The Use of Methotrexate in Rheumatoid Arthritis
`Andrea T. Borchers, Carl L. Keen, Gurtej S. Cheema, and M. Eric Gershwin
`
`Objective: To address the long-term efficacy and toxicity issues related to
`methotrexate (MTX) and compare it with other disease-modifying antirheu-
`matic drugs (DMARDs).
`Methods: Review of the international literature on the clinical use of MTX in
`rheumatoid arthritis (RA) disease.
`Results: MTX has emerged as a relatively safe and effective treatment for RA
`that compares favorably with other therapies, particularly because of its
`considerably longer median drug survival. The toxicity profile of MTX is well
`established and includes serious and sometimes fatal liver disease, pneumo-
`nitis, and cytopenias. Hence, regular and careful monitoring of patients taking
`MTX is essential, particularly when MTX is combined with other DMARDs.
`Folate supplementation can reduce some of the most common side effects of
`MTX, but it has not yet been established whether this translates into a reduced
`risk of serious disease. Another potential approach to reducing the toxicity of
`MTX is therapeutic drug monitoring and dose individualization. However,
`correlations between pharmacokinetics and clinical response have been ad-
`dressed in only a few studies and with conflicting results.
`Conclusions: MTX is an effective DMARD with a relatively safe profile
`compared with other therapies. Folate supplementation can significantly re-
`duce the risk of MTX toxicity. Finally, it is essential that patients be monitored
`carefully to reduce the potential serious toxicities of MTX.
`Semin Arthritis Rheum 34:465-483. © 2004 Elsevier Inc. All rights reserved.
`
`INDEX WORDS: Bone marrow suppression; liver toxicity; methotrexate; rheu-
`matoid arthritis.
`
`RHEUMATOID ARTHRITIS (RA) is a sys-
`
`temic autoimmune disease characterized by
`chronic inflammation of synovial joints resulting in
`destruction of cartilage and, ultimately, bone.
`There is no cure for RA, and treatment aims at
`limiting joint damage, preventing loss of function,
`and decreasing pain. The armamentarium of drugs
`used for these purposes includes nonsteroidal an-
`tiinflammatory drugs, disease modifying antirheu-
`matic drugs (DMARDs), and corticosteroids. Un-
`fortunately, joint destruction can often progress
`despite treatment, leading to deformity and disabil-
`ity in a substantial number of patients. In recent
`years, several studies have shown that a greater
`impact on slowing disease progression can be
`achieved if patients with recent-onset RA are
`treated with DMARDs earlier than had previously
`been recommended (1). Hence, the current Amer-
`ican College of Rheumatology (ACR) guidelines
`recommend initiation of DMARD treatment within
`3 months of diagnosis (2). According to these
`guidelines, methotrexate (MTX) is the standard by
`
`which new DMARDs are evaluated, and MTX as
`monotherapy, or
`in combination with other
`DMARDs, should be instituted if a satisfactory
`
`From the Division of Rheumatology, Allergy and Clinical
`Immunology, and the Department of Nutrition, University of
`California at Davis, Davis, CA.
`Andrea T. Borchers, PhD: Visiting Postdoctoral Scholar,
`Division of Rheumatology, Allergy and Clinical Immunology,
`University of California at Davis, Davis, CA; Carl L. Keen,
`PhD: Professor and Chair, Department of Nutrition, University
`of California at Davis, Davis, CA; Gurtej S. Cheema, MD:
`Assistant Clinical Professor of Medicine, Division of Rheuma-
`tology, Allergy and Clinical Immunology, University of Cali-
`fornia at Davis, Davis, CA; M. Eric Gershwin, MD: Professor
`of Medicine, Division of Rheumatology, Allergy and Clinical
`Immunology, University of California at Davis, Davis, CA.
`Address reprint requests to M. Eric Gershwin, MD, Division
`of Rheumatology, Allergy and Clinical Immunology, University
`of California at Davis School of Medicine, TB 192, Davis, CA
`95616. E-mail: megershwin@ucdavis.edu
`© 2004 Elsevier Inc. All rights reserved.
`0049-0172/04/3401-0000$30.00/0
`doi:10.1016/j.semarthrit.2003.12.003
`
`Seminars in Arthritis and Rheumatism, Vol 34, No 1 (August), 2004: pp 465-483
`
`465
`
`Medac Exhibit 2013
`Koios Pharmaceuticals v. Medac
`IPR2016-01370
`Page 00001
`
`

