Clinical Therapeutics/Volume 36, Number 7, 2014

Randomized Controlled Trial Comparing 2 Different StartingDoses of Methotrexate in Rheumatoid Arthritis

Varun Dhir, MD, DM; Mandeep Singla, MD; Nidhi Gupta, MSc; Palvi Goyal, MD;Vinay Sagar, MD; Aman Sharma, MD; Shefali Khanna, MD; and Surjit Singh, MD

Department of Internal Medicine (Rheumatology Unit), Post Graduate Institute of MedicalEducation and Research, Chandigarh, India

Accepted for publication May 31, 2014.http://dx.doi.org/10.1016/j.clinthera.2014.05.0630149-2918/$ - see front matter

& 2014 Elsevier HS Journals, Inc. All rights reserved.

ABSTRACT

Purpose: Methotrexate (MTX) remains the goldstandard disease-modifying antirheumatic drug for thetreatment of rheumatoid arthritis (RA). Few studieshave compared different starting doses of MTX inRA. We hypothesized that starting with a higher MTXdose may be more effective but associated with moreadverse effects. We compared a starting dose of 7.5versus 15 mg per week of MTX followed by similarfast escalation.

Methods: This was an open-label (blinded asses-sor), parallel-group, randomized controlled trial thatincluded RA patients aged 18 to 65 years, not onMTX, and having active disease (Disease ActivityScore for 28 joints using 3 variables [DAS28(3)]Z5.1). Patients were randomized to receive MTX ata starting dose of 7.5 mg (group 1) or 15 mg (group 2)per week. The dose of MTX was escalated by 2.5 mgevery 2 weeks to a maximum of 25 mg. Patients wereseen every 4 weeks, and dose escalation was continuedif DAS28(3) was 42.6 and there were no laboratoryabnormalities (transaminitis [42 � upper limit ofnormal] or cytopenia). The primary endpoint waschange in disease activity at 12 weeks (assessed byusing the DAS28[3]). Secondary endpoints were pa-tient withdrawals and episodes ofcytopenia or trans-aminitis. Adverse effects were ascertained by using aquestionnaire. Both intention-to-treat and per-protocol analyses were performed.

Findings: We enrolled 100 patients (female:maleratio, 78:22) with a mean (SD) age of 43.6 (10.8)years and a disease duration of 4.7 (4.8) years. Atbaseline, patients had a mean DAS28(3) of 6.2 (0.7)and a Health Assessment Questionnaire score of 1.3(0.6). Group 1 (7.5 mg) and group 2 (15 mg) included47 and 53 patients, respectively, with no significantdifferences in baseline characteristics. At 12 weeks, the

July 2014

mean dose of MTX reached was 17.3 (4.6) mg ingroup 1 and 23.6 (3.0) mg in group 2 (P o 0.001).The 2 groups had a similar number of patient with-drawals. The mean change in DAS28(3) at 12 weeksin group 1 (–0.47 [0.86]) and group 2 (–0.55 [0.79])was not significantly different (P ¼ 0.60). The changein the Health Assessment Questionnaire score wasalso similar in the groups. The frequency of episodesof transaminitis (6 and 7; P ¼ 0.8) and cytopenia (1and 2; P ¼ 0.9) did not differ significantly betweengroups 1 and 2, respectively. Results remained thesame according to the per-protocol analysis. Amongadverse effects, nausea was more common in group 2compared with group 1 (relative risk, 1.6 [95% CI,1.1–2.2]).

Implications: There were no significant differencesin efficacy between the 2 starting doses of MTX. Thefast escalation of dose in both groups may have bluntedany advantage of starting at a higher dose. Nauseaoccurred more commonly in patients started on 15 mgof MTX. We suggest longer trials to confirm our find-ings. ClinicalTrials.gov identifier: NCT01404429. (ClinTher. 2014;36:1005–1015) & 2014 Elsevier HS Jour-nals, Inc. All rights reserved.

Key words: methotrexate, randomized controlledtrial, rheumatoid arthritis, starting dose.

INTRODUCTIONRheumatoid arthritis (RA) is a chronic systemic auto-immune disease predominantly affecting synovial jointsthat has a prevalence of 0.5% to 1% worldwide.1

1005

Clinical Therapeutics

Methotrexate (MTX), a folic acid antagonist andderivative of aminopterin, was first used in thetreatment of RA in the 1960s.2 Subsequently, rando-mized controlled trials conducted nearly 3 decadesago3,4 and long-term studies5 have confirmed thedrug’s superior efficacy and better safety profilecompared to other disease modifying antirheumaticdrugs. These studies led to the widespread use ofMTX by the 1990s. Currently, MTX is the mostprescribed disease-modifying antirheumatic drug(DMARD) in RA and is recommended by contempo-rary guidelines as the first line of therapy.6–8 ThisMTX-first strategy has also been recently validated inpoor-prognosis RA.9

The accumulation of decades of experience has ledto the realization that the effective dose of MTX liesbetween 15 and 25 mg per week.10 At the same time,there has been an urgency, on the part of physicians,to achieve this effective dose by 12 weeks, in keepingwith the strategy of “targeting” early and rapidcontrol of disease activity to minimize damage.11

