错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Fracture-Related Safety Reporting of JAK Inhibitors: An Analysis from the WHO Global VigiBase

  • Adrian Martinez de la Torre,
  • Andreas Bech Clausen,
  • Andrea M. Burden,
  • Stefan Weiler

摘要

Introduction

The Janus kinase (JAK) inhibitors are treatment options for autoimmune diseases. Numerous safety concerns have been raised. The European Medicines Agency updated the product information of tofacitinib to include the risk of fractures—but not for other JAK inhibitors. We conducted a global pharmacovigilance analysis of previously investigated JAK inhibitors to investigate a potential class effect.

Methods

Individual case safety reports (ICSRs) for all licensed JAK inhibitors were identified from the global WHO pharmacovigilance database. The primary outcome of interest was a bone fracture. Disproportionality analyses using reporting odds ratios (RORs) were conducted.

Results

We identified 122,037 ICSRs for tofacitinib, 27,786 ICSRs for upadacitinib, 14,616 ICSRs for baricitinib, 830 for filgotinib, and 350 for abrocitinib. Among the ICSRs, we identified 2198 (1.8%), 634 (2.3%), and 144 (1.0%) reports, where a bone fracture was reported for tofacitinib, upadacitinib, and baricitinib, respectively. Few reports were available for the newest drugs filgotinib (10) and abrocitinib (1). JAK inhibitors were associated with increased reporting for fracture: tofacitinib (ROR 3.34, 95% confidence interval [CI] 3.20–3.48), upadacitinib (ROR 4.23, 95% CI 3.80–4.48), baricitinib (ROR 1.80, 95% CI 1.52–2.11) and filgotinib (ROR 2.24, 95% CI 1.11–3.94). Patients with bone fractures were more often female, older and had a higher number of co-reported medications. They were more likely to use glucocorticoids, opioids, and bisphosphonates.

Conclusion

The results from this pharmacovigilance analysis, based on a spontaneous reporting database associated with inherent limitations, suggest a potential risk of fractures with JAK inhibitors, indicating that a class effect cannot be ruled out.