Introduction

The use of early, aggressive treatment strategies, as well as the development of biologic therapies, has substantially improved outcomes in patients with rheumatoid arthritis (RA) [1]. Although there is no cure for RA [1], potential therapies include the Janus kinase (JAK) inhibitors, which target various cytokine signalling pathways and suppress the activation and proliferation of inflammatory cells [2]. Two JAK inhibitors are currently approved in the USA, Europe and Asia for treatment of RA in adults: tofacitinib (a pan-JAK inhibitor) and baricitinib (a JAK1 and JAK2 selective inhibitor) [3,4,5,6,7,8]. A third JAK inhibitor, peficitinib (a pan-JAK inhibitor), is approved in Japan [9]. Peficitinib, an orally bioavailable, once-daily JAK inhibitor that inhibits JAK1, JAK3, and TYK2 with greater specificity than JAK2, may modulate crucial cytokine receptor pathways in RA [10, 11].

The clinical efficacy and safety of peficitinib have been demonstrated in phase II, randomized, double-blind, placebo-controlled trials of patients with moderate-to-severe RA [12,13,14]. In a Japanese phase II trial, peficitinib 50 mg, 100 mg, and 150 mg monotherapy for 12 weeks provided dose-dependent and significantly higher American College of Rheumatology 20% criteria (ACR20) response rates compared with placebo and had an acceptable safety profile [14]. Two global, phase IIb, randomized, double-blind, placebo-controlled trials assessed peficitinib for 12 weeks. In the first trial [‘without methotrexate (MTX)’], peficitinib 50 mg, 100 mg, or 150 mg once-daily plus limited csDMARDs provided significant, dose-dependent improvements in ACR20 compared with placebo [12]. In the second trial (‘with MTX’), significant improvements in ACR20 response with peficitinib plus MTX were observed with peficitinib 50 mg compared with placebo plus MTX [13]. In both phase IIb trials, peficitinib was well tolerated over the 12-week treatment periods.

Patients enrolled in this long-term extension study had previously completed treatment with either placebo or peficitinib in a global phase IIb, double-blind trial of adults with moderate-to-severe RA. In the ‘without MTX’ trial, 289 patients with long-term refractory disease and inadequate response or intolerance to csDMARDs received placebo or peficitinib 25–150 mg once daily in combination with limited csDMARDs [12]. In the ‘with MTX’ trial, 378 patients with an inadequate response to MTX received placebo or peficitinib 25–150 mg once daily added to existing MTX treatment [13]. After 12 weeks of double-blind treatment in either study, all eligible patients entered the extension study and received open-label peficitinib 100 mg once daily for 2 years. The primary objective was to assess the long-term tolerability and safety profile of peficitinib in adults with moderate-to-severe RA.

Methods

Design and Patients

This global open-label, single-arm, non-comparative, long-term extension study of peficitinib in adults with RA included patients who had completed one of two double-blind, placebo-controlled, 12-week phase IIb trials [12, 13]. The long-term extension study was conducted in 51 centres in the USA, Poland, Hungary, Colombia, Czech Republic, Mexico, Bulgaria, and Belgium (NCT01711814). Patients aged ≥ 18 years were eligible for the open-label extension study if they continued to fulfil the inclusion criteria for the double-blind trials [12, 13].

In the double-blind trials, patients received placebo or peficitinib 25–150 mg once daily in combination with limited csDMARDs (‘without MTX’; NCT01565655) or in combination with MTX (‘with MTX’; NCT01554696). This extension study included patients from both trials. All patients from both clinical trials were continued on 100 mg peficitinib daily for 105 weeks. There were 13 visits between the extension study baseline and week 105, with the end-of-study visit 30 days later.

An Institutional Review Board/Independent Ethics Committee-approved written informed consent form was obtained from each patient or from a legally authorized representative prior to the initiation of any study-specific procedures. This study was conducted in compliance with the Declaration of Helsinki, Good Clinical Practice, International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use guidelines, the EU Clinical Trials Directive, and applicable laws and regulations. The central IRB for the USA was Quorum IRB. Details of individual IRBs are included in Supplement 7.