This strategy has led to faster escalation protocols,from the traditional 5 mg every 3 months to 5 mgevery month. This policy has been supported by theresults of a research study.12 In addition, internationalguidelines have recommended starting MTX at higherdoses of 10 to 15 mg per week.8,13,14 Although asystematic review also concluded that starting with ahigher dose (15 mg) was optimal, this recommenda-tion was based on weak evidence.13,15 On performinga literature search, we found only a limited number ofstudies, many dating back 2 to 3 decades, which hadactually compared fixed MTX doses.16,17 There hasbeen no study comparing starting doses of 7.5 and 15mg of MTX. We hypothesized that starting at a higherdose may have an incremental efficacy benefit, but wewere concerned about more adverse effects leading tointolerance and withdrawals. In an earlier study, wehad found that nearly 30% of Indian patients expe-rienced adverse effects with MTX (mean dose, 17 mg);although these effects were “minor” (eg, nausea), theyled to noncompliance.18 Other studies have alsodemonstrated more gastrointestinal adverse effectswith higher doses of MTX.16,19,20

Finding the optimum initial dose of MTX wouldhelp achieve the right balance between efficacy andtolerability. This may lead to more patients remainingon MTX, thus obviating the need for more expensiveor toxic therapies. Thus, the goal of the present study

1006

was to compare differences in efficacy and toxicitybetween 2 dosage regimens of oral MTX consisting ofdifferent starting doses (7.5 or 15 mg per week)followed by similar fast escalation over 12 weeks.

PATIENTS AND METHODSEthics and Conduct

This trial was conducted from May 2011 to June2012 in a rheumatology clinic at a university center inNorth India. The study complied with the guidingprinciples for human research per the Declaration ofHelsinki and was approved by the institutional ethicscommittee. Written informed consent was obtainedfrom all patients. The trial (DMIRA [Different Me-thotrexate Dosages in Rheumatoid Arthritis]) wasregistered in the Clinical Trials Registry of India(CTRI/2011/07/001924) as well as ClinicalTrials.gov(NCT01404429).

This was a parallel-group, open-label (assessor-blinded), randomized controlled trial comparing 2starting doses of MTX (7.5 mg per week and 15 mgper week) followed by similar fast escalation (2.5 mgevery 2 weeks) over 12 weeks in patients with RA.

PatientsPatients with RA attending the rheumatology clinic

were enrolled if they provided written informedconsent and fulfilled the inclusion and exclusioncriteria as given here.

Inclusion criteria included the following: (1) fulfill-ment of the 1987 American College of Rheumatology(formerly American Rheumatism Association) crite-ria21; (2) age between 18 and 65 years; (3) havinghighly active disease (Disease Activity Score for 28joints using 3 variables [DAS28(3)] Z5.1)22; (4)permitted to be on corticosteroids if dose was stablefor Z1 week before enrollment and dose o10 mg/dof prednisolone; and (5) permitted to be on otherDMARDs such as sulfasalazine, leflunomide, andhydroxychloroquine, if their dose was stable Z2weeks before enrollment.

Exclusion criteria included the following: (1) ongoingor recent treatment with MTX (2 months); and (2)currently pregnant or breastfeeding or plans to becomepregnant in the next 6 months. Also excluded werepatients diagnosed with any of the following: (1) chronicliver disease or renal failure; (4) leukopenia (white bloodcell count o4000/μL) or thrombocytopenia (plateletcount o150,000/μL); (5) interstitial lung disease; and

Volume 36 Number 7

V. Dhir et al.

(6) active infection (tuberculosis, HIV, and hepatitis B orC virus). Although not specified a priori, patients withprevious severe reactions to MTX were excluded. Thisincluded patients with previous suspected lung toxicityor severe cytopenia.

InterventionPatients were randomized into 2 groups; group 1

started oral MTX at 7.5 mg (1 tablet of 7.5 mg) andgroup 2 started oral MTX at 15 mg (1 tablet of 15mg) per week. The dose of MTX was escalated byadding 1 tablet of 2.5 mg to the starting dose every 2weeks. Patients were seen every 4 weeks in the clinic,and dose escalation was continued if the DAS28(3)was 42.6 and there were no laboratory abnormalitiesthat were out of range of predefined limits (detailed inassessment on follow-up visits). The dose of MTXcould be escalated to a maximum dose of 25 mg perweek. All patients received folic acid at a dose of 5 mgtwice a week, which could be taken on any day exceptthe day on which MTX was taken. One intramuscularinjection of depot methylprednisolone acetate 80 mgwas allowed per patient during the first 8 weeks ofthe study.

Outcome Measures/EndpointsThe primary endpoint was decline in disease activ-

ity (DAS28[3]) at 12 weeks and the proportion ofpatients with a good response (defined according tothe European League Against Rheumatism [EULAR]response criteria).22 The secondary endpoints were:(1) patients who withdrew due to any cause; (2)patients who developed adverse effects in the formof cytopenia (white blood cell count o4000 orplatelet count o100 � 103/μL) or transaminitis(alanine aminotransferase [ALT] or aspartate amino-transferase [AST] 42 � the upper limit of normal);and (3) patients who withdrew due to intolerance.