`

`466
`
`BORCHERS ET AL
`
`response has not been achieved with hydroxychlo-
`roquine (HCQ) or sulfasalazine (SSZ). In patients
`with very active disease or a poor prognosis, it may
`even be advisable to use MTX as the initial
`DMARD. This has been the practice among rheu-
`matologists in the United States even before pub-
`lication of the updated guidelines (3). For patients
`with severe RA, MTX is the DMARD of first
`choice for rheumatologists in the United States and
`in Europe (England and The Netherlands), and in
`both regions it is the anchor drug in therapy based
`on the combination of 2 DMARDs (4,5).
`
`METHODS
`A Medline search of published studies with key
`words
`“methotrexate,”
`“rheumatoid arthritis,”
`“DMARDs,” and “amethopterin” from 1950 to
`2003 was conducted. From these studies and their
`accompanying references, a total of 1114 studies
`dating from 1952 to the present were reviewed and
`131 were determined to be pertinent to our discus-
`sion. These manuscripts included a variety of sub-
`jects, including pharmacology, therapy of connec-
`tive-tissue diseases, and use of DMARDs. The
`manuscripts were reviewed with particular refer-
`ence to the usage of MTX in RA, with emphasis on
`double-blinded studies and on mechanisms of ac-
`tion.
`
`Pharmacology
`MTX (amethopterin) is a methyl-derivative of
`aminopterin, a drug that was first used for the
`treatment of RA in 1951. Both aminopterin and
`MTX are folate analogues, but they are also folate
`antagonists in that they compete with folate for
`cellular uptake and intracellular polyglutamation
`and they reversibly inhibit dihydrofolate reductase
`(DHFR). DHFR participates in single carbon trans-
`fers necessary for methionine, purine, and thymi-
`dylate synthesis and, ultimately, DNA synthesis.
`The inhibition of other folate-dependent enzymes
`by polyglutamated MTX further contributes to this
`reduction of nucleotides available for cellular pro-
`liferation. Through partial substrate depletion and
`inhibition of thymidylate synthase, MTX polyglu-
`tamates block the de novo synthesis of thymidy-
`late. In addition, they interfere with de novo purine
`synthesis by directly blocking the activity of
`5-amino-imidazole-4-carboxamide ribonucleotide
`(AICAR) transformylase and by resulting in the
`accumulation of dihydrofolate polyglutamates,
`
`inhibitors of this
`
`which are even more potent
`enzyme (6).
`The inhibition of DNA synthesis rapidly affects
`proliferating cells, such as malignant cells, and
`also lymphocytes. This is likely the major mecha-
`nism of action of high-dose MTX in leukemia and
`other malignancies. “High-dose” refers to MTX
`therapy consisting of between 200 mg/m2 and 30
`g/m2 body surface administered intravenously over
`a period of 24 hours, first at weekly intervals and
`then at much longer intervals. However, patients
`with RA and psoriasis are treated with low-dose
`MTX (ie, doses ranging between 7.5 and 20 mg per
`week). However, we caution that “low-dose” refers
`to the comparative use of higher dosages in cancer
`and does not imply the doses used are nontoxic.
`The mechanism(s) of action of low-dose MTX
`have still not been fully elucidated. It has recently
`been reported that low concentrations of MTX
`either inhibit proliferation or induce apoptosis in
`activated, but not resting, peripheral blood lym-
`phocytes (7). These findings suggest that antipro-
`liferative and immunosuppressive effects contrib-
`ute to the effectiveness of
`low-dose MTX.
`However, somewhat conflicting results have been
`reported concerning the effect of MTX on prolif-
`erative responses, partly caused by the use of
`different methodologies for measuring prolifera-
`tion (8).
`MTX exerts a variety of antiinflammatory activ-
`ities (6,9). These are at least partly mediated via
`the increased release of adenosine as a conse-
`quence of AICAR transformylase inhibition by
`MTX polyglutamates. Inhibition of this enzyme
`results in the accumulation of AICAR, which in
`turn inhibits adenosine deaminase, thereby increas-
`ing the intracellular adenosine concentration and
`the subsequent release of adenosine into the extra-
`cellular fluid. Adenosine exhibits a variety of an-
`tiinflammatory activities, including the inhibition
`of leukocyte accumulation and neutrophil-medi-
`ated endothelial injury at sites of inflammation; the
`reduction in tumor necrosis factor (TNF)-␣synthe-
`sis; and inhibition of a variety of natural killer,
`monocyte/macrophage, and T-cell activities (6,9-
`11).
`In addition to its effects on de novo purine
`synthesis,
`low-dose MTX therapy was recently
`reported to be associated with decreased activity of
`several enzymes participating in the salvage path-
`way of purine synthesis (12). This effect was seen
`
`Page 00002
`
`

