REVIEW

New Rheumatoid Arthritis Treatments for ‘Old’Patients: Results of a Systematic Review

Roberto Caporali . Walid K. H. Fakhouri . Claudia Nicolay .

Harriet J. Longley . Serena Losi . Veronica Rogai

Received: May 22, 2020 / Published online: July 23, 2020� The Author(s) 2020

ABSTRACT

Introduction: In the last 20 years, biologic andtargeted synthetic disease-modifying antirheu-matic drugs (DMARDs) have become availablefor treating rheumatoid arthritis (RA), and atreat-to-target strategy has been introduced. Wehypothesise that these advances should haveresulted in changes to the characteristics ofpatients with RA participating in clinical trialsof the newest therapies. This study determinedwhether the baseline characteristics of patients

with RA enrolled in clinical trials have changedin the past decade versus patients participatingin earlier RA studies.Methods: This secondary analysis was based onrandomised controlled trials (RCTs) identifiedin a systematic literature review. Baseline char-acteristics of patients with RA with inadequateresponse to conventional synthetic DMARDswere compared between RCTs published in1999–2009 and those published in 2010–2017using random-effects meta-analyses.Results: Forty RCTs were analysed: 22 from1999–2009 and 18 from 2010–2017. No signifi-cant difference between the two timeframes andno obvious trend over time were observed forage, gender, disease duration, rheumatoid factorstatus, tender and swollen joint counts, physi-cian and patient global assessments of diseaseactivity, and pain scores. Variability betweenRCTs was high. Similar results were observed forDisease Activity Scores and Health AssessmentQuestionnaire-Disability Index scores, but withlow variability between RCTs.Conclusion: The baseline characteristics ofpatients with RA participating in RCTs do notappear to have changed in the last decadedespite the availability of new treatments and adifferent treatment approach. Further researchshould determine the impact of baseline patient

Digital Features To view digital features for this articlego to https://doi.org/10.6084/m9.figshare.12593558.

Electronic Supplementary Material The onlineversion of this article (https://doi.org/10.1007/s12325-020-01435-6) contains supplementary material, which isavailable to authorized users.

R. CaporaliDepartment of Clinical Sciences and CommunityHealth, University of Milan, Milan, Italy

R. Caporali (&)Rheumatology Unit, ASST Gaetano Pini-CTOHospital, Milan, Italye-mail: roberto.caporali@unimi.it

W. K. H. Fakhouri � H. J. LongleyEli Lilly and Company, Windlesham, Surrey, UK

C. NicolayLilly Deutschland GmbH, Bad Homburg, Germany

S. Losi � V. RogaiEli Lilly Italy S.p.A., Sesto Fiorentino, Italy

Adv Ther (2020) 37:3676–3691

https://doi.org/10.1007/s12325-020-01435-6

characteristics on patients’ response to RAtreatments.

PLAIN LANGUAGE SUMMARY

In the last 20 years, new treatments and a newtreatment approach (called treat-to-target) havebeen introduced for rheumatoid arthritis (RA).Consequently, the characteristics of patientswith RA participating in clinical trials of thenewest therapies should have changed com-pared with those of patients who participated inclinical trials of older therapies. This is impor-tant as patient characteristics may influencepatients’ response to drug treatment. To deter-mine whether characteristics of patients withRA have changed over time, we compared thebaseline characteristics (e.g. age, gender, diseaseduration, measures of disease activity, and painscores) of patients with RA between 22 clinicaltrials published in 1999–2009 and 18 publishedin 2010–2017. No significant differencebetween the two timeframes and no obvioustrend over time were observed for any baselinecharacteristic of patients with RA, includingphysician and patient measures of diseaseactivity, and patient measures of physicalfunction and pain. Thus, the baseline charac-teristics of patients with RA participating inclinical trials do not appear to have changed inthe last decade despite the introduction of newtreatments and the treat-to-target approach.Further research is needed to determine theimpact of baseline patient characteristics onpatients’ response to RA treatments.

Keywords: Disease-modifying antirheumaticdrugs; Patient characteristics; Randomisedcontrolled clinical trials; Rheumatoid arthritis;Rheumatology; Systematic review

Key Summary Points

Why carry out this study?

The introduction of new treatments and atreat-to-target strategy for rheumatoidarthritis (RA) in the last 20 years shouldhave resulted in changes to thecharacteristics of patients with RAparticipating in clinical trials of thenewest therapies.

