New therapies for rheumatoid arthritis

Abstract

Rheumatoid arthritis (RA) is a chronic, systemic inflammatory disease, which continues to cause significant morbidity in affected persons. In the past few years, a number of new exciting therapeutic options have become available. These reflect the application of knowledge obtained from advancements in understanding of disease pathogenesis and underlying molecular mechanisms. A number of these therapies are outlined in the following review, including the various biological modifiers, in particular, anti-tumour necrosis factor-α agents and interleukin-1 (IL-1) receptor antagonists, which have been developed in recognition of the role of pro-inflammatory cytokines in RA. Also notable, is the current interest centring on the development and trials with B cell depletion therapies, specifically rituximab, in patients with RA. This demonstrates acknowledgment for a more significant role for B cells in the aetiology of RA, in contrast to the long held view that RA was a predominantly T cell mediated disease. To evaluate this therapeutic option for RA, salient features from recent rituximab trials have been collated. Finally, a selection of other therapeutic alternatives, including anti-IL-6 receptor monoclonal antibody and tacrolimus, and newer anti-rheumatic therapies presently in development are summarized.

Rheumatoid arthritis

Rheumatoid arthritis (RA) is a chronic, systemic inflammatory disease which targets synovial joints, and is often accompanied by an array of extra-articular manifestations. Its prevalence is estimated at 0·8–1·0% of adults [1], and the disease continues to cause significant morbidity and premature mortality [2]. Landré Beauvois provided the first good clinical description of rheumatoid arthritis in 1800, yet despite significant gains in knowledge of the immunopathology, the exact aetiology of RA remains uncertain. Immunopathogensis of RA is multifactorial. Evidence suggests that interaction between an unknown exogenous or endogenous antigen via antigen presenting cells and CD4+ T helper cells are involved in the induction of the immune response in RA. Subsequent recruitment and activation of monocytes and macrophages occurs with secretion of pro-inflammatory cytokines, in particular TNF-α and IL-1 into the synovial cavity. Release of these cytokines mediates tissue destruction by activation of chondrocytes and fibroblasts which release collagenases and metalloproteinases with resultant cartilage loss and bone erosion. B lymphocyte dysregulation with production of rheumatoid factor and other autoantibodies, formation of immune complexes and release of destructive mediators also contribute to this process (Fig. 1). A more prominent role for B cells in RA pathogenesis is being increasingly accepted and will be further addressed in relation to B cell depletion therapy. Advancements in understanding of these mechanisms have facilitated the development of many new biological modifiers, which form the basis of this review.

Fig. 1.

Diagramatic overview of immunopathological processes in rheumatoid arthritis.

It has now been identified that joint damage occurs very early in RA, and by 2 years approximately 60% of patients will demonstrate some radiographic evidence of erosions [3]. Whilst the advantages of early management of rheumatoid arthritis with disease modifying antirheumatic drugs (DMARDs) are now well recognized [4–6], until recently, treatment options were limited to mono- or combination therapy with a relatively restricted therapeutic armament. These comprise immunosuppressives and DMARDs such as prednisolone, azathioprine, cyclosporin A, gold, sulphasalazine and methotrexate [7–11]; the last of these has generally become the standard of care. Many of these medications reduce disease activity and radiological progression of bony erosion [12–15], and clinical experience has shown them to be adequate in many patients. The first new antirheumatic medication released in nearly 20 years was leflunomide (1999), a competitive dihydroorotate dehydrogenase inhibitor which inhibits the rate-limiting enzyme for de novo synthesis of pyrimidine required by activated T lymphocytes. In doing so, it hinders lymphocyte ability to initiate the pro-inflammatory processes in patients with RA. Studies of up to five years treatment have demonstrated efficacy is comparable to those of sulphasalazine and moderate dose methotrexate [16,17]. It appears to be most effective in combination with methotrexate [18], although this may exacerbate the potential for hepatic enzyme disturbances [19]. Additional side-effects include weight loss [20], diarrhoea [21], skin rash and alopecia and usage can be complicated by its long half-life (approximately 21 days). Further discussion of leflunomide and these conventional RA therapies is beyond the scope of this article, and the reader is directed to several excellent reviews in the literature [22–29].

