Immunological Checkpoints Blockade Reaction Predictor in Gastrointestinal Leukemia

Pike Don^*
*Correspondence: Pike Don, Institute of Translational Medicine, Semmelweis University, 1094 Budapest, Hungary, Email:

INTRODUCTION

Immunotherapy is a revolutionary method for suppressing tumours that uses the immune system, and it has improved the quality of life of cancer patients by reducing the side effects of anti-cancer treatment. A more comprehensive view of a cancer patient results in fundamental changes in the assessment of therapeutic efficacy and toxicity. The high cost of drugs, however, and their ineffectiveness in some patients, are significant drawbacks. Many studies were conducted to find markers predicting immune checkpoint blockade (ICB) response in order to overcome these limitations. Although T lymphocytes have long been recognised for their role in cancer immunosurveillance, the discovery that cancer cells can eventually evade a response from tumor-reactive T cells has sparked efforts to improve anti-tumor immune response effectiveness. Cytotoxic T lymphocyte antigen 4 and programmed cell death are the most effective T cell immunological checkpoint molecules. Tumor-associated macrophages and fibroblasts play an important role in immunosuppression in the inflammatory tumour microenvironment. Among the new medications being studied for gastric cancer treatment are margetuximab, ZW25, and combination strategies involving chemotherapy, human epidermal growth factor receptor 2-targeted treatments, and programmed cell death protein 1. Recently, multi-omics data were used to predict drug responses using machine learning, and gastric cancer data were used to predict ICB responses using a patient stratification algorithm. Using multi-omics data, we previously predicted intra-tumoral heterogeneity using machine learning algorithms.

Description

Despite the discovery of several markers, there are no official standards for predicting immune checkpoint blockade (ICB) responses. The main disadvantages of immunotherapy are expensive drugs and different reactivities for each patient. Gastric cancer is resistant to immunotherapy because it is stem-like in nature. We wanted to find a gene that predicts ICB response in gastric cancer and find a drug target for non-responders in this study [1]. To predict ICB response in gastric cancer patients, we built and tested a model using four machine learning algorithms on two cohorts of bulk and single-cell RNA seq data. Recently, multi-omics data were used to predict drug responses using machine learning, and gastric cancer data were used to predict ICB responses using a patient stratification algorithm. Using multi-omics data, we previously predicted intra-tumoral heterogeneity using machine learning algorithms. Machine learning algorithms offer a novel approach to integrating and analysing omics data, allowing for the discovery of new biomarkers [2].

Deep learning algorithms have recently emerged as one of the most promising approaches in multi-omics data analysis due to their predictive performance and ability to capture nonlinear and hierarchical characteristics [3]. The signature genes have the advantage of predicting the ICB response of patients with gastric cancer and providing researchers with general insights into the mechanism. Using bulk and single-cell expression data, we identified a signature gene that distinguishes between ICB responses for each algorithm [4]. The signature genes from the bulk sample, in particular, were linked to tumour necrosis factor and naba core matrix. The high expression of these genes was associated with a poor prognosis. Signature genes were chosen for each molecular subtype. High expression of the signature gene was associated with poor prognosis in stem-like and mixed stroma types, but this did not differ significantly between the groups [5].

Conclusion

Our study predicted ICB response based on single-cell and bulk types. Tumor cells performed no better than other cell types in terms of ICB response, and other cell types were more suitable for the predictive model. The random forest model of proliferating cells performed best among single-cell types, while endothelial cells performed poorly; however, the relevance to the ICB response prediction model was weak. VCAN functions as a mesenchymal stem cell marker, with significantly higher expression in the non-responder group than in the responder group. In summary, using gastric cancer bulk and single-cell expression data, we identified a signature that can predict ICB response using a basic machine learning algorithm. This will allow for more effective precision medicine treatment of patients.

References

1. Barbalinardo, Marianna, Federico Caicci, Massimiliano Cavallini and Denis Gentili. “Protein Corona Mediated Uptake and Cytotoxicity of Silver Nanoparticles in Mouse Embryonic Fibroblast.” Small 14 (2018): 1801219.

Google Scholar, Crossref, Indexed at

2. Prelaj, Arsela, Rebecca Tay, Roberto Ferrara and Nathalie Chaput, et al. “Predictive biomarkers of response for immune checkpoint inhibitors in non-small-cell lung cancer.” Eur J Cancer 106 (2019): 144-159.

Google Scholar, Crossref, Indexed at

3. Philip, Daizy, C. Unni, S. Aswathy Aromal and V. K. Vidhu. “Murraya Koenigii leaf-assisted rapid green synthesis of silver and gold nanoparticles. Spectrochim.” Acta Part A Mol Biomol Spectrosc 78 (2011): 899-904.

Google Scholar, Crossref

4. Wang, Chao, Yeon Ju Kim, Priyanka Singh and Ramya Mathiyalagan, et al. “Green synthesis of silver nanoparticles by Bacillus methylotrophicus, and their antimicrobial activity.” Artif Cells Nanomed Biotechnol 44 (2016): 1127-1132.

Google Scholar, Crossref, Indexed at

5. Zhang, Zhibin, Ying Zhang, Shiyu Xia and Qing Kong, et al. “Gasdermin E suppresses tumour growth by activating anti-tumour immunity.” Nature 579 (2020): 415-420.

Google Scholar, Crossref, Indexed at