Discontinuation Criteria

The patient discontinuation criteria, including a failure to achieve either low disease activity or remission, are shown in Supplement 1. Patients were free to withdraw from the study treatment and/or study for any reason and at any time without giving a reason. Patients were discontinued from the study if they withdrew consent, were non-compliant with the protocol in a way that might affect evaluation of treatment, or had an ongoing adverse event (AE) or an unresolved laboratory result. If a patient discontinued because of an AE, the event was followed until it resolved [15].

There were six protocol amendments during the study, including the criteria for withdrawals that were adopted to be consistent with EULAR and ACR treatment guidelines [16]. Amendment 4 (2 April 2013) changed the discontinuation criterion from no ACR20 response by week 13 to patients without a 20% improvement in tender and swollen joints. Protocol amendment 6 (24 March 2015) superseded amendment 4 and mandated patients to discontinue from the study for low disease activity or remission at their most recent visit, or if at their most recent visit they had recurrence of moderate or high disease activity for two consecutive visits after a previous response with low disease activity or remission.

Objectives

The primary objective was to evaluate 2-year safety by assessing adverse events (AEs) and clinical laboratory evaluations, respectively. The safety endpoints were AEs, serious AEs (SAEs), AEs of special interest, pregnancy outcomes, and laboratory assessments.

All AEs were coded using the Medical Dictionary for Regulatory Activities version 14.0; investigators designated the causal relationship between study treatment and AEs.

The exploratory objective was to evaluate 2-year effectiveness. The endpoints were ACR20/50/70 response, DAS28 (CRP) and DAS28 (ESR) scores, DAS28 (CRP) remission, Simplified Disease Activity Index (SDAI), and Clinical Disease Activity Index (CDAI).

Statistical Analyses

All safety analyses were descriptive. Values were tabulated and provided as number and percentage reporting the event. Rates of AEs [per 100 patient-years (PYs)] were also calculated as the number of AEs for all patients in the treatment cohort divided by the total PYs in the treatment cohort, multiplied by 100.

The full analysis set included all enrolled patients who had completed a double-blind trial and who received at least one dose of open-label peficitinib. The double-blind trials have been previously published [12, 13].

ACR20/50/70 responses and the maintenance of ACR20 response in patients who achieved an ACR20 response at week 12 in the double-blind trials were analysed using observed data and non-responder imputation.

As the control group is not available in the long-term extension study, Kaplan-Meier analyses were performed. Survival plots were based on prior treatment arms and included time to withdrawal from the study due to AEs or lack of efficacy (LOE) (including failure to meet ACR20 and failure to achieve LDAS) for all patients, patients from the preceding phase IIb trial with MTX add-on, and patients from the preceding phase IIb trial without MTX add-on. For each patient who withdrew from the study because of AEs or LOE (i.e., event = 1), the time to withdrawal from the study due to AEs or LOE was defined as the number of days from the date of initial dose of study drug to the date of withdrawal from the study due to AEs or LOE and calculated as: date of withdrawal from the study due to AEs or LOE—date of initial dose of study drug + 1. For each patient who did not withdraw from the study because of AEs or LOE (i.e., event = 0), the event time (in days) was censored at the date of last dose of study drug.

Results

Patients

A total of 611 patients entered the long-term extension study and received ≥ 1 dose(s) of peficitinib 100 mg and were included in the FAS (Fig. 1). Baseline demographic and disease characteristics were similar across the study arms in the double-blind trial (Table 1). A total of 319 (52.2%) patients completed the study; among the 292 patients who discontinued, the main reasons were withdrawal of patient consent (n = 96; 32.9%), failure to achieve low disease activity (n = 62; 21.2%), AE not including death (n = 41; 14.0%), and lack of efficacy (n = 32; 11.0%).

