sign in sign up
<<
Home , Janus kinase inhibitor, TNF inhibitor, Tofacitinib

Review The mechanism of action of – an oral inhibitor for the treatment of J.A. Hodge1, T.T. Kawabata2, S. Krishnaswami2, J.D. Clark3, J.-B. Telliez3, M.E. Dowty3, S. Menon2, M. Lamba2, S. Zwillich2

1Pfizer Inc, New York, NY, USA ABSTRACT flammatory and degradative 2Pfizer Inc, Groton, CT, USA Rheumatoid arthritis (RA) is a chron- mediators, and subsequent joint dam- 3 Inc, Cambridge, MA, USA ic inflammatory age (1, 2). Immune and inflammatory Jennifer A. Hodge, PhD characterised by infiltration of immune responses are the driving forces in RA Thomas T. Kawabata, PhD cells into the affected synovium, release and transform the synovial membrane Sriram Krishnaswami, PhD of inflammatory cytokines and degra- into an inflammatory tissue capable of James D. Clark, PhD Jean-Baptiste Telliez, PhD dative mediators, and subsequent joint invading and destroying adjacent carti- Martin E. Dowty, PhD damage. Both innate and adaptive arms lage and bone (3, 4). Sujatha Menon, PhD of the immune response play a role, Manisha Lamba, PhD with activation of immune cells lead- Cytokines: regulators of immune Samuel Zwillich, MD ing to dysregulated expression of in- and inflammatory responses Please address correspondence to: flammatory cytokines. Cytokines work Numerous cytokines, involved in both Jennifer A. Hodge, within a complex regulatory network in innate and adaptive immunity, are im- Pfizer Inc., RA, signalling through different intra- plicated in the pathogenesis of RA (5). 235 E 42nd Street, cellular kinase pathways to modulate 10017 New York, NY, USA. Cytokines are small protein messengers E-mail: [email protected] recruitment, activation and function of that mediate communication between Received on June 10, 2015; accepted in immune cells and other leukocytes. As cells (5). Cytokines regulate a variety revised form on October 16, 2015. our understanding of RA has advanced, of bodily processes, including haemat- Clin Exp Rheumatol 2016; 34: 318-328. intracellular signalling pathways such opoiesis, and are particularly important as Janus kinase (JAK) pathways have in the regulation of immunity and in- © Copyright Clinical and Experimental 2016. emerged as key hubs in the flammation (6). Cytokines bind to cell- network and, therefore, important as surface receptors to initiate a cascade of therapeutic targets. Tofacitinib is an signalling events that transmit informa- Key words: tofacitinib, rheumatoid oral JAK inhibitor for the treatment tion intracellularly and coordinate the arthritis, Janus kinase, cytokines, of RA. Tofacitinib is a targeted small cellular response (7). Type I and type II immune response molecule, and an innovative advance in cytokine receptors lack intrinsic kinase RA therapy, which modulates cytokines activity and rely on associated tyrosine critical to the progression of immune kinases, such as Janus kinases (JAKs), and inflammatory responses. Herein for intracellular signalling (8). we describe the mechanism of action of Cytokines are involved in each step of tofacitinib and the impact of JAK inhi- the pathogenesis of RA. They promote bition on the immune and inflammatory autoimmunity, maintain chronic inflam- responses in RA. matory synovitis and drive the destruc- tion of joint tissue (2). Key cytokines Introduction in RA pathogenesis include: Rheumatoid arthritis: a chronic (IFN)α, IFNβ; interleukin (IL)-1, IL-6, inflammatory disease characterised IL-7, IL-10, IL-12, IL-15, IL-17, IL-18, by a dysregulated and autoimmune IL-21, IL-23; transforming growth fac- response tor (TGF)-β; and tumour necrosis factor Competing interests: the studies described Rheumatoid arthritis (RA) is a chronic (TNF) (Fig. 1a) (2, 9). A subset of these in this article were sponsored by Pfizer Inc. inflammatory, systemic autoimmune cytokines, important to both innate and Editorial assistance was provided, disease that damages the joints of the adaptive immune responses in RA, uti- under the direction of the authors, by Jason Gardner of Complete Medical body. An inflamed synovium is the lise JAK pathways to transmit signals. Communications and was funded by Pfizer hallmark of RA, characterised by syno- These cytokines include: IFNα, IFNβ, Inc. All authors are employees and vial intimal lining hyperplasia with in- IL-6, IL‑7, IL‑10, IL-12, IL-15, IL-21 shareholders of Pfizer Inc. filtration of immune cells, release of in- and IL-23 (Fig. 1a) (2, 8, 9).

