Predicting Risk of Post-Treatment Relapse in Patients with Inflammatory Bowel Disease Based on Intestinal Microbiota.

Background: Inflammatory bowel disease (IBD) is a chronic inflammatory disorder of the gastrointestinal tract. Post-treatment relapse is a major clinical challenge, and gut microbiota dysbiosis is hypothesized to be involved. Methods: We enrolled 88 patients with IBD (46 UC, 42 CD) to investigate gut microbiota features associated with post-treatment relapse. Fecal samples collected before and after therapy were analyzed by 16S rRNA sequencing. A random forest (RF) model was developed to evaluate the predictive value of microbiota signatures for recurrence. Results: At baseline (pre-treatment), no differences were observed in gut microbiota diversity between patients with UC and CD. However, significant compositional differences were observed, with Fusobacterium and Parabacteroides enriched in CD patients, and Anaerostipes and Enterococcus enriched in UC patients. Post-treatment, there was no significant difference in the α-diversity across IBD patients; however, β-diversity exhibited significant alterations, marked by enrichment of Akkermansia and Lachnoclostridium. Patients maintaining remission exhibited significant post-treatment beta-diversity shifts and enrichment of Erysipelatoclostridium, Delftia, Tyzzerella, Sphingomonas, Subdoligranulum, Proteus, and Enterococcus. Conversely, patients experiencing recurrence showed a significant reduction in Shannon alpha-diversity post-treatment and enrichment of UCG-002, Odoribacter, Delftia, Flavonifractor, and Erysipelotrichaceae_UCG-003. Post-treatment microbiota composition differed significantly between recurrent and non-recurrent patients, with higher alpha-diversity in the non-recurrent group. Non-recurrent patients exhibited enrichment of Eubacterium_hallii_group, Clostridioides, UCG-002, Paraprevotella, Bilophila, Desulfovibrio, Butyricimonas, Clostridium_sensu_stricto_1, Megamonas, Romboutsia, Parabacteroides, and Enterococcus, while Delftia was predominantly enriched in recurrent patients. The RF model, built using differentially abundant genera to distinguish recurrence status, achieved an area under the curve (AUC) of 0.721 in the validation set and 0.861 in the test cohort, indicating good predictive performance. Conclusion: Our findings suggest that gut microbiota composition may hold clues for predicting IBD relapse. The RF model is a proof-of-concept that warrants external validation in prospective, multi-center studies before clinical application.