Diagnostic performance of contrast-enhanced dynamic and diffusion-weighted MR imaging in the assessment of tumor response to neoadjuvant therapy in muscle-invasive bladder cancer

Objectives To evaluate the diagnostic performance of DCE MRI and DWI in the assessment of pathologic complete response (pCR) after neoadjuvant chemotherapy (NAC) in patients with muscle-invasive bladder cancer (MIBC). Methods This prospective study included 90 patients with MIBC who finished NAC. Two radiologists independently assessed MRI for the determination of semi-quantitative parameters (wash-in rate and wash-out rate) and apparent diffusion coefficient (ADC) value. The correlation between pCR and wash-in rate, wash-out rate, ADC value were analyzed. The area under the ROC curve (AUC) was used to evaluate the diagnostic performance for detecting pCR. Inter-reader agreement was assessed using the ICC statistics. Results On cystectomy specimens, pCR was confirmed in (43.3%, 39/90). pCR is negatively correlated with wash-out rate (r = - 0.701, p = 0.01) and ADC value (r = - 0.621, p = 0.01). ADC value is positively correlated with wash-out rate (r = 0.631, p = 0.001). The diagnostic accuracy of ADC value (cut-off value: 0.911 x 10(-3)mm(2)/s) and wash-out rate (cut-off value: 0.677 min(-1)) in the identification of pCR was (92% for reader 1, 91% for reader 2), and (90% for reader 1, 88% for reader 2), respectively. The sensitivity, specificity for predicting pCR using ADC value + washout rate cut off values were 95.4%, 97.7% for reader 1, and 96%, 97% for reader 2, respectively. AUC was 0.981 for reader 1, 0.971 for reader 2. The overall reproducibility of the mean ADC value and wash out rate was excellent (ICC = 0.83-0.90). The ICC values for the mean ADC value, washout rate was 0.89 (95% CI 0.84-0.89) and 0.87 (95% CI 0.86-0.91), respectively. Conclusion Semi-quantitative parameter (wash-out) derived from DCE-MRI and ADC has the potential to assess the tumor's complete pathologic response. The two parameters using together can offer the best possibility to identify complete response to NAC in MIBC.