Low expression of TOX predicts poor prognosis of patients with breast cancer in the real world: a retrospective study

Abstract

Background: TOX is a transcription factor that is implicated in the regulation of T cell exhaustion in tumors. TOX has been proven to have prognostic value in some malignant tumors. We aim to analyze the expression of TOX in breast cancer patients, and the association between TOX and prognostic significance in patients with breast cancer. Methods: 313 breast cancer patients were enrolled into the current study. The expression of TOX was determined by the immunohistochemistry assay in breast cancer tissues. The relationships between TOX and clinical pathological variables in breast cancer were performed via chi-square test. Survival curves were performed by Kaplan-Meier method and log-rank test. The potential independent factors were assessed by Cox model regression analyses. Nomogram models, calibration curve, decision curve analyses were applied to analyze the clinical utility of the predictive models. Results: According the semi-quantitative scoring under electron microscope, 129 breast cancer patient samples were classified into low group, and 184 breast cancer patient samples were classified into high group. Patients with high expression of TOX had survived longer than those with low expression of TOX (DFS: 71.70 vs. 64.05 months; OS: 81.03 vs. 73.72 months), and the difference in survival time between the two groups was statistically significant (DFS: χ2=11.6300, P=0.00065; OS: χ2=11.4200, P=0.00073). Patients with high expression of TOX had survived longer than those with low expression of TOX in patients received adjuvant therapies, especially in patients with chemotherapy (DFS: 72.85 vs. 64.53 months, P=0.00029; OS: 82.06 vs. 74.29 months, P=0.00033). Based on the Cox model regression analyses for DFS and OS, multivariate analysis indicated that TOX [DFS, hazard ratio (HR): 0.412, 95%CI: 0.248-0.684, P=0.001); OS, HR: 0.395, 95%CI: 0.237-0.660, P＜0.0001] was the potential prognostic factor. The established nomogram was used to predict survival and the calibration curve analysis had performed that the predicted line was well-matched with base line in postoperative 1-, 3-, and 5-year survival rate. Conclusions: The expression of TOX is a potential prognostic factor, and can be a promising biomarker for predicting survival and guiding treatment strategy in breast cancer patients.