`

`CLINICAL USE FOR METHOTREXATE
`
`467
`
`after 48 weeks, but not after 6 weeks, of MTX
`treatment, was independent of folate supplementa-
`tion, and was not associated with measures of
`MTX efficacy or toxicity. An exception was 5⬘-
`nucleotidase, which exhibited decreased activity
`only in those patients who discontinued MTX
`because of hepatotoxicity.
`A further mechanism of action may involve an
`MTX-induced reduction in the synthesis of poly-
`amines via inhibiting the conversion of homocys-
`teine to methionine, the precursor for the methyl-
`donor S-adenosyl-methionine that is required for
`polyamine synthesis. Polyamines may play a role
`in the pathogenesis of RA and other inflammatory
`diseases (9).
`
`Efficacy in RA
`MTX is among the most commonly used
`DMARDs in the treatment of patients with RA. It
`is taken once a week in doses commonly ranging
`between 7.5 and 20 mg. Whereas in some clinical
`trials the weekly dose was divided into 3 equal
`aliquots taken in 12-hour intervals (13,14), patients
`with RA generally take the entire dose at once. The
`most common route of administration is oral, but
`parenteral (intramuscular, subcutaneous, and occa-
`sionally intravenous) routes are preferred for some
`patients. Although bioavailability of oral MTX is
`similar to that of parenteral MTX (15-17), there are
`reports of patients who respond to intramuscular
`injections but not to oral MTX. A possible expla-
`nation for this observation is that the bioavailabil-
`ity of oral MTX relative to intramuscular MTX
`was 13.5% lower at the usual maintenance dose
`(mean, 17 mg) than at the 7.5-mg dose commonly
`used at the initiation of MTX therapy (18).
`The efficacy of MTX in the treatment of RA has
`been shown in a number of randomized, controlled
`trials (RCTs) (19,20). A Cochrane systematic re-
`view concluded that, despite a fairly high (22%)
`withdrawal rate caused by adverse events, short-
`term MTX therapy provided substantial benefit to
`patients with RA and this benefit was both clini-
`cally and statistically significant (20). Compared
`with placebo, MTX therapy induces significant
`improvement in the number of tender and swollen
`joints, pain, physician and patient global assess-
`ment, and functional status. Approximately 60% to
`70% of patients have a significant response to
`MTX, although the estimates depend to some ex-
`tent on how “significant response” is defined (19).
`
`The onset of MTX-induced improvement is gener-
`ally within 3 months in the majority of patients
`who will eventually respond. After an initial period
`of rapid improvement, a plateau in the response is
`often reached after 6 to 12 months (21,22).
`In direct comparisons of traditional DMARDs in
`RCTs,
`their efficacy is generally quite similar,
`although MTX has been found to be superior to
`azathioprine and auranofin (19,23). The duration of
`most clinical trials is between 6 months and 2
`years, which is short compared with the natural
`history of RA in the majority of patients. In addi-
`tion, only select patients are included in RCTs and
`a variety of aspects of patient treatment, including
`dose adjustments and drug termination, are han-
`dled differently in RCTs than in common clinical
`practice. Long-term observational
`studies, al-
`though frequently not randomized and therefore
`liable to some bias, can offer a different perspec-
`tive on the efficacy and toxicity of a drug and on
`the reasons for its discontinuation.
`With few exceptions (24), long-term observa-
`tional studies indicate that toxicities–rather than
`lack of efficacy–are the most common cause of
`discontinuing MTX (22,25-32). As will be dis-
`cussed later, folic acid supplementation can reduce
`some of the toxicities of MTX. Many of the avail-
`able data stem from a time when patients treated
`with MTX did not routinely receive folate supple-
`mentation, and, in some of the more recent studies,
`information on the extent of supplementary folate
`intake was unavailable. This makes it difficult to
`assess whether the now-widespread use of folate
`has had an impact on the frequency of MTX
`discontinuations and the reasons for them.
`Studies comparing different second-line drugs
`show that discontinuations are less frequent
`in
`patients taking MTX than in those taking other
`DMARDs (24,26,27). Thus, MTX therapy has the
`highest median drug survival among DMARDs
`and the greatest cumulative probability of being
`continued for 5 years or even longer. The median
`time on MTX of 4.25 years was more than twice as
`long as that of other DMARDs (1.6 to 2 years)
`(26). Others have reported similar, occasionally
`even longer, median survival times (24,25,27). The
`percentage of patients still taking MTX after 5
`years ranges between 45% and 65% (25,27,31,33),
`but it has been reported to be as high as 73% in 1
`study (22) and only 25% in another (29). Even
`after 10 years, 30% of patients with RA continued
`
`Page 00003
`
`

`

`468
`
`BORCHERS ET AL
`
`to take MTX at a tertiary care center in Alabama
`(30). At last follow-up, 53% of patients with RA
`treated in 6 Australian community practices were
`still taking MTX after 12 years, but this percentage
`decreased to 38% if intermittent discontinuations
`of ⬎3 months were considered as terminations
`(32). It should be emphasized that careful moni-
`toring is required no matter how long a patient
`takes MTX.
`Notwithstanding its long median survival times,
`MTX is associated with some serious toxicities,
`discussed in more detail later. Thus, despite its
`proven effectiveness, it is important to examine
`whether newer DMARDs offer similar or even
`greater effectiveness with fewer toxicities. In re-
`cent years, several new agents for the treatment of
`RA have been developed, including leflunomide
`(Arava; Aventis, Bridgewater, NJ) and anticyto-
`kine agents, such as etanercept, infliximab, and
`anakinra.
`Leflunomide is a new DMARD approved for the
`treatment of RA by the Food and Drug Adminis-
`tration (FDA) in 1998. It is an isoxazolyl derivative
`that is converted by first-pass metabolism in the
`liver and gut to the active metabolite, A77 1726.
`This metabolite inhibits the rate-limiting enzyme
`in pyrimidine synthesis, dihydroorotate dehydro-
`genase, and thereby prevents cellular proliferation,
`particularly that of CD4⫹ T cells, which are known
`to proliferate rapidly during the initiation of RA.
`Recent comparisons between MTX and lefluno-
`mide indicate that they are similarly effective in
`treating RA (34,35), equally slowing radiographic
`progression (36) and both showing sustained clin-
`ical benefit over at least a 2-year period (35,37)
`(Tables 1 and 2). In 1 study, the clinical response
`rates to leflunomide were equivalent to those seen
`with MTX during the first year (34) but signifi-
`cantly higher after the second year (37). Con-
`versely, in another study, MTX resulted in greater
`improvements did than leflunomide in all outcome
`measures after 1 year, but the benefits from the 2
`drugs were similar at the end of the second year in
`the subgroup of patients who chose to continue the
`double-blind treatment (35). The frequency of ad-
`verse events, including serious ones, was similar
`with the 2 drugs, although the proportion of with-
`drawals because of toxicity tended to be higher in
`the leflunomide group in 1 trial (35). However, in
`the MTX group, there were 2 treatment-related
`deaths, whereas no deaths occurred in the lefluno-
`
`mide group. Asymptomatic transaminase eleva-
`tions resulting in withdrawal were more than twice
`as frequent in the leflunomide group than in the
`MTX group in 1 trial (7.1% vs 3.3%, respectively)
`(34), but half as frequent in another (1.6% vs 3.2
`%) (35).
`The effect of the addition of leflunomide to a
`stable, but clinically unsatisfactory, dose of MTX
`was assessed in an open-label trial (38) and an
`RCT (39). In the open-label trial, 20% of subjects
`met the ACR20 response criteria after 1 month of
`leflunomide treatment; this percentage increased to
`57% after 9 months, with a similar response rate
`maintained at 12 months. Similar results were
`reported in the RCT (39). Although generally well
`tolerated,
`the combination of
`leflunomide and
`MTX was associated with elevated liver enzymes
`in 63% of patients (38). Elevations of alanine
`aminotransferase (ALT) ⬎2 times the upper limit
`of normal were observed in 34% of the patients
`and elevations of aspartate aminotransferase (AST)
`were observed in 24%. For comparison, in a ran-
`domized study comparing MTX and leflunomide,
`serum transaminase elevations 2 times the upper
`limit of normal occurred in 9.3% of patients treated
`with MTX and 11% of patients treated with le-
`flunomide (34).
`The first report of serious liver disease (early
`cirrhosis) associated with the use of a combination
`of leflunomide and MTX was published in 2000
`(40). Since then, a number of serious hepatic ab-
`normalities have been described in postmarketing
`reports, resulting in considerable controversy over
`the safety of leflunomide (41). This issue was
`recently addressed by an FDA-appointed advisory
`panel. Based on the currently available evidence,
`this panel deemed the safety of leflunomide ade-
`quate and acceptable and its benefits greater than
`its potential safety risks.
`
`Anticytokine Agents
`Based on the recognition that TNF-␣ and inter-
`leukin (IL)-1 play important roles in the pathogen-
`esis of RA (42), several biologic agents that block
`the activity of these two proinflammatory cytokines
`have been developed. These include etanercept, a
`soluble TNF receptor (p75):Fc fusion protein; in-
`fliximab, a chimeric human-murine monoclonal
`anti–TNF-␣antibody; and anakinra, a recombinant
`human IL-1 receptor antagonist (IL-1ra).
`Etanercept (Enbrel; Immunex, Seattle, WA) has
`
`Page 00004
`
`

`

`CLINICAL USE FOR METHOTREXATE
`
`469
`
`Table 1: MTX Versus Leflunomide or Etanercept in 12-month RCTs
`
`MTX (34)
`
`Leflunomide
`
`MTX (35)
`
`Leflunomide
`
`MTX* (48)
`
`Etanercept
`
`7.5 mg/wk,
`increased
`to 15 mg/
`wk if active
`disease
`persisted
`
`Loading dose
`of 100 mg/
`d for 3 d;
`20 mg/d
`thereafter
`
`7.5 mg/wk in
`wks 1-4, 10
`mg/wk in
`wks 2-12,
`then 10-15
`mg/wk
`
`Loading dose
`of 100 mg/
`d for 3 d:
`20 mg/d
`thereafter
`
`46
`23
`9.5 (6.5) wk
`
`52
`34
`8.6 (7.4) wk
`
`64.8†
`—
`101 ⫾ 92.5 d†
`
`50.5
`—
`74 ⫾ 80 d
`
`25 mg sub-
`cutan-
`eously
`twice
`weekly
`
`7.5 mg/wk,
`increased
`to 15 mg/
`wk at wk 4,
`and to 20
`mg/wk at
`wk 8
`72
`65
`MTX ⬍ Etanercept†
`MTX ⬎ Etanercept†
`
`⫺6.6 ⫾ 7.6
`⫺5.4 ⫾ 5.5
`0.88 (3.3)†
`
`⫺7.7 ⫾ 7.8
`⫺5.7 ⫾ 6.5
`0.53 (4.5)
`
`⫺9.7 ⫾ 7.9†
`⫺9.0 ⫾ 7.3†
`—
`
`⫺8.3 ⫾ 7.9
`⫺6.8 ⫾ 7.3
`—
`
`—
`—
`1.59
`
`—
`—
`1.00
`
`—
`
`—
`
`0.03
`
`0.03
`
`—
`
`—
`
`0.48 (1.8)
`
`0.23 (2.2)
`
`—
`
`—
`
`1.03†
`
`0.47
`
`10‡
`
`24‡
`
`22
`
`17
`
`15‡
`
`3‡
`
`19
`
`7
`
`11†
`
`4
`
`5
`
`5
`
`Dose
`
`ACR20 (%)
`ACR50 (%)
`Time to reach
`ACR20
`Joint count
`Tender
`Swollen
`Change in
`Sharp
`score
`(units)
`Change in
`Larson
`score
`(units)
`Change in
`erosion
`score
`(units)
`Discontinuation
`Because of
`toxicity
`Because of
`lack of
`efficacy
`
`*Some data were presented as figures (ie, actual numbers were not provided).
`†Statistically significantly different from treatment comparison.
`‡No statistical evaluation was presented.
`
`been approved by the FDA for RA as monotherapy
`or in combination with MTX. The approval was
`based on the significant clinical benefit, as assessed
`by ACR criteria, that this agent exhibited either
`alone or in addition to MTX in multi-center, dou-
`ble-blind, placebo-controlled trials (43-45). Clini-
`cal trials of longer duration have further confirmed
`the efficacy of etanercept and have added the
`findings that the benefit is sustained over at least a
`2-year period and that etanercept can slow, or even
`halt, radiographic progression (46,47).
`A recent RCT directly compared MTX and et-
`
`anercept (48). The weekly MTX dose was in-
`creased from the initial 7.5-mg dose to 20 mg in all
`patients, but the dose had to be reduced in 15% of
`them because of adverse events or serum transam-
`inase elevations. Etanercept was administered sub-
`cutaneously twice a week at a dose of either 10 mg
`or 25 mg. Compared with MTX, the higher dose of
`etanercept was found to induce an ACR20 re-
`sponse faster and in a greater proportion of patients
`during the first 6 months of the 12-month trial.
`Thereafter, the percentages were similar in the 2
`groups. Several factors are likely to account for the
`
`Page 00005
`
`

`

`470
`
`BORCHERS ET AL
`
`Table 2: MTX Versus Leflunomide or Etanercept in 24-month RCTs
`
`MTX (37)
`
`Leflunomide
`
`MTX (35)
`
`Leflunomide MTX (47) Etanercept
`
`Dose
`
`7.5-15 mg/wk
`
`20 mg/d
`
`10-15 mg/wk
`
`20 mg/d
`
`ACR20 (%)
`ACR50 (%)
`Joint count
`Tender
`Swollen
`Change in Sharp score
`(units)
`Change in Larson score
`(units)
`Change in erosion score
`(units)
`Discontinuation because
`of adverse events
`Toxicity
`Lack of efficacy
`
`67*
`43†
`
`79
`56
`
`71.7
`—
`
`64.3
`—
`
`⫺8.8 ⫾ 9.0
`⫺7.7 ⫾ 6.7
`1.2 ⫾ 3.8
`
`⫺10.1 ⫾ 6.7 ⫺10.9 ⫾ 8.2
`⫺8.2 ⫾ 6.1 ⫺10.3 ⫾ 7.3*
`1.6 ⫾ 4.2
`—
`
`⫺10.5 ⫾ 8.1
`⫺9.1 ⫾ 7.7
`—
`
`—
`
`—
`
`MTX better than
`leflunomide‡
`
`0.6 ⫾ 1.7
`
`1.0 ⫾ 3.1
`
`—
`
`17
`—
`
`27
`—
`
`4
`6
`
`—
`
`6
`8
`
`20 mg/wk 25 mg twice
`weekly
`72
`49
`
`59*
`42
`
`—
`—
`3.2*
`
`—
`
`1.9*
`
`12*
`—
`
`—
`—
`1.3
`
`—
`
`0.7
`
`7
`—
`
`*Statistically significantly different from treatment comparison.
`†P ⫽ .053.
`‡In the second year, there was no further increase in joint damage in lefunomide and a small improvement in MTX, resulting in a
`small but statistically significant treatment difference.
`
`observation that etanercept gave more favorable
`results than MTX in the early phase of the trial.
`One is that MTX is started at a low dose and
`gradually increased, another is the previously men-
`tioned finding that it takes approximately 6 months
`for the response to MTX to plateau.
`The trial was continued with 512 of the original
`632 patients in an open-label manner for another
`12 months (47). During this time, patients contin-
`ued to receive the therapy to which they had been
`randomized. The benefits from both treatments
`were sustained, and etanercept continued to pro-
`vide a greater clinical response so that 72% of
`patients in the 25-mg etanercept group, after 24
`months, met ACR20 improvement compared with
`59% of the MTX group. Although the subjects
`participating in this trial had RA for ⱕ3 years, the
`eligibility criteria were designed to insure that only
`patients at high risk of radiographic progression
`were recruited. It is noteworthy that, at the end of
`the trial, significantly more patients in the 25 mg
`etanercept group than in the MTX group showed
`no radiographic progression from baseline (no in-
`crease in total Sharp score in 63% vs 51% and no
`increase in erosions in 70% versus 58% of pa-
`
`tients). In addition, the mean changes in total Sharp
`score and erosion score were significantly lower in
`the 25-mg etanercept group (1.3 and 0.7 units,
`respectively) than in the MTX group (3.2 and 1.9
`units, respectively).
`infliximab, particularly
`The effectiveness of
`when combined with MTX, has also been shown in
`multicenter RCTs (49,50) and infliximab has been
`approved by the FDA for use in RA. Infliximab has
`not been compared directly with MTX but it has
`been evaluated in combination with MTX. In an
`RCT involving patients with RA who were not, or
`only partially, responsive to MTX, addition of
`infliximab in 4 different doses to their stable dose
`of MTX was shown to result in a rapid improve-
`ment that was sustained over 30 weeks (50). Total
`ACR20 response rates were significantly higher in
`the MTX plus infliximab groups than in the MTX
`plus placebo group, and substantially more inflix-
`imab-treated patients achieved ACR50 and ACR70
`responses. The clinical benefit was sustained over
`24 weeks and, importantly, radiographic progres-
`sion was essentially halted in the patients receiving
`MTX plus infliximab, whereas a 10% increase in
`the total radiographic score was noted in the group
`
`Page 00006
`
`

