摘要
Objective： The indication of adjuvant chemotherapy recommendation（ACR） in breast cancer patients with intermediate recurrence score（RS） is controversial. This study investigated the relationship between routine clinicopathological indicators and ACR, and established a nomogram for predicting the probability of ACR in this subset of patients.Methods： Data for a total of 504 consecutive patients with intermediate RS from January 2014 to December2016 were retrospectively reviewed. A nomogram was constructed using a multivariate logistic regression model based on data from a training set（378 cases） and validated in an independent validation set（126 cases）. A Youdenderived cut-off value was assigned to the nomogram for accuracy evaluation.Results： The multivariate logistic regression analysis identified that age, histological grade, tumor size, lymph node（LN） status, molecular subtype, and RS were independent predictors of ACR. A nomogram based on these predictors performed well. The P value of the Hosmer-Lemeshow test for the prediction model was 0.286. The area under the curve（AUC） values were 0.905 [95% confidence interval（95% CI）： 0.876–0.934] and 0.883（95%CI： 0.824–0.942） in the training and validation sets, respectively. The accuracies of the nomogram for ACR were84.4% in the training set and 82.1% in the validation set.Conclusions： We developed a nomogram to predict the probability of ACR in breast cancer patients with intermediate RS. This model may aid the individual risk assessment and guide treatment decisions in clinical practice.
Objective： The indication of adjuvant chemotherapy recommendation（ACR） in breast cancer patients with intermediate recurrence score（RS） is controversial. This study investigated the relationship between routine clinicopathological indicators and ACR, and established a nomogram for predicting the probability of ACR in this subset of patients.Methods： Data for a total of 504 consecutive patients with intermediate RS from January 2014 to December2016 were retrospectively reviewed. A nomogram was constructed using a multivariate logistic regression model based on data from a training set（378 cases） and validated in an independent validation set（126 cases）. A Youdenderived cut-off value was assigned to the nomogram for accuracy evaluation.Results： The multivariate logistic regression analysis identified that age, histological grade, tumor size, lymph node（LN） status, molecular subtype, and RS were independent predictors of ACR. A nomogram based on these predictors performed well. The P value of the Hosmer-Lemeshow test for the prediction model was 0.286. The area under the curve（AUC） values were 0.905 [95% confidence interval（95% CI）： 0.876–0.934] and 0.883（95%CI： 0.824–0.942） in the training and validation sets, respectively. The accuracies of the nomogram for ACR were84.4% in the training set and 82.1% in the validation set.Conclusions： We developed a nomogram to predict the probability of ACR in breast cancer patients with intermediate RS. This model may aid the individual risk assessment and guide treatment decisions in clinical practice.
