{
  "abstract": "Background Gut microbial dysbiosis plays a pivotal role in the initiation and progression of colorectal cancer (CRC). Clinical staging remains the cornerstone for prognostic assessment and therapeutic decision-making in CRC patients. However, the dynamic interplay between intratumoral microbiota and host genes across different clinical stages and its causal role in tumor advancement remains largely elusive. This study aims to characterize the stage-specific features of intratumoral microbiota–host transcriptome interactions and identify robust biomarkers for precise CRC staging.Methods We collected 139 samples, including paired tumor tissues and adjacent normal mucosa, from 31 CRC patients through multi-point sampling, stratifying them into early-stage (I/II) and advanced-stage (III/IV) cohorts. Intratumoral microbiota composition and host gene expression were profiled using 16S ribosomal RNA gene sequencing and mRNA transcriptome sequencing. Integrated bioinformatics analyses, including microbial diversity and network analysis, differential abundance analysis, Procrustes analysis, regularized canonical correlation analysis, and machine learning modeling, were employed. Two-sample Mendelian randomization validated causal relationships between key microbial taxa and CRC progression. External validation was conducted in two independent public CRC cohorts.Results We identified disease stage and tumor size as primary phenotypic factors shaping intratumoral microbial community composition. Advanced-stage CRC exhibited significantly higher microbial α-diversity, a more complex co-occurrence network, and distinct functional gene expression signatures. A robust global correlation between intratumoral microbiota and host gene expression was revealed, with more coordinated interactions in early-stage CRC. We identified ten high-confidence microbiota–gene interaction pairs at the genus level and six at the species level. Notably, the oral-derived pathobiont Alloprevotella emerged as a key effector driving CRC progression, showing the strongest negative correlation with tumor suppressor gene MDH1. Mendelian randomization substantiated a causal effect of Alloprevotella enrichment on CRC progression. Furthermore, a multi-omics classifier integrating microbial and host gene signatures achieved an area under the receiver operating characteristic curve exceeding 0.9 for distinguishing CRC stages, demonstrating robust performance in independent validation cohorts.Conclusions Our study unveils the stage-specific remodeling of the intratumoral microbiota and its dynamic interplay with host genes during CRC progression. We identify a novel causal Alloprevotella-MDH1 axis driving malignant advancement and provide a high-performance multi-omics biomarker panel for precise CRC staging and risk stratification.",
  "authors": [
    {
      "affiliations": [
        "Department of Gastroenterology, Second Affiliated Hospital of Fujian Medical University, China"
      ],
      "name": "Yiqun Hong"
    },
    {
      "affiliations": [
        "Department of Gastroenterology, Xiang’an Hospital of Xiamen University, China"
      ],
      "name": "Zhiyuan Chen"
    },
    {
      "affiliations": [
        "Department of Gastroenterology, Xiang’an Hospital of Xiamen University, China"
      ],
      "name": "Wenxuan Gan"
    },
    {
      "affiliations": [
        "Department of Gastroenterology, Second Affiliated Hospital of Fujian Medical University, China"
      ],
      "name": "Taiyong Fang"
    },
    {
      "affiliations": [
        "Department of Gastroenterology, Second Affiliated Hospital of Fujian Medical University, China"
      ],
      "name": "Yifan Lian"
    }
  ],
  "title": "IDDF2026-ABS-0206 Stage-specific intratumoral microbiota-host gene interaction in colorectal cancer and its implications for staging prediction",
  "uid": "05d0efbe-9274-5a44-9b1b-20ff30f0d75a"
}