Identification of the Immune Cell Infiltration Landscape in Pancreatic Cancer to Assist Immunotherapy

Abstract

Background: A malignant tumor's immune environment, including infiltrating immune cell status, can be critical to patient outcomes. Recent studies have shown that immune cell infiltration (ICI) in pancreatic cancer (PC) is highly correlated with the response to immunotherapy and patient prognosis. Therefore, we aimed to create an ICI score that accurately predicts patient outcomes and immunotherapeutic efficacy. Methods: The ICI statuses of patients with PC were estimated from the publicly available The Cancer Genome Atlas (TCGA) pancreatic ductal adenocarcinoma and GSE57495 gene expression datasets using two computational algorithms (CIBERSORT and ESTIMATE). ICI and transcriptome subsets were defined using a clustering algorithm, and survival analysis was also performed. Principal component analysis was used to calculate the novel ICI score, and gene set enrichment analysis was performed to identify the pathways underlying the defined clusters. The tumor mutational burden (TMB) was further explored in TCGA cohort, and survival analysis was used to assess the capability of the ICI and TMB scores to predict overall survival. Additionally, common driver gene mutations and their differential expression in the different ICI score group were investigated. Results: The ICI landscapes of 240 patients were generated using the devised algorithm, revealing three ICI and three gene clusters whose use improved the prediction of overall survival (p = 0.019 and p < 0.001, respectively). Crucial immune checkpoint genes were differentially expressed among these subtypes; the RIG-I-LIKE and NOD-LIKE receptor signaling pathways were enriched in samples with low ICI scores (p < 0.05). We also found that the TMB scores could predict survival outcomes, whereas the ICI scores also could predict prognoses independent of TMB. Notably, ICI scores could effectively predict responses to immunotherapy. KRAS, TP53, CDKN2A, SMAD4 and TTN remained the most commonly mutated genes in PC; moreover, KRAS and TP53 mutation rates were significantly different between the two ICI score groups. Conclusions: We developed a novel ICI score that could independently predict the response to immunotherapy and survival of patients with PC. Evaluation of the ICI landscape in a larger cohort could clarify the interactions between these infiltrating cells, the tumor microenvironment and response to immunotherapy.

Lay abstract

Pancreatic cancer (PC) is a lethal malignancy with a higher mortality rate. Currently, immunotherapy is increasingly interesting to clinical researchers and considered a novel and efficient treatment. However, in clinical practice, immunotherapy has not demonstrated consistent therapeutic responses across all patients. Thus, to identify the immunotherapy-sensitive subgroup of advanced PC patients is important based on immune cell infiltration. In this study, we downloaded and processed transcriptomic data from TCGA-PAAD and GEO databases and used CIBERSORT and ESTIMATE algorithms to reveal the immune cell infiltration landscape of pancreatic cancer. According to consensus clustering results, we identified three ICI and gene clusters for guiding the identification of immune-subtype in future clinical treatments. Finally, we calculated an ICI score for each subject to describe their tumor immune landscape and performed the risk grouping for all patients and multiomics analysis. In sum, the ICI score and clusters could be used in the future to assist clinicians in identifying patients with the greatest chance of responding to immunotherapy.

Keywords:

• immune cell infiltration
• immunotherapy
• pancreatic cancer
• prognosis

Supplementary data

To view the supplementary data that accompany this paper please visit the journal website at: www.tandfonline.com/doi/suppl/10.2217/fon-2021-0495

Author contributions

ZZ Wang, WB Zou and F Wang contributed equally to this work and are considered co-first authors. MG Hu, R Liu, WB Zou, ZZ Wang and F Wang participated in design and conception of this study and performed the bioinformatics software analysis. G Zhang and K Chen downloaded the data form corresponding database and assisted bioinformatics analysis. All authors participated in writing the manuscript. MG Hu and R Liu proposed the concept and revised the manuscript. All authors read and approved the final manuscript.

Acknowledgments

The authors acknowledge TCGA and GEO databases for providing their platforms and the contributors for uploading their meaningful datasets.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.

Availability of data and materials

The datasets downloaded for supporting the results of this article are publicly available at The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/) and Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/).