BRCA1/2 Germline Mutations Risk Prediction Nomogram Modles Establishment And Comparison In Early-onset Breast Cancer In China

Object: Breast cancer is the most common type of cancer and the most common cause of cancer deaths among females worldwide.In China,the incidence and mortality of breast cancer is rising rapidly.Early screening and diagnosis of breast cancer has been paid more and more attention.In clinical work,BRCA1/2 mutations often appear in family heredity,and the carriers are prone to suffer from early-onset breast cancer,and there are special precautions and special treatments for the carriers.Therefore,it is necessary to select possible carriers in the population and carry out BRCA1/2 mutation detection.Currently the most widely used BRCA risk calculator models are the BRCAPRO and Myriad models.However,numbers of independent post-verification studies show that there are significant differences in the performance of these risk prediction models in different environments,regions and ethnicities.Especially,the prediction of these two models is not satisfactory in the verification of Asian.Therefore,we retrospectively collected epidemiological and clinicopathological data of early-onset breast cancer female patients in Southeast China,to determine the relationship between these factors and the occurrence of BRCA1/2 mutation,and explore and establish new BRCA1/2 risk prediction nomogram models.Methods: Our study included 1014 female patients of early onset breast cancer.First of all,a epidemiological survey were carried out among all participants,including demographic data,menstrual,reproductive and breastfeeding history,hormone replacement therapy history,previous benign breast disease history,and their family history of breast cancer.Subsequently,a 3-ml peripheral blood sample was collected to extract genomic DNA,and and The whole BRCA1/2 exon of these samples were screened by using the next generation sequencing(NGS).The associations between BRCA1/2 and the risk of breast cancer were assessed bycomputing ORs and 95% CIs using multivariate logistic regression models,and then shown in the form of nomogram.The ROCs of the nomogram model and the BRCAPRO and Myriad models are drawn and the ROCs are compared with the DeLong’s test.Results: The areas under the ROC were 0.698(95%CI,0.632-0.763)and0.525(95%CI,0.444~0.606)for BRCA1 /2,respectively,for BRCAPRO and Myriad.The result of single Logistic regression of the factors of BRCA 1/2 germline mutation probability in Chinese Han patients with early onset breast cancer: age at diagnosis(>35y,OR 0.626,95%CI 0.365~1.073),menstrual period(≥21y,OR 0.531,95%CI0.306-0.922),Her-2(positive,OR0.203,95%CI0.073~0.568),Ki67(≥15%;OR2.794,95%CI1.25-6.245),Triple-Negative Breast Cancer(TNBC)(OR 5.735,95% CI2.032~16.182),Luminal B(HER 2-)(OR 3.484,95%CI 1.332-9112),Medullary(OR15.559,95%CI 1.521-159.185),Other Histologic Subtype(including Mixed)(OR10.593,95%CI 1.349~83.203).According to the results of logistic regression,we have established three mutation risk prediction nomogram models for BRCA1,BRCA2,and BRCA1\2 germline mutation.the AUC of the three modles are 0.83,0.782 and 0.738.The results are better than the BRCAPRO and Myriad models’,and DeLong’s showed that there was no statistically significant difference between our nomograms and BRCAPRO,and there was a significant difference with Myriad.Conclusion: The BRCAPRO model has a reasonable sensitivity and specificity to the prediction of BRCA risk in early onset breast cancer patients in China,and The Myriad model is not sufficient to predict the BRCA risk in Chinese patients.we have established three germline mutation risk prediction nomogram models for BRCA1,BRCA2,and BRCA1\2,and the accuracy of the three models is moderate on the training set and the testing set.Comparing our nomogram model with BRCAPRO and Myriad models,it is not worse than the BRCAPRO model,and is obviously better than the Myriad model.The nomograms established in this study is suitable for Chinese breast cancer population,and it needs a larger sample to verify and improve it.In addition,our data suggest that the BRCA1/2 germline mutation rate of early breast cancer patients with Han women in the southeast China is smallt(5.52%).If we pickout patients who were diagnosed less than 35 years old,TNBC or Luminal B(HER-2-)breast cancer,to perform BRCA1/2 test,The possibility of pathogenic mutation carriers will be higher.