Identification of a prognostic gene signature in patients with cisplatin resistant squamous cell lung cancer

Background: In the absence of targeted mutations and immune checkpoints, platinum-based chemotherapy remains a gold standard agent in the treatment of patients with lung squamous cell carcinoma (LUSC). However, cisplatin resistance greatly limits its therapeutic efficacy and presents challenges in the treatment of lung cancer patients. Therefore, the potential clinical needs for this research focus on identifying novel molecular signatures to further elucidate the underlying mechanisms of cisplatin resistance in LUSC. A growing body of evidence indicates that alternative splicing (AS) events significantly influence the tumor progression and survival of patients with LUSC. However, there are few systematic analyses of AS reported in LUSC. This study aims to explore the role of messenger RNA (mRNA), microRNA (miRNA), and AS in predicting prognosis in patients with cisplatin-resistant LUSC and provide potential therapeutic targets and drugs. Methods: Gene expression and miRNA expression, using RNA sequencing (RNA-seq), and SpliceSeq data were downloaded from The Cancer Genome Atlas (TCGA) database. The least absolute shrinkage and selection operator (LASSO) Cox regression analysis were used to construct predictive models. Kaplan- Meier survival analyses were used to evaluate patients' prognosis. Single-sample gene set enrichment analysis (ssGSEA) conducted via the R package "GSEAbase" was used to evaluate the immune-related characteristics. Immunohistochemistry was used to examine protein expression. The Connectivity Map (CMap) database was used to screen for potential drugs. The 3-(4,5)-dimethylthiahiazo (-z-y1)-3,5-di-phenytetrazoliumromide (MTT) assay was used to determine and calculate the half-maximal inhibitory concentration (IC50) 50 ) of the drugs, sulforaphane and parthenolide. Results: In this study, bioinformatics were used to identify mRNAs, miRNAs, and AS events related to response to cisplatin and to establish an integrated prognostic signature for 70 patients with LUSC and cisplatin resistance. The prognostic signature served as an independent prognostic factor with high accuracy [hazard ratio (HR) =2.346, 95% confidence interval (CI): 1.568-3.510; P<0.001], yielding an area under the curve (AUC) of 0.825, 0.829, and 0.877 for 1-, 3-, and 5-year survival, respectively. It also demonstrated high predictive performance in this cohort of patients with LUSC, with an AUC of 0.734, 0.767, and 0.776 for 1-, 3-, and 5-year survival, respectively. This integrated signature was also found to be an independent indicator among conventional clinical features (HR =2.288, 95% CI: 1.547-3.383; P<0.001). In addition, we analyzed the correlation of the signature with immune infiltration and identified several small-molecule drugs that had the potential to improve the survival of patients with LUSC. Conclusions: This study provides a framework for the mRNA-, miRNA-, and AS-based evaluation of cisplatin response and several potential therapeutic drugs for targeting cisplatin resistance in LUSC. These findings may serve as a theoretical basis for the clinical alleviation of cisplatin resistance and thus help to improve treatment responses to chemotherapy in patients with LUSC.