Establishment of the model for selection of dominant advanced triple negative breast cancer patients with chemotherapy under the guidance of individualized gene detection

  Objective  To establish a model for therapeutic prediction of advanced triple negative breast cancer (TNBC) patients with chemotherapy under the guidance of individualized gene detection.

  Methods  Training set included 97 patients from May 2013 to May 2015 in Breast Center of Cancer Hospital of Henan University, they all participated in a clinical study on "Efficacy of advanced TNBC patients with chemotherapy based on the guidance of individualized gene detection". All the patients were conducted with chemotherapy, and the expressions of ERCC1, TOP2A, TUBB3 and TYMS genes were detected. Based on this group of people, the efficacy prediction model was established and verified internally. Logistic method was used to screen the efficacy-related predictive factors, and the predictive equation model and Nomogram were established finally.

  Results  Univariate analysis showed that histological grade, brain metastasis, Ki-67 index, score of Performance Status (PS), and expressions of ERCC1, TOP2A, TUBB3 and TYMS genes were correlated with efficacy (P < 0.05 or P < 0.01). Multivariate regression analysis showed that histological grade, brain metastasis, and the expressions of ERCC1, TOP2A and TUBB3 genes were correlated with efficacy (P < 0.05 or P < 0.01). Based on the results of multivariate regression, the efficacy prediction model was established and the Nomogram was drawn. The area under curve (AUC) of the model was 0.861, 95% CI was 0.789 to 0.933, and the best cut-off value was 0.739. Chi-square goodness-of-fit test showed no significant difference between predicted probability and measured probability (χ²=1.698, P=0.975).

  Conclusion  The model established in this study can predict the dominant advanced TNBC patients with chemotherapy and provide references for clinical decision-making.