cc bladder 81.6 (IQR 76.9-85.7 Gy), rectum 62.2 (IQR 57.9-65.3 Gy), sigmoid 67.4 (IQR 60.8-71.1 Gy), and bowel 55 (IQR 49.9-63.2 Gy), (all EQD23). The 2, 5 and-10-year OS were 98%, 80% and 75%. The 2, 5 and 10-year DFS were 98%, 75% and 60%. Late toxicity ≥G3 was 9% GU, 6% GI and 3% vaginal. There was a significant relationship between EQD2 and ≥G3 GU toxicity (OR: 1.11, 95% CI: 1.01-1.25; P = 0.04). When comparing those who received EQD2≥ 85Gy versus <85Gy, higher doses were associated with a greater proportion of ≥G3 GU toxicity (13% vs 6%). No significant predictors of ≥ G3 GI toxicity were observed. We observed excellent LC and OS. A significant relationship was found between EQD2 >85 Gy and bladder toxicity, although ≥G3 toxicity was low. We did not identify predictors of bowel toxicity. New predictors of bowel toxicity are required. Mean EBRT dose, D1.0 cc rectum, sigmoid and bowel are being investigated further.