Prediction of the Effectiveness of Neoadjuvant Polychemotherapy in Patients with Triple-Negative Breast Cancer

This study was aimed at developing an algorithm for predicting the effectiveness of neoadjuvant chemotherapy in the treatment of patients with triple-negative breast cancer. For prognosis, the patient’s clinical and hematological parameters obtained before treatment are used. The development of the prognosis algorithm is based on statistical analysis of retrospective data containing 81 cases of the disease. As a result of using a logistic regression model, it turned out that out of 19 indicators participating in the study, only 3 (index Ki-67, tumor size, and neutrophil to lymphocyte ratio) are significant predictors for prognosis. Two complementary forms of prognosis: probabilistic and binary are proposed. The influence of significant predictors on the prognosis of treatment effectiveness is demonstrated by several examples. The calculated characteristics of the prognosis quality are given. Their numerical values indicate the low quality of the prognosis, confirming the difficulties of predicting the effectiveness of treatment for patients in this category.