Novel adenoma-immune phenotypes are associated with risk of metachronous polyps and colorectal cancer in a bowel screening cohort

This study demonstrates that the immune contexture of screen-detected adenomas, specifically characterized by T cell and macrophage density and spatial distribution, serves as a novel biomarker that independently predicts the risk of metachronous colorectal neoplasia, offering a potential tool to refine current surveillance guidelines and reduce misclassification of patient risk.

The Gist

The Big Picture: Why Do Some People Get New Polyps?

Imagine your colon is a garden. Every now and then, weeds (polyps) start growing. Doctors usually cut these weeds out (polypectomy) to stop them from turning into poisonous plants (cancer).

However, there's a problem with the current gardening rules. Based on how big the weed was or how ugly it looked, doctors guess who will get new weeds later. But they get it wrong about 40% of the time.

• Some people are told to come back for frequent check-ups, but they never get new weeds (wasting time and money).
• Other people are told they are "safe," but they grow dangerous weeds again (missing a chance to catch cancer early).

The Question: Is there a hidden clue inside the removed weed that tells us if the garden is likely to grow more weeds soon?

The Discovery: The "Garden Guard" Theory

The researchers looked at the immune system cells inside the removed polyps. Think of these immune cells as security guards patrolling the garden.

Usually, we think "more guards = better safety." But this study found something surprising and counter-intuitive:

• The Counter-Intuitive Finding: In the first round of testing, they found that polyps with lots of T-cells (a type of security guard) actually belonged to patients who were more likely to grow new polyps later.
• The Twist: It turns out it's not just about how many guards there are, but how they are standing and who they are talking to.

The Three "Garden Climates"

To solve the mystery, the team used high-tech microscopes to group the polyps into three distinct "neighborhoods" or clusters based on how the guards and other cells were arranged.

1. The "Adaptive" Neighborhood (Cluster 1) – The Low Risk Group

• The Scene: There are lots of T-cell guards, but the "bad guys" (macrophages, which can sometimes help tumors) are staying away. The guards are doing their job alone.
• The Result: These patients had the lowest risk of growing new polyps.
• The Analogy: Imagine a neighborhood where the police are patrolling effectively, and the criminal gangs are too scared to hang out nearby. The system is working.

2. The "Immune Cold" Neighborhood (Cluster 2) – The High Risk Group

• The Scene: The guards are nowhere to be found. The area is empty. It's "cold."
• The Result: These patients had the highest risk of growing new polyps.
• The Analogy: This is a ghost town with no police. If a weed tries to grow, there is no one to stop it. The garden is wide open for trouble.

3. The "Mixed" Neighborhood (Cluster 3) – The Medium Risk Group

• The Scene: There are lots of guards, but they are hanging out with the "bad guys" (macrophages) and construction workers (fibroblasts). They are all crowded together.
• The Result: These patients had a medium risk.
• The Analogy: Imagine the police are patrolling, but they are stuck in traffic jams with the criminals. They are too close together to do their job effectively. The criminals are influencing the police, or the police are confused.

The "Spatial" Secret: It's All About Distance

The most exciting part of the study used a technology called Spatial Transcriptomics. Think of this as a high-definition map that shows exactly where every single cell is standing in relation to its neighbors.

They discovered that in the High Risk group (Cluster 3), the "bad" macrophages were standing right next to the T-cell guards.

• The Metaphor: It's like a criminal whispering in a police officer's ear, convincing the officer to stand down or ignore the weeds. The immune system is present, but it's being hijacked or confused by the tumor environment.

In the Low Risk group (Cluster 1), the guards and the "bad guys" were far apart. The guards could do their job without interference.

What Does This Mean for You?

Currently, doctors look at the size and shape of a polyp to decide your future check-up schedule. This study suggests that in the future, they might look at the immune "personality" of the polyp.

• Better Predictions: Instead of guessing, they could say, "Your polyp had a 'Ghost Town' immune profile, so you need to be watched closely," or "Your polyp had a 'Clean Patrol' profile, so you can relax and wait longer."
• New Treatments: If we understand why the guards are confused (like in Cluster 3), we might be able to give patients a pill or a shot to "wake up" their immune system and stop new polyps from forming in the first place.

The Bottom Line

This research is like upgrading from a simple "weed size" rule to a sophisticated "security camera system" for your colon. By understanding the invisible immune landscape inside a polyp, we can finally predict who is truly at risk of getting colorectal cancer again, saving lives and stopping unnecessary hospital visits.

Note: This study is a preprint, meaning it is new research that hasn't been fully checked by other scientists yet, but the findings are very promising for the future of bowel cancer screening.

Technical Summary

1. Problem Statement

Current post-polypectomy surveillance guidelines (e.g., British Society of Gastroenterology) rely on pathological features such as polyp size, number, and dysplasia to stratify the risk of developing metachronous lesions (new polyps or colorectal cancer [CRC] after the initial removal). However, these guidelines are imprecise, misclassifying risk in approximately 40% of patients. This leads to unnecessary surveillance for low-risk individuals and false reassurance for high-risk individuals. There is a critical need for novel biomarkers that can more accurately predict the risk of metachronous neoplasia independent of standard pathological characteristics.

2. Methodology

The study utilized the INCISE cohort, a large, highly annotated dataset from the Scottish Bowel Screening Programme (NHS Greater Glasgow and Clyde, 2009–2016).

