P365 Identification of predictive factors of AZA/6-MP treatment outcome in paediatric luminal Crohn’s disease: a multicentre study of the paediatric IBD Porto group of ESPGHAN
T. LERCHOVA1, O. Hradsky1, M. Kulich2, G. Veres3, J.A. Dias4, M. Sladek5, S. Kolacek6, S. Van Biervliet7, J. Melek8, D. Serban9, K. Winther10, T. de Meij11, J. Schwarz12, K.L. Kolho13, J.C. Escher14, J. Bronsky1
1Department of Paediatrics, 2nd Faculty of Medicine- Charles University and Motol University Hospital, Prague, Czech Republic, 2Department of Probability and Mathematical Statistics, School of Mathematics- Charles University, Prague, Czech Republic, 3Pediatric Institute-Clinic, University of Debrecen, Debrecen, Hungary, 4Department of Paediatrics, Hospital de Sao Joao, Porto, Portugal, 5Department of Pediatrics- Gastroenterology and Nutrition, Polish-American Children’s Hospital- Jagiellonian University Medical College, Cracow, Poland, 6Department of Paediatrics, Children’s Hospital Zagreb Faculty of Medicine, Zagreb, Croatia, 7Department of Pediatrics and Medical Genetics, Ghent University- Ghent University Hospital, Ghent, Belgium, 8Department of Paediatrics, Charles University Teaching Hospital, Hradec Kralove, Czech Republic, 9Department of Gastroenterology, 2nd Clinic of Pediatrics- ‘Iuliu Hatieganu’ University of Medicine and Pharmacy- Emergency Clinical Hospital for Children, Cluj-Napoca, Romania, 10Department of Paediatrics, Hvidovre Hospital, Copenhagen, Denmark, 11Department of Paediatrics, Amsterdam UMC- Location VUMC, Amsterdam, The Netherlands, 12Department of Paediatrics, Faculty of Medicine in Pilsen, Charles University, Pilsen, Czech Republic, 13Department of Gastroenterology, Children’s Hospital and University of Helsinki, Helsinki, Finland, 14Department of Gastroenterology, Erasmus MC−Sophia Children’s Hospital, Rotterdam, The Netherlands
Background
According to current guidelines, most paediatric patients in Europe diagnosed with Crohn′s disease (CD) are prescribed long-term immunosuppressive therapy with azathioprine (AZA). This study aimed to develop a predictive model allowing to stratify patients who will not benefit from AZA maintenance treatment and who require a more intensive therapeutic approach early after diagnosis.
Methods
The study was designed to develop a clinical prediction rule using retrospective data analysis from patients included to prospective inception cohort, the EUROKIDS-IBD until 2017. In total, 1190 CD patients using AZA were selected from the registry in 13 European centres. Of these, 441 patients who responded to induction treatment, started AZA treatment within 6 weeks from the time of diagnosis and maintained remission at week 12 were entered into this multicentre study. The primary outcome was time to clinical relapse defined as a necessity of re-induction of remission. A sequence of Cox models was fitted to predict the risk of relapse. Variables appearing to be significant in univariate risk analyses were added to the model one by one and non-significant terms were dropped.
Results
More than 50 % of patients did not experience a clinical relapse within 2 years of AZA treatment (Figure 1). The median time to relapse was 2.11 (CI 1.59–2.46) years. Out of all tested parameters available at the time of diagnosis,8 were predictive in the context of early relapse. Only 4 of these variables, namely centre (
Conclusion
In paediatric CD patients achieving clinical remission after induction treatment at week 12, AZA was effective as maintenance treatment in over 50% of patients for 2 years. Due to low values of HR, the possibility to predict relapse on AZA treatment appears to be limited.
Acknowledgement: This work was supported by the Grant Agency of Charles University in Prague [grant number 364617] and by Ministry of Health, Czech Republic-conceptual development of research organisation, Motol University Hospital, Prague, Czech Republic [00064203].