P307. Age at diagnosis 40-years is not an accurate predictor of disease outcome in patients with Crohn's disease
B.D. Lovasz1, L. Lakatos2, P.A. Golovics1, A. Horvath3, I. Szita2, T. Pandur2, M. Mandel1, Z. Vegh1, G. Mester4, M. Balogh4, C. Molnar5, E. Komaromi6, L.S. Kiss1, P. Lakatos1, 1Semmelweis University, 1st Department of Medicine, Budapest, Hungary, 2Csolnoky F. Province Hospital, 1st Department of Medicine, Veszprem, Hungary, 3Csolnoky F. Province Hospital, Department of Pediatrics, Veszprem, Hungary, 4Grof Eszterhazy Hospital, Papa, Hungary, 5Magyar Imre Hospital, Ajka, Hungary, 6Municipal Hospital, Varpalota, Hungary
An age <40-years at diagnosis has been identified as a marker of disabling Crohn's disease (CD) disease in previous studies. However, the definition of disabling disease was variable. The aim of this study was to analyze the predictive value of age at onset <40 years to predict the evolution of disease behavior, need for resective surgery, time to azathioprine (AZA) or steroid exposure in the population-based Veszprem province database.
Data of 506 incident CD patients were analyzed (median age at diagnosis: 31.5; SD: 13.8 years). Both in- and outpatient records were collected and comprehensively reviewed.
Patients with an age at diagnosis >40 years presented more frequently with colonic (48.2% vs 32.6%, p < 0.001) and complicated (50%, vs 41.1%) disease compared to patients with an age at diagnosis <40 years. However, in a Kaplan–Meier analysis the probability to develop complicated (<40-years: 54.6% and 65.8% vs >40 years: 59.2% and 60.4% after 5 and 10 years), penetrating disease (<40-years: 36.8% and 49.1% vs >40 years: 31% and 37.7% after 5 and 10 years, pLogRank = 0.15) or need for surgery (<40-years: 13.6%, 33.3% and 44.8% vs. >40 years: 15.9%, 33% and 43.3% after 1, 5 and 10 years, pLogRank = 0.97) was not significantly different in patients with an age at onset < or > 40 years. In contrast, overall AZA (50.6% vs. 28.6%, OR: 2.56, 95% CI: 1.62–4.05), anti-TNF (10.3% vs. 4.5%, p = 0.058), steroid exposure (72.2% vs. 56.3%, OR: 2.02, 95% CI: 1.31–3.13) and perianal disease (27.8% vs. 17%, OR: 1.88, 95% CI: 1.10–3.24) were more frequent in patients with an age at diagnosis <40 years despite similar length of follow-up (mean: 12.6 years vs 11.7 years). In a Kaplan–Meier analysis, the probability of AZA use was significantly higher in patients with an age at diagnosis <40 years (32.8%, 39.2% and 51% vs >40 years: 17.1%, 25%, 31.1%, pLogRank = 0.001, HR: 1.89, 95% CI: 1.27–2.82) after 1, 5 and 10 years of disease duration.
Despite subtle differences in disease phenotype at diagnosis, age at diagnosis <40 years was not associated with clinically important disabling outcomes (e.g. need for surgery, development of complicated or penetrating disease). In addition, the association between age at diagnosis and drug exposures (e.g. need for or time to AZA, anti-TNF or steroids) rather represents a difference in treatment strategy and should not necessarily be interpreted as a disabling outcome.