Abstract
PredictSURE IBD is a prognostic blood test that classifies newly diagnosed, treatment-naïve Inflammatory Bowel Disease (IBD) patients into 'IBDhi' (high-risk) or 'IBDlo' (low-risk) groups (risk of future aggressive disease). We evaluated this assay in a multinational cohort and explored the effect of concomitant corticosteroids on its discrimination.
One hundred thirty-six (71 Ulcerative colitis [UC], 65 Crohn's Disease [CD]) and 41 (15 UC, 26 CD) patients with active IBD were 'unexposed' and 'exposed', respectively, to corticosteroids at baseline blood sampling. The number of treatment escalations, time to first escalation, and need for repeated escalations were compared between the biomarker subgroups. Another 20 patients (13 UC, 7 CD) were longitudinally sampled over 6 weeks after commencing corticosteroids.
In corticosteroids-naïve UC and CD patients, all bowel surgeries (n = 6) and multiple therapy escalations (n = 10) occurred in IBDhi patients. IBDhi UC patients required significantly more treatment escalations, had a shorter time to first escalation, and a greater need for multiple escalations than IBDlo patients. No statistically significant differences were observed among CD patients. In corticosteroid-exposed patients, 66.6% of 'misclassifications' were IBDlo patients who required escalations. Among corticosteroid-treated patients with longitudinal sampling, 81.3% of those classified as IBDhi before steroids switched to IBDlo during therapy.
No significant differences in treatment escalations were observed between biomarker-defined subgroups in CD. However, IBDhi UC patients required significantly earlier and more frequent therapy escalations, highlighting the need to further investigate PredictSURE IBD in UC. Notably, the discrimination ability of the biomarker was unreliable in patients receiving corticosteroid therapy.