Therapeutic dilemmas frequently arise however, in a significant number of patients with only partially responsive or treatment refractory disease. Excitement has therefore surrounded application of recent research advances which have resulted in development of a number of new therapeutic options, in particular anti-tumour necrosis factor-alpha (TNF-α) agents, interleukin-1 (IL-1) receptor antagonists, B cell depletion regimes and other targeted cytokine immunotherapies. These new therapies herald an exciting period for rheumatologists and their patients and will be discussed in this review.

New therapies in rheumatoid arthritis

Efficacy and response to new rheumatic medications is generally defined by an outcome measure of the American College of Rheumatology (ACR) [30] (Table 1). This assesses the percentage improvement from baseline with regards; – number of tender and swollen joints, patient pain (Visual Analogue Scale), global assessments by patient and physician (Visual Analogue Scales), self assessed physical disability and levels of acute phase reactants. ACR20 is most often used, although ACR50 and ACR70 (reflecting larger percentage improvements from baseline) are being increasingly utilized and generally considered more clinically relevant.

Table 1.

American College of Rheumatology Preliminary Definition of 20 Percent Improvement in Rheumatoid Arthritis (ACR20).

Anti-tumour necrosis factor-alpha therapies

TNF-α is an inflammatory cytokine that plays a pivotal role in the pathogenic mechanisms of RA [31–33]. Importance of this cytokine in RA is supported by the over expression of TNF-α in RA synovium [34], data from in vitro synovial cell cultures with use of anti-TNF-α antibody [35], and animal studies which demonstrated development of disease in mice expressing the transgene for TNF-α and amelioration after treatment with anti-TNF-α agents [36,37]. TNF-α binds to two widely expressed receptors, type 1 (p55) and type 2 (p75), and soluble receptors also influence activity of the cytokine [38]. There are three agents currently available which inhibit the action of TNF-α; – infliximab, etanercept and adalimumab (Table 2). Salient features of each, and published trials are summarized.

Table 2.

Anti-TNF-α antagonists for the treatment of rheumatoid arthritis.

Drug	Primary action	Route of adminstration	Usual dose	Half life
Infliximab	Chimeric anti-TNF-α antibody	Intravenous infusion	3 mg/kg body weight at 0, 2, 6 week, then 8 weekly	9 days
			Gradual increase to 10 mg/kg if incomplete response
Etanercept	Soluble TNF-α fusion protein	Subcutaneous injection	25 mg twice/wk	4 days
	Binds TNF-α and β		or 50 mg once/wk
Adalimumab	Human anti-TNF-α antibody	Subcutaneous injection	40 mg every second week	2 wk

Adapted from Olsen & Stein [100] with kind permission from the publisher. TNF, Tumour necrosis factor; wk, Week.

Infliximab

Infliximab is a chimeric IgG1 anti-TNF-α antibody consisting of the antigen-binding region of mouse antibody and the constant region of the human antibody. The antibody binds soluble and membrane bound TNF-α, thereby impairing binding to its receptor. In addition, the antibodies also mediate killing of cells expressing TNF-α. Infliximab (3–10 mg/kg) is administered as a 4–8 weekly infusion with methotrexate to minimize formation of anti-infliximab antibodies [39]. A study of 428 patients with active RA despite methotrexate therapy, showed achievement of ACR20 at 54 weeks in 42–59% depending on the dosage regimen [40]. This improvement was greater than for methotrexate and placebo with all schedules. Patients receiving 10 mg/kg every 4–8 weeks achieved ACR50 more frequently than patients dispensed lower dosages [40]. Another multicentre study demonstrated the combination of infliximab and methotrexate was superior to methotrexate alone in improving clinical response and radiological progression [41].

