Skip to main content


Springer Nature is making SARS-CoV-2 and COVID-19 research free. View research | View latest news | Sign up for updates

Table 1 Immune signatures and their development

From: Immune approaches to the treatment of breast cancer, around the corner?

Immune signature Signature development
Immune response (IR) module [23] A subclass of estrogen receptor-negative (ER) tumors that overexpress IR genes and that have a good prognosis compared with the rest of ER breast tumors independently of lymph node status or lymphocytic infiltration was identified. Subsequently, an associated module of complement and IR genes that define prognostic markers was identified and validated in over 240 ER samples.
STAT1 module [22] On the basis of the literature, genes to act as ‘prototypes’ for different biological processes - ER for ER signaling, HER2 for HER2 signaling, AURKA for proliferation, CASP3 for apoptosis, VEGF for angiogenesis, PLAU for tumor invasion/metastasis, and, in this case, STAT1 for immune response - were selected. A comparison of linear models was then applied to generate modules of genes specifically associated with each of the prototype genes but not with the other prototypes.
B-cell metagene [7] Gene expression patterns of 200 patients who did not receive systemic treatment and co-regulated genes related to proliferation, steroid hormone receptor expression, and B-cell and T-cell infiltration were identified after hierarchical cluster analysis was performed. Metagenes were calculated as a surrogate for all genes contained within a particular cluster and their expression was correlated to time to metastasis. The B-cell metagene showed independent prognostic information in carcinoma with high proliferative activity.
IgG, HCK, MHC-I, MHC-II, LCK, STAT1, and IFN metagenes [24] Unsupervised hierarchical clustering of genes in 12 primary invasive breast cancer datasets as well as combined datasets revealed a large cluster of genes with functions in immune cells. Among this cluster, clusters that contained a minimum number of elements and a minimal average correlation were selected, and seven metagenes were derived. Each metagene then was associated with a cell type or immunological state or both.
HRneg/Tneg signature [25] A cohort of patients with node-negative, adjuvant treatment-naïve hormone receptor-negative (HRneg), and triple-negative (Tneg) breast cancer has been used to define and validate genes predictive for distant metastatic relapse. A composite HRneg/Tneg signature index was able to identify cases likely to remain free of metastatic relapse with high accuracy. Of note, significant positive correlation was observed between the HRneg/Tneg index and three independent immune-related signatures (STAT1, IFN, and IR), and network analysis showed that the signature was linked to immune/inflammatory cytokine regulation.
Support Vector Machine (SVM) classifier [26] Gene expression data of 2,145 invasive early breast adenocarcinomas were collected and used to test and validate the predictive performance of an SVM classifier based on a 368-gene expression signature associated with medullary breast carcinoma (MBC), which displays a basal profile but has good prognosis. The SVM model accurately classified all MBC samples in the learning and validation sets and was able to separate 466 cases of basal breast cancers into two subgroups (subgroup 1 and subgroup 2) containing, respectively, good- and poor-prognosis tumors. Ontology analysis revealed, among other features, effective IR in the good-prognosis subgroup.
  1. AURKA, aurora kinase A; CASP3, caspase 3; HCK, hemopoietic cell kinase; IFN, interferon; IgG, immunoglobulin G; LCK, lymphocyte-specific kinase; MHC, major histocompatibility complex; PLAU, Urokinase-type plasminogen activator; STAT1, signal transducer and activator of transcription 1; VEGF, vascular endothelial growth factor.