318 Mechanism of action of tofacitinib / J.A. Hodge et al. REVIEW

most specific, associating with only the common γ-chain (γc) receptor subunit and JAK1. The γc subunit pairs with a variety of other subunits to form the re- ceptors for IL-2, IL-4, IL-7, IL-9, IL-15 and IL-21 (8, 11). JAK2 is associated with many receptors, including those for , , IFNg, IL-3, IL-5, growth hormone and granulocyte/macrophage colony stimu- lating factor (GM-CSF) (8, 11). It fre- quently associates with itself (JAK2/ JAK2). JAK1 associates with the re- ceptors for IFNs and IL-10-related cy- tokines, γc cytokines, and IL-6, as well as other cytokine receptors containing the gp130 subunit (8, 11). It forms pairs with any of the three other JAKs. Final- ly, TYK2 transmits signalling by type I IFNs (IFNα, IFNβ), IL-12 and IL-23, amongst others (8, 11). Binding of a cytokine to its receptor activates the receptor-associated JAKs (12). The activated JAKs phosphoryl- ate specific tyrosine residues in the cytoplasmic domains of the subunits, which then act as docking sites for Signal Transducer and Activator of Transcription (STAT) pro- teins (12, 13). The STAT family of tran- scription factors consists of seven pro- teins: STAT1, STAT2, STAT3, STAT4, STAT5a, STAT5b and STAT6 (12). Af- ter docking to tyrosine-phosphorylated cytokine receptor subunits, the STATs themselves are, in turn, tyrosine-phos- Fig. 1. Key cytokines in the pathogenesis of RA and cytokine signalling through JAK/STAT combinations. phorylated by the receptor‑associated A. Key cytokines in the pathogenesis of RA. An important subset of proinflammatory cytokines signal JAKs (12). Phosphorylated STATs then through JAK pathways in RA. *Type II cytokine receptors such as those for the gp130 subunit sharing dissociate from the receptor subunits, receptors IL-6 and IL-11 as well as IL-10, IL-19, IL-20 and IL-22 mainly signal through JAK1, but also associate with JAK2 and TYK2. B. Cytokine signalling through JAK/STAT combinations. Type I and II dimerise with each other, and translo- cytokine receptors lack intrinsic enzymatic activity and utilise associated JAK combinations to signal. cate to the . Once in the nu- *γ-chain cytokines: IL-2, IL-4, IL-7, IL-9, IL-15 and IL-21. ¥Type II cytokine receptors such as those cleus, STATs function as transcription for the gp130 subunit sharing receptors IL-6 and IL-11 as well as IL-10, IL-19, IL-20 and IL-22 mainly factors to regulate gene transcription signal through JAK1, but also associate with JAK2 and TYK2. EPO: Erythropoietin; GM-CSF: Granulocyte/macrophage colony stimulating factor; IFN: Interferon; (12). IL: Interleukin; JAK: Janus kinase; RA: Rheumatoid arthritis; STAT: Signal Transducer and Activator In RA, B cells, T cells, macrophages of Transcription; TGF: Transforming growth factor; TPP: Thrombopoietin; TYK: Tyrosine kinase. and other leukocytes infiltrate the syn- ovium in response to pro-inflammatory JAK pathways: perpetuating Each JAK protein has specificity for a cytokines and chemokines, leading to a chronic cycle of different set of cytokine receptors; the inflammation and tissue destruction The JAK family consists of four non- function of the JAK protein is thereby (Fig. 2). Cytokine signalling via JAK receptor protein tyrosine kinases: linked to the function of the cytokines pathways leads to further induction of JAK1, JAK2, JAK3 and tyrosine kinase that bind the receptors (Fig. 1b) (8, 11). inflammatory gene expression, which 2 (TYK2) (10). JAK pathways are nor- Each cytokine receptor requires at least continues the loop of inflammatory sig- mally involved in growth, survival, de- two associated JAKs in order to signal. nalling. Inhibiting cytokine signalling velopment and differentiation of a vari- JAKs may work in pairs of identical by inhibiting the JAK pathways may, ety of cells, but are crucially important JAKs (e.g. JAK2/JAK2) or of different therefore, interrupt the cycle of leuko- for immune and haematopoietic cells. JAKs (e.g. JAK1/JAK3). JAK3 is the cyte recruitment, activation and pro-in-

319 REVIEW Mechanism of action of tofacitinib / J.A. Hodge et al.

Fig. 2. Perpetuating the chronic cycle of inflammation in RA: JAK pathways. In response to chemokines and cytokines, B cells, T cells, macrophages and other leukocytes infiltrate the synovium (A). These activated immune cells produce cytokines (B) which signal through the JAK pathways (C) leading to production of new pro-inflam- matory mediators such as cyto- kines (D) propagating a cycle of chronic inflammation. JAK: Janus kinase; RA: rheumatoid arthritis.

flammatory cytokine expression at sites vation of STATs, and thus the activation Tofacitinib: non-clinical studies of inflammation (5, 14). of gene transcription (Fig. 3). This leads Previous studies have established the to decreased cytokine production and effects of tofacitinib on murine innate Review modulation of the immune response. and adaptive immune responses (16). Tofacitinib: an oral JAK inhibitor Tofacitinib is a potent inhibitor of the Specifically, tofacitinib inhibited the for the treatment of RA JAK family of kinases with a high acute lipopolysaccharide-induced in- Tofacitinib is an oral JAK inhibitor for degree of selectivity against other ki- flammatory response in a murine mod- the treatment of RA. It is a targeted nases in the human kinome. With in el of innate immunity which is depend- synthetic small molecule (molecular vitro kinase assays, tofacitinib inhibits ent upon IFNγ and STAT1. Tofacitinib weight 312.4 Da; 504.5 for the citrate JAK1, JAK2, JAK3 and, to a lesser ex- dosed 5 mg/kg inhibited TNF and IL-6 salt), not a biologic. Unlike targeted bi- tent, TYK2. In cellular settings, where production, whilst the production of the ologic therapies, which are directed at JAKs signal in pairs, tofacitinib prefer- anti- IL-10 was extracellular targets such as individual entially inhibits signalling by cytokine enhanced. These results were inter- soluble cytokines (e.g. TNF), cytokine receptors associated with JAK3 and/or preted as consistent with inhibition of receptors (IL-1R, IL-6R) or other cell- JAK1 with functional selectivity over IFNγ signalling by tofacitinib through surface receptors (CD20, CD80, and receptors that signal via pairs of JAK2 blockade of JAK1. In studies of the CD86), tofacitinib works at the intra- (15). Specifically, in human whole adaptive immune response, specifically cellular level. Tofacitinib possesses blood cellular studies, tofacitinib po- the differentiation of T helper cells, to- high in vitro passive permeability prop- tently inhibits IL-15-induced phospho- facitinib blocked the differentiation of erties consistent with intracellular en- rylation of STAT5 with a half maxi- type 1 T helper (Th1) cells and type 2 T try by transcellular diffusion. mal inhibitory concentration (IC50) of helper (Th2) cells, and interfered with Tofacitinib is a reversible, competitive 56 nM, and IL-6-induced phosphoryla- the generation of pathogenic type 17 T inhibitor that binds to the adenosine tion of STAT1 and STAT3 with IC50s helper (Th17) cells (16). triphosphate (ATP) binding site in the of 54 and 367 nM, respectively (15). These results were extended in stud- catalytic cleft of the kinase domain of Both IL-15 and IL-6 are dependent on ies of tofacitinib in vivo. In a murine JAK. The structure of tofacitinib mim- JAK1, and IL-15 is also dependent on collagen-induced arthritis model, to- ics that of ATP without the triphosphate JAK3. In contrast, tofacitinib inhibits facitinib produced significant dose- group. As a result of binding to the ATP GM-CSF-induced phosphorylation of dependent attenuation of inflammatory site, tofacitinib inhibits the phospho- STAT5, a JAK2-mediated signalling paw swelling and cell influx (16, 17) rylation and activation of JAK, thereby event, with lower potency and an IC50 when dosed before (prophylactically) preventing the phosphorylation and acti- of 1377 nM (15). or after (therapeutically) disease. Plas-