Etanercept

Etanercept is a soluble TNF-receptor fusion protein, composed of 2 dimers, each with the ligand binding portion of the type 2 receptor (p75) linked to the Fc portion of human IgG1. The protein binds to both TNF-α and TNF-β, preventing each from interacting with respective receptors. Dose escalation studies report that both 10 mg and 25 mg twice weekly were significantly more effective than placebo (ACR20 51–59% versus 11%) [42]. These reports also confirmed similar ACR20 responses with both doses, but a more rapid response and more frequent achievement of the more clinically relevant ACR50 with the higher dose (24%versus 40%) [42]. A double blind, randomised study studied the response to etanercept (10 mg or 25 mg twice weekly) versus methotrexate (dose escalated to 20 mg/week over 8 weeks) [43]. Those receiving higher dose etanercept had a more rapid response than those on methotrexate, but at 12 months, no differences in patient ACR20 responses were evident between the etanercept 25 mg/twice weekly or methotrexate groups – 72% versus 65% (P = 0·16) [43]. Changes in radiologic scores were also not significantly different. In patients with ongoing disease activity despite methotrexate, Weinblatt et al. [44] reported that addition of etanercept resulted in significant improvement in disease control. Etanercept and methotrexate in combination were shown to be significantly better in reduction of disease activity and delay of radiographic progression than either therapy alone [45]. This study is important as it is the first to assess concurrent initiation or combination usage of etanercept and methotrexate compared to either drug alone. In addition it is unique as it studied patients with RA who had failed previous therapy with DMARDs other than methotrexate, which allowed fairer comparison of the effects of methotrexate with etanercept. 686 patients were randomly assigned to treatment regimens and methotrexate dosages were aggressively escalated to mean dose 17·2 mg weekly. At 52 weeks, 43% of patients in the combination group achieved ACR70 compared to 19% in the methotrexate group (P < 0·0001) and 24% in the etanercept group (P < 0·0001). Notably, combination therapy resulted in negative progression of the modified Sharp radiological score compared to baseline, suggesting that repair of previous joint damage might be possible. No new safety data were recorded.

Adalimumab

Adalimumab is a recombinant human IgG1 monoclonal antibody that binds TNF-α, thereby precluding binding to its receptor. The monoclonal antibody also lyses cells expressing the cytokine on their surface. Being a human recombinant product, formation of anti-chimeric antibodies is avoided with adalimumab. Adalimumab 40 mg second weekly plus methotrexate induced an ACR20 response rate of 67% compared to 15% in the methotrexate plus placebo group [46].

Contraindications and adverse effects of anti-TNF-α antagonists

Multiple adverse events attributed to the anti-TNF-α agents have been described in case reports or small patient series [47], although the exact relationships are yet to be defined. More common reactions include headaches, nausea, minor infusion reactions and local skin irritation, usually of insufficient severity to result in treatment cessation.

Patients with active infection should not be commenced on anti TNF-α agents, and drug should be ceased if serious infection occurs. Serious bacterial infections, tuberculosis, atypical mycobacterial infections, aspergillosis, histoplasmosis, Pneumocystis carinii pneumonia and other opportunistic infections have been reported following therapy with all agents in this drug class [48–52]. Reactivation of latent tuberculosis [53] and the occurrence of infectious granulomatous diseases are particularly increased, thought in part to be secondary to the importance of TNF in granuloma formation [54]. The rate of tuberculosis in patients with RA treated with infliximab was 24·4 cases per 100 000 as compared to background rate in RA patients of 6·2 cases per 100 000 patients [55]. TNF-α treated patients more frequently suffer re-activated latent disease complicated by extra-pulmonary manifestations. This observation underscores the importance of pretreatment tuberculsosis screening.

