Abstract

Background: Inflammatory bowel disease (IBD) medications like thiopurines and anti-tumor necrosis-alpha (anti-TNF) agents have been associated with a risk of non-melanoma skin cancer (NMSC). While tofacitinib (JAK inhibitor) has shown similar risks in trials, real world data are sparse. Less is known about NMSC risk with newer biologics like ustekinumab (anti-interleukin 12/23) and vedolizumab (anti-integrin).

Objectives: This study measures NMSC risk in patients with IBD administered tofacitinib, ustekinumab, and vedolizumab compared to thiopurines and anti-TNFs.

Methods: Using commercial health plan data in the Healthcare Integrated Research Database, we identified patients with prevalent IBD between 03/25/2009 and 04/30/2023. Treatment episodes were established using outpatient pharmacy dispensing and infusion codes with a 183-day washout and 183-day blackout. The therapies of interest were tofacitinib, ustekinumab, and vedolizumab; anti-TNFs and thiopurines were comparators. NMSC was identified using diagnosis codes and procedure codes for interpretation of pathology, destruction, or excision of lesions on the same day. Cohorts were weighted using a stabilized inverse propensity score (PS) for treatment and adjusted for any unbalanced baseline variables like prior treatment, demographics, region, and comorbidities. Crude incidence rates (IRs), adjusted hazard ratios (HRs), and 95% confidence intervals (CIs) are presented. Results were stratified by Crohn’s disease (CD) or ulcerative colitis (UC).

Results: We identified 63,860 patients with IBD with 196,505 treatment episodes. Mean age was 40 years. Patients were largely non-Hispanic white (88%) living in the southern census region (35%). Compared to thiopurines, IRs of NMSC (per 1000 person-years) were 18.5 for tofacitinib, 7.4 for ustekinumab, 9.9 for vedolizumab, and 24.7-25.5 for thiopurines. Corresponding HRs were 1.23 (CI: 0.45 – 3.38) for tofacitinib, 0.30 (CI 0.18 – 0.50) for ustekinumab, and 0.36 (CI 0.26 – 0.51) for vedolizumab. For comparisons to anti-TNF, IRs of NMSC (per 1000 person-years) were 19.0 for tofacitinib, 7.2 for ustekinumab, 9.9 for vedolizumab, and 9.3 for anti-TNF. Corresponding HRs were 1.63 (0.76 – 3.52) for tofacitinib, 0.77 (0.46 – 1.29) for ustekinumab, and 0.72 (0.54 – 0.97) for vedolizumab. The direction and magnitude of the results of all analyses were similar when stratified by CD or UC.

Conclusions: In patients with IBD, treatment with ustekinumab or vedolizumab was associated with a lower risk of NMSC compared to thiopurine or anti-TNF therapy. Tofacitinib was associated with a higher risk of NMSC as compared to anti-TNF or thiopurines. This information can help guide patients and clinicians in IBD treatment selection and help inform skin cancer screening recommendations.