320 Mechanism of action of tofacitinib / J.A. Hodge et al. REVIEW

Fig. 3. Tofacitinib mechanism of action. Under normal conditions, binding of a cytokine to its specific cell-surface receptor causes the receptor chains to polymerise and activate the associated JAKs. (Before) Activated JAKs phosphorylate specific residues in the cytoplasmic domains of the cytokine receptor chains, which then act as docking sites for STAT proteins. Once they have docked, STATs are phosphorylated by the activated receptor-associated JAKs. Phosphorylated STATs then dissociate from the receptor chains, dimerise with each other, and translocate to the cell nucleus where they activate gene tran- scription. Tofacitinib binds in the catalytic cleft in the kinase domain of JAK. (After) This prevents activation of JAK and thus of STAT phosphorylation and translocation to the nucleus to activate gene transcription. JAK: Janus kinase; STAT: Signal Transducer and Activator of Transcription. ma cytokine (e.g. IL-6) and chemokine periments evaluated the effects of to- ≥1 TNF inhibitor (23), and those who (e.g. chemokine [C-X-C motif] ligand facitinib on the production of recep- were -naïve (24). Tofaci- 10 [CXCL10]; IFNγ induced protein tor activator of nuclear factor-kappaB tinib demonstrated greater 10 [IP-10]) mRNA and protein levels (RANK) and RANK ligand (RANKL). compared with placebo across clinical, were reduced to pre-disease levels but Tofacitinib led to a dose-dependent functional and structural measures of not completely inhibited in tofacitinib- decrease in human T disease severity (19-24). treated mice, demonstrating normalisa- RANKL production, whereas osteo- The safety profile of tofacitinib was tion of chemokine and cytokine levels. clast differentiation and function were consistent across the Phase 3 stud- Finally, tofacitinib reduced macrophage not directly affected. These data pro- ies (19-24). Common adverse events and infiltration in inflamed paw vide a mechanistic explanation for the (AEs) with tofacitinib included naso- tissue of treated mice as assessed by observed inhibition of structural joint pharyngitis, upper respiratory tract in- histology and immunohistochemistry damage by tofacitinib. fection, and diarrhoea, and (16) as well as bone resorption (18). the most common serious AEs report- Similarly, in a rat model of adjuvant- Tofacitinib: ed were serious (including induced arthritis (AIA), tofacitinib was clinical studies – overview pneumonia, cellulitis, herpes zoster, given at the peak of disease and dosed The clinical efficacy and safety of to- and urinary tract ) (25). Ma- for 1 week (18). Treatment led to a re- facitinib 5 mg and 10 mg twice daily lignancies (excluding non-melanoma duction in oedema, in osteoclast-medi- (BID) as monotherapy or in combina- skin ), and gastroin- ated bone resorption, and in monocyte/ tion with conventional synthetic dis- testinal perforations have been report- macrophage and T cell infiltration into ease-modifying anti-rheumatic drugs ed in patients receiving tofacitinib (25). the joint and surrounding tissue. The (DMARDs) for the treatment of RA Changes in safety laboratory test val- rise in plasma and/or paw tissue lev- have been reported in six Phase 3 clini- ues included increases in high-density els of several inflammatory cytokines, cal studies (19-24). The Phase 3 study lipoprotein, low-density lipoprotein, chemokines and adhesion molecules populations included patients with RA total cholesterol, serum creatinine and associated with AIA was inhibited by who previously had an inadequate re- transaminases (25). Changes in tofacitinib. sponse to either methotrexate (19, granulocyte and lymphocyte numbers To better understand the mechanism 20), ≥1 biologic DMARD or conven- and immune function are described in behind reduced bone resorption, ex- tional synthetic DMARD (21, 22), or detail below.