Malignancy, lupus-like autoimmune disease, demyelinating disorders, vasculitis, liver disease and haematologic abnormalities have all been described in association with TNF-α antagonists [56–59]. Persistent anxiety surrounds the possible long-term safety of the anti-TNF-α agents. Debate in particular, relates to whether reports of lymphoma and TNF-α antagonists reflect a definite drug effect or the known increased incidence in patients with RA [60,61]. Lymphoma rates in patients with RA treated with anti-TNF-α agents are estimated at 2·3–6·4 that of the general population [56]. Other subtypes of malignancy have not been found to be increased  [58,62].  Antibodies  to  etanercept  were  reported in 3% of patients, although their clinical significance is unknown [43]. Human antichimeric antibodies to infliximab have been well recognized [63], hastening drug clearance and increasing infusion reactions. As with infliximab, the incidence is significantly reduced with use of methotrexate [39]. Antinuclear and antidouble stranded DNA antibodies were detected in patients treated with infliximab, etanercept and adalimumab [40,42,46]. Drug induced lupus however, is rare [57,64]. TNF-α agents are contraindicated in patients with previous multiple sclerosis. Exacerbations of the latter and new onset demyelinating neurological diseases have been reported [58].

Therapy with anti TNF-α agents should be avoided in patients with congestive cardiac failure [65]. A summary of monitoring recommendations is shown in Table 3.

Table 3.

Monitoring recommendations for anti-TNF-α antagonists and Anakinra.

Anti-TNF-α agents

″Screen all patients prior to commencing treatment for previous exposure to tuberculosis.

″Be clinically alert for infectious diseases, in particular, tuberculosis, histoplasmosis, Pneumocystis carinii pneumonia; symptoms of congestive cardiac failure and demyelinating disease

″No regular laboratory tests are required unless patients are on conventional disease modifying antirheumatic drugs. In this case, appropriate routine screening must continue

Anakinra

″Baseline complete blood count, monthly for 3 months and 3 monthly thereafter (in particular to monitor for neutropenia)

″Pre treatment screening for asthma

″Be clinically alert for development of infections, particularly pneumonia

Other anti-TNF-α antagonists

Studies with a novel polyethylene glycol linked (PEGylated) anti-TNF-α Fab′ fragment, CDP870 have suggested it is therapeutically effective and well tolerated in patients with active RA. Clinical improvement with CDP870 was shown to be comparable to etanercept with 75% patients reaching ACR20 at 4 weeks, with sustained effect to 8 weeks [66]. Advantages of this method of TNF-α neutralization include long plasma half life, reduced cost and thus theoretically improved availability for patients. Larger studies examining longer term efficacy and safety are underway.

Interleukin-1 receptor antagonist: anakinra

Interleukin 1 is a pro-inflammatory cytokine produced by stimulated monocytes, macrophages and some specialiazed synovial lining cells [67]. IL-1 receptor antagonist competes with IL-1 for binding to the receptor, subsequently down regulating IL-1 actions. By stimulating the release of matrix metalloproteinases and increasing bone resorption by effects on osteoclasts, IL-1 has been shown to have a significant role in RA pathogenesis, particularly in regards cartilage and bone erosion. Mice deficient of IL-1 receptor antagonist, demonstrate the development of inflammatory arthritis similar to RA [68]. In addition, patients with RA have been shown to have lower IL-1 receptor antagonist levels than anticipated for the level of IL-1 in the joint [69,70]. It was hypothesized that addressing this imbalance, with a recombinant IL-1 receptor antagonist could be beneficial in RA and anakinra was thus developed for this purpose.

Anakinra is administered by daily subcutaneous injection. It is primarily renally cleared and thus dosage adjustments are necessary for patients with renal impairment [71]. A large randomised controlled study in patients with RA, demonstrated at 24 weeks that anakinra is more effective than placebo (ACR20 43%versus 27%) [72]. The long-term extension trial demonstrated data from the first 24 weeks to be robust, however, by 48 weeks ACR50 and ACR70 responses were 18% and 3%, respectively [73]. Radiological progression was shown to be significantly delayed in a multicentre analysis [74]. The most frequent adverse effect is dose-dependent injection site irritation (50–80% patients) with smaller numbers reporting a more severe allergic reaction [75]. Infection risk appears to be increased, although a large multicentre study did not identify increased risk of opportunistic infection [76]. Patients should be monitored for reversible thrombocytopenia and neutropenia, the latter exacerbated by concomitant methotrexate use (Table 3). Anakinra should not be used in conjunction with anti-TNF-alpha antagonists.

B cell depletion therapy

Contrary to the long held view that RA is a predominantly T cell mediated disease, the important role of B cells in disease aetiology is supported by development and trials with B cell depletion therapies notably rituximab. B cells contribute to the pathogenesis of RA via a number of proposed mechanisms including; presentation of antigen complexed with IgG to T cells, and T cell independent generation of TNF-α by tissue macrophages after stimulation by oligomeric IgG rheumatoid factor (RF) immune complexes [77,78]. In addition, the ability of IgG RF B cells to self perpetuate due to secretion of own antigen, provided rationale for the proposal that elimination of the RF B cell clones may result in prolonged disease remission [79].

B cell development is defined by series of changing surface phenotypes. CD20 is expressed at intermediate stages and is lost during terminal differentiation to the immunoglobulin producing plasma cells. The level of expression of CD20 on memory cells remains undetermined. Rituximab, a chimeric monoclonal antibody against human CD20, is thought to act via complement mediated- and antibody dependent cell mediated-cytotoxicity, induction of apoptosis and inhibition of cell growth [80]. It contains the human IgG1 heavy-chain and kappa light-chain constant region sequences and murine variable region sequences [80]. Rituximab treatment results in rapid depletion of CD20 positive B cells in the peripheral blood, and normal B cells are subsequently replenished by stem cells in most patients 3–12 months after therapy [81]. The drug was initially licensed in 1997 for treatment of relapsed low grade B cell follicular NHL and following this success, experimental use in autoimmune disorders was undertaken with initial promise demonstrated in chronic idiopathic thrombocyopenic purpura (ITP) [82,83].

More recently, rituximab has been trialled in patients with treatment resistant, erosive rheumatoid arthritis [84]. In the preliminary study, 5 patients resistant to at least 5 DMARDS received B cell depletion with rituximab (plus cyclophosphamide 2 × 750 mg IVI and 60 mg prednisolone daily for 11–22 days then tapered). All had ACR50 responses at 6 months, 3 had ACR70 and benefit persisted for approximately 12 months. B cells were undetectable shortly after infusion and remained depressed for 6 months. This study was extended to include 22 seropositive RA patients, with disease resistant to multiple DMARDs [85]. Patients were treated with varying protocols of rituximab, cyclophosphamide and/or high dose prednisolone to further assess dose–response and requirement for additive therapies. Optimal treatment was determined as rituximab = 600 mg/m² plus cyclophosphamide. Apart from infrequent mild infusion reactions, no major adverse events were reported. Rheumatoid factor titres were depressed in all for 3–6 months, whilst total immunoglobulin declined modestly [85]. A randomised double blind controlled study published this year reported that 43% and 41% of patients with active RA despite methotrexate, achieved ACR50 at six months after one course of full dose rituximab with either methotrexate or cyclophosphamide, respectively [86]. This was compared to 13% with methotrexate monotherapy.

Anti-interleukin-6 receptor antibody

IL-6 produced by lymphocytes, monocytes, fibroblasts, synoviocytes and endothelial cells, is a pleiotropic cytokine with roles in inflammation and haematopoiesis [87,88]. Whilst the cytokine appears to have both pro- and anti-inflammatory roles in vivo, overproduction of IL-6 is thought to play a crucial role in RA pathogenesis via activation of auto-reactive T cells and osteoclasts, stimulation of autoantibody production and release of inflammatory mediators [89]. Levels of IL-6 and soluble IL-6 receptor, have been shown to be higher in patients with RA compared with controls [90], and to correlate with RA disease activity and radiological joint damage [91,92]. Open labelled studies with anti-IL-6 receptor monoclonal antibody resulted in rapid clinical improvement and normalization of acute phase reactants [93] and an initial phase I/II study examining the efficacy and safety of anti-IL6 therapies was promising [94]. More recently, ACR50 response rates of 40% at 3 months were reported in patients receiving four weekly 8 mg/kg infusions with anti-IL-6 receptor monoclonal antibodies [95]. Mild transient abnormalities in liver function tests and leucopenia were frequent, and elevations in total cholesterol, HDL cholesterol and triglycerides were reported in up to 44% of patients. Further studies are necessary to determine the safety and role of these therapies in long-term management of RA.

Tacrolimus

Tacrolimus, an orally available immunomodulatory and anti-inflammatory agent was studied in a long-term trial of 896 patients with RA [96]. It acts by inhibiting calcineurin, thereby diminishing its ability to initiate gene transcription for the synthesis of inflammatory cytokines, in particular TNF-α, interleukin-2 and interferon-γ. At the end of 12 months therapy with 3 mg/day of tacrolimus, ACR20, ACR50 and ACR70 response rates were 38·4%, 18·6% and 9·0%, respectively. Adverse events occurred in 59% of patients and included diarrhoea, nausea and tremor. Hypertension was reported in 5·4% of patients. Importantly, mean creatinine levels were noted to increase ≥ 30% in a significant number of patients, although in most cases, creatinine remained within the normal range. Overall, although responder rates are not as compelling as with the TNF-α agents, for those who respond, it offers another effective and well tolerated oral treatment option for RA.

Conclusion

The last few years have been very exciting in the management of rheumatoid arthritis, with dramatic gains in understanding of RA pathogenesis and subsequent translation to effective treatment options. The therapies reviewed are all relatively safe, more effective than placebo and slow the progression of disease as measured radiologically. In particular, anti-TNF-α agents are becoming more widely available and offer rapid onset of effective treatment in many, although frustratingly, there as yet remains no adequate explanation for some patients poor response to this drug class. Advancement in this area is eagerly awaited. At present there are few direct comparison studies between the various TNF-α agents or TNF-α drugs and DMARDs and comparative trials are needed. Research into newer pharmacological therapies is ongoing, with multiple compounds in development (Table 4).

Table 4.

Selection of newer anti-rheumatic therapies in development.

		Development phase
Compound (developer)	Mechanism of action	USA	EU
TNF Inhibitors
″CDP-870 (UCB)	Anti-TNF antibody fragment	III	III
″Pegsunercept (Amgen)	Pegylated soluble TNF receptor type I	II
″ISIS-104838 (Isis Pharmaceuticals)	TNF-α antisense inhibitor	II
″AGIX4207 (AtheroGenics)	TNF inhibitor (oral)	II	II
Interleukin-based therapies
″Atlizumab (MRA) (Roche/Chugai)	Humanized anti-IL-6 receptor monoclonal antibody	II	I
″HuMAX-IL-15/AMG-714 (Immunex/Genmab)	Anti-IL-15 monoclonal antibody	II	II
″ABT-874/J-695 (Abbott/Cambridge Antibody Technology)	Anti-IL-12 monoclonal antibody	II	II
Cell Adhesion Molecule Inhibitors
″Natalizumab (Elan/Biogen Idec)	Humanized 4–1 and 4–7 monoclonal antibody	II	II
Co-stimulation inhibitors
″CTLA4-Ig (Bristol-Myers Squibb)	CD28/B7 pathway inhibitor	III
″Alefacept (Biogen Idec)	CD2 Antagonist	II
MAP Kinase Inhibitors
″Scio-469 (Scios)	p38 MAP kinase inhibitor (oral)	IIb
Ion Channel Blockers
″AZD-9056 (AstraZeneca)	P2 X 7 receptor antagonist	II	II
Other therapies
″CCI-779/Temsirolimus (Wyeth)	Cell cycle inhibitor	II
″Belimumab (Human Genome Sciences/Cambridge Antibody Technology)	Antibody against B lymphocyte stimulator protein	II
″AT-001/dnaJp1(Androclus Therapeutics)	Heat shock derived protein	II

In conclusion, there are compelling data to suggest that the combination of earlier use of disease modifying treatments [4–6,97], attention to coexisting conditions such as atherosclerosis and osteoporosis [98,99] and evolution and refinement of the newer therapies will allow more patients to realistically strive for disease remission and return of function in the near future.