This is important as patient characteristicsmay influence patients’ response to drugtreatment.

To determine whether characteristics ofpatients with RA have changed over time,baseline characteristics were comparedbetween 22 clinical trials published in1999–2009 and 18 published in2010–2017.

What was learned from the study?

No significant difference between the twotimeframes and no obvious trend overtime were observed for any baselinecharacteristic of patients with RA,including physician and patientassessments of disease activity, and HealthAssessment Questionnaire-DisabilityIndex and pain scores.

The baseline characteristics of patientswith RA participating in clinical trials donot appear to have changed in the lastdecade; further research is needed todetermine the impact of baseline patientcharacteristics on patients’ response to RAtreatments.

INTRODUCTION

Rheumatoid arthritis (RA) is a chronic, autoim-mune, inflammatory arthritis associated withpain, disability and an increased risk of mor-tality [1–3]. Globally, the prevalence of the

Adv Ther (2020) 37:3676–3691 3677

disease is in the range 0.5–1.1% [4, 5]. Con-ventional synthetic disease-modifying anti-rheumatic drugs (csDMARDs), such asmethotrexate, leflunomide and sulfasalazine,provided the standard of care for RA for manyyears. However, the choice of treatments for thedisease has expanded over the last 20 years withthe introduction of biologic DMARDs(bDMARDs), such as abatacept, adalimumab,certolizumab pegol, etanercept, golimumab,infliximab, rituximab, sarilumab and tocilizu-mab, and, more recently, targeted synthetic(ts)DMARDs, such as baricitinib and tofacitinib.

Guidelines for the management of RA fromthe European League Against Rheumatism(EULAR) and the American College ofRheumatology (ACR) recommend early use ofDMARDs, with the goal of remission or lowdisease activity [6, 7]. This treat-to-targetapproach, first introduced in 2010 [8], hasimproved the care and outcomes of patientswith RA [9, 10]. However, several studies suggestthat many patients with RA are not reaping thebenefits of this treat-to-target approach or thebenefits derived from newer DMARDs becauseof underuse of these drugs. A review of datafrom the US National Ambulatory Care MedicalSurvey over a 12-year period (1996–2007)showed that only 47% of visits to a physicianfor RA were associated with a DMARD pre-scription [11], while a later retrospective reviewof data over a 3-year period (2005–2008)showed that only 63% of 93,143 Medicareenrollees with RA in the USA received aDMARD. Prescribed DMARDs varied withdemographic factors, socioeconomic status andgeographic location. Cited reasons for the lowproportion of DMARD prescriptions includedlow income, for-profit health plans, patientsrefusing treatment and increased comorbiditiesin the elderly resulting in contraindications toavailable drugs [12].

A systematic review of 127 studies of patientswith RA highlighted significant differences incertain patient characteristics, including age,disease duration, number of DMARDs previ-ously used and Disease Activity Score for28-joint count (DAS28), between patientsenrolled in randomised controlled clinical trials(RCTs) and those enrolled in registries [13]. In

addition, baseline DAS28 and Health Assess-ment Questionnaire-Disability Index (HAQ-DI)scores for patients with RA prescribedbDMARDs (etanercept, rituximab or tocilizu-mab) decreased over time for RCTs and obser-vational studies published between 2004 and2014 [13], probably owing to the introductionof the treat-to-target strategy during this period[14]. This trend was consistent with the resultsof an observational study conducted in theNetherlands, in which symptom duration andinflammatory activity at presentation werefound to have decreased over a 23-year period(1993–2015) among patients with RA attendinga local rheumatology department. Paradoxi-cally, however, patient-reported outcomes(pain, fatigue, disease activity and globalhealth) worsened over this time period, possiblyas a result of increased societal and patientexpectations [10].

Given the change in treatment strategy andthe introduction of new DMARDs with differentmechanisms of action in recent years (e.g. theJanus kinase inhibitors baricitinib and tofaci-tinib, and the interleukin-6 inhibitor sar-ilumab), the characteristics of patientsparticipating in RA clinical trials might beexpected to have changed over time. This isimportant as patient characteristics may influ-ence the effects of drug treatment [13].According to data from the systematic literaturereviews (SLRs) conducted by Kilcher et al. [13],the characteristics of patients with RA enrolledin RCTs who show an inadequate response tocsDMARDs might be expected to change overtime, and this may influence the rate ofresponse to new treatments as well as making itdifficult to compare drugs evaluated duringdifferent time periods. To date, however, nosuch analysis has been published. To addressthis important issue and inform the design offuture RCTs evaluating new RA treatments, weconducted a study to determine whether thebaseline characteristics of patients with RA withan inadequate response to csDMARDs partici-pating in RCTs between 1999 and 2017 havechanged over time.

