Objective: Extracellular RNAs (exRNAs), including in serum, show potential as noninvasive biomarkers. However, their (patho)physiological role is still not fully understood. In this study, we characterized serum exRNAs in patients with inflammatory bowel diseases (IBDs) and evaluated their ability to differentiate ulcerative colitis (UC) from Crohn's disease (CD).

Methods: We profiled serum exRNAs from 26 IBD patients (15 UC, 11 CD) with active endoscopic disease. Using Small Input Liquid Volume Extracellular RNA-sequencing, we investigated co-expression networks followed by randomized generalized linear modeling. An independent cohort (n = 109 UC, n = 150 CD) was studied to assess the exRNA reflection within the inflamed colonic and ileal mucosa.

Results: We detected 41 910 exRNAs, capturing 80.9% of genes expressed in intestinal tissue. Network analysis identified 69 clusters, of which one correlated with the distinction between CD and UC (r = -0.70, adjusted P = .006), featuring GNA12 as a hub gene. Serum GNA12 showed no association with fecal calprotectin, disease duration, age, or sex, but did correlate with UC-specific encoding genes (P < .05). Modeling within this upregulated UC-cluster prioritized a signature of 8 exRNAs including GNA12, distinguishing UC from CD with an accuracy [95% CI] of 95.8% [86.7-100.0%]. This elevated 8-marker signature was mirrored in the inflamed UC colon, when compared to the inflamed CD colon and ileum, also when evaluated along the disease location spectrum (P < .05).

Conclusions: Liquid biopsies may represent a novel, noninvasive approach in IBD biomarker development. Although larger in-depth studies are needed, this proof-of-concept study bridges noninvasive measurements with established IBD mechanisms, offering a promising tool for accurately distinguishing CD from UC.