`

`CLINICAL USE FOR METHOTREXATE
`
`471
`
`receiving MTX plus placebo (51). The frequency
`and types of adverse events,
`including serious
`ones, did not differ significantly between the MTX
`plus infliximab and the MTX plus placebo groups.
`Anakinra (recombinant IL-1ra) has been re-
`ported to significantly improve measures of clini-
`cal disease activity and slow the rate of radio-
`graphic progression in patients with RA (52-54),
`and it has been approved by the FDA for the
`treatment of RA. A direct comparison of this bio-
`logic agent with MTX has not yet been conducted,
`but the combination of the 2 in patients with only
`suboptimal responses to MTX alone induced an
`ACR20 response faster and in a greater proportion
`of patients than did MTX plus placebo (55). The
`response to anakinra was dose-dependent, and the
`changes seen with doses of 1.0 mg/kg and 2.0
`mg/kg reached statistical significance relative to
`placebo.
`The cost of biologic agents is considerably
`higher than that of most
`traditional DMARDs.
`According to the ACR, the estimated annual drug
`cost for usual maintenance doses in 2001 were
`$697 to $1859 for MTX, $509 to $763 for SSZ,
`$2938 for leflunomide, $15436 for Etanercept, and
`$13,940 to $36,694 for infliximab (not including
`physician’s office visit fees or laboratory costs
`associated with monitoring) (2). Cost effectiveness
`increasingly enters into physicians’ decisions of
`which drug to prescribe (56). Because of the lack
`of longitudinal data that allow for the calculation
`of the cost effectiveness of these newer treatment
`options, only some preliminary estimates are avail-
`able (2,57).
`
`Combination Therapy
`In addition to combinations of biologic agents
`and traditional DMARDs, regimens combining 2
`or 3 traditional DMARDs are being used with
`increasing frequency in clinical practice. This trend
`has developed despite the limited number of RCTs
`that have addressed the issue of whether combina-
`tions of DMARDs provide greater benefit than
`monotherapies without increasing toxicity. A sys-
`tematic review of 20 studies meeting the inclusion
`criteria covered 8 trials of MTX combined with
`other DMARDs (58). Whereas the double therapy
`consisting of MTX with azathioprine, auranofin, or
`SSZ did not provide greater benefit than MTX
`alone, the combination of MTX with cyclosporin
`was superior to MTX alone. Triple or quadruple
`
`therapy involving MTX plus HCQ plus SSZ with
`or without prednisone has been reported to result in
`higher response rates and less radiographic pro-
`gression than MTX alone or MTX combined with
`only 1 other DMARD (59-62). Direct comparisons
`of the results of such studies are often hampered by
`significant differences in study design, particularly
`with respect to the type of patients recruited, dos-
`ages and permitted dosage increases, primary out-
`come measures, and the criteria used to define
`them (Table 3).
`Whereas many of the published studies used a
`parallel or step-up approach, the step-down (from
`aggressive high dosages to less aggressive lower
`dosages) approach has also been investigated in
`only 1 study by comparing SSZ alone with a
`combination of SSZ, prednisolone, and MTX in
`patients with early but very active RA (63). The
`initial prednisolone dose of 60 mg/d was tapered
`rapidly to 7.5 mg/d by week 7, maintained at that
`level until week 28, then gradually tapered and
`stopped. MTX therapy consisted of 7.5 mg/wk
`until week 40, followed by gradual tapering and
`termination by the end of week 46. Patients in the
`combined-treatment group improved faster and
`achieved a higher response (ACR20) rate after 28
`weeks than the SSZ group. Once prednisone was
`discontinued, some of the initial benefit was lost,
`whereas the SSZ group continued to slightly im-
`prove. After 56 weeks, the differences in measures
`of clinical efficacy were no longer significant.
`However, joint destruction and erosion scores were
`significantly lower in the combined group at 28,
`56, and 80 weeks. This research study suggests that
`prednisolone can significantly slow radiographic
`progression. Others have reported similar findings
`(64,65), but follow-up in 1 of the study cohorts
`indicated that joint destruction resumed on with-
`drawal of steroids (66). This contrasts with the
`findings after 4 to 5 years of follow-up in the SSZ
`and combined SSZ/MTX/prednisolone groups de-
`scribed previously (63) that were recently pub-
`lished (67). Compared with the group treated with
`SSZ alone, the combined-therapy group continued
`to exhibit a significantly reduced rate of radio-
`graphic progression (3.7 points per year less, even
`after adjustment for differences in treatment and
`disease activity during follow-up).
`Although the results of some of these clinical trials
`with DMARD and/or prednisolone combinations are
`promising, it appears that the effects are additive
`
`Page 00007
`
`

