PANoptosis-related molecular subtype and prognostic model associated with the immune microenvironment and individualized therapy in pancreatic cancer

• Published in: Frontiers in oncology Vol. 13; p. 1217654
• Main Authors: Zhang, Biao, Huang, Bingqian, Zhang, Xiaonan, Li, Shuang, Zhu, Jingyi, Chen, Xu, Song, Huiyi, Shang, Dong
• Format: Journal Article
• Language: English
• Published: Switzerland Frontiers Media S.A 14.07.2023
• Subjects: molecular subtypes; Oncology; pan-cancer analysis; pancreatic cancer; PANoptosis; prognostic model; tumor microenvironment; molecular subtypes; drug sensitivity; pancreatic cancer; prognostic model; tumor microenvironment; PANoptosis; pan-cancer analysis
• ISSN: 2234-943X; 2234-943X
• DOI: 10.3389/fonc.2023.1217654

PANoptosis is an inflammatory type of programmed cell death regulated by PANopotosome. Mounting evidence has shown that PANoptosis could be involved in cancer pathogenesis and the tumor immune microenvironment. Nevertheless, there have been no studies on the mechanism of PANoptosis on pancreatic cancer (PC) pathogenesis. We downloaded the data on transcriptomic and clinical features of PC patients from the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus databases. Additionally, the data on copy number variation (CNV), methylation and somatic mutations of genes in 33 types of cancers were obtained from TCGA. Next, we identified the PANoptosis-related molecular subtype using the consensus clustering analysis, and constructed and validated the PANoptosis-related prognostic model using LASSO and Cox regression analyses. Moreover, RT-qPCR was performed to determine the expression of genes involved in the model. We obtained 66 PANoptosis-related genes (PANRGs) from published studies. Of these, 24 PC-specific prognosis-related genes were identified. Pan-cancer analysis revealed complex genetic changes, including CNV, methylation, and mutation in PANRGs were identified in various cancers. By consensus clustering analysis, PC patients were classified into two PANoptosis-related patterns: PANcluster A and B. In PANcluster A, the patient prognosis was significantly worse compared to PANcluster B. The CIBERSORT algorithm showed a significant increase in the infiltration of CD8 T cells, monocytes, and naïve B cells, in patients in PANcluster B. Additionally, the infiltration of macrophages, activated mast cells, and dendritic cells were higher in patients in PANcluster A. Patients in PANcluster A were more sensitive to erlotinib, selumetinib and trametinib, whereas patients in PANcluster B were highly sensitive to irinotecan, oxaliplatin and sorafenib. Moreover, we constructed and validated the PANoptosis-related prognostic model to predict the patient's survival. Finally, the GEPIA and Human Protein Atlas databases were analyzed, and RT-qPCR was performed. Compared to normal tissues, a significant increase in and (associated with the model) expression was observed in PC tissues. We first identified the PANoptosis-related molecular subtypes and established a PANoptosis-related prognostic model for predicting the survival of patients with PC. These results would aid in exploring the mechanisms of PANoptosis in PC pathogenesis.