Patient AssessmentIn all patients, a recent hemogram, liver and renal

function tests, radiograph (of the hands and chest),and rheumatoid factor status were obtained at base-line. Rheumatoid factor status was defined as positiveaccording to the latex fixation test (420 IU) or Rose-Waaler test (41:16) scores. A 28-joint assessment wasconducted, and the modified disease activity scoreusing 3 variables [DAS28(3)] was calculated by usingthe formula DAS28(3) ¼ [0.56*√ (TJC28) þ 0.28*√

July 2014

(SJC28) þ0.70*Ln (ESR)] *1.08 þ 0.16.22 (TJC28 =Tender Joint count 28, SJC28 = Swollen joint count28 and ESR = Erythrocyte sedimentation rate.) Thefunctional status of patients was measured by using avalidated Indian version of the Health AssessmentQuestionnaire.23

Assessment at follow-up visits was conducted every4 weeks. Disease activity was determined using DAS28(3) by an assessor (V.D.) blinded to the allocationsequence. Patients were also asked to rate the im-provement in symptoms at every follow-up visit on ascale of 0 to 100 (numerical verbal scale). Laboratorytests were conducted at every visit, including a hemo-gram and serum transaminase levels. In case oflaboratory abnormalities in the form of mild trans-aminitis (defined as ALT or AST 42 but o3 � upperlimit of normal) or mild cytopenia (defined as whiteblood count between 3500 and 4000 or platelet countbetween 90 and 100 � 103/μL), further MTX escala-tion was stopped, and the same dose of MTX wascontinued. In case of laboratory abnormalities in theform of severe transaminitis (defined as ALT or AST43 � upper limit of normal) or severe cytopenia(defined as white blood count o3500 or plateletcount o90 � 103/μL), MTX was stopped. In bothof these scenarios, patients were called early for thenext follow-up visit (after 2 weeks instead of 4 weeks),and the laboratory investigations were repeated. If thelaboratory abnormality had normalized by the nextvisit, MTX was resumed at the same dose at which itwas discontinued, and dose escalation was restarted.

Minor adverse effects (symptoms) were determinedby using a questionnaire (used in a previous study).18

Briefly, the questionnaire asked patients to indicate(from a list) the presence of any new symptom. Inaddition, patients were required to rate the severity asmild, moderate, or severe (mild, not interfering indaily activities; moderate, interfering with daily acti-vities; severe, unable to perform any daily activity) aswell as note the duration (in days per week) of theadverse effect. In case a patient missed a visit, he orshe was contacted by telephone to ascertain the rea-son. On the last visit (12 weeks), the functional statusof patients was again measured by using the Indianversion of the Health Assessment Questionnaire.

Sample SizeWe presumed a difference in the final DAS28(3)

between the 15-mg group and the 7.5-mg group to be

1007

Clinical Therapeutics

1.0 (SD, 2.0). For a power of 75% and an α error of0.05, it was required to recruit total of 100 patients.A power of 75% was considered acceptable due tofeasibility.

Randomization, Allocation Concealment,and Blinding

Simple randomization was conducted by usingrandom number tables by a colleague not involvedin the study. Patient allocation was concealed by usingserially numbered, opaque sealed envelopes (num-bered 0–100). A single assessor (V.D., blinded toallocation group), who has formal training in jointcounts, examined patients at all visits. At baseline,after obtaining consent, the patient’s name was writ-ten on the allocation envelope (starting from envelopenumbered 1). The patient was given a prescriptionthat read “Starting dose of MTX (1 tablet) per weekfor 2 weeks, followed by starting dose of MTX (1tablet) þ 2.5 mg MTX (1 tablet) per week for the next2 weeks.” A research assistant in another roomopened the envelope and dispensed 4 tablets of thestarting dose and 2 tablets of 2.5 mg. At every follow-up visit, the patient was first seen and examined by theassessor. The assessor calculated the patient’s diseaseactivity and looked for the presence of any laboratoryabnormalities. The research assistant in another roomthen dispensed the 4 tablets of the starting dose andthe additional multiples of 2.5-mg tablets.

Statistical AnalysisAnalysis was according to an intention-to-treat

basis. In patients who were lost to follow-up, efficacymeasures were imputed from the last observation (lastobservation carried forward). In addition, a per-protocol analysis for patients who completed thestudy (completers) was also conducted for efficacymeasures. The differences in the disease activity andfrequency of adverse effects among the groups werecompared by using Student’s t test and the χ2 test,respectively. SPSS version 15 (IBM SPSS Statistics,IBM Corporation, Armonk, New York) and Graph-Pad Prism (GraphPad Software, Inc, La Jolla, Cal-ifornia) were used for analysis.

RESULTSThis study enrolled 100 patients with RA who wererandomized into 2 groups: group 1 (7.5 mg per week)had 47 patients and group 2 (15 mg per week) had

1008

53 patients. Patients in these 2 groups had similarbaseline characteristics (Table I). The mean (SD) doseof MTX reached at 12 weeks was 17.3 (4.6) and 23.6(3.0) mg per week, respectively (P o 0.001). Inpatients who completed the study, the mean dose ofMTX reached was 19.2 (1.8) and 24.5 (1.7) mg perweek (P o 0.001). The number of patients whoreceived a single depot steroid injection was notstatistically different (group 1, n ¼ 22; group 2, n ¼28 [P ¼ 0.5]) between groups. The number of patientswho withdrew from the study did not differsignificantly between groups (group 1, n ¼ 9; group2, n ¼ 7 [P ¼ 0.4]). The number of patients whospecifically withdrew due to adverse effects (2 and 2)was also not significantly different (P ¼ 0.9) betweengroups 1 and 2, respectively (Figure 1).