• Cohort: 2,642 patients who underwent polypectomy for conventional adenomas and had at least one surveillance colonoscopy between 6 months and 6 years later. Serrated lesions, IBD, and polyposis syndromes were excluded.
• Primary Outcome: Development of metachronous polyps or CRC during the surveillance period.
• Multi-Modal Analysis Approach:
  1. Chromogenic Immunohistochemistry (IHC): Performed on 2,466 whole-slide FFPE adenoma sections to quantify CD3+ T-cell density in the epithelium and lamina propria using Visiopharm software.
  2. Tissue Microarrays (TMAs): A sub-cohort (n=1,155) was used for TMA construction (4 cores per patient). IHC for CD3, CD4, and CD8 was performed and quantified using QuPath.
  3. Multiplex Immunofluorescence (mIF): A 6-plex panel (CD3, CD8, FOXP3, CD68, αSMA, Pan-CK) was applied to a sub-cohort (n=334) to define spatial immune cell densities. Hierarchical and K-means clustering were used to identify immune phenotypes.
  4. Bulk RNA Sequencing (RNAseq): TempO-Seq technology was used on FFPE samples to analyze gene expression and estimate immune/stromal cell abundance (MCPcounter).
  5. Spatial Transcriptomics: 6000-plex single-cell spatial transcriptomics (CosMx SMI) was performed on 7 adenoma cores to analyze cell-cell interactions and gene expression at single-cell resolution.
  6. Mutational Analysis: Agilent SureSelect CancerPlus panel was used for variant calling.
• Statistical Analysis: Survival analysis (Kaplan-Meier, Log-rank), multivariable logistic regression (adjusting for age, sex, site, size, dysplasia), and clustering algorithms (Ward's method, K-means).

3. Key Contributions

• Novel Biomarker Identification: The study identifies adenoma immune contexture (specifically T-cell and macrophage spatial distribution) as an independent predictor of metachronous lesion risk, surpassing traditional pathological risk factors.
• Discovery of Immune Phenotypes: Through mIF clustering, the authors defined three distinct immune phenotypes in adenomas:
  • Cluster 1 ("Adaptive"): High T-cell density, low macrophage/fibroblast density.
  • Cluster 2 ("Immune Cold/Excluded"): Low density of all immune and stromal markers.
  • Cluster 3 ("Mixed"): High density of T-cells, macrophages, and αSMA (fibroblasts).
• Spatial Mechanism Elucidation: Using spatial transcriptomics, the study revealed that the spatial co-localization of macrophages and T-cells is a critical differentiator. Cluster 3 showed significant short-range association between CD68+ macrophages and CD8+ T-cells, whereas Cluster 1 showed spatial separation.
• Validation Across Modalities: The findings were validated across multiple technologies (IHC, mIF, Bulk RNAseq, and Spatial Transcriptomics), confirming that immune exclusion and specific macrophage-T cell interactions drive future risk.

4. Key Results

• CD3+ Density Paradox: Contrary to the expectation that high T-cell infiltration implies better immune surveillance, high CD3+ T-cell density in both the epithelium and lamina propria was significantly associated with a higher risk of metachronous lesions (HR 1.43, p<0.001) after adjusting for all standard risk factors.
• Cluster Risk Stratification:
  • Cluster 1 (High T, Low Macrophage): Lowest risk of metachronous lesions (22%).
  • Cluster 2 (Immune Excluded): Highest risk (41%).
  • Cluster 3 (High T, High Macrophage, High αSMA): Intermediate-High risk (36%).
  • Statistical Significance: Differences in risk between clusters were significant (p=0.032).
• Genomic and Transcriptomic Correlates:
  • Mutations: Cluster 3 showed significantly higher rates of HLA-A and KRAS mutations compared to Cluster 1, suggesting an immune-suppressed environment.
  • Gene Expression: Cluster 2 (High Risk) showed upregulation of immunoglobulin genes and genes associated with poor prognosis (e.g., POU2AF1, DERL3).
  • Spatial Transcriptomics: Macrophages in Cluster 3 overexpressed genes associated with an immune-suppressive/pro-tumor microenvironment (e.g., APOC1, APOE, CTSD). CD4+ T-cells in Cluster 3 overexpressed antigen presentation genes (MHC II), suggesting a state of chronic antigen exposure or exhaustion.
• Independence: These immune signatures remained significant predictors of risk even when controlling for age, sex, lesion site, size, dysplasia, and histological subtype.

5. Significance and Implications

• Refining Surveillance: The study proposes that immune profiling could refine post-polypectomy surveillance guidelines, potentially reducing unnecessary colonoscopies for low-risk patients (Cluster 1) and intensifying surveillance for high-risk patients (Cluster 2 & 3) regardless of polyp size or dysplasia.
• Mechanistic Insight: The findings suggest that immune exclusion (lack of immune cells) and dysfunctional immune interactions (macrophages co-located with T-cells in a suppressive state) are fundamental drivers of field cancerization and future lesion development, not just the progression of the specific removed polyp.
• Clinical Translation: While current multi-omics technologies are not yet feasible for routine clinical use, the study highlights the potential to develop simplified IHC or H&E-based classifiers to identify these immune phenotypes in standard pathology workflows.
• Prevention Strategies: Understanding the "immune field effect" may open new avenues for chemoprevention strategies targeting the immune microenvironment of the colon to prevent metachronous lesions.

Limitations: The study is retrospective, single-center (predominantly Caucasian), and excluded serrated lesions. The spatial transcriptomics sample size was small (n=7), limiting statistical power for some cell-type comparisons. However, the multi-modal validation strengthens the robustness of the primary findings.