Multiparametric Magnetic Resonance Imaging as a Noninvasive Assessment of Tumor Response to Neoadjuvant Pembrolizumab in Muscle-invasive Bladder Cancer: Preliminary Findings from the PURE-01 Study

Background: In the PURE-01 study, pembrolizumab was given preoperatively before radical cystectomy in clinical T2-4aN0M0 patients. An accurate clinical response assessment may be useful for developing new perioperative strategies in these patients. Objective: To evaluate the association between bladder multiparametric magnetic resonance imaging (mpMRI) findings after pembrolizumab and the pathological complete response (CR; pT0). Design, setting, and participants: Patients were staged using bladder mpMRI whereby radiologists were asked to characterize the following parameters: residual disease at T1- and T2-weighted images (step 1: yes/no), presence of hyperintense spots within the bladder wall on diffusion-weighted imaging (step 2: yes/no), and presence of pathological contrast enhancement (step 3: yes/no), before and after three cycles of pembrolizumab. Examinations were internally assessed by two senior radiologists and externally evaluated by a third senior radiologist. Intervention: To evaluate bladder tumor response after neoadjuvant pembrolizumab, mpMRI was used. Outcome measurements and statistical analysis: The primary objective was to predict the pT0 after neoadjuvant pembrolizumab by relying on the mpMRI findings. Cohen's kappa statistics was used to assess interobserver variability. Univariable analyses for pT0 were performed including internal and external post-therapy mpMRI steps. Results and limitations: From February 2017 to October 2018, 82 patients (164 total mpMRI assessments) were analyzed. The agreement between the internal and external mpMRI assessments after therapy was acceptable (κ values ranging from 0.5 to 0.76). Each mpMRI step was significantly associated with pT0 in both internal and external assessments. In patients with CR/no evidence of residual disease (NED) in all internally evaluated mpMRI steps (N = 37), the pT0 was seen in 23 (62%), compared with 19 of 26 externally evaluated NED patients (73%). Conclusions: In post-pembrolizumab muscle-invasive bladder cancer, mpMRI sequence assessment had acceptable interobserver variability and represented the basis for the proposal of a radiological CR/NED status definition predicting the pT0 response to pembrolizumab. After validation of these findings with external datasets, we propose this tool for developing bladder-sparing immunotherapy maintenance therapies. Patient summary: Assessment of the extent of disease in patients with muscle-invasive bladder cancer using conventional imaging yields serious limitations. In the PURE-01 study, we evaluated the potential of bladder multiparametric magnetic resonance imaging (MRI) to predict the pathological complete response to neoadjuvant pembrolizumab. After validation with larger datasets, the proposed stepwise assessment incorporating multiparametric MRI sequences will be used at our center to develop bladder-sparing approaches in future studies. • In the PURE-01 study, multiparametric magnetic resonance imaging (mpMRI) of the bladder was used to stage and evaluate the response to pembrolizumab, before radical cystectomy. • Assessments of mpMRI sequence were externally reviewed, showing preliminary but promising reproducibility. • We proposed an mpMRI-based definition of complete response predicting the pathological complete response to pembrolizumab that was externally replicated.