IDDF2023-ABS-0250 Comprehensive Analyses of Microbiome and Metabolism Reveal Liver Metastasis-Specific Signature in Colorectal Cancer

Xiaohong Wang,Jingwan Zhang,Xiao‐Xu Hu,Biao Fan,Chi-Chun Wong,Haoqiang Ying,Lianhai Zhang,Ziyu Li,Jiafu Ji
DOI: https://doi.org/10.1136/gutjnl-2023-iddf.102
2023-01-01
Abstract:Background Accumulating experimental evidence suggests that the microbiome plays an important role in cancer metastasis; however, microbial features of human colorectal cancer (CRC) metastases remain largely unknown. Here, we performed an integrated metagenomic and metabolomic analysis of primary CRC and its associated liver metastases. Methods CRC patients (N=98) who underwent the resection of primary lesions and developed liver metastases were included. Metagenomic and metabolomic analysis was performed on paired CRC tumors, paracancerous normal colon (adj-CRC), liver metastases (LM), and paracancerous normal liver (adj-LM) to profile alteration in the gut microbiome and metabolites by DESeq2. Results Distinct microbiome profiles are observed in paired CRC tumor, adj-CRC, paired LM and adj-LM tissues. Colonic microbiota was dominated by the phyla Firmicutes and Bacteroidetes, whereas Proteobacteria was most abundant in the liver. LM and adj-LM were further separated into two distinct clusters, with significant enrichments of Bacteroides and Fusobacteria in LM compared to adj-LM (P<0.05). Within the Taxonomy >20% prevalence in CRC, the dysregulated proportion of Firmicutes, Actinobacteria and Fusobacteria was found in LM compared to adj-LM. In terms of metabolomic profiling, amino acids metabolism and energy production were dysregulated in LM. Integrative analyses revealed that many microbial-metabolite connections showed a similar trend among CRC and LM tissues. Notably, the abundance of Ruminococcus, Christensenellaceae, Fretibacterium, Ruminococcaceae, and Intestinimonas was significantly higher in CRC patients with post-operative recurrence compared to the patients with no recurrence. These microbes were also associated with patient poor survival. Besides, we identified 12-hydroxy-12-octadecanoylcarnitine and 3-hydroxy-9Z-octadecenoyl-carnitine as metabolites that were concomitantly enriched in both primary CRC and recurrent LM. These microbial and metabolite biomarkers, respectively, predict patient survival and recurrence of LM in CRC. Conclusions We identified the microbiome and metabolic signatures in primary CRC that are associated with LM, which may play an important role in driving liver metastases.
What problem does this paper attempt to address?