2. Overall efficacy of JAK inhibitors

Currently, the term tsDMARDs refers only to JAK inhibitors (JAKis), with five treatments, already available or pending. Baricitinib, filgotinib, tofacitinib, and upadacitinib have shown their efficacy in different subpopulations of adult patients with RA: early active RA, and active RA with an inadequate response to csDMARDs and/or bDMARDs. Peficitinib has only been approved in Japan where clinical trials revealed significant efficacy.

In all studies, the response was obtained rapidly and appeared to plateau after week 12 with a treatment benefit remaining stable over the study duration. JAKis combined with a csDMARD proved effective in patients refractory to bDMARDs, regardless of the number and type of prior bDMARDs used [3–6]. JAKis were also efficient in monotherapy, mostly in early RA, or when combined with csDMARD other than MTX [7].

Recent head-to-head trials evaluating the efficacy and tolerance of baricitinib [8], upadacitinib [9] and filgotinib [10], have shown the superiority of these drugs over adalimumab (that represents the standard-of-care biologic DMARD) for some endpoints. Tofacitinib demonstrated its non-inferiority with adalimumab [11]. In all studies, JAKis and adalimumab were compared in combination with methotrexate. Interestingly, these studies provide important information about patients rescued from adalimumab to JAKi or patients entering the long-term extension study with a potential switch from adalimumab to JAKi [12,13]. Patients who switched improved disease control. Because the switch from adalimumab to JAKis occurred without a washout period, and because adalimumab has a mean circulating half-life of approximately 14 days, patients would therefore have received several weeks of dual TNF and JAK inhibition after treatment change. The observation of added efficacy, without apparent acute safety signal during the weeks after the switch, is reassuring and compatible in clinical practice with a rapid introduction of the JAKis in bDMARDs inadequate responders.

The summary of product characteristic of baricitinib mentions the possibility of reducing the dose of baricitinib to 2 mg daily for patients in whom sustained control of disease activity has been achieved with 4 mg daily [14]. The majority of patients who lost their LDA or remission status after dose reduction could regain disease control after the dose was returned to 4 mg. This possibility is only mentioned for tofacitinib in the event of severe kidney failure or moderate liver failure, with no information on efficacy. To the best of our knowledge, tapering of filgotinib, upadacitinib, and peficitinib have not been described in any studies.

3. Current concerns about the safety of JAK inhibitors

The practical management of the various adverse effects (severe or not) observed on JAKi therapy has been recently described by a French group [15]. To date, compared to bDMARDs, JAKi carry a further caution of herpes zoster and thrombosis. While the management of herpes zoster should be facilitated in the future by worldwide availability of non-live recombinant vaccine, concerns persist about the cardiovascular risk.

Several randomized control trials (RCTs) evaluating the efficacy and safety of JAKi have reported a risk of venous thromboembolism (VTE) and pulmonary embolism. The number of events (in a selected population) was low, but in three of these RCTs, VTE occurred only in patients receiving JAKis [16]. Data compilation did not reveal a significant change in the risks of all cardiovascular events (CVEs), major adverse cardiovascular events (MACEs) and VTEs [17]. Data from tofacitinib and baricitinib long-term extension studies were also reassuring [18,19]. However, recently, tofacitinib 10 mg twice daily has been associated with increased risks of pulmonary embolism (PE) and death in the post marketing ORAL Surveillance safety trial comparing TNF blockers (etanercept or adalimumab) and tofacitinib in RA patients who were at least 50 years of age and older and had at least one cardiovascular risk factor (see most recent summary of product characteristic). Compared with TNF blockers, the hazard ratio (HR) for PE with tofacitinib 10 mg twice daily was 5.96 (1.75–20.33), and for 5 mg twice daily the HR was 2.99 (0.81–11.06). The HR for death was 3.28 (1.55–6.95) for tofacitinib 10 mg twice daily and 2.11 (0.96–4.67) for tofacitinib 5 mg twice daily, versus TNF blockers. The mortality was mainly due to cardiovascular events, infections, and malignancies. Final data of this study, and results in the same population with other JAKis will be important to better determine the risk of all CVEs. So far, long-term safety data for up to 9.5 years for tofacitinib and up to 5.5 years for baricitinib are available, but not for upadacitinib and filgotinib.

