Table_3_Gut Microbiota-Based Algorithms in the Prediction of Metachronous Adenoma in Colorectal Cancer Patients Following Surgery.docx
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Evaluating the risk of colorectal metachronous adenoma (MA), which is a precancerous lesion, is necessary for metachronous colorectal cancer (CRC) precaution among CRC patients who had underwent surgical removal of their primary tumor. Here, discovery cohort (n = 41) and validation cohort (n = 45) of CRC patients were prospectively enrolled in this study. Mucosal and fecal samples were used for gut microbiota analysis by sequencing the 16S rRNA genes. Significant reduction of microbial diversity was noted in MA (P < 0.001). A signature defined by decreased abundance of eight genera and increased abundance of two genera strongly correlated with MA. The microbiota-based random forest (RF) model, established utilizing Escherichia–Shigella, Acinetobacter together with BMI in combination, achieved AUC values of 0.885 and 0.832 for MA, predicting in discovery and validation cohort, respectively. The RF model was performed as well for fecal and tumor adjacent mucosal samples with an AUC of 0.835 and 0.889, respectively. Gut microbiota profile of MA still existed in post-operative cohort patients, but the RF model could not be performed well on this cohort, with an AUC of 0.61. Finally, we introduced a risk score based on Escherichia–Shigella, Acinetobacter and BMI, and synchronous-adenoma achieved AUC values of 0.94 and 0.835 in discovery and validation cohort, respectively. This study presented a comprehensive landscape of gut microbiota in MA, demonstrated that the gut microbiota-based models and scoring system achieved good ability to predict the risk for developing MA after surgical resection. Our study suggests that gut microbiota is a potential predictive biomarker for MA.
Read the peer-reviewed publication