`

`472
`
`BORCHERS ET AL
`
`Table 3: Endpoints Used in the Evaluation of DMARDs
`
`Description
`
`Comments
`
`To qualify as a responder, the patient
`must complete the trial and must
`meet ACR response criteria at the
`end of the trial; some researchers
`report the percentage of patients
`fulfilling the criteria at some point
`during the trial
`The actual number of joints should
`be 71 for tenderness and 68 for
`swelling, but this number varies
`greatly among studies
`A variety of different questionnaires
`have been used
`The possibility of tradeoffs between
`efficacy and toxicity does not enter
`into the ACR definition of response
`Not as widely used and accepted as
`the ACR response criteria
`
`Provide some measure of functional
`status, but are generally used only
`as secondary outcome measures
`This measure is particularly valuable
`for comparisons between drugs or
`between single drugs and drug
`combinations; unfortunately, the
`definitions used for what
`constitutes efficacy and toxicity
`necessitating withdrawal vary
`considerably among studies
`Radiologic assessment provides
`valuable information about the rate
`of disease progression
`
`Endpoint
`
`ACR 20, 50,
`70
`response
`
`ⱖ 20%, 50%, and 70% improvement
`in tender joint count and swollen
`joint count and in at least 3 of the
`following 5 criteria: patient global
`assessment, physician global
`assessment, pain intensity, HAQ,
`CRP, ESR
`
`Other
`composite
`measures
`Other
`individual
`measures
`
`Paulus criteria, DAS, DAS28, EULAR,
`WHO/ILAR criteria
`
`Duration of morning stiffness, 50-foot
`walking time, grip strength
`
`Withdrawal because of lack of
`efficacy or toxicity
`
`Radiographic
`progression
`
`Sharp
`Score
`Larson
`Score
`
`Assessment of joint erosion and joint
`space narrowing by using the
`Sharp or the Larsen scoring
`system or modifications thereof
`Counts the number of erosions from
`0 to 5 regardless of their size*
`Defines grades less precisely by
`using general terms and standard
`reference radiographs
`
`Abbreviations: CRP, C-reactive protein; DAS, Disease Activity Score; ESR, erythrocyte sedimentation rate; EULAR, European League
`Against Rheumatism; HAQ, health assessment questionnaire; WHO/ILAR, World Health Organization/International League Against
`Rheumatism.
`*The van der Heijde modification does take the size relative to the joint surface into account.
`
`rather than synergistic. No increase in toxicity was
`detected with combinations compared with single
`agents in a majority of the short term RCTs, but it
`remains to be established whether this will hold true
`
`for long-term combination treatment. Prednisolone
`therapy in RA, if used inappropriately, is fraught with
`potential toxicity. Further research is clearly needed
`to establish optimal combinations and dosages.
`
`Page 00008
`
`

`

`CLINICAL USE FOR METHOTREXATE
`
`473
`
`Radiographic Progression
`A meta-analysis of trials published until 1991
`indicated that, although MTX did not slow radio-
`graphic progression to a significantly greater extent
`than gold salts, it was superior to azathioprine (68).
`The inclusion in the meta-analysis of case-control
`and long-term observational studies along with
`RCTs may be questioned because treatment dura-
`tion in most of the RCTs was 6 to 12 months,
`whereas some of the observational studies fol-
`lowed up patients for up to 84 months. Data from
`long-term observational studies of patients treated
`with MTX show that radiographic progression oc-
`curs despite MTX therapy in a majority of patients
`(21,22,28,33). However, in all of these studies, a
`subset of patients did not experience radiographic
`worsening of lesions.
`A vast majority of patients included in these
`stud