There was no difference in disease activity at 4, 8,or 12 weeks between groups. The mean DAS28(3) at4 weeks (6.0 [0.7] and 6.0 [1.0], P ¼ 0.8), 8 weeks(5.9 [0.7] and 5.8 [1.1], P ¼ 0.8), and 12 weeks (5.8[1.0] and 5.6 [1.1], P ¼ 0.5) was not significantlydifferent between groups 1 and 2, respectively. Thus,at 12 weeks, the mean DAS28(3) had declinedsimilarly (P ¼ 0.60) in group 1 (–0.47 [0.86]) andgroup 2 (–0.55 [0.79]). The mean difference in DAS28(3) between the groups at 12 weeks was 0.14 (95%CI, –0.26 to 0.54) (Figure 2). There was only 1 patientwho had an EULAR-defined good response by 12weeks (in group 2). There were no significant differ-ences in the tender joint counts, swollen joint counts,erythrocyte sedimentation rate, or patient-rated im-provement between groups at any visit. The per-protocol analysis also did not find any significantdifferences between groups in these efficacy measures.At 12 weeks, the decline in the Health AssessmentQuestionnaire score in group 1 (–0.17 [0.48]) andgroup 2 (–0.28 [0.44]) was not significantly different(P ¼ 0.22).

There were no significant differences in the episodesof transaminitis (6 and 7; P ¼ 0.8) or cytopenia (1 and2; P ¼ 0.9) that occurred in patients in groups 1 and2. However, the proportion of patients who experi-enced nausea was higher in group 2 compared withgroup 1 (relative risk, 1.6 [95% CI, 1.1–2.2]) (Table II).The total number of episodes of nausea was alsohigher but not statistically different between the 2groups. Among the patients who experienced nausea,the proportion who rated it to be moderate or severewas not significantly different (P ¼ 0.9) between

Volume 36 Number 7

Table I. Baseline characteristics of all patients and comparison of patients included in the 2 study groups.Unless otherwise indicated, values are given as mean (SD) or number (%).

CharacteristicAll Patients(N ¼ 100)

Group 1: MTX7.5 mg (n ¼ 47)

Group 2: MTX15 mg (n ¼ 53)

P: Group 1 vsGroup 2

Age, y 43.6 (10.8) 44.5 (10.3) 42.8 (11.2) 0.4% Female 78 79 77 0.95Body mass index, kg/m2 23.5 (3.4) 23.6 (3.5) 23.4 (3.4) 0.7Duration of disease, y 4.7 (4.8) 4.8 (4.6) 4.7 (4.5) 0.9Disease activity score: DAS28(3)* 6.2 (0.7) 6.2 (0.7) 6.2 (0.8) 0.7Tender joint count (0–28) 13.3 (4.5) 13.4 (4.0) 13.2 (4.9) 0.8Swollen joint count (0–28) 7.9 (4.2) 7.7 (3.5) 8.0 (4.7) 0.8HAQ (0–3)† 1.3 (0.6) 1.3 (0.6) 1.3 (0.6) 0.99ESR, mm 1st hour‡ 63.9 (30.9) 65.4 (30.0) 62.6 (32.0) 0.6Rheumatoid factor positive 63 (63) 26 (55.3) 37 (69.8) 0.1ACPA positive, N positive/N tested (%) 71/82 (86.6) 29/34 (85.3) 42/48 (87.5) 0.8Erosions on hand radiograph 22 (22) 11 (23.9) 11 (22.4) 0.7On oral steroids at enrollment 13 (13) 5 (10.6) 9 (17.0) 0.4Previously taken MTX 11 (11) 7 (14.9) 4 (7.5) 0.2On other DMARDs at inclusion§ 3 (3) 2 (4.2) 1 (1.8) 0.6

ACPA ¼ anticitrullinated protein antibody; MTX ¼ methotrexate.*Disease Activity Score for 28 joints using 3 variables (DAS28[3]): calculated using tender joint count (28), swollen jointcount (28), and erythrocyte sedimentation rate (Westergren method).

†Scores on the Health Assessment Questionnaire (HAQ), using the Indian validated version, ranged from 0 to 3.‡Modified Westergren method.§Disease-modifying antirheumatic drugs (DMARDs) included hydroxychloroquine in 1 patient in both groups, andsulfasalazine in 1 patient in the 7.5-mg group.

V. Dhir et al.

group 1 (21.7%) and group 2 (25%). The duration ofnausea was also not significantly different (1.6 [0.9] vs1.8 [1.4] days; P ¼ 0.5) between groups. We alsocompared the frequency of nausea during the periodwhen patients in both the groups were on a similardose of MTX. For this purpose, episodes of nausea inthe last 8 weeks of group 1 (MTX dose, 12.5–20 mg/week) were compared with the first 8 weeks of group2 (MTX dose, 15–22.5 mg/week). Although episodesof nausea remained higher in group 2 compared withgroup 1, this finding was not statistically significant(24 vs 13; P ¼ 0.07).

