Journal of Experimental & Clinical Cancer Research, Journal Year: 2018, Volume and Issue: 37(1)

Published: Dec. 1, 2018

Abundant evidence shows that triple-negative breast cancer (TNBC) is heterogeneous, and many efforts have been devoted to identifying TNBC subtypes on the basis of genomic profiling. However, few studies explored classification specifically based immune signatures may facilitate optimal stratification patients responsive immunotherapy.Using four publicly available genomics datasets, we classified immunogenomic profiling 29 signatures. Unsupervised supervised machine learning methods were used perform classification.We identified three named Immunity High (Immunity_H), Medium (Immunity_M), Low (Immunity_L) demonstrated this was reliable predictable by analyzing multiple different datasets. Immunity_H characterized greater cell infiltration anti-tumor activities, as well better survival prognosis compared other subtypes. Besides signatures, some cancer-associated pathways hyperactivated in Immunity_H, including apoptosis, calcium signaling, MAPK PI3K-Akt RAS signaling. In contrast, Immunity_L presented depressed increased activation cycle, Hippo DNA replication, mismatch repair, adhesion molecule binding, spliceosome, adherens junction function, pyrimidine metabolism, glycosylphosphatidylinositol (GPI)-anchor biosynthesis, RNA polymerase pathways. Furthermore, a gene co-expression subnetwork centered around five transcription factor (TF) genes (CORO1A, STAT4, BCL11B, ZNF831, EOMES) significant subtype two TF (IRF8 SPI1) characteristic subtype.The identification has potential clinical implications for treatment.

Language: Английский