VEO-IBD AS MODELS FOR PATHOGENIC STUDIES AND DEVELOPMENT OF PRECISION THERAPIES

Multiple monogenic disorders present as very early onset inflammatory bowel disease (VEOIBD) or as IBD with severe and atypical features. Establishing a genetic diagnosis may change patients’ management and prognosis. In this study, we describe the diagnostic approach to suspected monogenic IBD in a real clinical setting, discussing genetic and phenotypic findings and therapeutic implications of molecular diagnosis. Monogenic VEOIBD diagnostic approach changed over time, especially after the advent of next generation sequencing (NGS) techniques. NGS should be preferred in patients with nonspecific phenotypes. Nevertheless, Sanger sequencing is still effective in patients with suggestive clinical and immunological findings. In a multicentric collaboration with Bambino Gesù Children’s hospital, we developed a target gene panel sequencing (TGPS) including the most common monogenic diseases presenting with IBD symptoms, as first line of genetic approach for patient with non-specific phenotypes and negativity to this panel, we performed WES with an in silico analysis of 400 genes responsible for primary immunodeficiencies. 94 patients were included, and 13 (14%) reached a genetic diagnosis. Candidate sequencing was performed in 47 patients (50%), and NGS was performed in 85 patients (90%). Candidate sequencing had a good diagnostic performance only when guided by clinical features specific for known monogenic diseases, whereas NGS helped finding new causative genetic variants and would have anticipated one monogenic diagnosis (XIAP) and consequent bone marrow transplant (BMT). Genetic diagnosis impacted patient management in 11 patients (92%), 7 of whom underwent BMT. Although we identified 14% of monogenic disease in our cohort, the majority of cases remains without a genetic diagnosis. We hypothesized that transcriptome analysis by RNA sequencing (RNAseq) could help grouping multifactorial cases and correlating profiles with those found in distinct monogenic forms. We proposed a disease-similarity method for patients’ stratification and the detection of possible biomarkers. 13 out of 94 patients, depending on RNA availability (4 monogenic and 9 nonmonogenic) described in the previous genetic workup, performed gene expression analysis by RNAseq of peripheral blood cells. We compared gene expression profile of the 4 monogenic IBD (XIAP, TTC37, DKC1, and LRBA) with nonmonogenic IBD and performed cluster analysis. The most evident impact on peripheral blood cells came from XIAP and DKC1. TTC37 and LRBA did not show enriched pathway probably due to wrong sampling. Few nonmonogenic patients that presented extraintestinal manifestations (feature suggestive of monogenic defect) had a hybrid expression profile between monogenic and nonmonogenic IBD. Cluster and machine learning analyses might be applied to group patients by gene expression patterns in an unbiased manner. We performed an unsupervised analysis including our monogenic IBD, the nonmonogenic IBD from the cohort of Trieste and the first 13 genetically undefined VEOIBD and EOIBD enrolled within the collaborative project with University of Brescia, whose clinical collection data and genetic investigations are in progress. However, this data should be complemented by clinical reports and therapeutic management at the time of sampling to get more precise results and evaluate the obtained functional subgroups. Nevertheless, the characterization of more monogenic forms is a crucial point to expand this analysis and obtain more reliable results. The implementation of this knowledge may allow the use of monogenic disorders as prototypical diseases for the stratification and the therapeutic management of likely multifactorial cases towards a tailored therapy.