Prediction model of axillary lymph node metastasis in HER2 negative breast cancer by PR gene expression combined with ultrasound image and its application
Objective To explore the risk factors for axillary lymph node metastasis(ALNM)in human epidermal growth factor receptor 2(HER2)negative breast cancer by combining the expression of progesterone receptor(PR)gene with conventional ultrasound images,and to construct an ultrasound prediction model and evaluate its value.Methods The clinical data of 281 HER2 negative breast cancer patients were retrospectively collected.According to the postoperative pathological re-sults,the patients were divided into metastatic group and non-metastatic group,and divided into training set and test set in the ratio of 8∶2.In the training set,the logistic regression analysis was used to screen out the independent risk factors associated with the outcome,and a prediction model was established.In the test set,the relevant metrics were calculated to evaluate the performance of the model.Results The results of univariate analysis and multivariate analysis showed that clustered calcifica-tion,hyperechoic halo,posterior acoustic attenuation and positive PR gene expression were independent risk factors for ALNM in HER2 negative breast cancer patients(P＜0.05).The prediction model was:Y=-5.28+0.90 × clustered calcification(0:absent;1:present)+1.22 × hyperechoic halo(0:absent;1:present)+2.06 × posterior acoustic attenuation(0:absent;1:present)-1.34 × Ki_67(1 at an expression level of 5％-30％,otherwise 0)+0.75× Ki_67(1 at expression level＞30％,otherwise 0)+3.79XPR gene expression(0:negative;1:positive)+0.71 × maximum tumor diameter(1 over 2 cm,other-wise 0).The Hosmer-Lemeshow goodness-of-fit test showed P=0.110.In the training set,the model consistency index(C in-dex)reached 0.870.When the model score was≥0.495,the presence of ALNM was considered.In the test set,the model C index was 0.812,the sensitivity,specificity,positive predictive value,negative predictive value and accuracy of the model in di-agnosing ALNM were 0.750,0.679,0.700,0.731 and 0.714,respectively.Conclusion Clustered calcification,hyperechoic halo,posterior acoustic attenuation and positive PR gene expression were independent risk factors for axillary lymph node metas-tasis of HER2 negative breast cancer,and the constructed risk prediction model can well predict ALNM of breast cancer.
breast canceraxillary lymph nodemetastasisPR geneultrasound
万方数据
维普