3678 Adv Ther (2020) 37:3676–3691

METHODS

Objective

This study aimed to compare the characteristicsof patients with an inadequate response tocsDMARDs participating in RA RCTs betweentwo different, predefined timeframes–an earliertimeframe (1999–2009) when bDMARDs werefirst introduced, and a later timeframe(2010–2017) after the introduction of treat-ments with different mechanisms of action andadoption of the treat-to-target strategy.

Systematic Literature Review

This secondary analysis was based on the resultsof a previously conducted SLR [15]. The SLRaimed to identify evidence for the efficacy andsafety of treatments for moderately to severelyactive RA in adults. Searches of Medline, Med-line in Process, Embase, Biosciences Informa-tion Service and the Cochrane Library wereperformed to identify RCTs published between1 January 1999 and 11 December 2017 usingsearch terms related to RA, associated inter-ventions and RCTs (Table S1 in the supple-mentary material). The original searches wereperformed on 17 June 2015 and updated sear-ches were performed on 10 August 2016 and11 December 2017. There were no languagelimits on the database searches. Conferenceabstracts (2013–2017), grey literature and thebibliographies of key articles were also reviewed.Data were extracted from relevant full-text publications and quality-checked by anindependent reviewer. The quality of each studywas also assessed using National Institute forHealth and Care Excellence (NICE) guidelines[16]. The SLR was conducted according to Pre-ferred Reporting Items for Systematic Reviewsand Meta-Analyses (PRISMA) guidelines [17].

Current Analysis

The analysis reported here focused on studiesfrom the SLR involving patients with aninadequate response to csDMARDs, includingmethotrexate, hydroxychloroquine, leflunomide

andsulfasalazine.As such, themainanalysis studypoolonly included studies thatdidnotallowpriorbDMARD use. The following baseline patientcharacteristics of interest extracted from thestudies were analysed: age, gender, disease dura-tion, rheumatoid factor status, DAS28 based onerythrocyte sedimentation rate (DAS28-ESR) orC-reactive protein (DAS28-CRP), HAQ-DI score,swollen joint count (SJC), tender joint count(TJC), Physician’s Global Assessment of DiseaseActivity (PGA), Patient’s Global Assessment ofDiseaseActivity (PtGA), painvisual analogue scale(VAS) score, previous use of non-steroidal anti-inflammatory drugs (NSAIDs) and number ofpreviously used csDMARDs.

Additional baseline characteristics that wereextracted from the studies but could not beanalysed because of lack of or limited dataavailability included weight, comorbidities,smoking status, Clinical Disease Activity Index(CDAI) and Simplified Disease Activity Index(SDAI) scores, Routine Assessment of PatientIndex Data-3 (RAPID-3) score, erosion score,joint space narrowing score, van der Heijdemodified total Sharp score (mTSS) and Euro-QoL-5 Dimensions (EQ-5D) questionnaire score.

The analysis reported in this article is basedon previously conducted studies and does notinvolve any studies with human participants oranimals performed by any of the authors.

Statistical Analysis

If not already presented as such, baselinepatient characteristics from different treatmentarms within a given study were combined intoan overall study result. Where necessary, medi-ans and ranges from each study were convertedto means and standard deviations using themethods of Wan et al. [18]. For binary data, thevariance estimate (v) for proportions (p) wasused to derive the standard error using the for-mula v = p(1 - p). The results from the variousstudies were compared between the two studypublication timeframes (1999–2009 vs2010–2017) using random-effects meta-analy-ses. Between-study variance was assessed usingrestricted maximum-likelihood estimation andwas presented as an I2 value, showing the

Adv Ther (2020) 37:3676–3691 3679

proportion of the observed variance that reflectsreal difference between studies (i.e. it is not dueto random error), where higher percentagesrepresent higher variance; and a Tau-squaredvalue, showing between-study variation, wherehigher values represent higher variation. Thelevel of significance was taken as p B 0.05(unadjusted for multiple testing). Missing datawere not imputed for this analysis. Forest plotswere generated for all analyses, with studyresults ordered by year of publication to providea visual display of potential changes over time.Analyses were conducted using SAS version 9.4,R studio version 3.4, and metafor package ver-sion 2.0 [19].

