OP04 Endoscopic healing in paediatric IBD perpetuates a persistent signature characterized by pathogenic Th17 cells and tissue-associated molecular and microbial drivers of disease

Abstract

Abstract Background Endoscopic healing (EH) is the major long-term treatment target for inflammatory bowel diseases (IBD) in adults and paediatric patients. However, EH may not represent disease clearance. Hence, risk of relapse remains high and treatment discontinuation after reaching EH often results in disease exacerbation. We aimed to identify persistent cellular, molecular and microbial drivers of IBD under EH and longitudinally collected endoscopic biopsies from paediatric patients for analysis of single T-cell and bulk transcriptomics as well as mucosa-associated microbiota. Methods We followed disease trajectories of paediatric IBD patients treated according to international guidelines and collected mucosal biopsies from the terminal ileum (TI) and sigmoid colon (SC) at baseline (T1) and after approximately 3-12 months (T2) to assess EH. Non-IBD controls showed no macroscopic or histologic signs of inflammation. Host RNA and bacterial DNA were simultaneously extracted from single biopsies for bulk RNA-seq, and 16S rRNA seq. A second biopsy was used for isolation of lamina propria mononuclear cells (LPMCs), followed by CD45+, CD3+ cell sorting and single T-cell analysis (10X Genomics). Data were quality controlled and integrated. Results 217 biopsies (135 simultaneous extraction of host RNA and bacterial DNA; 82 single T-cell analysis) from 32 paediatric IBD patients (mean age 13 ± 3.5 years, 21 Crohn’s disease [CD], 11 ulcerative colitis [UC]; T1 = 32, T2 = 32, T > 2 = 6) and 5 non-IBD controls were analysed. Time between T1 and T2 averages 40 ± 22 weeks. At baseline, 15 patients (14 CD, 1 UC) were newly diagnosed and 31/32 patients had active mucosal inflammation. Twenty-two patients (71%) achieved EH at T2 (SES-CED ≤2 and absence of ulcerations), mostly on anti-TNF therapy. One patient (EH at T1) experienced relapsing disease within 10 weeks after discontinuing medication. In the bulk RNA-seq, we identified 299 differential expressed genes (DEGs; corrected p-value <0.05, FC >2.0), such as DUOX2, SAA2-SAA4, FCGR3B and NOS2, that are associated with active IBD and remained upregulated in EH in contrast to non-IBD controls. In addition single cell analysis revealed an IBD-specific pathogenic Th17 cluster that continued to exist in EH and showed high correlation with top DEGs after data integration. 16S rRNA gene seq highlighted highly individual mucosa-associated bacterial profiles and a reduced α-diversity in active and EH compared to non-IBD controls. Conclusion A persisting IBD signature in EH reflects ongoing cellular, molecular and microbial activity in comparison to non-IBD controls despite EH and mucosal regeneration. These markers may provide targets for future or sequential therapies.