P083. Atypical disease phenotypes in paediatric ulcerative colitis: 5‑year analyses of the EUROKIDS registry [EUROKIDS Porto IBD Working Group of ESPGHAN]
C. De Bie1, A. Levine2, D. Turner3, S. Cucchiara4, M. Sladek5, S. Murphy6, J. Escher1
1Erasmus MC-Sophia Children's Hospital, Paediatric Gastroenterology, Rotterdam, Netherlands; 2Wolfson Medical Center, Tel Aviv University, Pediatric Department, Holon, Israel; 3Shaare Zedak Medical Center, Pediatric Gastroenterology Unit, Jerusalem, Israel; 4Sapienza University of Rome, Paediatric Gastroenterology and Liver Unit, Rome, Italy; 5Jagiellonian University School of Medicine, Department Of Pediatrics, Krakow, Poland; 6University of Birmingham and Birmingham Children's Hospital, Institute of Child Health, Birmingham, United Kingdom
Background: Diagnosis of paediatric ulcerative (UC) and Crohn's colitis may be particularly challenging since isolated colitis is not unusual in early-onset Crohn's disease (CD). Our aim was to identify the prevalence of atypical disease patterns in a large unselected prospective cohort, in order to better facilitate accurate diagnosis of paediatric isolated colitis at onset.
Methods: Information was collected from the EUROKIDS registry, a prospective, web-based registry of newly diagnosed paediatric IBD patients in Europe and Israel that was initiated in May 2004. During the first five years, the registry has extended to allow inclusion of patients from 44 centres in 18 countries. Patients with new-onset UC were evaluated for extent of disease, rectal sparing, macroscopic patchiness, backwash ileitis, and upper gastrointestinal (UGI) involvement. Children with new-onset isolated Crohn's colitis were included for calculating positive predictive values (PPVs) of the findings, to take into account the prevalence of Crohn's colitis or UC in the presence of isolated colitis.
Results: 643 patients with UC (age 11.6±4.0 years; 50% males) met eligibility criteria, of whom 5% presented with proctitis, 18% with left-sided colitis, 9% with extensive colitis, and 69% with pancolitis. Patients with proctitis were significantly older at disease-onset compared with other patients (mean age 13.5±3.4 vs. 11.5±4.0 years, p = 0.011). Macroscopic rectal sparing was present in 5% of patients, who were significantly younger at diagnosis (p = 0.03) and had a trend towards more extensive disease (p = 0.08). The PPV of rectal sparing in diagnosing UC in the presence of isolated colitis was 58% (95% CI: 44% to 72%). Erosions or ulcerations anywhere in the UGI tract were present in 4% of children with UC (PPV 37%; 95% CI: 19% to 54%), but only in 0.8% in the oesophagus or duodenum (PPV: 25%; 95% CI: 0 to 55%). A caecal patch occurred in 2% of patients, macroscopic patchiness in 0.5%, and backwash ileitis in 10%.
Conclusions: There are multiple atypical phenotypes of paediatric UC at diagnosis. Although macroscopic rectal sparing and UGI involvement are uncommon in children with UC, their presence should not preclude this diagnosis. Gastric ulcerations and erosions are found in UC at diagnosis, but frank ulcerations in the oesophagus and duodenum of children with UC are rare. Recognition of atypical phenotypes is crucial for accurate diagnosis in early-onset UC at presentation.