Back

The infiltration pattern of microenvironmental cells and different immune escape mechanisms in colorectal cancer

Yihao, M.; Feng, Q.; Chang, W.; Lv, Y.; Xu, Y.; Chang, J.; Zheng, P.; Yu, S.; Zhang, Z.; Li, Z.; Lin, Q.; Tang, W.; Zhu, D.; Ji, M.; Ren, L.; Wei, Y.; He, G.; Xu, J.

2022-02-19 cancer biology
10.1101/2022.02.16.480691 bioRxiv
Show abstract

BackgroundThe tumour microenvironment (TME) plays a crucial role in tumour progression and metastasis. However, the infiltration pattern of TME cells in CRC samples and the immune escape mechanism underneath have not been studied sufficiently. MethodsTranscriptomic data from public datasets were retrieved online. In total, 1802 samples from the microarray dataset and 619 samples from the TCGA dataset were enrolled. The ssGSEA algorithm and unsupervised clustering were used for TME cells infiltration speculation and infiltration pattern recognition. ResultsCRC samples can be classified into three distinct TME cell subtypes. Subtype 1, the immune-active subtype, was characterised by high infiltration of activated adaptive immune cells. Subtype 2, the immune-desert subtype, featured high tumour purity and low infiltration of immune and stromal cells. And subtype 3, stroma-rich subtype, had high infiltration of stromal cells. The stroma-rich subtype conferred a significantly worse prognosis. Three subtypes had different immune escape mechanisms. The immune-active subtype has the highest immune checkpoint expression level. In comparison, the immune-desert subtype had the lowest immunogenicity and the defective antigen presentation. And the stroma-rich subtype lacked activated immune cells. ConclusionsDistinct TME cell subtypes and immune escape mechanisms may provide inspiration and direction for further researches on CRC immunotherapy.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.