Abstract
BACKGROUND: Gut microbiome plays an indispensable role in the occurrence and progression in various diseases. The incidence of pancreatic cancer (PC) and liver metastasis (PCLM) are high, most of them are found in advanced stage. Therefore, it is particularly necessary to search for predictive biomarkers, which are helpful for early detection and treatment, and thus improve the survival rate and quality of life of PC patients. METHODS: We retrospectively analyzed 44 pancreatic cancer patients (P group, n = 44) and 50 healthy people (N group, n = 50) from March 21, 2021 and August 2, 2022. Among all PC patients, we divided them into liver metastasis group (LM group, n = 27) and non-liver metastasis group (non-LM group, n = 17). DNA was extracted and 16S ribosomal RNA (16S rRNA) gene sequencing was performed. SPSS was used for statistical analyses and all bioinformatics analyses were based on QIIME2, p < 0.05 were considered statistically significant. RESULTS: The microbial richness and diversity of group P and LM were higher than that of group N and non-LM. LEfSe analysis found that Streptococcus was a significantly different microorganism, which was further identified by random forest (RF) model, and its ability to predict PC and PCLM was verified by ROC curve. CONCLUSION: We demonstrated significant differences in intestinal microbiome composition between PC patients and healthy people, and found that Streptococcus is a potential biomarker for early prediction of PC and PCLM, which is critical for early diagnosis of diseases.