Diffusion-Weighted Magnetic Resonance Imaging as a Predictor of Outcome in Cervical Cancer After Chemoradiation

Purpose: To determine whether apparent diffusion coefficient (ADC) value is predictive of survival after definitive chemoradiation for cervical cancer independent of established imaging and clinical prognostic factors. Methods and Materials: Between 2011 and 2013, the pretreatment MRI scans for 69 patients treated with definitive chemoradiation for newly diagnosed cervical cancer were retrieved. Scans were acquired with a 1.5-T magnetic resonance scanner, including diffusion-weighted imaging sequences. Mean ADC value was measured within a region of interest in the primary cervical cancer on the baseline MRI scan. Baseline tumor maximum standardized uptake value on positron emission tomography/computed tomography was determined by the reading radiologist. Treatment included external beam radiation therapy to the pelvis followed by brachytherapy in 97%, and with concurrent weekly cisplatin in 99% of patients. Univariate and multivariate analyses were done to investigate the association of clinical and imaging variables with disease control and survival endpoints using a Cox proportional hazard test. Results: Median follow-up was 16.7 months (range, 3.1-44.2 months). The 1-year overall survival, locoregional recurrence-free survival, and disease-free survival rates were 91%, 86%, and 74%, respectively. The median ADC value was 0.941 x 10(-3) mm(2)/s (range, 0.256-1.508 x 10(-3) mm(2)/s). The median standardized uptake value in the primary tumor was 15 (range, 6.2-43.4). In multivariate analysis, higher ADC value (hazard ratio [HR] 0.36, 95% confidence interval [CI] 0.15-0.85, P = .02), higher stage (HR 2.4, 95% CI 1.1-5.5, P = .033), and nonsquamous histology (HR 0.23, 95% CI 0.07-0.82, P = .024) were independent predictors of disease-free survival. Conclusions: The mean ADC value of the primary tumor on pretreatment MRI was the only imaging feature that was an independent predictor of disease-free survival in cervical cancer patients treated with chemoradiation. Further validation will be needed to determine whether ADC values may prove useful in identifying cervical patients at high risk of recurrence. (C) 2016 Elsevier Inc. All rights reserved.