321 REVIEW Mechanism of action of tofacitinib / J.A. Hodge et al.

Tofacitinib: placebo control groups) and 38% of pa- unpublished observations). This moder- clinical studies – immune impact tients had moderate-severe lymphope- ate decrease in T cell counts precluded The inhibition of JAK pathways is a nia (0.5–1.5 x103/mm3) (Pfizer unpub- a definitive assessment of reversibility novel mechanism of action that results lished observations). Mean increases after drug discontinuation. in a pattern of partial inhibition of the from baseline in lymphocyte levels The reversibility of tofacitinib’s effect intracellular effects from several in- were seen with tofacitinib and adali- on T cell function was assessed in a flammatory cytokines and overall mod- mumab at Month 1, which persisted for study of stable kidney transplant recipi- ulation of the immune and inflamma- , but not tofacitinib (ap- ents (n=8) treated with tofacitinib (30 tory response. Further characterisation proximate 10% decrease from baseline mg BID) for 29 days (33). Peripheral of the impact of immunomodulation by over 12 months) (Fig. 4b). The occur- blood mononuclear cells (PBMC), pre- tofacitinib was undertaken. Doses from rence of lymphopenia did not appear to pared from blood samples collected on 1 to 30 mg BID, inclusive, and 20 mg be dose-related and the frequency of Day 1 (before tofacitinib treatment) and once daily (QD) were studied in Phase lymphopenia in the Phase 3 studies was Day 29, were stimulated in vitro with 2 of the RA development programme similar between tofacitinib and placebo allogeneic cells (mixed lymphocyte re- (26-30) and the 5 and 10 mg BID doses patients at Months 3 and 6. The moder- sponse). The proliferative response of were advanced into Phase 3 (19-24). ate decrease in lymphocyte counts pre- PBMC from Day 1 was similar to that cluded definitive assessment of revers- observed with PBMC from Day 29. ibility after drug discontinuation. Since tofacitinib is washed away dur- A dose-dependent mean decrease in ab- ing PBMC preparation, this indicates solute counts, which reached Lymphocyte subsets that tofacitinib does not have a residual a plateau within 3 months of treatment, The effects of tofacitinib on changes in effect on T cell function at 6-times the was observed in tofacitinib-treated RA circulating numbers of lymphocyte sub- approved clinical dose. patients (Fig. 4a; Pfizer unpublished sets including CD3+ cells (pan T lym- B cell counts showed dose-dependent observations). Neutrophil counts large- phocytes), CD4+ cells (helper T cells), increases in all studies (Fig. 4g; Pfizer ly remained within the normal reference CD8+ cells (cytotoxic T cells), CD19+ unpublished observations). The mecha- range throughout the duration of tofaci- cells (B cells) and CD16+/CD56+ cells nism and clinical significance of in- tinib treatment. These data are consist- (natural killer [NK] cells) were as- creases in circulating B cell numbers in ent with exposure-response modelling sessed in early studies in RA patients response to JAK inhibition are unclear. of Phase 2 data predicting a steady (Pfizer unpublished observations). Humans with JAK3 deficiencies (JAK3 state neutrophil nadir by 6 to 8 weeks Dose-response plots indicated that -/-) also have increased numbers of cir- and a median decrease from baseline of changes in CD3+, CD4+ and CD8+ counts culating B cells but are accompanied approximately 1.0 and 1.5 x 103/L for were small and variable with no consist- by severe reductions in serum immuno- tofacitinib 5 and 10 mg, respectively ent dose-response pattern across studies globulin (Ig) levels (34). With tofacitin- (Pfizer unpublished observations). Af- (Fig. 4c-e). In contrast, NK cell counts ib, only modest decreases in serum IgG, ter cessation of tofacitinib, neutrophil showed dose‑dependent decreases (Fig. IgM and IgA levels are observed in RA counts recovered in a dose-dependent 4f). Model-based characterisation of patients (see below) which may be due manner by 6 weeks (31). NK cell counts indicated that reductions to changes in disease activity rather Modest decreases in neutrophils are ex- were predicted to reach nadir (steady than a direct effect on B cell function. pected with efficacious RA treatment. state) by 8 to 10 weeks after initiation of Four Phase 2 studies evaluated serum Mean neutrophil count decreases, of a tofacitinib treatment, with approximate- immunoglobulin levels in tofacitinib- similar magnitude as those with tofaci- ly 36% and 47% peak mean reductions treated RA patients (Pfizer unpublished tinib, were also observed in adalimum- for 5 and 10 mg doses, respectively (25). observations). Decreases in median ab-treated patients in one of the Phase 3 Similar changes in subsets IgG, IgM and IgA levels (<10 to 20% studies, which included adalimumab as were observed in tofacitinib‑treated pso- change from baseline) were observed an active control arm (NCT00853385) riasis patients (32). after 12 (Studies NCT00603512 and (20), with corresponding reductions Based on the pharmacological proper- NCT00687193 in Japanese patients) or in acute phase reactants. Therefore, ties of tofacitinib, T cell-mediated im- 24 weeks (Studies NCT00413660 and changes in neutrophils may be primar- munity may be altered by inhibiting NCT00550446, global studies) of to- ily related to decreasing inflammation, cytokine signalling required for T cell facitinib treatment when compared to rather than specific to the mechanism of development and T cell function after baseline levels (Fig. 5). The median im- action of tofacitinib. antigen stimulation. T cell development munoglobulin concentrations after to- was minimally altered in most patients, facitinib treatment were within normal Lymphocytes based on the findings that circulating T reference ranges (35-37). These chang- In the Phase 3 studies, mean lympho- cell counts showed little change with es in immunoglobulin levels may be cyte levels at baseline were below the short-term treatment and decreased secondary to a decrease in generalised lower reference range (2 x103/mm3) in by less than 20% on average (CD8+ T inflammation associated with tofacitin- all groups (including adalimumab and cells) with long-term treatment (Pfizer ib treatment. It has been reported that

322 Mechanism of action of tofacitinib / J.A. Hodge et al. REVIEW

Fig. 4. Immune cells and lymphocyte subsets. A. Mean (standard error) neutrophil levels in Phase 3 Studies (0 to 12 Months) B. Dose-dependent changes in total lymphocytes were measured in Phase 3 studies (maximum duration of 12 months); lymphocyte subsets were analysed in Phase 2 stud- ies in RA patients after 6 weeks to 24 weeks (6 months) of tofacitinib treatment. Percentage change from baseline in the indicated cell type is depict- ed for the varying doses of tofacitinib (mg BID). C. CD3+ T cells; D. CD4+ T cells; E. CD8+ T cells; F. NK cells; G. B cells. Solid horizontal line within each box represents median; lower and upper horizontal lines of each box represent 1st and 3rd quartiles, respectively; whiskers represent 1.5 x inter-quartile range; lines outside the whiskers are outliers (individual data points). Grey shading represents 95% confidence interval of the median. BID: twice daily; NK: natural killer; Pbo: placebo; RA: rheumatoid arthritis.

323 REVIEW Mechanism of action of tofacitinib / J.A. Hodge et al. active RA is associated with increased immunoglobulin levels (38-40). Additional information on lymphocyte subsets was collected in a substudy of one of the open-label long-term exten- sion studies which enrolled eligible patients from Phase 2 and Phase 3 stu- dies. Subsets were assessed in appro- ximately 150 patients, most of whom had completed tofacitinib treatment in Phase 3 studies and had been treated for a median duration of approximately 22 months. Thus, these patients did not have baseline (prior to treatment initia- tion) values to formally assess magni- tude of change. Nevertheless, a compa- rison of the distribution of subset levels based on a cross‑sectional analysis in different groups of patients suggests that after 6 months of tofacitinib, NK cell counts decrease, with a return to baseline after 22 months of treatment. B cell counts remain elevated after 6 and 22 months of treatment. There were slight decreases in total T cells (10.2%), Fig. 5. Serum immunoglobulins. The effect of increasing doses of tofacitinib on serum immuno- + CD4 T cells (3.9%) and CD8+ T cells globulin levels was evaluated in four Phase 2 studies. Decreases in median IgG, IgM and IgA levels (18.1%) after 22 months of tofacitinib (<10 to 20% change from baseline) after 12 or 24 weeks of tofacitinib treatment were observed when treatment (Fig. 6). The median absolute compared to baseline levels. BID: Twice daily; Ig: Immunoglobulin. CD3+ (total), CD4+, and CD8+ T cells, B cells and NK cells in RA patients after long-term tofacitinib exposure are si- milar to those reported in healthy adult populations from Germany (41) and Switzerland (42), and contrast sharply with measures of CD3+, NK and B cells reported in patients with Severe Combi- ned Immune Deficiency (34).