Sensitivity Analyses

Two sensitivity analyses were performed toassess the effect of changing the patient popu-lation on results. One sensitivity analysisexcluded Asia–Pacific studies, since patientsfrom these countries (mainly Japanese) werelikely to have been treated with a low dose ofmethotrexate (\7.5 mg/week), which couldpotentially impact the extent of methotrexatefailure among trial populations. In addition tostudies included in the main analysis, the sec-ond sensitivity analysis included studies thatallowed prior bDMARD use in up to 20% ofpatients, since patients receiving bDMARDswere more likely to be further advanced in theRA treatment algorithm. The value of 20% wasselected as many studies allowing some prioruse of bDMARDs cited this value as the cut-off.

RESULTS

Study Numbers

A total of 147 primary studies including patientswith an inadequate response to csDMARDs wereidentified in the SLR, of which 94 were excluded(Fig. 1). Of the remaining 53 studies, 13 allowedprior use of bDMARDs in up to 20% of patientsand were consequently excluded from the mainanalysis. Thus, the main analysis included a totalof 40 studies. Of these, 22 were published in the

earlier timeframe and 18 in the later timeframe.The 13 studies allowing prior use of bDMARDs inup to 20% of patients were combined with the 40studies from the main analysis into a sensitivityanalysis; details of these 53 studies are provided inTable S2 in the supplementary material [20–72].Treatments evaluated in the studies includedcsDMARDs (hydroxychloroquine, methotrexate,sulfasalazine), bDMARDs (abatacept, adalimumab,certolizumab pegol, etanercept, golimumab,infliximab, rituximab, sarilumab, tocilizumab),and tsDMARDs (baricitinib, tofacitinib). Not allidentified studies were included in each specificcharacteristic analysis as some did not reportrelevant information or reported data in a man-ner thatdidnot allowdata tobeconverted (e.g. tomeans and standard deviations).

Main Analysis

There was no statistically significant differencebetween the two timeframes and no obvioustrend over time for age, gender, disease duration(mean difference between timeframes approxi-mately 1 year), rheumatoid factor status, PGA,PtGA and pain VAS scores (Figs. 2 and S1).However, there was high variability (hetero-geneity) between studies. Similarly, there wasno statistically significant difference betweentimeframes and no obvious trend over time forDAS28-ESR, DAS28-CRP and HAQ-DI scores; inthese cases, heterogeneity was low (Fig. 3). SJC(score range 0–66) and TJC (score range 0–68)were not significantly different between time-frames and there was no obvious trend overtime; heterogeneity was also high. Differentinclusion criteria were observed between studiesfor the minimum number of swollen or tenderjoints. We therefore conducted further analysesfocusing on the minimum number of swollen ortender joints used as the inclusion criterion ineach study, and on the most common inclusioncriteria for minimum number of swollen andtender joints (B 6 for SJC, B 6 and C 8 for TJC).However, this did not change the results to anygreat extent (Figs. 4 and S2).

Information on the number of csDMARDspreviously used was not reported in a unifiedmanner, with very few studies (n = 3 in

3680 Adv Ther (2020) 37:3676–3691

Records excluded at data extrac�on n = 28Study type n = 3Popula�on n = 5Interven�on n = 4Outcomes n = 4Other n = 12

DATA EXTRACTION(Further records excluded)

n = 350

Records excluded at level 2n = 612Study type n = 284 Popula�on n = 58 Interven�on n = 88Outcomes n = 74Other n = 108

SCRE

ENIN

G

Records excluded at level 1n = 9,046 Study type n = 6,183Popula�on n = 1,285Interven�on n = 1,213Outcomes n = 175Other n = 190

ELIG

IBIL

ITY

INCL

UDE

DID

ENTI

FICA

TIO

N

Duplicates excluded n = 6,130

LEVEL 2 SCREEN(Full texts screened)

n = 962

Records iden�fied through database searches

n = 14,632

Total records iden�fied a�er elimina�on of duplicatesn = 10,008

(Database searches = 8,517; Internet searches = 1,458; Hand searches = 33)

