P790 Assessment of steroid use in patients with active ulcerative colitis who initiated a new Janus kinase inhibitor or tumour necrosis factor inhibitor using data from a United States claims database

M C Dubinsky,M Gianfrancesco,G Gauthier,L Fallon,G R Lichtenstein,N H Khan,G Y Melmed,S Hanauer,T Ritter,G Bell,Y C Lee,N Kulisek,A Yndestad,D T Rubin
DOI: https://doi.org/10.1093/ecco-jcc/jjad212.0920
2024-01-01
Journal of Crohn's and Colitis
Abstract:Abstract Background In 2021, the United States (US) FDA issued a label update limiting the use of Janus kinase inhibitors (JAKi) to after tumour necrosis factor inhibitors (TNFi). We examined the rate of steroid use and treatment failure among patients (pts) with ulcerative colitis (UC) initiating a JAKi vs TNFi using data from a US claims database. Methods A database of adjudicated medical and pharmacy claims (IQVIA PharMetrics Plus; 2007–2022) was utilised to select pts with UC starting either a new JAKi/TNFi on/after 30 May 2018. Pts were followed from index date to the end of the study period, event of interest or treatment switch (whichever came first). The study assessed steroid use within 90 days of index date, and also treatment failure over the first 6 months after index date, defined as a composite of any: hospitalisation related to UC/colectomy (inpatient/emergency room)/switch to another advanced treatment (AT)/steroid use ≥90 days after index treatment initiation. Individual components of treatment failure were also analysed. Stabilised inverse probability treatment weights (sIPTW) were calculated using 19 confounders. Cox proportional hazards models with sIPTW were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). Additional analyses stratifying by AT-naïve vs AT-experienced pts were conducted. Results In total, 6019 pts were included; 763 initiated a JAKi and 5256 a TNFi. More pts initiating a TNFi had prior use of non-biologic conventional treatments and were on concomitant conventional treatments at index date, compared with pts initiating a JAKi. Prior AT was more prevalent among JAKi- (74.3%) vs TNFi- (15.3%) treated pts; baseline steroid use was generally similar across groups (Table). Overall, there was a significantly lower risk of steroid use within 90 days in pts initiating a JAKi vs a TNFi (HR 0.75 [95% CI 0.65, 0.87]), and among AT-naïve and AT-experienced pts (HR 0.79 [95% CI 0.64, 0.97] and HR 0.69 [95% CI 0.59, 0.82], respectively; Figure a). There was a significantly lower risk of treatment failure with JAKi vs TNFi (HR 0.80 [95% CI 0.68, 0.94]; Figure b). A significantly lower risk of steroid use ≥90 days was found amongst JAKi- vs TNFi-treated pts overall and within AT-naïve and AT-experienced subgroups (Figure b–d). Conclusion In this large US-based claims analysis, pts with UC treated with a JAKi had significantly less steroid use than those treated with a TNFi; this was consistent in AT-naïve and AT-experienced pts. Limitations included confounding of factors uncontrollable in claims data and inherent bias due to approved use of JAKi in the US only in pts with inadequate response to TNFi. Study sponsored by Pfizer. Medical writing support provided by C Duncan, CMC Connect; funded by Pfizer.
gastroenterology & hepatology
What problem does this paper attempt to address?