4. Expert opinion

JAK inhibitors are particularly attractive as they offer the advantage of targeting multiple cytokines at once and hence may have a broad spectrum of activity compared with neutralizing antibodies against single cytokines or specific immune cells, even if JAKis, in contrast to bDMARDs, do not completely block their pathways. Thus, JAKi could theoretically provide new benefits as compared with current targeted DMARDs. RCTs have demonstrated that JAKi monotherapy or in combination with MTX can lead to significant improvements of RA, including in those patients who failed to respond to previous bDMARDs. Although there are no studies to state this with certainty, their rapid onset of action could prevent the use of steroids as bridging therapy. Interestingly, in patients who have an inadequate response to MTX, baricitinib, upadacitinib, and filgotinib are associated with significant clinical improvements as compared with adalimumab, all in combination with MTX.

The future place of JAKi in the management of RA patients will depend on several factors. The first overarching principles of last EULAR recommendations [2] states that ‘treatment of patients with RA should aim at the best care and must be based on a shared decision between the patients and the rheumatologist.’ Since bDMARDs and JAK inhibitors have on average similar efficacy, patient preferences may one of the main drivers of treatment choice. Some studies showed that most patients preferred the oral route of administration over parenteral routes as well as drugs that do not have to be combined with MTX [20]. Most JAK inhibitors available in RA offer these advantages. The clinical benefits of JAKis monotherapy need, however, to be confirmed, especially in populations with long-term established RA. Whether JAKi monotherapy may be able to inhibit structural joint damage should also be confirmed. To date, these data are mostly available in csDMARDs-naive patients [21–23]. Furthermore, data from RA-BEAM [8], SELECT-COMPARE [9], and FINCH 1 [10] trials suggest that baricitinib, upadacitinib, and filgotinib could be more effective than adalimumab, all in combination with MTX. However, since the fifth EULAR overarching principles [2] states that ‘RA incurs high individual, medical and societal costs, all of which should be considered in its management by the treating rheumatologist,’ some rheumatologists, in specific countries, favor biosimilars of infliximab, etanercept, and adalimumab. But, several new information may influence prescribing decisions in favor of JAKis in the next few years: availability of additional long-term safety data, notably with recent drugs such as upadacitinib and filgotinib; a similar or lower price of JAKis compared with biosimilars; the provision of strategy trials that will provide better insights on the best use of JAKis; and extension indications in various pathologies, such as systemic lupus, spondyloarthritis, and inflammatory bowel diseases.

Strategies that simultaneously target different pathways involved in the pathogenesis of RA may enhance treatment responses in patients with RA. JAKis inhibit many of the pro-inflammatory cytokines involved in the pathogenesis of RA but does not directly block signaling downstream of TNF, even if an indirect effect on TNF production is likely to occur. Data from the association of JAK inhibitors with bDMARDs are limited. However, combination of tofacitinib with various biologics has been reported in a sample of five RA patients and one psoriatic arthritis patient [24]. No serious adverse effects (SAEs) were reported over a mean of approximately 11 months of therapy. The clinical improvement was mild but noticeable in these refractory RA patients. Furthermore, clinical outcomes in patients whose treatment was changed from adalimumab to JAKis have recently been published [12,13]. As described above, within the first few weeks following the change in treatment, patients may have received dual inhibition of TNF and JAK. The observed clinical improvements, without any increase of SAEs, may reflect a benefit provided by the combination. Altogether, these warrant further investigation into combination therapy with JAKis and biologics when treating refractory RA, and toxicity should not preclude further investigation. Interestingly, the last EULAR recommendations [2] mentioned in its research agenda this topic through two questions: ‘How safe and efficacious is the combination of a JAKi with a bDMARD, such as a TNF inhibitor?’ and ‘For active patients with RA who have failed multiple drugs, are there combinations that may be more successful such as JAKi with bDMARD?’

Finally, these drugs, known to increase the risk of infection, and notably herpes zoster and simplex infection, have been proposed as therapy of COVID-19 infection [25]. Several trials have been registered on ClinicalTrial.gov (3 with baricitinib, and 1 with tofacitinib, 12 April 2020). Data from these studies will provide important information about the safety of these drugs in particular frail patients and will be informative for their use in regular indications.