Predicting survival and immunotherapy response in triple-negative breast cancer using a platinum resistance-related gene signature

Triple-negative breast cancer (TNBC) is a highly aggressive malignancy in women with a poor prognosis. Platinum resistance-related genes are intricately associated with the treatment of TNBC, but their potential role in the prognosis of TNBC patients remains incompletely understood. The training set and the validation set of TNBC patients were obtained from the TCGA database and the GSE58812 dataset, respectively. Platinum resistance-related differentially expressed genes were enriched, and a prognostic risk signature for TNBC was constructed. The model was validated through KM survival analysis and an ROC curve. By integrating the risk scores and relevant clinical traits, a nomogram was created to enhance the clinical efficacy of our prediction tool. In addition, the correlations between the risk scores and the immune microenvironment as well as drug sensitivity in TNBC were investigated. Furthermore, we conducted gain- and loss-of-function experiments on a risk gene to verify its impact on the malignant biological behaviors of TNBC cells. A four-gene signature related to platinum resistance successfully classified TNBC patients into high- and low-risk groups. The established nomogram demonstrated superior prognostic prediction performance for TNBC compared to other individual predictors and models. In addition, in the high-risk group, the infiltration levels and functions of immune cells were decreased, the expression of immune checkpoint genes was weakened, and the TIDE scores were elevated. The high-risk group was less sensitive to nine chemotherapeutic drugs. In vitro experiments have confirmed that SLC27A2 and EXO1 inhibit the proliferation and invasion of TNBC cells, which is consistent with our finding that their activity is suppressed in tumors. In contrast, TMEM88 and TGFB1 exhibited oncogenic properties by promoting these malignant behaviors. The platinum resistance-related gene signature can serve as a tool to predict the outcomes and select a more appropriate treatment regimen for patients with TNBC.