From Paris to Montreal: disease regression is common during long term follow-up of paediatric Crohn's disease

Introduction: Paediatric Crohn's disease (PCD) often presents with extensive and a frequent pan-enteric phenotype at onset. However, its long term evolution into adulthood, especially since the widespread use of biological agents, is not well characterised. We conducted a single centre cohort study of all PCD patients transitioned to adult care to assess the long term disease evolution in the era of biologic therapy. Methods: We conducted a retrospective observational, study of all PCD patients who were subsequently transferred to the care of an adult gastroenterology unit and had a minimum follow up of 2 years. We examined the case notes for evolution of disease location and behaviour. Disease location and behaviour was characterised using Paris classification at diagnosis and Montreal classification at last follow-up. In addition, we examined variables associated with complicated disease behaviour and the need for CD related intestinal resection. Results: In total, 132 patients were included with a median age at diagnosis of 13 (IQR 11-14) and a median follow up of 11 years (range 4-14). At diagnosis, 23 (17.4%), 39 (29.6%) and 70 (53%) patients had ileal, colonic and ileocolonic disease respectively. In addition, 31 (23.5%) patients had L4a or L4b disease at diagnosis (proximal or distal to the ligament of treitz respectively) and 13 patients (9.8%) had both whilst 27 (20.4%) patients had perianal disease. At diagnosis, 27 (20.4%) patients had complicated disease behaviour but 83 (62.9)% of patients had an extensive 'pan-enteric' phenotype. Of these patients only 55 (66.3%) retained the pan-enteric phenotype at last follow-up (p = .0002). Disease extension was noted in 25 (18.9%) of patients and regression was noted in 47 (35.6%) of patients, whereas upper GI disease was noted in significantly fewer patients at last follow-up (21, 15.9%) (p = .0001). More patients had complicated disease behaviour (46 patients, 34.9%, p = .0018) at last follow-up. There was a high exposure to both thiopurines 121 (91.7%) and biologics 84 (63.6%). The cumulative probability (95% CI) of surgery was 0.05 (0.02, 0.11) at 1 year, 0.17 (0.11, 0.24) at 3 years and 0.22 (0.15, 0.30) at 5 years. Neither disease location nor behaviour were associated with the need for intestinal resectional surgery. Conclusions: Over the course of an extended follow-up period, there appeared to be changes in both disease location and behaviour in PCD. Interestingly, a significant proportion of patients had disease involution which may be related to a high rate of exposure to thiopurines and biologics. We were unable to identify any variables associated with complicated disease course or the need for intestinal surgery.