Analysis Of Influencing Factors And Predictive Model Construction Of Breast Cancer Response To Neoadjuvant Chemotherapy

Objective:To explore the influencing factors of neoadjuvant chemotherapy for breast cancer,we establish a clinical prediction model for the efficacy of neoadjuvant chemotherapy,early assess the sensitivity of patients to chemotherapy,and guide individualized treatment.Methods:A retrospective analysis of the clinicopathological data of 124 patients with neoadjuvant chemotherapy for breast cancer who were newly treated at the First Affiliated Hospital of Nanchang University from January 2014 to December 2019 and August 2017 to May 2020 at the Second Affiliated Hospital of Nanchang University,according to the pathological response after neoadjuvant chemotherapy,they are divided into effective group and ineffective group.The χ2 test was used to conduct single-factor analysis on the clinicopathological indicators of the two groups:statistically significant indicators were included in the multivariate logistic regression analysis,independent predictors were screened out,and a clinical prediction model of neoadjuvant chemotherapy efficacy was constructed.The receiver operating characteristic curve(ROC)was used to evaluate the predictive performance of the model.Results:1.A total of 124 patients with neoadjuvant chemotherapy were included in this study,of which 73 were in the effective group and 51 were in the ineffective group.Single factor analysis showed that the two groups in tumor size,ER status,molecular classification,Ki-67 expression status and histological grade difference was statistically significant(P < 0.05),while in age,menstrual status,lymph node status,lesions calcification,clinical stage,chemotherapy cycle,BMI,PR status and the state of HER-2 expression,there was no statistically significant difference.2.Multivariate analysis showed that Ki-67 expression status,molecular classification,and histological grade before chemotherapy were independent predictors of the efficacy of neoadjuvant chemotherapy(P<0.05).The effective rate of chemotherapy for patients with Ki-67 high expression,high histological grade and non-Luminal type is higher than that of Ki-67 low expression,low histological grade and Luminal type patients.3.The area under the curve(AUC)of Ki67,histological grade and molecular classification alone predicting the efficacy of chemotherapy are 0.574,0.598 and0.636 respectively;the AUC of the prediction model is 0.732.The best cutoff value of this model is 0.6795,the sensitivity is 49.3%,the specificity is 86.3%,the positive predictive value is 83.7%,and the negative predictive value is 54.3%.4.In evaluating the efficacy of neoadjuvant chemotherapy,the Kappa value of the consistency test of ultrasound evaluation and pathology evaluation is 0.510;the sensitivity of ultrasound evaluation is 83.6%,the specificity is 66.7%,and the accuracy is 76.6%.Conclusion:1.Ultrasound evaluation of chemotherapy response is generally consistent with pathological evaluation.Ultrasound imaging alone cannot fully and accurately evaluate the true response of chemotherapy.2.Ki-67 expression status,molecular classification,and histological grade before chemotherapy are independent predictors of the efficacy of neoadjuvant chemotherapy for breast cancer;patients with high Ki-67 expression,high histological grade and non-Luminal type are more sensitive to neoadjuvant chemoth erapy,It is easier to benefit from chemotherapy.3.The predictive model has a general predictive ability for the efficacy of neoadjuvant chemotherapy,which can provide a reference for guiding patients’ individualized treatment.