Venous thromboembolism risk with Janus kinase inhibitors: Is it a class wide effect?

1. Introduction

Janus kinase (JAK) inhibitors are a widely used therapy in the treatment of inflammatory arthritis. In 2019, the Food and Drug Administration (FDA), prompted by the initial results of a postmarketing study, issued a black box warning about an increased risk of thromboembolism and death with tofacitinib 10 mg twice a day. The FDA then applied this boxed warning to baricitinib and upadacitinib, agents not included in the postmarketing study. In this editorial, we review the evidence for venous thromboembolism (VTE) risk in JAK inhibitors and whether this risk also extends to the second-generation JAK inhibitors.

2. Tofacitinib

The ORAL surveillance study compared the safety of tofacitinib 5 mg and 10 mg twice daily regimes versus TNF inhibitor therapy (adalimumab and etanercept) in rheumatoid arthritis (RA) patients aged 50 years and over, with at least one cardiovascular risk factor. Incidence rates (IRs) per 100 patient-years for pulmonary embolism (PE) in the tofacitinib 10 mg twice daily, 5 mg twice daily, and TNF inhibitor arms were 0.5, 0.3, and 0.1, respectively. The significant imbalance in risk of pulmonary embolism (PE) and all-cause mortality in the tofacitinib 10 mg twice daily arm led to a discontinuation of this dose regime and the issuing of a safety signal [1]. Currently, tofacitinib at 10 mg twice daily is only approved for ulcerative colitis.

By contrast, a post hoc safety analysis, including RA, psoriatic arthritis (PsA), and psoriasis (PsO) patients, showed similar IRs for thromboembolic events in the two tofacitinib dose regimes (10 mg and 5 mg twice daily), but IRs were higher for all tofacitinib doses in those with baseline cardiovascular (CVD) or VTE risk factors versus those without. IRs in patients with and without VTE risk factors were 0.21 and 0.07 per 100 patient-years for deep vein thrombosis (DVT), 0.18 and 0.04 per 100 patient-years for PE, and 0.35 and 0.10 per 100 patient-years for VTE, respectively [2]. In the RA programme, IRs for patients with baseline CVD risk factors were broadly similar to those in ORAL surveillance, except for PE, which was higher in ORAL surveillance (0.54 per 100 patient-years (0.32–0.87)) than the RA programme (0.24 per 100 patient-years (0.13–0.41)) for patients receiving tofacitinib 10 mg twice daily [2].

In the US Corrona RA registry, real-world data were collected, and therefore, included patients were ineligible for randomized control trials due to comorbidities. Patients on tofacitinib predominantly received the 5 mg twice daily regime, and VTE rates were similar for tofacitinib (IR 0.29 per 100 patient-years) and biologic DMARD (IR 0.33 per 100 patient-years) groups. The mean follow-up for tofacitinib patients was 1.58 years [3]. Among patients with CVD risk factors, VTE rates were comparable to those in the RA tofacitinib program [2]. In addition to the FDA warning, the European Medicines Agency (EMA) recommends using tofacitinib in RA patients ‘above the age of 65 only when there is no alternative treatment.’

3. Baricitinib

In a post hoc safety analysis for baricitinib, six VTE events occurred in 997 patients treated with a 4 mg daily regime during the 24-week placebo-controlled period versus zero VTEs in the 1070 placebo-treated patients. However, all six patients had VTE risk factors. Also, this was not replicated during the initial 24 weeks of treatment when patients from the placebo group were transferred to baricitinib 4 mg daily. IRs were similar between the 2 mg and 4 mg regimes during extended observations. In the All-bari-RA, a data analysis set including all baricitinib exposures at any dose, the IR of DVT/PE was 0.5 per 100 patient-years and stable over time [4]. The authors placed this in context by stating that this IR was in the reported range for RA patients in North America and Europe [4,5].

4. Upadacitinib

For upadacitinib, in an integrated analysis from the SELECT phase III clinical programme, including 3834 treated patients from five studies, IRs of adjudicated VTE in patients receiving 15 mg and 30 mg daily regimes were similar in both dose regimes, as well as placebo, adalimumab, and methotrexate. The mean exposure for upadacitinib 15 mg and 30 mg was 53 and 59 weeks, respectively [6]. Body mass index and prior VTE were factors associated with VTE in both upadacitinib 15 mg and 30 mg groups. Age and NSAID use were associated with VTE in the 15 mg but not the 30 mg group [7].

5. Filgotinib

With filgotinib, an integrated safety analysis with 3691 patients from seven studies showed that no VTEs were reported in the placebo-controlled group, but exposure-adjusted IRs for VTEs were 0.2 and 0 per 100 patient-years for filgotinib 200 mg and 100 mg dose regimes, respectively. The median duration of filgotinib exposure was 1.6 years. In the placebo-controlled analysis set, the median placebo exposure was 0.5 years. All nine patients developing VTE had at least one VTE risk factor [8].

6. Peficitinib

Peficitinib is approved in Japan, Korea, and Taiwan. In a long-term extension study assessing the safety of peficitinib in Asian patients, the IR was 0.1 per 100 patient-years; these two cases were not considered related to peficitinib [9].

7. Expert opinion

It remains unknown whether JAK inhibitors play a direct causal role in the development of VTEs or whether these VTEs occur in the context of a higher baseline risk in RA [10]. One meta-analysis, including 42 studies and 6,542 JAK inhibitor and 1,578 placebo patient exposure years, concluded that the pooled VTE risk for JAK inhibitors as a whole does not confirm any increased VTE risk [11]. However, as Cohen [12] pointed out, the majority of VTEs occur in the long-term extensions of clinical trials, an aspect missed if only the placebo-controlled parts of the trials are included. This can lead to an underestimated VTE risk. Furthermore, patients with significant cardiovascular risk factors are often excluded from such clinical trials, potentially skewing the outcomes [13].

Although the data appear more promising for the second-generation JAK inhibitors, there are less patient years of exposure than for the first-generation agents, as well as a paucity of data comparing VTE risk between the JAK inhibitors. Registry data and extension studies are required to evaluate the second-generation agents further [14]. It also remains unanswered how the blocking of a JAK-STAT pathway could increase prothrombotic activity and whether such a mechanism is bypassed with agents that have more focused JAK selectivity.

When prescribing JAK inhibitors, clinicians should consider the patient’s VTE risk profile, including increased age and cardiovascular risk factors such as obesity, diabetes, hypertension, hyperlipidemia, and smoking. In each patient, the weight of each individual risk factor should be considered, as well as the cumulative weight of the risk factors combined [13].

Owing to the limitations of the current data, we believe that it is too early to conclude that the second-generation JAK inhibitors have a lower VTE risk and caution in patients with VTE risk factors is required.

Declaration of interests

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers of this manuscript have no relevant financial or other relationships to disclose.

Author contributions

Ameen Jubber, Jonathan Woodward, Hasan Tahir, and Arumugam Moorthy wrote or contributed to the writing of the manuscript.

Additional information

Funding

This paper was not funded.