Development and validation of a programmed cell death index to predict the prognosis and drug sensitivity of gastric cancer

Aim: Programmed cell death (PCD) critically influences the tumor microenvironment (TME) and is intricately linked to tumor progression and patient prognosis. This study aimed to develop a novel prognostic indicator and marker of drug sensitivity in patients with gastric cancer (GC) based on PCD.

Methods: We analyzed genes associated with 14 distinct PCD patterns using bulk transcriptome data and clinical information from TCGA-STAD for model construction with univariate Cox regression and LASSO regression analyses. Microarray data from GSE62254, GSE15459, and GSE26901 were used for validation. Single-cell transcriptome data from GSE183904 were analyzed to explore the relationship between TME and the newly constructed model, named PCD index (PCDI). Drug sensitivity comparisons were made between patients with high and low PCDI scores.

Results: We developed a novel twelve-gene signature called PCDI. Upon validation, GC patients with higher PCDI scores had poorer prognoses. A high-performance nomogram integrating the PCDI with clinical features was also established. Additionally, single-cell transcriptome data analysis suggested that PCDI might be linked to critical components of the TME. Patients with high PCDI scores exhibited resistance to standard adjuvant chemotherapy and immunotherapy but might benefit from targeted treatments with NU7441, Dasatinib, and JQ1.

Conclusion: The novel PCDI model shows significant potential in predicting clinical prognosis and drug sensitivity of GC, thereby facilitating personalized treatment strategies for patients with GC.