Records iden�fied through Internet searches

n = 1,473

LEVEL 1 SCREEN(Titles/abstracts screened)

n = 10,008

Records iden�fied through hand searches

n = 33

Total records included in systema�c review n = 322

(Total studies n = 147)

Records excluded n = 94Study type n = 5Pa�ent popula�on n = 56Interven�on n = 22Outcomes n = 5Other n = 6Studies of pa�ents with inadequate

response to MTX or other csDMARD n = 53

Studies included in main analysis n = 40

Records excluded n = 13Studies allowing prior bDMARD use in up to 20% of pa�ents were included in a sensi�vity analysis

Fig. 1 PRISMA diagram for study identification. bDMARD, biologic DMARD; csDMARD, conventional syntheticDMARD; DMARD, disease-modifying antirheumatic drug; MTX, methotrexate

Adv Ther (2020) 37:3676–3691 3681

1999–2009; n = 5 in 2010–2017) providingmeans and standard deviations. For studiesproviding the latter, a statistically significantdifference between timeframes was observedwith low to moderate heterogeneity (I2 33%,Tau2 0.06, p\ 0.001; Fig. S3). The remainingstudies either did not report this information orreported categories, which could not be con-verted into means and standard deviations. Forprior NSAID exposure, most studies (24 out of28; 86%) reported NSAID use in 100% ofpatients; therefore, no meta-analysis was run.

Sensitivity Analyses

The sensitivity analysis excluding Asia–Pacificstudies included 28 studies: 16 from the earliertimeframe and 12 from the later timeframe. Theresults of this analysis were consistent withthose of the main analysis (Fig. S4).

The sensitivity analysis allowing prior use ofbDMARDs in up to 20% of patients included 53studies: 26 from the earlier timeframe and 27from the later timeframe. Again, the results ofthis analysis were consistent with those of themain analysis (Fig. S5).

DISCUSSION

With the introduction of the treat-to-targetstrategy in RA and the availability of newtreatments, the inclusion criteria for RA clinicaltrials might be expected to have changed in thepast decade. However, the results of our analysissuggest that this is not the case: the character-istics of patients participating in recent clinicaltrials do not appear to have changed comparedwith those of patients participating in RCTs10–20 years ago.

Patient characteristics are an importantconsideration in clinical trials as they mayinfluence the effects of treatment. For example,older age is associated with decreased responserates in patients treated with etanercept ortocilizumab [73, 74], while male sex, beingrheumatoid factor-positive, having a low HAQ-DI score and being a non-smoker predict a bet-ter response to various bDMARDs [73, 75, 76].In addition, a study comparing the baseline

characteristics of patients with RA betweenRCTs and observational studies showed thatpatients participating in RCTs had better prog-nostic factors than those participating inobservational studies, which could result inoverestimation of the treatment effect [13].

In this analysis, the lack of a change inDAS28 and HAQ-DI scores at baseline betweenthe two timeframes was surprising given thatreal-world data from registries or observationalstudies suggest that baseline disease activityamong patients with RA has decreased overtime [77–79]. In addition, the aforementionedstudy by Kilcher et al. [13] comparing thebaseline characteristics of patients with RAbetween RCTs and observational studiesshowed that baseline DAS28, HAQ-DI, ESR andCRP significantly decreased in patients partici-pating in RCTs over the time period 1999–2015.An SLR of patients with RA receiving anti-tu-mour necrosis factor treatment in clinical trialsover a 16-year period (1993–2008) showed asimilar decrease in baseline CRP over timeamong patients previously treated withmethotrexate but not among those with noexperience of this drug [80]. However, the cur-rent analysis suggests that baseline diseaseactivity among patients with RA participating inRCTs has not decreased over time, possiblybecause patients participating in RCTs tend tohave more severe disease at baseline than thosein routine clinical practice [13, 81]. The lack ofchange in disease activity in the current analysismay also be due to barriers in adopting thetreat-to-target strategy in clinical practice com-pared with clinical trials, such as a lack ofphysician understanding of this treatmentstrategy, the feeling that the disease activityscore may be falsely high due to symptoms orinflammation unrelated to RA, physician resis-tance to algorithm-based treatment, or a lack oftime at clinic visits [82–84]. The lack of changein baseline disease activity over time likelyreflects a lack of change in the strict inclusioncriteria for patients participating in an RCT. Inview of this, different methodology to that usedin our analysis and measurement of differentclinical/laboratory parameters might be neces-sary to detect any change over time in patient

3682 Adv Ther (2020) 37:3676–3691

characteristics; this could be a subject for futureresearch.

This analysis suggests that the mean numberof previously used csDMARDs in patients par-ticipating in RCTs has decreased over time,

possibly reflecting adoption of the treat-to-tar-get approach. However, this result was based ononly a few studies and was not observed in thesensitivity analysis that included studies allow-ing prior bDMARD use in up to 20% of patients.

50 55 60 65 70 75 80

Mean score

BUILD (2017)

BEAM (2017)

EXXELERATE (2016)

AMPLE (2014)

Li (2013)

ADACTA (2013)

GO−FORTH (2012)

JESMR (2010)

GO−FORWARD (2009)

CNTO 148 (2008)

CHANGE (2008)

Kim (2007)

START (2006)

De (2006)

AIM (2006)

van (2004)

Keystone (2004)

Edwards (2004)

STAR (2003)

ARMADA (2003)

63.53 [60.76, 66.30]

65.07 [62.73, 67.40]

63.55 [61.51, 65.59]

58.80 [56.42, 61.18]

59.50 [56.75, 62.25]

66.60 [63.93, 69.27]

56.20 [52.95, 59.45]

58.20 [54.62, 61.78]

58.75 [55.80, 61.70]

60.50 [56.29, 64.71]

75.15 [72.19, 78.11]

63.85 [60.72, 66.98]

62.50 [59.43, 65.57]

59.50 [55.24, 63.76]

67.70 [65.40, 70.00]

67.75 [64.75, 70.75]

61.65 [59.11, 64.19]

66.33 [62.32, 70.35]

59.75 [57.51, 61.99]

58.80 [55.56, 62.04]

63.58 [60.03, 67.14]1999−2009

62.81 [59.78, 65.84]2010−2017

63.15 [60.64, 65.66]All Studies

I2 90 %Tau2 17.95p value: 0.759

Study Mean [95% CI]

(a) Physician's Global Assessment of Disease Activity (scale 0−100)

40 50 60 70 80

Mean score

MONARCH (2017)BUILD (2017)BEAM (2017)EXXELERATE (2016)MOBILITY (2015)AMPLE (2014)Li (2013)ADACTA (2013)GO−FORTH (2012)JESMR (2010)GO−FORWARD (2009)CNTO 148 (2008)CHANGE (2008)Kim (2007)START (2006)De (2006)AIM (2006)van (2004)Keystone (2004)Edwards (2004)STAR (2003)ARMADA (2003)

67.86 [65.19, 70.54]60.70 [57.65, 63.75]62.57 [59.92, 65.22]63.10 [60.83, 65.37]64.80 [62.24, 67.36]61.30 [58.70, 63.90]64.00 [63.02, 64.98]72.30 [69.46, 75.14]48.40 [44.78, 52.02]58.05 [54.34, 61.76]56.50 [55.27, 57.73]67.50 [65.95, 69.05]67.90 [64.39, 71.41]61.45 [57.87, 65.03]56.50 [55.47, 57.53]66.83 [61.92, 71.74]62.75 [60.14, 65.36]72.15 [68.97, 75.33]53.50 [50.61, 56.39]64.87 [60.14, 69.59]53.40 [50.80, 56.00]57.45 [53.57, 61.33]

59.13 [54.64, 63.62]1999−200963.15 [58.57, 67.72]2010−201761.68 [58.40, 64.96]All Studies

I2 96 %Tau2 36.38p value: 0.22

Study Mean [95% CI]

(b) Patient's Global Assessment of Disease Activity (scale 0−100)

45 50 55 60 65 70 75

Mean score

MONARCH (2017)BUILD (2017)BEAM (2017)EXXELERATE (2016)MOBILITY (2015)Machado (2014)AMPLE (2014)Li (2013)GO−FORTH (2012)APPEAL (2012)JESMR (2010)GO−FORWARD (2009)CNTO 148 (2008)CHANGE (2008)Kim (2007)START (2006)De (2006)Combe (2006)AIM (2006)van (2004)Keystone (2004)Edwards (2004)STAR (2003)ARMADA (2003)

