Combination of individual tumor gene expression profiles and quantitative ultrasound and mammography imaging to understand the response to neoadjuvant chemotherapy among Hispanic-Latina women with early stages of triple-negative breast cancer.

Abstract

e12605 Background: Triple-negative breast cancer (TNBC) is a heterogeneous subtype of breast cancer (BC) characterized by the absence of targetable receptors. Traditionally, neoadjuvant chemotherapy (NACT) has been used to downstage the tumors and increase the chance for breast-conserving surgery. The pathological complete response (pCR) has been traditionally considering the best predictive marker for the disease recurrence. Patients with residual disease (RD) have a poor prognosis with a high risk of recurrence, and therefore additional chemotherapy was recommended. Therefore it is an important task for clinical researchers to identify markers to predict the individual tumor response to chemotherapy and avoid in patients potentially resistant tumors. Instead, a surgical approach can be used or combined approach with chemotherapy and immunotherapy. It is not clear yet which approach is optimal for those patients with chemotherapy-resistant tumors since there is no clinical data available and no clinical tool that helps predict the individual tumor response. In this study, we examined breast ultrasound(US) images of patients before and after the completion of NACT and correlated with response to chemotherapy. To better understand the biology of resistance to chemotherapy, we also analyzed the gene expression profile of 15 patients with RD after NACT. Methods: In this study, we retrospective analyzed breast US data from 37 Hispanic patients diagnosed with TNBC and treated with NACT. Patients underwent breast US before and after NACT with documentation of clinical complete response (cCR) or clinical residual disease (cRD). Post-operatively, the pathologic response was defined as the absence of tumor cells (pCR) or presence of residual invasive tumor (RD). A multivariable logistic regression model assessed the influence of patient- and tumor-associated covariates as predictors for pCR. Also, we analyzed formalin-fixed paraffin-embedded tumor samples from 15 patients with RD after NACT. Results: Seventeen patients (45.9%) achieved pCR, and twenty (54.1%) had RD after NACT. The most common US findings connected with RD was the deposition of calcium before NACT six (30%) patients. Gene expression analysis of RD samples identified 446 upregulated and 275 downregulated genes. Among commonly upregulated genes related to cancer, we identified GLI1, IGF1, SERPINE1, ATF3, KLK 5; 7, and TUBB2b, and genes belonging to pathways encoding extracellular matrix–related proteins, DNA-damage response proteins, and pathways related to resistance to chemotherapeutic agents such as Taxol. Conclusions: Our data suggested that gene expression profiling in combination with imaging study can be used to identify patients with TNBC potentially resistant to chemotherapy.