Exploring program-cell death patterns to predict prognosis and sensitivity of cervical cancer immunotherapy via multi-omics analysis and clinical samples.

BACKGROUND

Cervical cancer (CC) progression and therapeutic resistance are driven by metastatic dissemination and immune evasion. Although immunotherapy has emerged as a promising strategy, current biomarkers fail to adequately predict patient prognosis or immune checkpoint inhibitor (ICI) responsiveness. Programmed cell death (PCD) pathways are intricately linked to tumor-immune crosstalk, yet their systematic integration into predictive models remains unexplored in CC.

METHODS

We constructed a prognostic gene model for PCD by mining the Cancer Genome Atlas (TCGA), GEO, and Genecards databases. The predictive capability of the model was assessed using Kaplan-Meier (K-M) analysis and Receiver Operating Characteristic (ROC) curve analysis. A nomogram was generated through Cox regression. The model was validated in both training and testing cohorts. Real-time quantitative PCR (qRT-PCR) and immunohistochemistry were used to verify the expression of the model genes. Finally, functional analysis of low- and high-risk groups based on the median risk score was performed, including immune infiltration, genomic mutations, tumor mutational burden (TMB), and drug sensitivity.

RESULTS

We established a prognostic model based on six PCD-related genes: CD46, TFRC, PGK1, GNG5, GAPDH, and PLAU. The risk score demonstrated good performance, with area under the curve (AUC) values indicating strong predictive ability (TCGA: AUC 1-year = 0.761, AUC 3-year = 0.754, AUC 5-year = 0.803; GEO: AUC 1-year = 0.702, AUC 3-year = 0.632, AUC 5-year = 0.579). Higher risk scores were associated with poorer overall survival (OS). Additionally, low-risk patients exhibited increased immune cell infiltration, higher IPS scores, enhanced expression of PDCD1 and CTLA4, and greater sensitivity to Niraparib, Paclitaxel, and Cisplatin. qRT-PCR confirmed overexpression of CD46, TFRC, PGK1, GNG5, and PLAU in cervical cancer cell lines and tissues, with consistent findings in immunohistochemistry (IHC).

CONCLUSION

This study establishes CDI as the PCD-based immune signature for CC, enabling precise prognosis prediction and ICI candidate selection. The CDI framework provides actionable insights for combination therapies targeting PCD-immune interplay, with translational potential for personalized oncology.