Predictive value of circulating tumor cells FTH1 gene on the efficacy of neoadjuvant chemotherapy in non-metastatic breast cancer.

Abstract

e12599 Background: Based on epithelial-mesenchymal transition (EMT) markers, circulating tumor cells (CTCs) can be classified into two main subtypes: epithelial CTCs(E-CTCs) and EMT-CTCs. CTCs with high expression of the FTH1 gene (HEF) may be related to poor efficacy of neoadjuvant chemotherapy (NAC). The relation between the CTCs with HEF and pathological complete response (pCR) after NAC is unclear. Methods: 83 non-metastatic breast cancer patients who completed NAC and surgery were enrolled in this study, and each patient had at least one result of CTC test before surgery. Expression of the FTH1 gene on CTCs was achieved by multiplex RNA in situ hybridization. Mann-Whitney U tests was conducted to examine the differences between the non-pCR and pCR groups in the numbers of total CTCs, E-CTCs and EMT-CTCs with HEF. Predictive ability of total and subtypes of CTCs with HEF to pCR rate was compared by the area under receiver operating characteristic curves (AUC). Univariate and multivariate binary logistic regression were used to select the variable and construct the predictive model. Results: Among 64 patients who underwent CTC test before NAC, CTCs were found in 52 patients, of which 20 patients achieved pCR. There was no significant difference in the numbers of total CTCs (9.9±10.9 vs 7.9±7.1, p=0.828) and E-CTCs with HEF (0.6±0.9 vs 0.3±0.7, p=0.256). Compared with the patients in the non-pCR group, those in pCR group had fewer EMT-CTCs with HEF (2.5±5.1 vs 1.0±2.5, p<0.01). 56 patients had CTC test after NAC, CTCs were detected in 53 patients, of which 18 patients gained pCR. There was no significantly difference in the numbers of total CTCs (11.7±13.8 vs 7.1±6.2, p=0.386), E-CTCs with HEF (0.5±1.0 vs 0.3±0.6, p=0.699) and EMT-CTCs with HEF (3.6±7.9 vs 1.3±3.1, p=0.323) between the two groups. In patients with CTC tests before NAC, EMT-CTCs with HEF have the highest predictive ability (AUC=0.718, 95%CI=0.567-0.869) than total CTCs (AUC=0.518, 95%CI=0.359-0.677) and E-CTCs with HEF (AUC=0.575, 95%CI=0.417-0.733). Univariate binary logistic regression analysis showed that HR status (OR=4.667, 95%CI= 1.177-18.506), HER2 status (OR=0.059, 95%CI=0.014-0.255) and EMT-CTCs with HEF status (<1 vs ≥1) (OR=8.800, 95%CI=2.336-33.152) were significantly related with the pCR rate. These variables remain significant in the multivariate logistic analysis. The AUC value of the pCR rate prediction model constructed based on the multivariate binary logistic regression results was significantly higher than the model combining HR status and HER2 status (AUC=0.930, 95%CI=0. 861-1.000 vs AUC=0.866, 95%CI=0.765-0.966; p<0.05), based on Delong test. Conclusions: EMT-CTCs with HEF before NAC has the predictive ability for the rate of pCR in patients with non-metastatic breast cancer. This will help clinicians personalize the treatment strategy for patients to achieve better survival.