Nomogram-derived prediction of pathologic complete response (pCR) in breast cancer patients treated with neoadjuvant chemotherapy (NCT)

Abstract

Abstract Background Previous research results on the predictive factors of neoadjuvant chemotherapy (NCT) efficacy in breast cancer are inconsistent, suggesting that the ability of a single factor to predict efficacy is insufficient. Combining multiple potential efficacy-related factors to build a model may improve the accuracy of prediction. This study intends to explore the clinical and biological factors in breast cancer patients receiving NCT and to establish a nomogram that can predict the pathologic complete response (pCR) rate of NCT. Methods We selected 165 breast cancer patients receiving NCT from July 2017 to May 2019. Using pretreatment biopsy materials, immunohistochemical studies to assess estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER-2), and Ki-67 expression. The correlation between biological markers and pCR was analyzed. These predictors were used to develop a binary logistic regression model with cross-validation and to show the established predictive model with a nomogram. Results The nomogram for pCR based on lymphovascular invasion, anemia (hemoglobin≤120 g/L), ER, Ki67 expression levels and NCT regimen had good discrimination performance (area under the curve [AUC], 0.758; 95% confidence interval [CI], 0.675–0.841) and calibration coordination. According to the Hosmer-Lemeshow test, the calibration chart showed satisfactory agreement between the predicted and observed probabilities. The final prediction accuracy of cross-validation was 76%. Conclusions We developed a nomogram based on multiple clinical and biological covariations that can provide an early prediction of NCT response and can help to quickly assess the individual benefits of NCT.