Construction Of A Prediction Model For Sentinel Lymph Node Metastasis In Early Breast Cancer Patients(cT1-2N0M0)

Research background and Purpose:In recent years,the incidence of breast cancer has been increasing every year.Breast cancer which threatens global women’s health is the most common malignant tumor,it is also the second leading cause of cancer death in women.Axillary lymph node metastasis is a risk factor for local recurrence and poor prognosis in breast cancer patients.Accurate identification of axillary lymph node status is important for prognosis and treatment decisions.Sentinel lymph node(SLN)is the first or first station lymph node of primary tumor metastasis,and its pathological status can predict the status of regional lymph nodes.Sentinel lymph node biopsy(SLNB)has replaced axillary lymph node dissection(ALND)as the gold standard for axillary staging and local control in patients with early breast cancer who have negative clinical axillary lymph node palpation.SLNB can provide sufficient axillary lymph node staging information to guide further regional and systemic treatment without affecting the regional control rate.Compared with ALND,although SLNB has fewer postoperative complications,SLNB can increase the duration of anesthesia and the cost of surgery,and even 2%-6 % of patients have upper limb lymphedema and 9% of patients have long-term sensory abnormalities.At the same time,about 60 to 70 percent of patients(c N0)with early breast cancer are SLN negative.For this group of patients,SLNB is not necessary and avoiding SLNB may largely reduce the associated complications,the length of the procedure and the associated costs.Therefore,we need a tool to help screen this group of early breast cancer patients so that SLNB can be safely and effectively avoided.For this purpose,this study constructed a non-invasive preoperative prediction model to predict the status of axillary lymph nodes in order to avoid unnecessary SLNB.Methods:Retrospective analysis of clinicopathological data of patients with early breast cancer(cT1-2N0M0)at our institution from January 1,2016,to December 1,2019.According to time of the patient’s diagnosis,patients were divided into a training set and a validation set.The related factors of sentinel lymph node metastasis in early breast cancer were analyzed by univariate and multivariate analysis,the nomogram was constructed,and the ability of the prediction model was verified by the validation set.P < 0.05 was statistically significant.Results:1.In 1010 patients with early breast cancer(cT1-2N0M0)with negative palpation of axillary lymph nodes,the age was 23-79 years and the median age was 52 years;the number of sentinel lymph nodes resected was 0-7 and the median number of resections was 3.The rate of sentinel lymph node metastasis was 32.7%.2.Univariate analysis of sentinel lymph node metastasis in early breast cancer showed that pathological type,tumor location,tumor size,menopausal status,ER,PR,KI-67,and axillary lymph node grading were associated with sentinel lymph node metastasis(all P < 0.05);age and HER-2were not associated with sentinel lymph node metastasis(all P > 0.05).3.Multivariate analysis of sentinel lymph node metastasis in early breast cancer showed that pathological type,tumor size,PR,KI-67 and axillary lymph node grading were independent influencing factors of sentinel lymph node metastasis(all P < 0.05).4.According to the results of multivariate Logistic analysis,a nomogram is constructed with statistically significant variables.The area under the operating curve was 0.885(95% confidence interval is 0.854-0.911)for subjects in the training set who predicted sentinel lymph node metastasis and0.874(95% confidence interval is 0.842-0.901)for subjects in the validation set who predicted sentinel lymph node metastasis.The calibration curve shows that the prediction model has good performance.Conclusion:1.Univariate analysis of sentinel lymph node metastasis in early breast cancer showed that pathological type,tumor location,tumor size,menopausal status,ER,PR,KI-67,and axillary lymph node grading were associated with sentinel lymph node metastasis(all P < 0.05),and age and HER-2 were not associated with sentinel lymph node metastasis(all P >0.05).2.Pathological type,tumor size,PR,KI-67,and axillary lymph node grading were independent influencing factors for sentinel lymph node metastasis(all P > 0.05).3.The prediction model of sentinel lymph node metastasis constructed by independent risk factors has high predictive value for sentinel lymph node metastasis in patients with early breast cancer(cT1-2N0M0).It is helpful to provide clinicians with diagnosis and treatment information and make reasonable axillary operation decisions.