Using TCGA Database To Establish A Risk Prediction Model Of Drug Resistance In HER2 Positive Breast Cancer Patients After Targeted Therapy Based On Ferroptosis

Objective: To determine the key ferroptosis genes in HER2 positive breast cancer patients with targeted drug resistance based on TCGA database.And to establish the correlated risk prediction model.Methods: the clinical data and gene expression profiles of HER2 positive breast cancer patients were collected through TCGA database.The differentially expressed genes between HER2-positive breast cancer targeted therapy drug-resistant patients and non-drug-resistant breast cancer patients were analyzed by limma R software package of R,and the differentially expressed genes were analyzed by KEGG enrichment function interpretation.The genes related to ferroptosis were obtained,and the differential genes in ferroptosis genes were identified by univariate analysis.According to the differential ferroptosis gene and the clinical characteristics of the patients,univariate and multivariate COX regression analysis was performed to identify the independent risk factors of drug resistance in HER2 positive breast cancer patients.Based on the results of multi-factor COX regression analysis,the Nomo map was drawn to establish the risk prediction model.Draw ROC curve and calculate the area under the curve,and verify the accuracy and clinical generalization ability of the prediction model by drawing calibration curve and DCA curve.Results: a total of 239 HER positive breast cancer patients with targeted therapy were obtained by TCGA database.According to whether the patients were drug resistant or not,the differentially expressed genes were analyzed.A total of 25 ferroptosis key genes were obtained by consulting the literature,and further analysis showed that 3 of the 25 ferroptosis related genes were significantly up-regulated in drug resistant patients(NFE2L2,NCOA4,LPCAT4).Univariate and multivariate COX analysis based on NFE2L2,NCOA4,LPCAT4 and clinical characteristics showed that age,ER,PR and NCOA4 were independent risk factors for targeted drug resistance in patients with breast cancer.The Nomo map was drawn according to the data of age,ER,PR,NCOA4,bone metastasis and brain metastasis to establish a risk prediction model.The ROC curve was drawn and the area under the curve was calculated,which showed that the area under the curve at 3 months,6 months and 12 months after targeted therapy was 0.673,0.693 and 0.774 respectively,indicating that the model had a good predictive effect.The calibration curves of 3 months,6 months and 12 months after targeted therapy showed that the predicted values coincided with the actual values,which proved that the model was in good agreement with the actual situation.Drawing DCA curve shows that the prediction model has a better net profit rate.Conclusion: NFE2L2,NCOA4 and LPCAT4 are all iron cell death related genes,which are significantly increased in targeted therapy and drug resistance of HER2 positive breast cancer.Age,ER,PR and NCOA4 are all independent risk factors for drug resistance in patients with breast cancer.The risk prediction model of drug resistance in HER2 positive breast cancer patients after targeted therapy based on ferroptosis gene expression is more effective.