The present categorization of inflammatory bowel disease (IBD) is based on clinical characteristics, which ignores the illness’s molecular foundation. The goal of this work was to use particular innate immune pathway profiling and unsupervised machine learning to stratify a treatment-naive juvenile IBD population (UML). In order to verify the molecular integrity of biological pathways involved in IBD, innate immune responses in 22 pediatric patients and 10 age-matched controls were examined at diagnosis. Using multiplex assays, peripheral blood mononuclear cells (PBMCs) were selectively stimulated to test the functioning of upstream activation receptors such as NOD2, toll-like receptor (TLR) 1-2, and TLR4, as well as downstream cytokine responses. To examine patient stratification, cytokine data were submitted to hierarchical clustering. When compared to controls, combined immune responses in patients across 12 effector responses were considerably lower, with “hypofunctional” TLR responses driving the majority of them. Three different groups of patients were formed as a result of hierarchical clustering, with the fourth group of “unclustered” people. There was no correlation found between the identified immunological clusters and the clinical illness profile.
Although there was no clinically beneficial consequence from hierarchical clustering, this work gives a reason for adopting a UML method to stratify patients. The study further emphasizes the importance of hypo-inflammatory innate immune responses in the etiology of IBD.
