Immune-Based Stratification of Breast Cancer Reveals Biologically Distinct Tumor Microenvironment Phenotypes with Prognostic Implications

Background: Conventional breast cancer molecular classification has improved therapeutic stratification but remains insufficient to capture tumor microenvironment (TME) heterogeneity. Even within aggressive subgroups, such as triple-negative breast cancer (TNBC), substantial immune variability persists without a robust microenvironmental stratification framework. We aimed to define reproducible, immune-based TME phenotypes using transcriptomic profiling and evaluate their biological, prognostic, and immunological relevance. Methods: Using bulk transcriptomic and clinical data from 1980 primary breast tumors in the METABRIC discovery cohort, relative abundances of 22 immune cell populations were estimated via CIBERSORTx, followed by silhouette-guided K-means clustering. Microenvironmental composition and tumor purity were quantified using the ESTIMATE algorithm, while differential gene expression (DGE) and gene set enrichment analysis (GSEA) characterized underlying transcriptional programs. Prognostic independence was evaluated using Kaplan–Meier and multivariable Cox proportional hazards models, and immune-evasion mechanics alongside predicted checkpoint blockade sensitivity were estimated using the Tumor Immune Dysfunction and Exclusion (TIDE) framework. Phenotypic reproducibility was evaluated via nearest-centroid projection in an independent TCGA-BRCA validation cohort (n = 1070). Results: Three distinct immune-based TME phenotypes (Clusters 1–3) were identified across conventional PAM50 molecular subtypes. Cluster 1 (immune-excluded-like) exhibited the highest tumor purity, reduced immune scores, enrichment of extracellular matrix and epithelial programs, elevated T cell exclusion scores, and the poorest overall survival, remaining an independent predictor of mortality after adjusting for age, histologic grade, and molecular subtype (multivariable HR = 1.21, 95% CI: 1.04–1.40, p = 0.013). Cluster 2 (immune-inflamed-like) displayed elevated immune and ESTIMATE scores, enrichment of interferon-γ, antigen presentation, and cytotoxic T cell signaling, and the lowest composite TIDE scores (adjusted p ≤ 0.0001 vs. Clusters 1 and 3), driven by elevated T cell dysfunction rather than exclusion. Cluster 3 (stromal-rich/immune-modulated) demonstrated the highest stromal scores alongside estrogen-responsive, adipogenic, and TGF-β-associated signatures, and shared elevated T cell exclusion with Cluster 1. Nearest-centroid projection in TCGA-BRCA reproduced these cellular and transcriptional profiles. Conclusions: Immune-based clustering uncovers a reproducible dimension of breast cancer biology that cuts across intrinsic molecular subtypes and provides independent prognostic stratification in the discovery cohort. While these transcriptomic and TIDE profiles generate the hypothesis that immune-inflamed tumors may be more amenable to immune checkpoint blockade whereas excluded or stromal-rich phenotypes may require microenvironment-modulating strategies, prospective clinical validation is required before clinical implementation.

Authors

Institutions

Publication Details

Journal
Cancers
Published
2026-11-19
DOI
https://doi.org/10.3390/cancers18223699
Primary Topic
Breast Cancer Treatment Studies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Immune-Based Stratification of Breast Cancer Reveals Biologically Distinct Tumor Microenvironment Phenotypes with Prognostic Implications

Abdallah Badou, Khalil Choukri, Hamza Baddi, Salma Jannane et al.
Cancers
Breast Cancer Treatment Studies
article

Immune-Based Stratification of Breast Cancer Reveals Biologically Distinct Tumor Microenvironment Phenotypes with Prognostic Implications

Abdallah Badou, Khalil Choukri, Hamza Baddi, Salma Jannane, Ilham Zanzoum, Dounia Souaf
article en

Abstract

Background: Conventional breast cancer molecular classification has improved therapeutic stratification but remains insufficient to capture tumor microenvironment (TME) heterogeneity. Even within aggressive subgroups, such as triple-negative breast cancer (TNBC), substantial immune variability persists without a robust microenvironmental stratification framework. We aimed to define reproducible, immune-based TME phenotypes using transcriptomic profiling and evaluate their biological, prognostic, and immunological relevance. Methods: Using bulk transcriptomic and clinical data from 1980 primary breast tumors in the METABRIC discovery cohort, relative abundances of 22 immune cell populations were estimated via CIBERSORTx, followed by silhouette-guided K-means clustering. Microenvironmental composition and tumor purity were quantified using the ESTIMATE algorithm, while differential gene expression (DGE) and gene set enrichment analysis (GSEA) characterized underlying transcriptional programs. Prognostic independence was evaluated using Kaplan–Meier and multivariable Cox proportional hazards models, and immune-evasion mechanics alongside predicted checkpoint blockade sensitivity were estimated using the Tumor Immune Dysfunction and Exclusion (TIDE) framework. Phenotypic reproducibility was evaluated via nearest-centroid projection in an independent TCGA-BRCA validation cohort (n = 1070). Results: Three distinct immune-based TME phenotypes (Clusters 1–3) were identified across conventional PAM50 molecular subtypes. Cluster 1 (immune-excluded-like) exhibited the highest tumor purity, reduced immune scores, enrichment of extracellular matrix and epithelial programs, elevated T cell exclusion scores, and the poorest overall survival, remaining an independent predictor of mortality after adjusting for age, histologic grade, and molecular subtype (multivariable HR = 1.21, 95% CI: 1.04–1.40, p = 0.013). Cluster 2 (immune-inflamed-like) displayed elevated immune and ESTIMATE scores, enrichment of interferon-γ, antigen presentation, and cytotoxic T cell signaling, and the lowest composite TIDE scores (adjusted p ≤ 0.0001 vs. Clusters 1 and 3), driven by elevated T cell dysfunction rather than exclusion. Cluster 3 (stromal-rich/immune-modulated) demonstrated the highest stromal scores alongside estrogen-responsive, adipogenic, and TGF-β-associated signatures, and shared elevated T cell exclusion with Cluster 1. Nearest-centroid projection in TCGA-BRCA reproduced these cellular and transcriptional profiles. Conclusions: Immune-based clustering uncovers a reproducible dimension of breast cancer biology that cuts across intrinsic molecular subtypes and provides independent prognostic stratification in the discovery cohort. While these transcriptomic and TIDE profiles generate the hypothesis that immune-inflamed tumors may be more amenable to immune checkpoint blockade whereas excluded or stromal-rich phenotypes may require microenvironment-modulating strategies, prospective clinical validation is required before clinical implementation.

CancersVol. 18(22)
University of Hassan II Casablanca (MA)
Breast Cancer Treatment Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon