Abstract
Strategies in rheumatoid arthritis (RA) based on ‘treat to target’ aim to control disease activity, minimize structural damage, and promote longer life. Several disease-modifying antirheumatic drugs (DMARDs) have been shown to be effective including biological DMARDs (bDMARDs). Treatment guidelines and recommendations for RA have also been published. According to those guidelines, conventional synthetic DMARDs (csDMARDs), as monotherapy or combination therapy, should be used in DMARD-naïve patients, irrespective of the addition of glucocorticoids (GCs). Combination therapies with bDMARDs are also essential for conducting treatment strategies for RA, because in every recommendation or guideline for the management of RA, combination therapies of csDMARDs with bDMARDs are recommended for RA patients with moderate or high disease activity after failure of csDMARD treatment. bDMARDs are more efficacious if used concomitantly with methotrexate (MTX) than with MTX monotherapy or bDMARD monotherapy. Thus, retention has been reported to be longer when combined with MTX. The superior efficacy of combination therapy compared with MTX monotherapy or bDMARD monotherapy could be because: (1) it could help to minimize MTX toxicity by reducing the dose of MTX, thus retention rate of the same therapeutic regimen would become high; (2) anti-bDMARD antibodies are observed at lower concentrations when using MTX concomitantly, so less clearance of bDMARDs via less formation of bDMARD and an anti-bDMARD immune complex; (3) of the additive effects of MTX to bDMARD, especially the combination of tumor necrosis factor inhibitors (TNFis) with MTX. Hence, evidence suggests that combination therapy with bDMARDs is more efficacious than monotherapy using a csDMARD or bDMARD, and that MTX is the best drug for this purpose (if MTX is not contraindicated). Finding the most effective drug regimen at the lowest cost will be the aim of RA treatment in the future.
Keywords
Introduction
In ‘treat-to-target’ (T2T) recommendations for rheumatoid arthritis (RA), remission (or at least low disease activity) was the aim [Smolen et al. 2016] and has become reality in recent years. T2T strategy is always discussed after taking methotrexate (MTX) and biological disease-modifying antirheumatic drugs (bDMARDs) into consideration. MTX was tested in 1951 in patients with RA and psoriasis, and then used more frequently for RA treatment in the late-1980s [Braun, 2011]. Previously, there was no treatment strategy using powerful drugs supported by solid evidence except for the ‘pyramidal plan’ [Smyth, 1972]. Thus, no treatment goal could be proposed to RA patients. MTX has been recognized as an ‘anchor drug’ for RA treatment, and much evidence of MTX efficacy has been accumulated. Another distinct turning point that shifted the paradigm of RA treatment was when bDMARDs were first used in 1998 in the USA. Today, five types of tumor necrosis factor inhibitors (TNFis) along with abatacept (ABT), tocilizumab (TCZ), rituximab (RTX), and anakinra (ANK) are available for RA treatment. By using these bDMARDs, remission has become a T2T in real practice.
Early diagnosis and a treatment strategy are also important to achieve T2T. The first treatment strategy was proposed in 1989 as a ‘step-down bridge’ concept [Wilske et al. 1989]. Along with the new drugs that have come onto the market, high-quality studies using a prospective cohort design have been published. In 1996, the American College of Rheumatology (ACR) published guidelines for RA management based on definitive evidences [American College of Rheumatology, 1996].
Studies have demonstrated the efficacy of conventional synthetic DMARDs (csDMARDs) and bDMARDs (and their combination) in RA treatment. Patients with early RA started on combined therapy of csDMARDs with bDMARDs showed earlier clinical improvement, less joint damage, with less toxicity at 5 years according to the BeSt study [Klarenbeek et al. 2011].
Numerous reports regarding efficacy of bDMARDs have been published in various clinical settings including early and established patient groups, patients with inadequate response for MTX, bio-naïve or use as a second bDMARD, or concomitant use of MTX or not. In this article, we review combination therapies with bDMARDs and discuss their benefits and disadvantages, focusing on RA in line with our clinical questions. A new nomenclature of DMARDs has been proposed in 2014 and we used the newest category in this article [Smolen et al. 2014b].
Which combination therapies are available?
According to the recommendations set by the ACR [Singh et al. 2016] and Japan College of Rheumatology (JCR) [Miyasaka et al. 2014], drug treatment starts with csDMARD monotherapy. Then, csDMARD combination therapy, with or without a glucocorticoid (GC), is used if monotherapy fails. Recommendations from the European League Against Rheumatism (EULAR) [Smolen et al. 2014a] allow combination therapy of csDMARDs from the beginning of treatment. EULAR recommendations are thought to have a sound therapeutic basis for abiding strictly to the T2T strategy. Indeed, recent studies have reported excellent results of triple DMARD therapy that is equivalent to that seen with bDMARDs [De Jong et al. 2014; Karlsson et al. 2013; Moreland et al. 2012; O’Dell et al. 2013; Matsuno et al. 2016]. However, evaluation of the results of triple therapy must be cautious because the prevalence of discontinuation of treatment is higher in triple therapy than combination therapy of MTX plus bDMARDs according to Moreland and colleagues (37% versus 30%) [Moreland et al. 2012] and Karlsson and colleagues (26% versus 10%) [Karlsson et al. 2013]. If an inadequate response is observed upon csDMARD combination therapy, combination therapy with bDMARDs can be considered.
GCs can also be used in combination therapy. EULAR recommendations state that low-dose GCs should be considered as part of the initial treatment strategy (in combination with one or more csDMARDs) for ⩽6 months [Smolen et al. 2014a]. This recommendation is because of their proven capacity to increase clinical, functional and structural efficacy if combined with csDMARDs [Bakker et al. 2012; Svensson et al. 2005; Kirwan, 1995]. This combination with low dose of prednisone (PSL) [Goekoop-Ruiterman et al. 2005] and with tapered high dose of PSL [Heimans et al. 2014] has similar efficacy when compared with TNFis plus MTX. For high risk early RA, MTX associated with a moderate step-down dose of GC was as effective in inducing remission at week 16 as DMARD combination therapies with moderate or high step-down GC doses, and it showed a more favorable short-term safety profile [Verschueren et al. 2015]. Of note, an aggressive treatment regimen with tapering PSL was shown to slow radiographic progression even in the patients with little response of clinical disease activity [Boers et al. 2013]. According to 2015 ACR guidelines, these state that, in cases of a flare, addition of GC is used at the lowest possible dose and for the shortest possible duration [Singh et al. 2016]. In a broad sense, intra-articular injection of triamcinolone with subcutaneous adalimumab (ADA) is also included in combination therapy with bDMARDs [Axelsen et al. 2015]. Axelsen and colleagues mentioned that a T2T strategy with oral MTX and an intra-articular injection of a GC, with or without subcutaneous ADA, in patients with early RA reduces synovitis, osteitis and tenosynovitis, and halts progression of structural damage (as judged by magnetic resonance imaging) [Axelsen et al. 2015].
Clinical results of combination therapy in comparison with monotherapy
bDMARDs for the treatment of RA patients include infliximab (IFX), etanercept (ETN), ADA, golimumab (GLM), certolizumab pegol (CZP), ABT, RTX, TCZ and ANK. According to EULAR recommendations [Smolen et al. 2014a] and ACR guidelines [Singh et al. 2016], ANK is not listed for the treatment of RA, so comments regarding ANK have been omitted in this review.
Several randomized clinical trials related to bDMARD therapy in RA patients have been classified into two categories according to disease duration (‘early’ or ‘established’). Then, patients with established RA have been subdivided according to whether they are resistant to csDMARD therapy, including MTX (usually bDMARD-naïve in this group) or to bDMARD therapy. To clarify the efficacy of combination therapies with bDMARDs, only studies in which populations were bDMARD-naïve were selected and are summarized in Table 1.
Summary of biologic disease-modifying antirheumatic drugs.
ACR, American College of Rheumatology; AS, ankylosing spondylitis; bDMARD, biological DMARD; CAPS, cryopyrin-associated periodic syndromes; CD, Crohn disease; CLL, chronic lymphocytic leukemia; COMB, combination therapy of MTX with bDMARD; CTLA, cytotoxic T-lymphocyte-associated protein; DMARD, disease-modifying antirheumatic drug; EU, European Union; HAQ-DI, Health Assessment Questionnaire Disability Index; HS, hidradenitis suppurative; Ig, immunoglobulin; IL-6, interleukin-6; JIA, juvenile idiopathic arthritis; Jpn, Japan; MONO, monotherapy of biologics or with MTX; mTSS, van der Heijde-modified total Sharp score; MTX, methotrexate; NHL, non-Hodgkin’s lymphoma; pCrohn, pediatric Crohn’s; pJIA, polyarticular onset juvenile idiopathic arthritis; PsA, psoriatic arthritis; PsO, plaque psoriasis; pUC, pediatric UC; RA, rheumatoid arthritis; REM, remission; sJIA, systemic JIA; TNF, tumor necrosis factor; UC, ulcerative colitis.
ABT, abatacept; ACR, American College of Rheumatology; ADA, adalimumab; bDMARD, biological DMARD; COMB, combination therapy of MTX with bDMARD; CZP, certolizumab pegol; DAS, Disease Activity Score; DMARD, disease-modifying antirheumatic drug; ETN, etanercept; GLM, golimumab, HAQ-DI, Health Assessment Questionnaire Disability Index; IFX, infliximab; MONO, monotherapy of biologics or with MTX; mTSS, van der Heijde-modified total Sharp score; MTX, methotrexate; PBO, placebo; REM, remission; RTX, rituximab; SDAI, Simple Disease Activity Index; TCZ, tocilizumab.
Among TNFis, use of IFX and GLM was restricted only with MTX. According to the ASPIRE study (early RA) [St Clair et al. 2004] and ATTRACT study (established RA) [Lipsky et al. 2000], MTX therapy combined with IFX was superior to MTX monotherapy in terms of clinical and radiologic outcomes at 54 weeks. In GO-BEFORE (early RA) [Emery et al. 2009] and GO-FORWARD (established RA) [Keystone et al. 2009] trials, the study population was allocated randomly into four arms to receive placebo plus MTX (group 1), GLM (100 mg) plus placebo (group 2), GLM (50 mg) plus MTX (group 3), or GLM (100 mg) plus MTX (group 4). At 24 weeks in early RA, combination of GLM plus MTX demonstrated a significantly better response compared with placebo plus MTX in terms of the achievement of at least 50% improvement according to the American College of Rheumatology criteria core set (ACR50), but GLM monotherapy was not inferior to MTX monotherapy. In established RA, identical results were reported in the GO-FORWARD study. That is, the clinical outcomes measured by the achievement of at least 20% improvement according to the American College of Rheumatology criteria core set (ACR20) at 14 weeks were significantly better in the combination therapy group than in the MTX monotherapy group. GLM monotherapy was not inferior to MTX monotherapy. In the GO-BEFORE study and GO-FORWARD study, radiologic examination was not conducted. In Japan, GLM can be used as monotherapy only at 100 mg based on the GO-MONO study [Takeuchi et al. 2013]. In the latter, 316 patients with established RA were randomized to receive placebo, GLM (50 mg) and GLM (100 mg) without concomitant MTX in all groups. At week 14, ACR20 response was significantly better in both GLM groups than in the placebo group. GLM monotherapy was also effective in Japanese patients with RA.
Therapeutic efficacy according to clinical and radiologic measurements of a combination of MTX with ETN and ADA has been reported to be better than MTX monotherapy in patients with early RA and established RA [Emery et al. 2008; Klareskog et al. 2004; Breedveld et al. 2006; Weinblatt et al. 2003]. In patients with early RA, even in those with poor prognostic factors (e.g. high levels of anticyclic citrullinated peptide; positive rheumatoid factor or bone erosion), CZP combined with MTX led to significant inhibition of structural damage and reduction in signs and symptoms at 52 weeks according to the C-OPERA study [Atsumi et al. 2016]. The RAPID study [Keystone et al. 2008] showed good clinical and radiologic outcomes in patients with established RA.
In a very large double-blind randomized controlled trial (FUNCTION) [Burmester et al. 2015b], remission of Disease Activity Score 28-Erythrocyte Sedimentation Rate (DAS28-ESR) in combination therapy and TCZ monotherapy were superior to MTX monotherapy. Comparison of the efficacy of adding TCZ to that of switching to TCZ monotherapy was made in inadequate responders to MTX (a situation often encountered in clinical practice) for the ACT-RAY study [Huizinga et al. 2015]. Clinically relevant superiority of the TCZ plus MTX regimen over the switch to TCZ monotherapy was not observed. However, Dougados and colleagues drew attention to the fact that structural damage on radiographs became advanced [Dougados et al. 2014]. Also, Kojima and colleagues emphasized the importance of MTX therapy concomitant with TCZ treatment in achieving better clinical outcomes and less structural damage for RA patients with high disease activity [Kojima et al. 2015].
ABT was also reported to elicit better results at 1 year if used with MTX in a MTX-naïve population with early RA and poor prognostic factors in the AGREE trial [Westhovens et al. 2009]. This study was assessed over 2 years after a double-blind period of 1 year [Bathon et al. 2011]. During the second year, the original ABT plus MTX group continued treatment, whereas ABT was initiated in the MTX-monotherapy group. Clinical outcomes in the original MTX-monotherapy group caught up with those of the ABT plus MTX group, but structural damage was significantly greater in the original MTX-monotherapy group than in the ABT plus MTX group. These results implied that early intervention using more aggressive treatment is important to lower structural damage. In established RA, better clinical and radiographic results were reported at 1 year using ABT plus MTX in the AIM trial [Kremer et al. 2006].
RTX has been approved for RA treatment if given in combination with MTX, but is not approved in Japan. If administered for early RA, the IMAGE trial showed the clinical and radiographic efficacy of RTX at the usual dose [Tak et al. 2011]. The SERENE trial [Emery et al. 2010] was conducted to determine the efficacy and safety of treatment with RTX plus MTX in patients with active RA who had an inadequate response to MTX and who were naïve to prior treatment with bDMARDs. The SERENE trial showed that RTX plus MTX significantly improved clinical outcomes at week 24, which were improved further by week 48, compared with MTX monotherapy.
In general, bDMARDs are more efficacious if used concomitantly with MTX than in bDMARD monotherapy, and retention has been reported to be longer if they are used with MTX [Zhang et al. 2015; Blum et al. 2011]. The strongest association between use of concomitant MTX and bDMARD persistence was observed for IFX [hazard ratio (HR) 1.8, 95% confidence interval (CI) 1.7–2.0], but not observed for ABT (HR 1.1, 95% CI 1.0–1.2), or TCZ (HR 1.1, 95% CI 0.9–1.3) [Zhang et al. 2015]. However, when considering the use of a bDMARD, MTX cannot always be used concomitantly because of patient factors (e.g. comorbidities, history of adverse reactions to MTX). In fact, up to one-third of RA patients are treated with a bDMARD as monotherapy [Gabay et al. 2015; Emery et al. 2013]. In this regard, TCZ is useful because TCZ monotherapy has been reported to be comparable with combination therapy with MTX [Emery et al. 2013].
Other conventional synthetic disease-modifying antirheumatic drugs that can be combined with biological disease-modifying antirheumatic drugs
csDMARDs other than MTX include sulfasalazine (SSZ), hydroxychloroquine (HCQ), and leflunomide (LEF). These agents are listed in ACR guidelines [Singh et al. 2016] and EULAR recommendations [Smolen et al. 2014a]. csDMARDs such as azathioprine, cyclosporine, minocycline and gold are not included in ACR guidelines or EULAR recommendations. In Japan, tacrolimus and bucillamine are often used. Most of the evidences on combination therapy with bDMARDs is related to MTX only. This situation arises because the protocols of most clinical studies on bDMARDs were planned with MTX (or for patients with an inadequate response to MTX), and recommendations or guidelines on clinical treatment have stated that MTX should be part of the first treatment strategy in patients with active RA. Combination of LEF with biologic agents has been reported to be efficacious, and responses may be even higher than those obtained with LEF plus MTX combinations [Kalden et al. 2005]. Singer and colleagues suggested that LEF in combination with a TNFi has comparable efficacy to that of MTX plus a TNFi [Singer et al. 2011]. No reports regarding combination therapy of LEF with ABT or TCZ were found in the literature searched with ‘leflunomide and abatacept or tocilizumab’ as keywords in the title through PubMed. However, little information has been obtained for other csDMARDs compared with MTX because most instances involve addition of a biologic agent to existing DMARD therapy. Hence, studies investigating the efficacy of other csDMARDs for combination with a bDMARD are lacking.
Combination of a bDMARD with another bDMARD is not well documented. Safety of RTX in combination with other bDMARDs (ADA, ETN, AVT or IFX) in RA was reported as an open-label study [Rigby et al. 2013]. Rigby and colleagues showed that no serious adverse events occurred within 24 h of any RTX infusion, and that efficacy improved at week 48 compared with that at week 24. Bispecific antibodies against TNFis and interleukin (IL)-17 have been reported to be more effective than single blockade in a model of arthritis in mice [Fischer et al. 2015]. Combination therapy using ETN plus ANK was reported to provide no additional benefit as well as an increased risk compared with ETN monotherapy, so was not recommended for RA treatment [Genovese et al. 2004]. Therefore, combination of two bDMARDs should be avoided because of the low benefit–risk ratio.
Why is concomitant use of methotrexate in biological disease-modifying antirheumatic drug therapy more efficacious than monotherapy?
In general, if ‘drug A’ is combined with ‘drug B’, intensification of the favorable effects of drug B or reduction of the adverse effects of drug B is expected, which increases the retention of drug B.
Use of MTX combined with bDMARDs could help to reduce the weekly dose of MTX, thereby minimizing MTX toxicity. Conversely, a high dose of MTX has been reported to be required to make combination therapy with bDMARDs most efficacious, but a maximal dose is not always necessary for this purpose. In 2015, the CONCERTO trial [Burmester et al. 2015a] revealed that in patients with early RA who were starting ADA combination therapy, administration of 10 and 20 mg/week MTX appeared to have an equivalent effect on treatment efficacy. This phenomenon could be because the percentage of patients with at least an anti-ADA antibody was lower in the high-dose group (6% and 6%; 10 and 20 mg/week of MTX, respectively) than in the low-dose group (21% and 13%; 2.5 mg and 5 mg/week of MTX, respectively). Therefore, if the efficacy of combination therapy of MTX with a bDMARD is not adequate and MTX dose is relatively low, the drug that should be increased in dose is MTX, but it would not have to be the maximum dose approved by the government. However, in the MUSICA study [Kaeley et al. 2014], the small differences in clinical, functional, and ultrasound outcomes between 7.5 mg/week of MTX with ADA and 20 mg/week of MTX with ADA in patients with moderate-to-severe RA, unresponsive to previous MTX at >15 mg/week, suggest that reduction in MTX dose should be considered when initiating ADA therapy in some patients with an inadequate response to MTX. Therefore, combination therapy with bDMARDs should enable dose reduction of MTX, thereby lowering the negative effects of MTX. Another reason for the superior outcomes of combination therapy is that MTX provides immunologic tolerance to bDMARDs [Maini et al. 1998; Bartelds et al. 2007]. Concomitant therapy with low-dose MTX was reported to diminish the appearance of anti-IFX antibody. This would consequently reduce clearance of bDMARDs via less formation of bDMARD and anti-bDMARD immune complex [Bartelds et al. 2007]. Thirdly, MTX would possibly potentiate the effects of bDMARDs in an additive fashion. MTX appeared to have a greater effect in terms of suppressing circulating IL-6 than on suppressing circulating TNF-α [Nishina et al. 2013]. In this regard, it makes sense that TNFis have a greater effect when used in combination with MTX than when used as monotherapy.
Future prospects
As shown above, several types of DMARDs are available for RA treatment, and more drugs are under development. There are three groups of patients for whom more evidence is required for RA treatment. The first group comprises individuals with comorbidities (e.g. renal dysfunction, diabetes mellitus, pulmonary disorders, liver dysfunction). Future treatment for this group must focus on minimal structural damages, by setting the treatment target to an alternative one in order not to aggravate the comorbidities, as the highest priority. For RA patients with comorbidities use of drugs should be limited, for example, tacrolimus and GC in diabetes, MTX in interstitial pneumonia or renal dysfunction. In clinical practice, we need deal with such patients individually, however, it is very helpful if there is guidance how to treat RA patients with comorbidity separately, like hepatitis B virus (HBV), tuberculosis in order to minimize structural damages. So we need evidence obtained by large-scale study to lead a better way to deal with such patients. Secondly, there are the patients who cannot afford the high cost of bDMARDs or targeted synthetic DMARDs (tsDMARDs). For such patients, four options are possible: (1) set the treatment target to an alternative target by accepting some degree of structural damage; (2) make full use of cheaper csDMARDs; (3) use bDMARDs or tsDMARDs at a reduced dose or for longer intervals; (4) use cheaper bDMARDs or tsDMARDs [i.e. biosimilar DMARDs (bsDMARDs)]. For option (2), triple therapy with MTX plus SSZ and HCQ is a good choice with clear evidence. Also one single intramuscular GC injection and a low-dose oral GC tapering scheme were suggested as sufficient bridging therapy [De Jong et al. 2014]. For option (3) several reports have focused on tapering studies for DMARDs. For option (4), combination therapy of MTX with bsDMARDs must be investigated. Thirdly, for patients sufficiently fit to undergo aggressive treatment, abiding strictly to the T2T strategy will result in immunologic remission beyond functional remission, which means that the patients are thought to be ‘cured’ of RA. Immunologic status of these patients would become double negative again in rheumatoid factor and anticyclic citrullinated peptide antibody, like the status as several years before the development of symptoms [Nielen et al. 2004]. It is unclear whether such patients need continue aggressive treatment, although Chan and colleagues warn about the possibility of dangers in dose reduction or cessation of bDMARD approaches especially for established disease [Chan et al. 2016].
Conclusion
Combination therapy with bDMARDs is more efficacious than monotherapy using a csDMARD or bDMARD. Evidence suggests that MTX is the best drug for this purpose (provided that MTX is not contraindicated). MTX can be expected to intensify the favorable effects of bDMARDs or reduce their adverse effects, thereby increasing retention of both drugs. In this regard, MTX should be used as an anchor drug. However, MTX can be contraindicated owing to comorbidities, but bDMARD monotherapy could have a role in such cases. Also, the cost of RA treatment for individual patients and governments is a limiting issue. Hence, finding the most effective drug regimen at the lowest cost will be the aim of RA treatment in the future.
Footnotes
Conflict of interest statement
The authors declare that there is no conflict of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.