Clinicopathologic Evaluation and Prediction Model for Stromal Tumor-Infiltrating Lymphocytes in the Development of Breast Cancer

OBJECTIVE: To investigate the correlation between clinicopathologic characteristics and the ratio of stromal tumor-infiltrating lymphocytes (sTILs) in patients with breast cancer, and to establish a prediction model for sTILs with which to evaluate breast cancer development. STUDY DESIGN: A total of 638 infiltrative breast cancer patients were recruited in this clinical trial. The correlation between clinicopathologic parameters and sTILs was statistically analyzed using R software. The calculated correlated parameters were utilized to establish an sTIL prediction model. The prediction accuracy of this model was assessed. RESULTS: The sTIL ratio was positively correlated with histologic grading and the expression of HER2, Ki-67, and P53, whereas the sTIL ratio was negatively correlated with estrogen receptor (ER) and progesterone receptor (PR) expression. The sTIL ratio in breast cancer patients with positive hormone receptor (HR) expression was the highest, followed by patients with positive HER2 and triple-negative breast cancer (TNBC). The sTIL ratio yielded the highest correlation with histologic grading (R = 0.4963), followed by Ki-67 (R = 0.3999). The sTIL decision tree and lymphocyte-predominant breast cancer (LPBC) classification tree prediction models were established. According to the simulation calculation, the degree of correlation be-tween the theoretical and actual values of sTILs was 0.574 (p < 0.001). The degree of correlation between the theoretical and actual values of LPBC was 0.373 (p < 0.001). CONCLUSION: We have shown that sTILs reflect the malignant tumor microenvironment. The prediction model established by clinicopathologic parameters might be suitable for prediction of LPBC.