Fig. 1
figure 1

Disposition of patients. *‘Discontinued treatment’ refers to any patient who discontinued at any time between receiving their first and last dose of peficitinib 100 mg. †Patients had not achieved either low disease activity or remission by their most recent visit, as determined by DAS28(CRP) < 3.2, DAS28(ESR) < 3.2 or SDAI ≤ 11 (if CRP was available) or CDAI ≤ 10 (if CRP was not available). Patients also met the discontinuation criteria if, after achieving low disease activity or remission, they experienced moderate or high disease activity for two consecutive visits as determined by DAS28(CRP) ≥ 3.2 (if CRP was available), DAS28(ESR) ≥ 3.2 (if ESR was available), and SDAI > 11 (if CRP was available) or CDAI > 10 (if CRP was not available). ‡Two patients discontinued treatment because of ‘miscellaneous’ reasons (one patient in the peficitinib 50 mg/100 mg group because of entering the study in error and one patient in the peficitinib 150 mg/100 mg group because of moving further from the study site). ACR20 American College of Rheumatology 20% improvement criteria, CDAI Clinical Disease Activity Index, CRP C-reactive protein, DAS28 Disease Activity Score in 28 joints, ESR erythrocyte sedimentation rate, LTE long-term extension, SDAI Simplified Disease Activity Index

Full size image
Table 1 Demographics and baseline disease characteristics
Full size table

During the study, 86.7% (n = 530) received MTX, 67.6% (n = 413) received folic acid, 27.2% (n = 166) received prednisone, and 23.9% (n = 146) received methylprednisolone. In the total population, the median (range) duration of peficitinib treatment was 720.0 days (1–771). Seven patients had a dose reduction during the study [they had previously received peficitinib 25 mg (n = 2), 50 mg (n = 3), 100 mg (n = 1), and 150 mg (n = 1)]: six patients discontinued because of AEs, and one patient fulfilled the pre-defined discontinuation criterion of absolute lymphocytopenia of < 500 cells/µl. Survival curves of patients who withdrew from the study are shown in Supplement 2.

Safety

Adverse Events

In total, there were 1813 AEs (182.4 AEs/100 PYs) (Table 2). Patients were grouped according to the dose received in the preceding phase II trial. The most frequently reported AEs per 100 PYs were upper respiratory tract infection (URTI) [n = 98 (9.9 events/100 PYs)], urinary tract infection (UTI) [n = 72 (7.2 events/100 PYs)], nasopharyngitis [n = 56 (5.6 events/100 PYs)], and bronchitis [n = 52 (5.2 events/100 PYs)]. There were no clinically meaningful differences in the frequency of any AEs between treatment groups and the number of AEs per 100 PYs was similar in patients from the ‘with MTX’ trial (373.2 events/100 PYs) and patients from the ‘without MTX’ trial (385.1 events/100 PYs). A total of 222 (36.3%) were considered by the investigator to have drug-related events, the most common being URTI (6.7%), hypercholesterolaemia (4.3%), and bacterial UTI (3.4%).

Table 2 Summary of AEs and those occurring with a frequency of ≥ 25 adverse events/100 PYs
Full size table

Serious Adverse Events

A total of 80 patients (13%) experienced an SAE; they had previously received placebo (12%), peficitinib 25 mg (10%), 50 mg (11%), 100 mg (20%), and 150 mg (12%) (Supplement 3). The most frequently reported SAEs were pregnancy [n = 8 (1%)], RA [n = 6 (1%)], ovarian cyst [n = 4 (1%)], UTI [n = 3 (< 1%)], anaemia [n = 3 (< 1%)], and acute renal failure [n = 3 (< 1%)]. Two patients who experienced SAEs died, one from a road traffic accident and the other a cardiac arrest. Incidences of pulmonary embolism and chest pain [n = 1 (< 1%)], pulmonary thrombosis and thrombosis [n = 1 (< 1%)], and deep vein thrombosis [n = 1 (< 1%)] were not considered by the investigator to be drug related. All events were resolved and the patients recovered.

A total of 78 patients (13%) had grade ≥ 3 AEs; they had previously received placebo (11%), peficitinib 25 mg (9%), 50 mg (12%), 100 mg (18%), and 150 mg (13%). The most frequently reported grade ≥ 3 AEs were RA [n = 5 (1%)], anaemia [n = 3 (< 1%)], appendicitis [n = 3 (< 1%)], headache [n = 3 (< 1%)], and UTI [n = 3 (< 1%)] (Table 3).

