An integrated tumor, immune and microbiome atlas of colon cancer

PMID: 37202560 · DOI: 10.1038/s41591-023-02324-5 · Journal: Nature Medicine (2023)

TL;DR

Roelands et al. built AC-ICAM (Sidra-LUMC Atlas and Compass of Immune-Cancer-Microbiome interactions), a multi-omics resource profiling 348 treatment-naive primary colon cancer patients with RNA-seq, whole-exome sequencing, TCR sequencing, 16S rRNA gene sequencing, and tumor whole-genome sequencing on paired tumor/normal colon tissue. They show that the 20-gene Immunologic Constant of Rejection (ICR) signature outperforms consensus molecular subtype (CMS) and MSI classification for prognosis, that quantifying genetic immunoediting further refines its prognostic value, and that a Ruminococcus bromii-driven microbiome signature combined with ICR (mICRoScore) identifies a subgroup with excellent survival (PMID:37202560).

Cohort & data

• 348 treatment-naive patients with primary colon adenocarcinoma from the Sidra-LUMC AC-ICAM cohort; fresh-frozen tumor and matched healthy colon tissue (PMID:37202560).
• Platforms: RNA-seq, whole-exome sequencing, deep TCRβ sequencing (immunoSEQ), 16S rRNA gene sequencing, and tumor whole-genome sequencing for microbiome characterization (PMID:37202560).
• An additional 42 tumor-only samples (ICAM42) were profiled with 16S rRNA sequencing (PMID:37202560).
• Median follow-up 4.6 years (PMID:37202560).
• TCGA-COAD (coadread_tcga) used as a comparator; limited survival follow-up and tumor-purity selection criteria motivated the new cohort (PMID:37202560).

Key findings

• Consensus clustering on the 20-gene ICR signature segregated AC-ICAM into ICR-high (hot), ICR-medium, and ICR-low (cold) subtypes (PMID:37202560).
• ICR was prognostic for overall survival: ICR-high vs ICR-low HR = 0.54 (95% CI 0.34–0.86, P = 0.0095); ICR-medium vs ICR-low HR = 0.63 (0.43–0.91, P = 0.014) (PMID:37202560).
• ICR outperformed CMS and MSI classifications for prognosis and retained prognostic value within the otherwise heterogeneous CMS4/mesenchymal subtype (PMID:37202560).
• NK and T cell subset abundances (ConsensusTME) were highest in ICR-high tumors across all CMS subtypes (PMID:37202560).
• Quantifying genetic immunoediting (fewer observed neoantigens than expected) refined ICR’s prognostic value (PMID:37202560).
• A microbiome signature driven by Ruminococcus bromii was associated with favorable outcome (PMID:37202560).
• The composite mICRoScore (ICR + microbiome signature) identified a subgroup with excellent survival probability and was validated within the resource (PMID:37202560).

Genes & alterations

• The 20-gene ICR panel includes T-helper-1 signaling genes (IFNG, TBX21, CD8A, CD8B, IL12B, STAT1, IRF1), CXCR3/CCR5 chemokine ligands (CCL5, CXCL9, CXCL10), effector cytotoxic genes (GNLY, PRF1, GZMA, GZMB, GZMH), and immunoregulatory counter-activation genes (CD274, CTLA4, FOXP3, IDO1, PDCD1) (PMID:37202560).
• No single driver gene is proposed as a biomarker; the paper’s gene-level claims are about the ICR module as a whole (PMID:37202560).

Clinical implications

• ICR-based immune classification is a stronger prognostic biomarker than CMS or MSI status in primary colon cancer and could be incorporated into risk stratification (PMID:37202560).
• Combining tumor immune status with gut microbiome features (mICRoScore) can identify very-good-prognosis patients, suggesting a rationale for microbiome-aware prognostic models (PMID:37202560).
• ICR also correlates with immunotherapy response in other tumor types referenced by the authors, suggesting potential predictive utility that this paper does not directly test in colon cancer (PMID:37202560).

Limitations & open questions

• AC-ICAM is a single multi-center European cohort of treatment-naive primary colon cancers; generalization to rectal cancer, metastatic disease, or diverse populations is not established (PMID:37202560).
• Microbiome composition is sensitive to sample handling, diet, and geography; the R. bromii signal requires external validation across cohorts (PMID:37202560).
• The paper validates mICRoScore internally; prospective validation and standardized assays for a fresh-frozen, multi-omics workflow would be required for clinical deployment (PMID:37202560).
• Immunoediting quantification depends on neoantigen prediction pipelines, which have known noise (PMID:37202560).

Citations from this paper used in the wiki

• “we performed comprehensive genomic analyses on fresh-frozen samples from 348 patients affected by primary colon cancer, encompassing RNA, whole-exome, deep T cell receptor and 16S bacterial rRNA gene sequencing on tumor and matched healthy colon tissue, complemented with tumor whole-genome sequencing” (Abstract).
• “A type 1 helper T cell, cytotoxic, gene expression signature, called Immunologic Constant of Rejection, captured the presence of clonally expanded, tumor-enriched T cell clones and outperformed conventional prognostic molecular biomarkers, such as the consensus molecular subtype and the microsatellite instability classifications” (Abstract).
• “We identified a microbiome signature, driven by Ruminococcus bromii, associated with a favorable outcome. By combining microbiome signature and Immunologic Constant of Rejection, we developed and validated a composite score (mICRoScore), which identifies a group of patients with excellent survival probability” (Abstract).
• “ICR high versus low, HR = 0.54 (0.34–0.86), P = 0.0095; ICR med versus low, HR = 0.63 (0.43–0.91), P = 0.014” (Fig. 1d).

This page was processed by crosslinker on 2026-04-08.