70.62 [67.87, 73.38]57.97 [54.85, 61.09]60.77 [58.07, 63.46]63.35 [61.07, 65.63]65.27 [62.67, 67.86]65.25 [62.34, 68.16]64.30 [61.72, 66.88]62.50 [59.31, 65.69]50.85 [47.19, 54.51]61.65 [58.54, 64.76]55.85 [52.05, 59.65]59.00 [55.36, 62.64]70.00 [65.82, 74.18]65.40 [61.84, 68.96]58.50 [54.96, 62.04]60.00 [56.69, 63.31]65.39 [60.81, 69.96]59.97 [56.00, 63.94]64.60 [62.03, 67.17]70.15 [67.02, 73.28]56.10 [53.22, 58.98]59.73 [55.37, 64.10]55.35 [52.73, 57.97]55.10 [51.29, 58.91]

60.39 [56.88, 63.89]1999−200962.65 [58.97, 66.33]2010−201761.82 [59.24, 64.40]All Studies

I2 90 %Tau2 24.1p value: 0.383

Study Mean [95% CI]

(c) Pain visual analogue scale (scale 0−100)

Fig. 2 Forest plots showing means and 95% confidenceintervals (CI; box and whisker plots) for the differentstudies according to year of publication for a Physician’sGlobal Assessment of Disease Activity score, b Patient’sGlobal Assessment of Disease Activity score and c PainVisual Analogue Scale score. If studies reported pain on ascale of 0–10, values were multiplied by 10. I2 and Tau2

values indicate a high degree of heterogeneity betweenstudies, while p values indicate no significant difference inmean values between studies published from 1999 to 2009(studies above dashed line) and those published from 2010to 2017 (studies below dashed line). Diamond shapesindicate 95% confidence intervals around the meanssummarised by timeframe and overall, respectively

Adv Ther (2020) 37:3676–3691 3683

This suggests that it was either a ‘chance find-ing’ based on a small number of studies or thatthe inclusion of patients with prior bDMARDuse corresponded to the inclusion of patientswith more severe disease and hence a higher

number of previously used csDMARDs, whichwould have diluted the effect over time. Veryfew studies reported the mean number of pre-viously used csDMARDs: most reported previ-ous csDMARD use as the percentage of patients

4 5 6 7 8

Mean score

MONARCH (2017)

BUILD (2017)

BEAM (2017)

RA−SCORE (2016)

EXXELERATE (2016)

Machado (2014)

ACT−RAY (2014)

ADACTA (2013)

REALISTIC (2012)

GO−FORTH (2012)

APPEAL (2012)

SERENE (2010)

JESMR (2010)

SATORI (2009)

GO−FORWARD (2009)

CNTO 148 (2008)

van (2004)

Edwards (2004)

6.80 [6.32, 7.28]

6.22 [5.57, 6.88]

6.43 [5.87, 6.99]

6.20 [5.22, 7.18]

6.50 [6.02, 6.98]

6.65 [6.13, 7.17]

6.34 [5.78, 6.91]

6.74 [6.12, 7.36]

6.40 [5.67, 7.13]

5.55 [4.76, 6.34]

6.10 [5.38, 6.82]

6.48 [5.76, 7.19]

6.05 [5.28, 6.82]

6.15 [5.37, 6.93]

6.11 [5.35, 6.87]

6.35 [5.34, 7.36]

7.08 [6.41, 7.75]

6.83 [5.87, 7.79]

6.54 [6.09, 6.98]1999−2009

6.40 [6.08, 6.73]2010−2017

6.41 [6.11, 6.72]All Studies

I2 3 %Tau2 0p value: 0.605

Study Mean [95% CI]

(a) DAS28 based on erythrocyte sedimentation rate

4 4.5 5 5.5 6 6.5 7

Mean score

MONARCH (2017)

BUILD (2017)

BEAM (2017)

RA−SCORE (2016)

MOBILITY (2015)

AMPLE (2014)

Li (2013)

REALISTIC (2012)

APPEAL (2012)

SERENE (2010)

GO−FORWARD (2009)

CNTO 148 (2008)

AIM (2006)

6.00 [5.40, 6.60]

5.55 [4.91, 6.19]

5.74 [5.19, 6.29]

5.50 [4.55, 6.45]

5.97 [5.43, 6.50]

5.50 [4.92, 6.08]

5.45 [4.73, 6.17]

5.70 [4.97, 6.43]

5.29 [4.57, 6.00]

5.87 [5.18, 6.57]

5.61 [4.84, 6.38]

5.30 [4.30, 6.30]