Table 3 Grade ≥ 3 AEs
Full size table

A total of 43 patients (7%) discontinued study treatment because of an AE. Of these, 9 patients had received placebo, and 9, 8, 11, and 6 patients had received peficitinib 25 mg, 50 mg, 100 mg, and 150 mg in the prior phase IIb clinical trials. Overall, the most frequent AEs leading to treatment discontinuation were pregnancy [n = 5 (< 1%)], blood creatine phosphokinase (CPK) increase > 1.5 × upper limit of normal [ULN; n = 4 (< 1%)], anaemia [n = 3 (< 1%)], and headache [n = 3 (< 1%)].

Adverse Events of Special Interest

The incidence of serious infections was 2.7 events per 100 PYs (Table 3). UTI had the highest incidence (0.4 events/100 PYs), followed by appendicitis (0.3 events/100 PYs). Diverticulitis, gastroenteritis, and pneumonia each had two occurrences (0.2 events/100 PYs), and there was one event each of bronchitis, bronchitis bacterial, cellulitis staphylococcal, chikungunya virus infection, clostridial infection, herpes zoster, herpes zoster ophthalmic, influenza, nasopharyngitis, pneumonia bacterial, pneumonia primary atypical, pneumonia staphylococcal, URTI, and viral infection (0.1 events/100 PYs). There were no reports of tuberculosis or opportunistic infections. The incidence of herpes zoster (including herpes zoster and herpes zoster ophthalmic) was 1.6 events per 100 PYs, and the incidence of malignancies was 0.6 events per 100 PYs. There were no thrombotic events observed.

The overall incidence of AE events/100 PYs was 182.4, which included URTI (9.9), UTI (7.2), nasopharyngitis (5.6), and bronchitis (5.2). The most frequent AEs were infections and infestations (48.4%), including upper respiratory tract (12.6%), urinary tract (8.7%), and bronchitis (6.9%). The incidence of AEs was similar in patients with and without concomitant MTX treatment.

Clinical Laboratory Evaluations

The reported incidence of liver enzyme changes was 2.3 events per 100 PYs. Hepatic enzyme increase had the highest incidence (0.6 events/100 PYs) followed by gamma-glutamyltransferase increase (0.4 events/100 PYs). Alanine aminotransferase (ALT) increase, abnormal liver function test, and transaminases increase each had three occurrences (0.3 events/100 PYs), and there were two occurrences of aspartate aminotransferase (AST) increase (0.2 events/100 PYs) and one occurrence of hyperbilirubinaemia and liver injury (0.1 events/100 PYs).

The shift from baseline to week 105 for clinical laboratory values was similar to that observed in both the phase IIb trials (‘with MTX’ and ‘without MTX’) (Table 4). Compared with baseline measurements, there were reductions in absolute neutrophil and lymphocyte counts and increases in fasting cholesterol and high-density lipoprotein levels at week 105 (Table 5).

Table 4 Summary of events of special interest and clinical laboratory values
Full size table
Table 5 Summary of shifts from baseline to week 105 for haematology and fasting lipid profile values
Full size table

Pregnancy Outcomes

Eight patients (1%) became pregnant during the study. Four reported abortions (one spontaneous, one induced, and two with no further information). Of the four pregnancies, one had a normal vaginal delivery, one had a caesarean delivery, and the outcome in two patients was lost to follow-up and was not reported. The study drug was discontinued in five patients with pregnancy.

Effectiveness

From baseline to week 105, the ACR20/50/70 response was 44% (n = 269), 32% (n = 195), and 21% (n = 128), respectively (Fig. 2a). In the extension study, comparable ACR20 response rates were observed at baseline and week 105 in all patients treated with peficitinib (Supplement 4).

Fig. 2
figure 2

a ACR20/50/70 responses, b maintenance of ACR20 responses, and c DAS28(CRP) and DAS28(ESR) over time. All data are relative to phase II trial baselines and were analysed using observed data. a ACR20/50/70 responses [percentages are calculated based on total number of patients in the FAS (N = 611)]. b Maintenance of ACR20 responses in patients who achieved an ACR20 response in their previous respective phase II trial. c Mean (± standard deviation) change from baseline DAS28(CRP) and DAS28(ESR). ACR20 American College of Rheumatology criteria for 20% improvement, ACR50 ACR 50% response, ACR70 ACR 70% response, BL baseline, DAS28(CRP) disease activity score in 28 joints using C-reactive protein DAS28(ESR) disease activity score in 28 joints using erythrocyte sedimentation rate

Full size image

ACR20 response was observed in 302 (49.4%) patients at week 12 in the double-blind trials, and all of those patients continued in the extension study, of which 171 (57%) patients had an ACR20 response at week 105 when analysed using non-responder imputation (Fig. 2b).

