P603 Patient characteristics and adverse effects in biologic treatment of Crohn’s Disease and Ulcerative Colitis: A nationwide Danish cohort study years 2015-2018.
Bjørn Jensen, C.(1);Jarlov Jensen, K.(1);Wennerström, C.(2);Sommer, K.J.(3);Burisch, J.(4);Petersen, J.(1);
(1)Frederiksberg Hospital- Frederiksberg- Denmark., Copenhagen Phase IV Unit- Center for Clinical Research and Prevention-, Frederiksberg, Denmark;(2)Janssen Cilag AB, Nordic Medical Operations, Solna, Sweden;(3)Janssen-Cilag A/S, Medical Affairs, Birkerød, Denmark;(4)Hvidovre Hospital, Gastrounit- Medical Division-, Hvidovre, Denmark
Background
Biological therapies have demonstrated effectiveness in the treatment of Crohn’s Disease (CD) and Ulcerative Colitis (UC), but comparative analyses of the various treatments remain limited.
Methods
Using Danish national registries, we performed a nationwide cohort study including all bio-naïve individuals diagnosed with UC or CD initiating treatment with infliximab (IFX), adalimumab (ADA), vedolizumab (VDZ), golimumab (GOL), or ustekinumab (UST) during 2015-2018. The hazard ratios (HR) of hospitalization, surgery, and start of corticosteroid therapy were explored using Cox regression adjusted for health-related and socio-economic parameters at treatment initiation.
Results
More than 90% of 1,836 CD and 1,886 UC patients had IFX as first-line treatment. Patients with ADA, VDZ, UST or GOL as first-line treatment were older at diagnosis and start of biologic treatment. A larger proportion were women, had higher education, and had previously had IBD surgery.
We observed a higher risk of all-cause hospitalization (HR: 1.56 (95% CI: 1.16; 2.10)) among CD patients with ADA as first-line treatment compared to IFX. No risk difference was observed of IBD hospitalization, IBD related surgery, or use of corticosteroids between ADA- and IFX-treated CD patients. We observed no risk difference between ADA- and IFX-treated UC patients for any outcomes. For both CD and UC patients, analyses of VDZ were based on very low number of cases and the results were inconclusive. Analyses of GOL and UST were not possible because of low numbers.
Second-line treatment with VDZ compared to ADA among CD patients previously treated with IFX showed no risk difference of all-cause hospitalization (1.32 (0.91; 1.92)), IBD hospitalization (1.27 (0.68; 2.35)), IBD related surgery (1.41 (0.69; 2.85)), or use of corticosteroids (0.70 (0.27; 1.82)). Among UC patients, second-line VDZ was associated with an increased risk of IBD related surgery (1.94 (1.01; 3.71)), but not of all-cause hospitalization (0.77 (0.48; 1.24)), IBD hospitalization (0.75 (0.38; 1.46)), or corticosteroid use (0.92 (0.37; 2.28)) compared to ADA.
Conclusion
We observed increased risk of hospitalization among CD patients treated with ADA compared to IFX, and increased risk of IBD-related surgery in UC patients receiving VDZ as second line treatment compared to ADA. The differences in key baseline characteristics patients initiating biologic treatment with ADA, VDZ, UST or GOL compared to IFX might reflect differences in disease state, that justify deviations from treatment guideline. It should be further investigated if the associations are an effect of differences in treatment efficacy and safety or confounding by differences in disease state at treatment start.