Effects on immune function: vaccination The effects of tofacitinib on immune function were primarily addressed by evaluating the responses of tofacitinib-treated RA patients vacci- nated with seasonal influenza and pneu- mococcal polysaccharide . Antibody responses to non-conjugated pneumococcal polysaccharide are considered to be T cell-independent whereas the response to influenza vac- cine requires T cells to provide help to B cells to produce anti-influenza anti- bodies (T cell-dependent antibody re- sponse). Fig. 6. Lymphocyte subsets long-term. Lymphocyte subsets after short- and long-term tofacitinib treatment: pre-treatment baseline, Month 1.5, Month 6 and Month 22. Baseline, Month 1.5 and Month Two studies were conducted in RA 6 data are from Phase 2 studies. Month 22 data predominantly represent Phase 3 patients who partici- patients: one evaluated vaccine re- pated in the long-term extension study, with a median of 22 months’ tofacitinib exposure. sponses over short-term (64 days; NK: Natural killer.

324 Mechanism of action of tofacitinib / J.A. Hodge et al. REVIEW

NCT01359150) tofacitinib treatment sulted in decreased levels of circu- through an indirect mechanism involv- in patients naïve to tofacitinib, and the lating IL-6 in RA patients (Studies ing IFNβ. Specifically, TNF induces the other vaccine responses during ongo- NCT00413660 and NCT00550446). production of IFNβ, which stimulates ing long-term (median duration of ap- However, circulating levels of IL-8, the production of various chemokines, proximately 22 months; NCT00413699 CXCL10 (IP-10), TNF, or IL-10 were and tofacitinib inhibits IFNβ signal- vaccine substudy) tofacitinib therapy. variable and inconsistent. In another ling (49). Inhibition of TNF-induced In both studies tofacitinib was admin- study (NCT00976599), levels of the chemokine and IL-6 production by istered with or without background circulating chemokine, CXCL10 (IP- synovial macrophages with tofacitinib methotrexate (MTX) therapy (43-45). 10) and acute phase proteins serum am- in vitro was also mediated by inhibiting Naïve patients treated with tofacitinib yloid A (SAA) and C‑reactive protein the IFN-β feedback loop (50). were vaccinated 4 weeks later and an- (CRP) were decreased after 4 weeks of In summary, the mechanism for the de- tibody responses measured 35 days tofacitinib (10 mg BID) treatment. In crease in circulating levels of IL-6 and post-vaccination. The percentage of pa- addition, in a study that evaluated early chemokines with tofacitinib may be due tients achieving a satisfactory response changes in CRP levels, decreases were to an indirect effect on T cells that pro- to influenza vaccine (≥4-fold increase observed within 1 to 2 weeks of tofaci- duce cytokines that stimulate IL-6 pro- above baseline in antibody titers for tinib treatment (NCT00147498) (Pfizer duction and the inhibition of signalling 2/3 antigens) was similar between the unpublished observations). by the IL‑6 family of cytokines and/ tofacitinib 10 mg BID and placebo The mechanistic basis for the observed or IFNα/β through a feedback loop. In groups after 64 days of treatment. The effects with tofacitinib was evaluated in contrast, the reductions in acute phase percentage of patients achieving a satis- vitro in human cells. In vitro tofacitinib proteins CRP and SAA are likely medi- factory response to pneumococcal vac- did not directly alter IL-6 and IL-8 pro- ated by a direct effect of tofacitinib on cine (≥2-fold increase above baseline in duction by synovial fibroblasts from RA IL-6 signalling. This is supported by antibody titers for 6/12 serotypes) was patients and monocytes (46). However, the findings that IL-6 stimulation of ex- similar between tofacitinib 10 mg BID tofacitinib decreased the production of pression of SAA by fibroblast-like syn- monotherapy and MTX alone groups. IL-6 and IL-8 by synovial fibroblasts oviocytes from RA patients and human However, the tofacitinib response was and monocytes stimulated with super- hepatocytes was inhibited with in vitro diminished in patients taking MTX in natant from anti-CD3 stimulated CD4+ tofacitinib exposure (48). combination with tofacitinib (45). Sim- T cells, suggesting that decreases may, ilarly, in patients with ongoing tofaci- in part, be indirectly mediated by inhib- Tofacitinib: tinib treatment (median duration ~22 iting T cell production of cytokines that partial and reversible inhibition months) greater than 60% of patients stimulate IL-6 and IL-8 production. T Tofacitinib is a partial and reversible treated with tofacitinib (with or with- cell cytokines such as IL-17 and IFNγ inhibitor of JAKs with a short pharma- out MTX) had satisfactory responses to are known to stimulate IL-6 production, cokinetic (PK) half-life. The PK profile influenza and pneumococcal vaccines. and production of IL-17 and IFNγ by of tofacitinib in humans is character- However, pneumococcal responses CD4+ T cells from RA synovium and ised by rapid absorption and elimina- were again lower in patients receiving peripheral blood were decreased with in tion, with a time to peak concentration MTX background therapy (44). Over- vitro tofacitinib exposure. This may, in (Tmax) of approximately 1 hour and a all, the data from the vaccine stud- turn, be due to inhibition of IL-2 auto- half-life (t½) of approximately 3 hours ies support that tofacitinib treatment crine stimulation of T cells and is con- (Fig. 7a). At the 5 mg BID dose, the (short- or long-term) does not have a sistent with findings that tofacitinib de- average inhibition of preferentially in- major effect on B cell and T cell func- creases the proliferative responses and hibited cytokines is partial (50–60%) tion required for humoral immune re- cytokine production (IL‑4, IL-17, IL- and declines to 10–30% at the trough or sponses to vaccines. 22 and IFNγ) of anti-CD3 stimulated Cmin values in each dosing period. The human peripheral blood CD4+ T cells clearance mechanisms for tofacitinib Tofacitinib pharmacodynamic profile: from healthy donors (47). Decreased in humans appear to be both non-renal biomarkers of inflammation circulating IL-6 levels with tofacitinib (hepatic metabolism, 70%) and renal Many of the pro-inflammatory cyto- may also be associated with the inhibi- (30%) of the parent drug. The kines involved in the pathogenesis and tion of (OSM) stimulated metabolism of tofacitinib is primar- maintenance of RA signal through JAK IL-6 production by fibroblast-like syn- ily mediated by cytochrome P450 3A4 and, therefore, may be modulated by oviocytes (48). OSM is a member of (CYP3A4; 53%) with a minor contribu- tofacitinib therapy. Studies in vitro and the IL-6 family of cytokines produced tion from CYP2C19 (17%) (51). in vivo in RA patients have investigated by macrophages and T cells and signals Since tofacitinib is metabolised by CY- these effects. through JAK. P3A4, interaction with drugs that inhibit TNF-induced production of the or induce CYP3A4 is likely. Tofacitinib Cytokine and chemokine effects chemokines, CXCL10, CCL5 and exposure is increased approximately 2 Tofacitinib therapy for a period of CCL2 by synovial fibroblasts was in- fold when co‑administered with potent 12 to 24 weeks in Phase 2 studies re- hibited by in vitro tofacitinib exposure, CYP3A4 inhibitors (e.g. ketocona-

