Infliximab dependency in children with Crohn’s disease

SummaryBackground  Recently, infliximab dependency has been described.Aim  To assess frequency of ID in 82 consecutive Crohn’s disease children treated with infliximab 2000–2006 and to describe clinical and genetic predictors of long‐term infliximab response.Methods  A phenotype model of infliximab dependency was used to assess treatment response: ‘immediate outcome’ (30 days after infliximab start) – complete/partial/no response. ‘Long‐term outcome’: (i) prolonged response: maintenance of complete/partial response; (ii) infliximab dependency: relapse ≤90 days after intended infliximab cessation requiring repeated infusions to regain complete/partial response or need of infliximab >12 months to sustain response. Polymorphisms TNF‐308 A>G, TNF‐857 C>T, Casp9 93 C>T, FasL‐844 C>T, LTA 252 C>T and CARD15 (R702W, G908R, 1007fs) were analysed.Results  Ninety‐four per cent of children obtained complete/partial response. In long‐term outcome, 22% maintained prolonged response, 12% had no response, while 66% became infliximab dependent. Perianal disease and no previous surgery were associated with infliximab dependency (OR 5.34, 95% CI: 1.24–22.55; OR 6.7, 95% CI: 1.67–26.61). No association was found with studied polymorphisms. The cumulative probability of surgery 50 months after starting infliximab was 10% in infliximab dependency, 30% in prolonged responders and 70% in nonresponders (P = 0.0002).Conclusions  Sixty‐six per cent of children became infliximab dependent. Perianal disease and no surgery prior to infliximab were associated with infliximab dependency phenotype.