Quantitative multiparametric MRI as a promising tool for the assessment of early response to neoadjuvant chemotherapy in bladder cancer

Abstract

PurposeMagnetic resonance imaging (MRI) has been applied to predict the neoadjuvant chemotherapy (NAC) response of bladder cancer at an early stage, but the performance remains poor. A tool for the selection of patients with muscle-invasive bladder cancer (MIBC) with high probability of benefitting from NAC is not yet available. We designed a prospective study to assess a quantitative MRI for predicting the early response of MIBC to NAC. MethodsAll individuals underwent a time-course MRI at pre-NAC, 24 h after initial cisplatin medication (24 h-NAC) and post-NAC.Chemosensitivity was evaluated according to pathological response.The transfer constant (Ktrans), plasma volume (Vp), extravascular extracellular space (Ve), and apparent diffusion coefficient (ADC) were quantitated based on dynamic contrast-enhanced and diffusion-weighted imaging.Quantitative RECIST criteria were constructed by modelling pre-NAC and 24 h-NAC MRI measures, and then compared with conventional RECIST by using pre- and post-NAC MRI measures. ResultsIn this pilot study, a total of 24 patients were enrolled into the study. Eight patients were pathologically confirmed to be NAC-responders. After a thorough evaluation of these parameters, different parameters showed good discrimination at different point in time. ROC curves showed quantitative MRI can predict the response to NAC, especially ADC_M (AUC = 0.859, P = 0.005) and ADC index (AUC = 0.844, P = 0.007) at pre-NAC timing and ADC_M (AUC = 0.816, P = 0.013) at 24 h-NAC timing. Then, a qRECIST model was established for predicting NAC sensitivity, with AUC of 0.91, TP rate of 0.8, and accuracy of 0.75. ConclusionThe diagnostic performance of mpMRI parameters for NAC response is excellent. The qRECIST derived between pre- and 24 h NAC MRI could predict the early response of MIBC to NAC and help for candidate selection.