325 REVIEW Mechanism of action of tofacitinib / J.A. Hodge et al.

Fig. 7. Partial and reversible inhibition; biochemical targets and immune cells. A. Tofacitinib partially and reversibly inhibits multiple JAK-dependent cytokine signalling pathways. Maximum plasma concentrations (Cmax) at 5 and 10 mg BID in RA patients occur approximately 1 hour after administration of tofacitinib. At Cmax, the inhibition of the preferentially inhibited cytokines is ~80–90% for 5 mg BID and ~89–95% for 10 mg BID. At Cmin, the inhibition is ~26–43% for 5 mg BID and ~42–60% for 10 mg BID. The average inhibi- tion for the preferentially inhibited cytokines during the dosing period is ~60–76% for 5 mg BID and ~75–86% for 10 mg BID. B. Tofacitinib is a partial and reversible inhibitor of JAK activity. Rapid reversibility (1 day to 2 weeks) of biochemical targets (left) and reversibility of immune cells within 2 to 6 weeks (right). pSTAT5: IL-15 stimulated pSTAT5 in CD8+ T cells. Solid horizontal line within each box represents median; lower and upper horizontal lines of each box represent 1st and 3rd quartiles, respectively; whiskers represent 1.5 x inter-quartile range; dots outside the whiskers are outliers (individual data points). Cmax: Maximum plasma concentration; Cmin: Minimum plasma concentration; CRP: C reactive protein; IP-10: induced protein 10; JAK: Janus kinase; NK: Natural killer; pSTAT: Phosphorylated Signal Transducer and Activator of Transcription. zole) or when administration of one or itinib. At clinical exposures, tofacitinib tide (OATP), organic anion transport- more concomitant results itself does not significantly inhibit the er (OAT), organic cation transporter in moderate inhibition of CYP3A4 and major CYP450 or uridine‑diphosphate- (OCT), or multidrug and toxic com- potent inhibition of CYP2C19 (e.g. flu- glucuronosyltransferase (UGT) hepatic pound extrusion (MATE). The potential conazole). Tofacitinib exposure is de- enzymes or a number of transporters, for tofacitinib to induce CYP3A4, 1A2 creased when co-administered with po- including P-glycoprotein, or 2B6 enzymes was also low at clini- tent CYP3A4 inducers (e.g. rifampin). resistance protein (BCRP), multidrug cally relevant exposures. Inhibitors of CYP2C19 or P-glycopro- resistance associated protein (MRP), The relationship between plasma con- tein are unlikely to alter the PK of tofac- organic anion transporting polypep- centrations and pharmacodynamic (PD)