DAS28(CRP) and DAS28(ESR) decreased during the first 9 weeks of the extension study relative to the start of the double-blind trials and then remained stable to week 105 (Fig. 2c). When analysed by previous treatment group, no differences were observed between treatment groups.

At baseline and week 105 in the extension study, 128 (20.9%) and 135 (22.0%) patients at baseline and week 105, respectively, achieved DAS28(CRP) < 2.6. The number of patients achieving DAS28(CRP) < 2.6 was similar when patients were stratified by previous treatment group.

CDAI and SDAI statuses at baseline and week 105 are shown in Supplements 5 and 6, respectively. At week 105, the proportions of patients achieving remission across the peficitinib groups were 95/323, 29.4% (SDAI ≤ 3.3) and 103/324, 31.8% (CDAI ≤ 2.8). Similarly for LDA, the proportions were 215/323, 66.6% (SDAI ≤ 11) and 215/324, 66.4% (CDAI ≤ 10).

Discussion

This phase IIb, open-label, single-arm global study showed that peficitinib 100 mg once daily had a safety profile that was consistent with JAK inhibitors licensed for the treatment of RA and sustained effectiveness over 2 years in adults with moderate-to-severe RA in individuals who remained in the study.

One unusual finding was the number of pregnancies that occurred in the extension study. There are no studies of JAK inhibitors in pregnant women, and licensed JAK inhibitors are contraindicated during pregnancy [1,2,3,4,5]. Women comprised 82.5% of this extension study population, and the protocol stated that women of childbearing potential had to be using highly effective contraception consisting of two forms of birth control. Eight (1%) patients became pregnant, and peficitinib treatment was stopped in five patients. Most of the pregnancies occurred towards the end of the study period, and it is possible that women became less diligent with contraception in the latter stages of the study. In this extension study, two pregnant women were lost to follow-up with ongoing pregnancy, but had stopped treatment before discontinuing; the known pregnancy outcomes were four spontaneous/elective abortions and two live births (both women had stopped study treatment).

Long-term RA outcomes based on various endpoints were evaluated in an exploratory analysis of the effectiveness of peficitinib from week 12 in the double-blind trials to week 105 of the extension study. Overall, 57% of patients who achieved an ACR20 response at week 12 continued with sustained and similar ACR20 responses at the end of the extension study. However, it should be noted that the main limitation of the study was the low completion rate (52%), which was influenced by the rate of patient withdrawals after the protocol amendment to be consistent with EULAR [15] and ACR treatment guidelines [16] was adopted. The latest amendment mandated that patients who did not achieve a low disease activity within 6 months of initiation of peficitinib therapy should be discontinued from the study. The amendment specified that dose intensification beyond 100 mg of peficitinib once daily in non-responders was not allowed. This amendment is one of the main reasons for the high level of discontinuation in this study. The study protocol was amended following a decision not to further investigate peficitinib outside of Japan and to ease the burden of study procedures on investigators and patients.

Additionally, patients who received placebo during the double-blind trials and switched to peficitinib during the extension trial showed greater improvements in ACR20 response to peficitinib compared with patients who received peficitinib during the double-blind trials. This difference was possibly due to higher disease activity in placebo-treated patients, in contrast to the peficitinib-treated patients, who had reported improvements during the double-blind trials or the open-label nature of the extension trial. Furthermore, although DAS28(CRP) and DAS28(ESR) decreased during the first 9 weeks of the extension study, as expected, responders were more likely to continue with treatment than were non-responders. Further limitations were the open-label design and the potential selection bias in the extension study for patients who responded to treatment and had acceptable safety during the double-blind trial.

Conclusion

This long-term study of open-label peficitinib in adults with moderate-to-severe RA showed that treatment had a favourable benefit-risk ratio, and although half of the population discontinued treatment during the 2-year study, among patients who completed, ACR20 responses at week 12 were sustained up to week 105.