Predictors of the need for second intestinal resection in children with Crohnʼs disease

Intestinal resection performed early in the course of disease was previously considered a poor prognostic indicator in Crohn's disease (CD). A recent population-based study from France demonstrated that surgery performed within the first year following diagnosis in children with CD was not associated with the need for a second surgery (1). The aim of this study was to assess whether earlier first surgery and younger age at diagnosis was prognostic of need for a second surgery. We used a large, population-based cohort of pediatric CD patients from Ontario, Canada (The Ontario Crohn's and Colitis Cohort) to validate the findings from France. A validated algorithm (2) was used to identify all children <17 years diagnosed 1994-2007 with CD and who underwent a first intestinal resection within Ontario's health administrative data comprising all legal residents of Canada's most populous province. Only patients with a minimum 2 year follow-up period were included. Physician billing and hospitalization data were used to identify whether patients underwent a second intestinal resection. Multivariable Cox proportionate hazard models assessed whether time from diagnosis to first surgery was predictive of undergoing a second surgery, controlling for the following a priori-determine confounders: age at diagnosis, gender, income quintile, and rural/urban status at diagnosis. Of 1917 children diagnosed with CD from 1994 to 2007, 28.8% (n=553) underwent a first intestinal resection in the follow-up period. Median age of diagnosis was 14 years (interquartile range (IQR) 11-15), median time to first surgery was 2.1 years (IQR 0.5-4.6) and median follow-up duration was 5.4 years (IQR 2.8-8.5) following first surgery. Of patients undergoing first surgery, 29% (n=163) patients underwent a second surgery within the follow-up period, with cumulative risk of 14% at two years, 25% at five years, and 36% at 10 years of undergoing either second surgery or censoring. In multivariable regression, earlier time from diagnosis to first surgery was not associated with increased risk of undergoing second surgery (HR for a 1-year decrease 1.05, 95% CI 0.99-1.12). Of covariates included in the model, only younger age at diagnosis (HR for a 1-year decrease 1.06, 95% CI 1.01-1.11) and rural residence at diagnosis (HR 1.58, 95% CI 1.05-2.37) were associated with undergoing a second surgery. Due to significant statistical interactions between follow-up time and the independent variables, analysis was stratified by years of follow-up. Earlier time to first surgery was only a predictor of second surgery at the second (HR 1.10, 95% CI 1.01-1.19) and third (HR 1.11, 95% CI 1.02-1.21) years of follow-up, and then became non-significant in years 4-14 of follow-up. Earlier age at diagnosis was predictive of second surgery in all subgroup analyses, except in those with only one year of follow-up. Rural status at diagnosis was predictive of second surgery only in those with 11-14 years of follow-up (HR range 1.58-1.59), but not shorter follow-up time (HR range 1.16-1.48). To our knowledge, this is the largest study to report on rates and predictors of second intestinal resection surgery in CD. Resection performed earlier in the course of disease was not associated with a higher likelihood of undergoing a second surgery. However, younger age at diagnosis and rural residence were significantly associated with likelihood of second surgery, when controlling for other sociodemographic confounders. Unfortunately, due to the limitations of health administrative data, clinical disease factors could not be assessed as potential predictors. This study confirms the findings from France that earlier age at diagnosis (and not earlier first resection) was a poor prognostic marker, and indicates that patients in rural Ontario were at increased risk for second surgical resection.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.