326 Mechanism of action of tofacitinib / J.A. Hodge et al. REVIEW effects, as reflected in changes in bio- matory cytokines and overall modula- 14. FIRESTEIN GS: Evolving concepts of rheu- markers, may be indirect, resulting in tion of the immune and inflammatory matoid arthritis. Nature 2003; 423: 356-61. 15. MEYER DM, JESSON MI, LI X et al.: longer duration of PD activity relative response. Tofacitinib also indirectly Anti-inflammatory activity and neutrophil to a short PK half-life. PD reversibil- modulates the production of inflamma- reductions mediated by the JAK1/JAK3 in- ity was evaluated over a range of bio- tory cytokines through autocrine and hibitor, CP-690,550, in rat adjuvant-induced markers, from STAT phosphorylation, paracrine feedback loops. As a conse- arthritis. J Inflamm (Lond) 2010; 7: 41. 16. GHORESCHI K, JESSON MI, LI X et al.: a direct read-out of JAK inhibition, to quence of these mechanisms, effects on Modulation of innate and adaptive immune changes reflecting downstream events, the such as moderate responses by tofacitinib (CP-690,550). such as levels of circulating chemokine decreases in total lymphocyte counts, J Immunol 2011; 186: 4234-43. 17. DOWTY ME, JESSON MI, GHOSH S et al.: Pre- CXCL10 and CRP levels, and lympho- dose-dependent decreases in NK cell clinical to clinical translation of tofacitinib, a cyte subset and neutrophil counts (Fig. counts, and increases in B cell counts , in rheumatoid arthritis. 7b) (31). have been observed. Despite these ob- J Pharmacol Exp Ther 2014; 348: 165-73. Phosphorylated STAT5 levels in ex vivo servations, and modest decreases in 18. LABRANCHE TP, JESSON MI, RADI ZA et al.: + JAK inhibition with tofacitinib suppresses IL-15 stimulated CD8 T cells, serum serum immunoglobulin levels, the im- arthritic joint structural damage through CXCL10 levels, CRP levels and B cell mune response as measured by vacci- decreased RANKL production. Arthritis counts reverse to baseline levels over a nation to both T cell-independent and Rheum 2012; 64: 3531-42. 2‑week washout period in RA patients -dependent antigens was intact and PD 19. van der HEIJDE D, TANAKA Y, FLEISCH- MANN R et al.: Tofacitinib (CP-690,550) in who received tofacitinib for 4 to 12 effects of short- or long-term tofacitinib patients with rheumatoid arthritis receiv- weeks. Reversibility in STAT phos- treatment were reversible after approxi- ing methotrexate: twelve-month data from a phorylation is seen as early as 24 hours mately 2 weeks. Therefore tofacitinib twenty-four-month phase III randomized ra- after cessation of tofacitinib treatment. provides a novel, innovative approach diographic study. Arthritis Rheum 2013; 65: 559-70. NK cell and neutrophil counts partial- to modulating the immune and inflam- 20. van VOLLENHOVEN RF, FLEISCHMANN R, ly reverse to baseline over 2–6 weeks matory response in RA. COHEN S et al.: Tofacitinib or adalimumab (31). This reversibility in RA patients versus placebo in rheumatoid arthritis. may reflect the effects of inflammatory References N Engl J Med 2012; 367: 508-19. 21. FLEISCHMANN R, KREMER J, CUSH J et al.: 1. LEE DM, WEINBLATT ME: Rheumatoid ar- cytokines and chemokines in elevating Placebo-controlled trial of tofacitinib mono- thritis. Lancet 2001; 358: 903-11. therapy in rheumatoid arthritis. N Engl J Med baseline NK cell and neutrophil counts, 2. McINNES IB, SCHETT G: Cytokines in the 2012; 367: 495-507. rather than residual PD effects of to- pathogenesis of rheumatoid arthritis. Nat Rev 22. KREMER J, LI ZG, HALL S et al.: Tofacitinib in Immunol 2007; 7: 429-42. facitinib. combination with nonbiologic disease-modi- The totality of these PD data dem- 3. OTERO M, GOLDRING MB: Cells of the syn- ovium in rheumatoid arthritis. Chondrocytes. fying antirheumatic drugs in patients with ac- onstrates reversibility of PD effects, Arthritis Res Ther 2007; 9: 220. tive rheumatoid arthritis: a randomized trial. Ann Intern Med 2013; 159: 253-61. generally within 14 days of tofacitinib 4. SCHETT G, STACH C, ZWERINA J, VOLL R, 23. BURMESTER GR, BLANCO R, CHARLES-SCH- discontinuation, after both short- and MANGER B: How antirheumatic drugs pro- tect joints from damage in rheumatoid arthri- OEMAN C et al.: Tofacitinib (CP-690,550) in long-term treatment. This is consistent tis. Arthritis Rheum 2008; 58: 2936-48. combination with methotrexate in patients with tofacitinib being a reversible in- 5. McINNES IB, LIEW FY: Cytokine networks - with active rheumatoid arthritis with an in- hibitor of JAK, as demonstrated by in towards new therapies for rheumatoid arthri- adequate response to tumour necrosis factor inhibitors: a randomised phase 3 trial. Lancet vitro assays, and having a short plasma tis. Nat Clin Pract Rheumatol 2005; 1: 31-9. 6. O’SHEA JJ: Targeting the Jak/STAT pathway 2013; 381: 451-60. elimination half-life. This also provides for . Ann Rheum Dis 24. LEE EB, FLEISCHMANN R, HALL S et al.: a scientific rationale for a 14-day dis- 2004; 63 (Suppl. 2): ii67-ii71. Tofacitinib versus methotrexate in rheumatoid continuation of tofacitinib while man- 7. MAVERS M, RUDERMAN EM, PERLMAN H: arthritis. N Engl J Med 2014; 370: 2377-86. Intracellular signal pathways: potential for 25. PFIZER INC.: Xeljanz prescribing informa- aging adverse events such as infections therapies. Curr Rheumatol Rep 2009; 11: tion. Available at http://labeling.pfizer.com/ and laboratory abnormalities. 378-85. ShowLabeling.aspx?id=959. Last updated Conclusions 8. O’SULLIVAN LA, LIONGUE C, LEWIS RS, May 2014. Accessed 22 July 2014. 26. KREMER JM, BLOOM BJ, BREEDVELD FC RA is a destructive, chronic inflamma- STEPHENSON SEM, WARD AC: Cytokine re- ceptor signaling through the Jak-Stat-Socs et al.: The safety and efficacy of a JAK in- tory disease mediated by multiple cy- pathway in disease. Mol Immunol 2007; 44: hibitor in patients with active rheumatoid tokines and cell types of the innate and 2497-506. arthritis: Results of a double-blind, placebo- adaptive immune system. Tofacitinib 9. LUBBERTS E: Th17 cytokines and arthritis. controlled phase IIa trial of three dosage lev- Semin Immunopathol 2010; 32: 43-53. els of CP-690,550 versus placebo. Arthritis is an oral JAK inhibitor for the treat- 10. GHORESCHI K, LAURENCE A, O’SHEA JJ: Rheum 2009; 60: 1895-905. ment of RA. Tofacitinib is a targeted Janus kinases in immune cell signaling. 27. KREMER JM, COHEN S, WILKINSON BE et small molecule that works intracellu- Immunol Rev 2009; 228: 273-87. al.: A phase IIb dose-ranging study of the oral larly to partially and reversibly inhibit 11. PESU M, LAURENCE A, KISHORE N et al.: JAK inhibitor tofacitinib (CP-690,550) ver- Therapeutic targeting of Janus kinases. sus placebo in combination with background JAK‑dependent cytokine signalling. Immunol Rev 2008; 223: 132-42. methotrexate in patients with active rheuma- Unlike biological treatments, that target 12. SHUAI K, LIU B: Regulation of JAK-STAT toid arthritis and an inadequate response to one cytokine pathway in the inflamma- signalling in the immune system. Nat Rev methotrexate alone. Arthritis Rheum 2012; 64: 970-81. tory network, JAK inhibition results in Immunol 2003; 3: 900-11. 13. O’SHEA JJ, MURRAY PJ: Cytokine signaling 28. TANAKA Y, SUZUKI M, NAKAMURA H, a pattern of partial inhibition of the in- modules in inflammatory responses. Immu- TOYOIZUMI S, ZWILLICH SH: Phase II study tracellular effects from several inflam- nity 2008; 28: 477-87. of tofacitinib (CP-690,550) combined with

