RESUMO
BACKGROUND: In a considerable proportion of anaemic children with inflammatory bowel disease (IBD), haemoglobin (Hb) does not normalise after iron therapy. We evaluated the added value of novel iron markers (hepcidin and soluble transferrin receptor [sTfR]) as compared to traditional iron markers (ferritin and transferrin saturation [TSAT]) to determine the best strategy for the prediction of non-responsiveness to iron suppletion. METHODS: In this secondary analysis of prospectively collected data, we measured iron markers in anaemic children (Hb Z-score < -2.0) with IBD at baseline and one month after the initiation of iron therapy. Non-responsiveness was defined as an increase in Hb Z-score of less than 1 within a month. Logistic regression analysis was used to construct multi-biomarker prognostic models. RESULTS: Of 40 anaemic paediatric IBD patients, sixteen (40%) were non-responsive to iron therapy after one month. Hb Z-score and hepcidin Z-score had the highest predictive ability (area under the ROC curve [AUROC] 0.80) providing sensitivity of 69% and specificity 92%. In a post-hoc analysis we defined hepcidin cut-off values to predict iron non-responsiveness. CONCLUSION: A diagnostic strategy that involves baseline Hb Z-score and hepcidin Z-score in anaemic children with IBD reliably identifies those who will not respond to iron therapy. IMPACT: Non-response to oral and intravenous iron suppletion therapy is high in paediatric IBD and should be identified early. Prediction models using baseline hepcidin demonstrated higher sensitivity and specificity to predict iron non-response compared to models using baseline traditional iron indicators (ferritin and transferrin saturation). In a post hoc analysis, we defined cut-off values for hepcidin to facilitate the correct timing of iron treatment in young anaemic patients with chronic inflammatory bowel disease.