Nomogram based on quantitative dynamic contrast-enhanced magnetic resonance imaging, apparent diffusion coefficient, and clinicopathological features for early prediction of pathologic complete response in breast cancer patients receiving neoadjuvant chemotherapy

BACKGROUND: The aim of this study was to develop two nomograms for predicting pathologic complete response (pCR) after neoadjuvant chemotherapy (NACT) for breast cancer based on quantitative dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), apparent diffusion coefficient (ADC), and clinicopathological characteristics at two time-points: before and after two cycles of NACT, respectively. METHODS: 3.0 T MRI scans were performed before and after 2 cycles of NACT in 215 patients. A total of 74 female patients with stage II–III breast cancer were included. According to univariate and multivariate logistic regression analysis, nomogram model 1 and nomogram model 2 were developed based on the independent predictors for pCR before and after 2 cycles of NACT, respectively. Nomogram performance was assessed with the area under the receiver operating characteristic curve (AUC) and calibration slope. RESULTS: The independent predictors of pCR were different at the two time points. Both nomograms were found to effectively predict pCR: nomogram model 2 based on Ki67, ΔK(trans)%, and ΔADC% after 2 cycles of NACT showed better predictive discrimination [AUC =0.900 (0.829, 0.970) vs. 0.833 (0.736, 0.930)] and calibration ability (mean absolute error of the agreement: 0.017 vs. 0.051) compared to nomogram model 1 based on pre-NACT HER2, Ki67, and K(trans). CONCLUSIONS: Nomograms based on quantitative DCE-MRI parameters, ADC, and clinicopathological characteristics can predict pCR in breast cancer and facilitate individualized decision-making for NACT.