P091 Usefulness of transcriptional whole blood biomarkers as a non-invasive surrogate marker of mucosal healing and endoscopic response in ulcerative colitis
Planell N.*1,2, Masamunt M.C.1, Franco Leal R.3, Rodríguez L.4, Esteller M.1, Lozano J.J.2, Ramírez A.1, Setsuko Ayrizono M.d.L.3, Saddy Rodrigues Coy C.3, Alfaro I.1, Ordás I.1, Visvanathan S.5, Ricart E.1, Guardiola J.4, Panés J.1, Salas A.1
1IDIBAPS, Hospital Clínic, CIBER-EHD, Gastroenterology, Barcelona, Spain 2CIBER-EHD, Bioinformatics Platform, Barcelona, Spain 3IBD Research Laboratory, University of Campinas, Surgery, Sao Paulo, Brazil 4Hospital Universitari de Bellvitge - Institut d'Investigaciό Biomèdica de Bellvitge, Gastroenterology, Barcelona, Spain 5Boehringer-Ingelheim Pharmaceuticals Inc, Ridgefield, United States
The importance of objective monitoring of inflammatory activity in IBD is becoming increasingly clear since the amelioration of symptoms alone is not a reliable predictor of mucosal healing. In the context of clinical trials that rely on clinical scores to assess disease activity for patient selection or outcome measurements, this may lead to high placebo response rates, as well as unreliable or un-reproducible results. While endoscopy remains the gold standard, it carries a cost and risk to patients, negatively affecting patient recruitment. The objective of this study was to identify new accurate non-invasive biomarkers based on whole blood transcriptomics that would predict the presence of mucosal lesions in UC patients independently of clinical symptoms.
A total of 152 UC patients were recruited at the time of endoscopic examination. Endoscopic activity was recorded (Mayo sub-score=0–3) and whole blood RNA obtained for transcriptional analysis. Transcriptional changes in the blood of UC individuals (n=25) and non-IBD controls (n=20) was analyzed using microarrays. Genes that correlated with endoscopic disease activity were validated using qPCR in an independent cohort of UC patients recruited at 3 different centers (n=111). Using this cohort, a prediction model for mucosal lesions was evaluated. Responsiveness to treatment of the identified biomarkers was evaluated in a third cohort of UC patients (n=16) that started anti-TNF and were followed-up for 14 weeks.
Microarray analysis identified 120 genes that were significantly altered in the blood of endoscopically active (endoscopic Mayo score≥1) UC patients. A significant correlation with the degree of endoscopic activity and inflammatory burden (endoscopic Mayo sum of segments) was observed in 20 and 18 genes, respectively, using qPCR in an independent cohort. Within the identified genes, the ones that showed better correlation and ability to discriminate between the different degrees of inflammation were HP, CD177, GPR84 and S100A12. Using HP as a predictor of endoscopic disease activity, an accuracy of 67.3% was observed, proving superior to current biomarkers, including CRP, ESR and platelet count (accuracies 52.4, 45.2 and 30.3%, respectively). Finally, at 14 weeks of treatment, anti-TNF induced alterations in the blood transcripts of HP, CD177, GPR84 and S100A12 that correlated with changes in endoscopic activity.
Transcriptional changes of selected genes in whole peripheral blood of UC patients correlate with endoscopic disease activity. Importantly, transcriptional changes in response to treatment in UC patients are sensitive to endoscopic improvement and appear to be an effective and non-invasive way to monitor patients over time.