IBD: Bacterial composition and function predict severity of single intestinal nutrition treatment for pediatric Crohn's disease

Backgroundthe intestinal microbiome is widely involved in the mitigation of pediatric Crohn's disease (CD) patients by affecting the efficacy of single intestinal nutrition (EEN)In this follow-up study of pediatric CD patients treated with EEN, the researchers used machine learning models trained on baseline gut microbiome data to distinguish between patients who achieved and sustained remission (SR) and those who did not achieve remission (non-SR) or relapsedmethod sed
methods The researchers took a total of 139 fecal samples from 22 patients (ages 8 to 15) and followed up to 96 weeksThe classification of the gut microbiome is evaluated through 16S rRNA gene sequencing, and the function of the flora is evaluated through macro genome sequencingUse standard indicators of diversity and taxonomics to quantify the differences between SR and non-SR patients and cortication with changes in gut microbes with fecal calcium protein (FCP) and the severity of the diseaseIn addition to the clinical metadata in the Random Forest (RF) model, the researchers used microbial data sets to classify therapeutic responses and predict FCP levelsresults
There was no change in microbiodiversity after eEN use, but the low FCP sample (250 ?g / g) had lower species richnessUsing microbial abundance, species abundance and the RF model of the Paris Classification of Disease, the RF model is the most able to classify therapeutic responses (area under the curve (AUC) s 0.9)KEGG Pathways also significantly classified treatment responses by adding the same clinical data (AUC - 0.8)The main features of the RF model are consistent with previously identified IBD clusters, such as ruminococcaceae and Ruminococcus gnavusconclusions
This study shows that different baseline microbiomes and clinical data can be used to distinguish between clinical remission in children receiving single intestinal nutrition therapy