Session Item

To analyze the number and structural changes of the bacteria in the occurrence and development of radiation enteritis (RE) and severe acute radiation enteritis (SARE), and grasp the microbiome characterization and biomarkers of RE/SARE associated with VMAT for cervical cancer.

50 patients who received cervical cancer radiotherapy in our department and 15 healthy control women were enrolled. The stool samples of patients at 4 time points during radiotherapy (baseline period of radiotherapy (T0), 2 weeks after the start of radiotherapy (T2), 4 weeks after the start of radiotherapy (T4), the end of radiotherapy i.e., 6 weeks after the start of radiotherapy (Tf)), and at the baseline in the control group were collected. The fecal supernatant samples were analyzed by 16S rDNA amplicon sequencing, and the bioinformatics analysis process was used to clarify the change process and functional prediction of bacterial community structure and diversity. Microbial taxonomy, the ratio of Bacteroidetes to Firmicutes (B/F) and enterotype were used to perform cluster analysis and to extract valid biomarkers of RE.

1. The top 10 species in abundance included Bacteroidetes, Blautia, Escherichia-Shigella, Faecalibacterium, Ruminococcus gnavus, Lachnoclostridium, Unclassified, Roseburia, Hortella, Megamonas. The ROC analyses were performed on the genus, and most of the AUCs were greater than 0.5, but all of them lacked statistical significance.

2. The ratio of B/F in cervical cancer VMAT patients gradually decreased with dose/time, but there was no significant difference between groups. The B/F ratio of RE and SARE were higher than those of non-RE and non-SARE, especially in the second week of radiotherapy and at the end of radiotherapy, which were increased significantly. However, the B/F ratio could not be used as a predictor of RE or SARE alone.

3. Blautia was defined as enterotype-1, Escherichia.Shigella as enterotype-2, and Faecalibacterium as enterotype-3 through the JSD cluster scoring method. There were significant differences in the enterotype between cervical cancer patients and healthy control women, including the decrease of Blautia, the increase of Escherichia.Shigella and Faecalibacterium which related to the occurrence of cervical cancer. The enterotype were changed during VMAT. Through the univariate Logistic regression analysis, it was found that Faecalibacterium enterotype increased the incidence of SARE at baseline (OR: 5.400, P=0.017), which increased the risk of SARE by 4.4 times compared with Blautia enterotype. It was indicated that Faecalibacterium enterotype could predict the occurrence of SARE. The pathogen Escherichia.Shigella was associated with the increased incidence of SARE during VMAT.

The single characteristic genus or the B/F ratio could not effectively predict RE or SARE. However, the enterotype classification could be the predictor of RE and SARE.