DISCUSSIONWe found no significant differences in efficacy, interms of change in DAS28(3), between the 2 dosageregimens of MTX. In addition, there were no signifi-cant differences in the functional status (measured by

July 2014

using the Indian version of the Health AssessmentQuestionnaire) or in patient-rated improvementbetween the groups at 12 weeks. A limited numberof studies have compared different starting doses ofMTX. We searched PubMed with the key words“MTX[TI] AND rheumatoid [TI] AND clinical trials”and found only 4 studies that directly compareddifferent dosage regimens of MTX. One more studywas found by cross-referencing (which had beenpresented as an abstract). None of these studies hadcompared the dosage regimens used in the presentstudy (Table III).

Among these 4 studies, 2 were conducted in the1980s and included only a small number of patients.Furst et al17 evaluated a dose of 5 mg/m2 per week(5–10 mg) and 10 mg/m2 per week (12.5–20 mg) in13 and 17 patients, respectively, over 18 weeksand found a significant dose–response relationship.

1009

115 patients screened forinclusion

100 patients randomized

Group1n = 47 allocated to MTX 7.5mg/week, all recieved treatment(all included in ITT analysis)

9 did not complete* 7 did not complete†

9 lost to follow-up 6 lost to follow-up2 lack of efficacy2 adverse effect2 shifted residence3 unknown cause

1 adverse effect1 marital problems1 shifted residence3 unknown cause

1 could not tolerate

Completed study, n = 38Analyzed, n = 47

Completed study, n = 46Analyzed, n = 53

Group2N = 53 allocated to MTX 15mg/week, all recieved treatment(all included in ITT analysis)

15 patients exluded7 Not meeting criteria

8 refused consend

Figure 1. Flow chart of patient disposition in the study. *Number of patients lost to follow-up at 4, 8, and 12weeks: 7, 1, and 1. †Number of patients lost to follow-up at 4, 8, and 12 weeks: 3, 3, and 1(Including patient who could not tolerate). ITT ¼ intention-to-treat.

Clinical Therapeutics

Thompson et al16 compared intramuscular doses of10 mg (17 patients) and 25 mg (14 patients) per weekover 12 weeks but found similar efficacy. However,there was no escalation of dose in either of thesestudies. Thus, rather than comparing different startingdoses, they confirmed that higher doses (up to 25 mg)were more efficacious. Another large study bySchnabel et al19 compared starting doses of 25 mgand 15 mg (escalated to 25 mg) per week of MTX in185 patients with RA. However, although they lookedat differences in adverse effects, these studies did notreport any efficacy measures. A recent pilot studyrandomized 19 patients to a starting dose of 15 mgper week (increased by 5 mg every 2 weeks) or 25 mgper week. They did not find any difference in theclinical response (using DAS28) at 16 weeks, which issimilar to our results (Table III).20 To explain ourresults, we hypothesized that the rapidity of doseescalation blunted any advantage of starting with ahigher dose. Indeed, by 12 weeks, even the group

1010

started on 7.5 mg per week of MTX had reached aneffective weekly dose of 20 mg. At a pharmacokineticslevel, it is well known that the lag in onset of actionof MTX is due to slow accumulation of MTXpolyglutamates intracellularly.24 This may also partiallyexplain the lack of benefit of a higher starting dose by12 weeks.

There were no significant differences in frequencyof cytopenia or transaminitis between the 2 groups. Asignificantly higher proportion of patients experiencednausea and vomiting in the 15-mg group (41.5%)compared with the 7.5-mg group (19.1%). However,there was no significant difference in the severity orduration of nausea between groups. There was also nosignificant difference in the number of patients whowithdrew from the study. Our results are similar tothose found by Schnabel et al,19 who compared 25and 15 mg (escalated to 25 mg) as the starting dose.They also found no difference in the frequency oftransaminitis or cytopenia between the groups but a

Volume 36 Number 7

7.0

15 16

15

14

13

12

11

10

10

9

8

7

6

5

4

14

TJC

TJC

(28

Join

ts)

SJC

SJC

13

12

11

10

9

8

7

6

5

0 4 Weeks 8 Weeks 12 Weeks 0 4 Weeks 8 Weeks 12 Weeks

0 4 Weeks 8 Weeks 12 Weeks 0 4 Weeks 8 Weeks 12 Weeks

030

40

50

60

ESR

70

80

0

20

40

60

ESR

80

4 Weeks 8 Weeks 12 Weeks 0 4 Weeks 8 Weeks 12 Weeks

0

10

Perc

ent

Impr

ovem

ent

Perc

ent

Impr

ovem

ent

20

30

40

50

4 Weeks 8 Weeks 12 Weeks

10

20

30

40

50

4 Weeks 8 Weeks 12 Weeks

IIT Completers

MTX 15 mg MTX 7.5 mg MTX 15 mg

MTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

MTX 15 mgMTX 7.5 mg

6.5

6.0

DA

S28(

3)5.5

5.00 4 Weeks 8 Weeks 12 Weeks

7.0

6.5

6.0

DA

S2(3

)

5.5

5.00 4 Weeks 8 Weeks12 Weeks

Figure 2. Measures of efficacy at 4, 8, and 12 weeks between patients started on 7.5 mg versus 15 mgmethotrexate (MTX) according to both intention-to-treat (ITT) and per-protocol (completers)analyses. Error bars represent SEM. DAS28(3) ¼ Disease Activity Score for 28 joints using 3variables; ESR ¼ erythrocyte sedimentation rate; SJC ¼ Swollen Joint Count; TJC ¼ TenderJoint Count.