327 REVIEW Mechanism of action of tofacitinib / J.A. Hodge et al.

methotrexate in patients with rheumatoid ar- 36. GONZALEZ-QUINTELA A, ALENDE R, GUDE vaccine responses in rheumatoid arthritis thritis and an inadequate response to metho- F et al.: Serum levels of immunoglobulins patients using tofacitinib. Arthritis Rheum trexate. Arthritis Care Res (Hoboken) 2011; (IgG, IgA, IgM) in a general adult popula- 2012; 64: S550. 63: 1150-8. tion and their relationship with alcohol con- 45. WINTHROP KL, NEAL J, HRYCAJ P et al.: 29. TANAKA Y, TAKEUCHI T, YAMANAKA H sumption, smoking and common metabolic Evaluation of Influenza and Pneumococcal et al.: Efficacy and safety of tofacitinib as abnormalities. Clin Exp Immunol 2008; 151: Vaccine Responses in Rheumatoid Arthritis monotherapy in Japanese patients with active 42-50. Patients using tofacitinib. Ann Rheum Dis rheumatoid arthritis: a 12-week, randomized, 37. STIEHM ER, FUDENBERG HH: Serum levels 2013; 72: 107. Phase 2 study. Mod Rheumatol 2015; 25: of immune globulins in health and disease: a 46. MAESHIMA K, YAMAOKA K, KUBO S et al.: 514-21. survey. Pediatrics 1966; 37: 715-27. The JAK inhibitor tofacitinib regulates syno- 30. FLEISCHMANN R, CUTOLO M, GENOVESE 38. BARDEN J, MULLINAX F, WALLER M: vitis through inhibition of interferon-gamma MC et al.: Phase IIb dose-ranging study of the Immunoglobulin levels in rhuematoid arthri- and interleukin-17 production by human oral JAK inhibitor tofacitinib (CP-690,550) tis: comparison with rhuematoid factor titers, CD4+ T cells. Arthritis Rheum 2012; 64: or adalimumab monotherapy versus placebo clinical stage and disease duration. Arthritis 1790-8. in patients with active rheumatoid arthritis Rheum 1967; 10: 228-34. 47. MIGITA K, MIYASHITA T, IZUMI Y et al.: with an inadequate response to disease-mod- 39. JORGENSEN C, ANAYA JM, COGNOT C, SANY Inhibitory effects of the JAK inhibitor ifying antirheumatic drugs. Arthritis Rheum J: Rheumatoid arthritis associated with high CP690,550 on human CD4(+) T lymphocyte 2012; 64: 617-29. levels of immunoglobulin a: clinical and bio- cytokine production. BMC Immunol 2011; 31. GENOVESE MC, KAWABATA T, SOMA K et logical characteristics. Clin Exp Rheumatol 12: 51. al.: Reversibility of pharmacodynamic ef- 1992; 10: 571-5. 48. MIGITA K, KOGA T, KOMORI A et al.: Influ- fects after short- and long-term treatment 40. VEYS EM, CLAESSENS HE: Serum levels of ence of Janus kinase inhibition on interleukin with tofacitinib in patients with rheumatoid IgG, IgM, and IgA in rheumatoid arthritis. 6-mediated induction of acute-phase serum arthritis. Arthritis Rheum 2013; 65: S193. Ann Rheum Dis 1968; 27: 431-40. amyloid A in rheumatoid synovium. J Rheu- 32. STROBER B, BUONANNO M, CLARK JD et 41. JENTSCH-ULLRICH K, KOENIGSMANN M, matol 2011; 38: 2309-17. al.: Effect of tofacitinib, a Janus kinase in- MOHREN M, FRANKE A: Lymphocyte sub- 49. ROSENGREN S, CORR M, FIRESTEIN hibitor, on haematological parameters during sets’ reference ranges in an age- and gender- GS, BOYLE DL: The JAK inhibitor CP- 12 weeks of treatment. Br J Derma- balanced population of 100 healthy adults--a 690,550 (tofacitinib) inhibits TNF-induced tol 2013; 169: 992-9. monocentric German study. Clin Immunol chemokine expression in fibroblast-like syn- 33. van GURP EA, SCHOORDIJK-VERSCHOOR 2005; 116: 192-7. oviocytes: autocrine role of type I interferon. W, KLEPPER M et al.: The effect of the JAK 42. BISSET LR, LUNG TL, KAELIN M, LUDWIG E, Ann Rheum Dis 2012; 71: 440-7. inhibitor CP-690,550 on peripheral immune DUBS RW: Reference values for peripheral 50. YARILINA A, XU K, CHAN C, IVASHKIV LB: parameters in stable kidney allograft pa- blood lymphocyte phenotypes applicable to Regulation of inflammatory responses in tients. Transplantation 2009; 87: 79-86. the healthy adult population in Switzerland. -activated and rheuma- 34. ROBERTS JL, LENGI A, BROWN SM et al.: Eur J Haematol 2004; 72: 203-12. toid arthritis synovial macrophages by JAK (JAK3) deficiency: clinical, 43. WINTHROP KL, SILVERFIELD J, RACEWICZ inhibitors. Arthritis Rheum 2012; 64: 3856- immunologic, and molecular analyses of 10 A et al.: The effect of tofacitinib on pneu- 66. patients and outcomes of stem cell transplan- mococcal and influenza vaccine responses 51. DOWTY ME, LIN J, RYDER TF et al.: The tation. Blood 2004; 103: 2009-18. in rheumatoid arthritis. Ann Rheum Dis 2015 , metabolism, and clear- 35. FURST DE: Serum immunoglobulins and risk Mar 20 [Epub ahead of print]. ance mechanisms of tofacitinib, a Janus ki- of infection: how low can you go? Semin 44. WINTHROP KL, RACEWICZ A, LEE EB et al.: nase inhibitor, in humans. Drug Metab Dis- Arthritis Rheum 2009; 39: 18-29. Evaluation of influenza and pneumococcal pos 2014; 42: 759-73.

328