OP20 Mucosal micoRNA profiles predict response to autologous stem-cell transplantation in Crohn’s disease
A. Lewis*1, R. Jeffrey1, T. Kumagai1, C. J. Hawkey2, M. M. Clark2, M. Allez3, J. Satsangi4, G. Rogler5, A. Silver1, J. O. Lindsay6
1Blizard Institute, Barts and The London School of Medicine and Dentistry, Centre for Genomics and Child Health, London, UK, 2Centre for Digestive Diseases, Queens Medical Centre, Nottingham, UK, 3Department of Gastroenterology, Hôpital Saint Louis, APHP, INSERM UMRS 1160, Paris Diderot, Sorbonne Paris-Cité Unversity, Paris, France, 4Translational Gastroenterology Unit, Nuffield Department of Medicine, University of Oxford, Oxford, UK, 5Department of Gastroenterology and Hepatology, University Hospital, Zurich, Switzerland, 6Blizard Institute, Barts and The London School of Medicine and Dentistry, Centre for Immunobiology, London, UK
The Autologous Stem Cell Transplantation for Crohn’s Disease (ASTIC) trial did not achieve its ambitious primary endpoint, but reported meaningful benefits in some CD patients previously refractory or intolerant to conventional therapies. However, the haematopoietic stem cell transplantation (HSCT) regimen used was associated with a high burden of adverse events. Consequently, there is a clear need to target this therapy to patients with the greatest chance of benefit. MicroRNAs (miRNAs) regulate cell signalling and their potential as biomarkers of disease is recognised. Here, we explore the ability of miRNAs to predict response to HSCT in CD patients.
miRNA profiles were analysed in RNA extracted from mucosal biopsies taken prior to HSCT from 14 CD patients enrolled in ASTIC. Clinical response to therapy was defined as CDAI <150 at 1 year; the cohort included seven ‘responders’ and seven ‘non-responders’. miRNA profiling was conducted using the miRCURY LNA microRNA Array (7th Gen). Natural groupings were explored using principal component analysis (PCA) and differences in miRNAs between groups determined by a two-tailed Student’s
PCA identified two natural groupings; Group 1 contained 6/7 of the responders and Group 2 contained 6/7 non-responders. Significant separation of responders and non-responders was identified along principal component 2 (
PCA separates responders and non-responders to HSCT in CD based on baseline miRNA profiles (A), along principal component 2 (B), which divides patients into two natural groupings (C).
The data indicate that miRNAs may act as predictive biomarkers of clinical response following HSCT. In particular, miR-155-5p, a well-characterised pro-inflammatory miRNA, was identified as a putative candidate biomarker. The results of the array now require independent validation.