V. Dhir et al.

July 2014 1011

Table II. Number of patients who had adverse effects. Unless otherwise indicated, values are given asnumber (%).

Adverse EffectGroup 1: MTX7.5 mg (n ¼ 47)

Group 2: MTX15 mg (n ¼ 53) P

Nausea or vomiting 9 (19.1) 22 (41.5) 0.02*Fatigue 4 (8.5) 6 (11.3) 0.64Loss of appetite 7 (14.9) 7 (13.2) 0.80Uneasiness 9 (19.1) 8 (15.1) 0.59Diarrhea 2 (4.3) 3 (5.7) 0.76Oral ulcers 3 (6.4) 0 (0) 0.06No. of episodes of deranged LFT* 6 7 0.8No. of episodes of thrombocytopenia or leukopenia† 1 2 0.9

LFT ¼ liver function test.*Defined as aspartate aminotransferase or alanine aminotransferase 42 � the upper limit of normal.†White blood count o4000 or platelet count o100 � 103/μL.

Clinical Therapeutics

higher incidence of nausea in the 25-mg group (28%and 17%). Similar to our results, they also did not findany difference in the patient withdrawals due toadverse effects (16% and 18%). A recent pilot study20

comparing 15 and 25 mg of MTX per week did notfind any difference in the frequency of nausea.However, that study included only a small numberof patients. It is difficult to compare our results withthe 2 older studies because they did not incorporateany dose escalation. Furst et al17 reported thatpatients receiving 10 mg/m2 (15–22 mg) comparedwith 5 mg/m2 (7.5–10 mg per week) of MTX hadhigher rates of nausea (24% and 8%) (Table III).However, this finding suggests a dose-toxicity rela-tionship, rather than a relationship of starting dosewith toxicity.

A limitation of our study is the short 12-weekduration of the trial. This duration was chosenbecause both the American College of Rheumatologyand EULAR (2010) guidelines had suggested 3months as a decision point, when other drugs couldbe added in the face of inadequate response toMTX.8,25 Even in SWEFOT (Swedish Pharmacother-apy study), 3 to 4 months was the cutoff used todefine nonresponders to MTX before the addition ofinfliximab or combination DMARDs.26 However, theEULAR 2013 update has now specified that the3-month period relates “solely to assessing improve-ment” and that “it must be borne in mind that

1012

maximal efficacy will not be seen before 6 months.”7

We concur with the same; in our study, we also founda relatively poor response by 12 weeks (averagechange in DAS28[3], –0.5). Indeed, in view of therelatively slow decline in disease activity, futurestudies may benefit from extending the follow-upperiod to 24 weeks.

The relatively small change in disease activity seenin our study can be explained by a combination offactors. These factors were the selection of highlyactive patients (mean DAS28[3], 6.2), having a longdisease duration (�5 years), and the fact that only afew patients were taking oral steroids (o15%). Wechose 7.5 and 15 mg as comparators because these arethe dosages most commonly used in major clinicaltrials. The 15-mg starting dose was used in 3 of the 4groups of the BeSt (Behandel Strategieen) trial,27

whereas 7.5 mg was used in the COBRA(Combinatietherapie Bij Reumatoide Artritis) andFIN-RACo (Finnish Rheumatoid Arthritis Combina-tion Therapy) trials, CAMERA (Computer AssistedManagement in Early Rheumatoid Arthritis Study),CIMESTRA (Cyclosporine, Methotrexate, Steroid inRA), and COMET (The combination of methotrexateand etanercept in early rheumatoid arthritis).12,28–31

We assessed efficacy of oral MTX only, and notparenteral MTX, to mimic our routine practice. Weused DAS28(3) rather than DAS28 (ie, excludedpatient global assessment/general health on a visual

Volume 36 Number 7

Table III. Comparison of studies using different starting doses of methotrexate.

Author, Year Duration

Groups, StartingWeekly Dose,

\Mode Of AdministrationEscalation,Slow/Fast

↑Efficacy HigherVersus Lower Dose

↑ Toxicity HigherVersus Lower Dose

Thompson et al,16 1984 6 weeks; N ¼ 48 Placebo, N ¼ 17 No No Yes (Nausea)10 mg/wk, N ¼ 1725 mg/wk, N ¼ 14Intravenously

Russel, 1985 (Abstract)* 6 weeks; N ¼ 64 10 mg No No —

25 mgIntramuscularly

Furst et al,17 1989 16 weeks; N ¼ 46 Placebo, N ¼ 16 No Yes Yes (nausea)5 mg/m2 (8 mg), N ¼ 1310 mg/m2 (18 mg), N ¼ 17Oral

Schnabel et al,19 1994 1 year; N ¼ 185 15 mg, N ¼ 91 Yes, slow 5 mg/3 mo NA Yes (GI adverse effects)25 mg, N ¼ 77Intravenously (3–4 wk)- oral

Hobl et al,20 2012 16 weeks; N ¼ 19 15 mg, N ¼ 10 Yes, fast 5 mg/2 weeks No No25 mg, N ¼ 9Oral

Present study 12 weeks; N ¼ 100 7.5 mg, N ¼ 47 Yes, fast 2.5 mg/2 weeks No Yes (nausea)15 mg, N ¼ 53Oral

GI ¼ gastrointestinal.*Cross referenced from Williams et al.4

V.Dhiret

al.

July

2014

1013

Clinical Therapeutics

analog scale) because this method was found (by thefirst author) in previous studies to be difficult toadminister to subjects. In addition, the assessment ofpatient global/general health by using a visual analogscale has been found to be the least important amongthe 4 parameters used in DAS28.32

CONCLUSIONSThis study found no significant difference in efficacybetween the 2 starting doses of MTX followed by fastescalation over 12 weeks. However, starting at ahigher dose was associated with more patients expe-riencing nausea. We suggest studies with a longerduration of follow-up to confirm our findings.

ACKNOWLEDGMENTSThe authors thank Mr. Amit Ajmera and ZydusSynovia Pharmaceuticals for providing the MTXtablets used in this study. No specific funding supportwas received; drugs were received as a gift from ZydusSynovia Pharmaceuticals. No specific funding wasreceived from the authors’ institution for this partic-ular study. Dr. Dhir was responsible for the studydesign, literature search, figures, data interpretation,and writing of the manuscript. Drs. Singla and Goyalwere responsible for the Data collection and interpre-tation. Dr. Gupta was responsible for the study designand literature search. Drs. Sharma, Khanna and Singhwere responsible for the data collection and writing ofthe manuscript.

The authors also thank Professor Amita Aggarwal,Sanjay Gandhi Post graduate Institute of MedicalSciences, Lucknow, and Dr Vivek Arya, RML Hospi-tal, New Delhi, for their proofreading, languagecorrection, and critical analysis of the manuscript.

CONFLICTS OF INTERESTThe authors have indicated that they have no conflictsof interest regarding the content of this article. Thepharmaceutical company, which provided the drug,has not given any financial support to any author.

REFERENCES1. Scott DL, Wolfe F, Huizinga TW. Rheumatoid arthritis.

Lancet. 2010;376:1094–1108.2. Weinblatt ME. Methotrexate in rheumatoid arthritis: a

quarter century of development. Trans Am Clin Climatol

Assoc. 2013;124:16–25.

1014

3. Weinblatt ME, Coblyn JS, Fox DA, et al. Efficacy of low-dose methotrexate in rheumatoid arthritis. N Engl J Med.1985;312:818–822.

4. Williams HJ, Willkens RF, Samuelson CO Jr., et al.Comparison of low-dose oral pulse methotrexate andplacebo in the treatment of rheumatoid arthritis. Acontrolled clinical trial. Arthritis Rheum. 1985;28:721–730.

5. Rau R, Herborn G. Benefit and risk of methotrexatetreatment in rheumatoid arthritis. Clin Exp Rheumatol.2004;22:S83–S94.

6. Sokka T, Kautiainen H, Toloza S, et al. QUEST-RA:quantitative clinical assessment of patients with rheuma-toid arthritis seen in standard rheumatology care in 15countries. Ann Rheum Dis. 2007;66:1491–1496.

7. Smolen JS, Landewe R, Breedveld FC, et al. EULARrecommendations for the management of rheumatoidarthritis with synthetic and biological disease-modifyingantirheumatic drugs: 2013 update. Ann Rheum Dis.2014;73:492–509.

8. Smolen JS, Landewe R, Breedveld FC, et al. EULARrecommendations for the management of rheumatoidarthritis with synthetic and biological disease-modifyingantirheumatic drugs. Ann Rheum Dis. 2010;69:964–975.

9. O’dell JR, Curtis JR, Mikuls TR, et al. Validation of themethotrexate-first strategy in patients with early, poor-pro-gnosis rheumatoid arthritis: results from a two-year random-ized, double-blind trial. Arthritis Rheum. 2013;65:1985–1994.

10. Jacobs JW. Lessons for the use of non-biologic anchortreatments for rheumatoid arthritis in the era of biologictherapies. Rheumatology (Oxford). 2012;51(Suppl 4):iv27–iv33.

11. Smolen JS, Aletaha D, Bijlsma JW, et al. Treating rheu-matoid arthritis to target: recommendations of an inter-national task force. Ann Rheum Dis. 2010;69:631–637.

12. Verstappen SM, Jacobs JW, Van Der Veen MJ, et al. Intensivetreatment with methotrexate in early rheumatoid arthritis:aiming for remission. Computer Assisted Management inEarly Rheumatoid Arthritis (CAMERA, an open-label strat-egy trial). Ann Rheum Dis. 2007;66:1443–1449.

13. Visser K, Katchamart W, Loza E, et al. Multinationalevidence-based recommendations for the use of methotrex-ate in rheumatic disorders with a focus on rheumatoidarthritis: integrating systematic literature research and expertopinion of a broad international panel of rheumatologistsin the 3E Initiative. Ann Rheum Dis. 2009;68:1086–1093.

14. Pavy S, Constantin A, Pham T, et al. Methotrexate therapyfor rheumatoid arthritis: clinical practice guidelines basedon published evidence and expert opinion. Joint Bone Spine.2006;73:388–395.

15. Visser K, Van Der Heijde D. Optimal dosage and route ofadministration of methotrexate in rheumatoid arthritis:a systematic review of the literature. Ann Rheum Dis.2009;68:1094–1099.

Volume 36 Number 7

V. Dhir et al.

16. Thompson RN, Watts C, Edelman J,Esdaile J, Russell AS. A controlledtwo-centre trial of parenteral metho-trexate therapy for refractory rheuma-toid arthritis. J Rheumatol. 1984;11:760–763.

17. Furst DE, Koehnke R, Burmeister LF,Kohler J, Cargill I. Increasing metho-trexate effect with increasing dose inthe treatment of resistant rheumatoidarthritis. J Rheumatol. 1989;16:313–320.

18. Dhir V, Aggarwal A. Methotrexate-related minor adverse effects in rheu-matoid arthritis: more than a nui-sance. J Clin Rheumatol. 2012;18:44–46.

19. Schnabel A, Reinhold-Keller E, Will-mann V, Gross WL. Tolerability ofmethotrexate starting with 15 or 25mg/week for rheumatoid arthritis.Rheumatol Int. 1994;14:33–38.

20. Hobl EL, Mader RM, Jilma B, et al. Arandomized, double-blind, parallel,single-site pilot trial to compare twodifferent starting doses of methotrex-ate in methotrexate-naive adult pa-tients with rheumatoid arthritis. Clin

Ther. 2012;34:1195–1203.21. Arnett FC, Edworthy SM, Bloch DA,

et al. The American Rheumatism Asso-ciation 1987 revised criteria for theclassification of rheumatoid arthritis.Arthritis Rheum. 1988;31:315–324.

22. Fransen J, Van Riel PL. The DiseaseActivity Score and the EULAR re-sponse criteria. Clin Exp Rheumatol.2005;23:S93–S99.

23. Kumar A, Malaviya AN, Pandhi A,Singh R. Validation of an Indianversion of the Health AssessmentQuestionnaire in patients with rheu-matoid arthritis. Rheumatology (Ox-

ford). 2002;41:1457–1459.24. Kremer JM. Toward a better under-

standing of methotrexate. Arthritis

Rheum. 2004;50:1370–1382.25. Singh JA, Furst DE, Bharat A, et al.

Update of the 2008 American Col-lege of Rheumatology recommenda-tions for the use of disease-modifying antirheumatic drugs andbiologic agents in the treatment ofrheumatoid arthritis. Arthritis Care

Res. 2012;64:625–639.

July 2014

26. Van Vollenhoven RF, Ernestam S,Geborek P, et al. Addition of inflix-imab compared with addition ofsulfasalazine and hydroxychloroquineto methotrexate in patients with earlyrheumatoid arthritis (Swefot trial): 1-year results of a randomised trial.Lancet. 2009;374:459–466.

27. Goekoop-Ruiterman YP, De Vries-Bouwstra JK, Allaart CF, et al. Clin-ical and radiographic outcomes offour different treatment strategies inpatients with early rheumatoid ar-thritis (the BeSt study): a random-ized, controlled trial. Arthritis Rheum.2005;52:3381–3390.

28. Mottonen T, Hannonen P, Leirisalo-Repo M, et al. Comparison of combi-nation therapy with single-drug ther-apy in early rheumatoid arthritis: arandomised trial. FIN-RACo trialgroup. Lancet. 1999;353:1568–1573.

29. Boers M, Verhoeven AC, MarkusseHM, et al. Randomised comparisonof combined step-down predniso-

lone, methotrexate and sulpha-salazine with sulphasalazine alonein early rheumatoid arthritis. Lancet.1997;350:309–318.

30. Hetland ML, Horslev-Petersen K.The CIMESTRA study: intra-articular glucocorticosteroids andsynthetic DMARDs in a treat-to-target strategy in early rheumatoidarhtritis. Clin Exp Rheumatol.2012;30:S44–S49.

31. Emery P, Breedveld FC, Hall S, et al.Comparison of methotrexate mono-therapy with a combination of me-thotrexate and etanercept in active,early, moderate to severe rheumatoidarthritis (COMET): a randomised,double-blind, parallel treatment trial.Lancet. 2008;372:375–382.

32. Van Der Heijde DM, Van ‘T Hof MA,Van Riel PL, et al. Judging diseaseactivity in clinical practice in rheuma-toid arthritis: first step in the develop-ment of a disease activity score. AnnRheum Dis. 1990;49:916–920.

Address correspondence to: Varun Dhir, MD, DM, Rheumatology Unit,Department of Internal Medicine, PGIMER, Chandigarh 160012, India.E-mail: varundhir@gmail.com

1015