6.40 [6.01, 6.79]

6.13 [5.35, 6.91]1999−2009

5.71 [5.37, 6.05]2010−2017

5.75 [5.38, 6.12]All Studies

I2 10 %

Tau2 0.01

p value: 0.11

Study Mean [95% CI]

(b) DAS28 based on C−reactive protein

0 0.5 1 1.5 2 2.5 3

Mean score

MONARCH (2017)BUILD (2017)BEAM (2017)RA−SCORE (2016)EXXELERATE (2016)MOBILITY (2015)Machado (2014)AMPLE (2014)ACT−RAY (2014)Li (2013)REALISTIC (2012)GO−FORTH (2012)APPEAL (2012)JESMR (2010)GO−FORWARD (2009)CNTO 148 (2008)CHANGE (2008)Kim (2007)START (2006)De (2006)AIM (2006)van (2004)Keystone (2004)Edwards (2004)STAR (2003)ARMADA (2003)ATTRACT (1999)

1.60 [1.11, 2.09]1.52 [1.00, 2.04]1.57 [1.10, 2.04]1.40 [0.61, 2.19]1.50 [1.11, 1.89]1.63 [1.18, 2.08]1.60 [1.08, 2.12]1.50 [1.04, 1.96]1.47 [1.02, 1.92]1.25 [0.68, 1.82]1.55 [0.96, 2.14]1.00 [0.39, 1.61]1.40 [0.83, 1.97]1.25 [0.56, 1.94]1.31 [0.69, 1.93]1.50 [0.75, 2.25]1.51 [0.87, 2.16]1.35 [0.71, 1.99]1.50 [0.89, 2.11]1.78 [0.83, 2.73]1.70 [1.25, 2.15]1.85 [1.29, 2.42]1.47 [0.98, 1.95]1.93 [1.16, 2.71]1.40 [0.97, 1.83]1.59 [0.95, 2.24]1.75 [1.16, 2.34]

1.55 [1.36, 1.74]1999−20091.53 [1.29, 1.77]2010−20171.53 [1.35, 1.72]All Studies

I2 0 %Tau2 0p value: 0.875

Study Mean [95% CI]

(c) Health Assessment Questionnaire−Disability Index score (scale 0–3)

Fig. 3 Forest plots showing means and 95% confidenceintervals (CI; box and whisker plots) for the differentstudies according to year of publication for a DiseaseActivity Score for 28-joint count (DAS28) based onerythrocyte sedimentation rate, b DAS28 based onC-reactive protein and c Health Assessment Question-naire-Disability Index score. Diamond shapes indicate 95%

confidence intervals around the means summarised bytimeframe and overall, respectively. I2 and Tau2 valuesindicate a low degree of heterogeneity between studies,while p values indicate no significant difference in meanscores between studies published from 1999 to 2009(studies above dashed line) and those published from 2010to 2017 (studies below dashed line)

3684 Adv Ther (2020) 37:3676–3691

using a certain number (e.g. 1, C 1, C 2, etc.),which could not be used in the current analysis.Thus, no definitive conclusions about previouscsDMARD use can be drawn.

RCTs are very different to real-world clinicalpractice in that patients who are eligible forclinical trials generally have more severe disease[81]. Results of the current analysis suggest thatpatients currently being enrolled in RA RCTs arenot receiving the correct treatment beforecommencement of the study. Although treat-to-target is the recommended approach for themanagement of RA [6, 7], RCTs are still beingperformed in which none of the participatingpatients have been treated accordingly. Thisbegs the question as to whether RCTs are asinformative as they were 15 years ago. In future,it would be interesting to design RCTs thatinclude patients who have been treatedaccording to treat-to-target recommendations.

To our knowledge, this is the first study toinvestigate changes over time in the character-istics of patients enrolled in RCTs of RA treat-ments. The main strength of this analysis is that

it was based on a comprehensive SLR. However,it should be noted that transformation ofmedian into mean values can introduce bias ifthe data summarised by a median value are notnormally distributed [13]. As it was not the aimof this analysis to evaluate outcomes, the risk ofbias was not assessed. Finally, although a highlysensitive search strategy with no language orgeographic limits was used, it cannot be guar-anteed that all relevant studies have beenincluded.

CONCLUSION

The results of this analysis suggest that thecharacteristics of patients included in currentRA clinical trials do not differ from those ofpatients included in trials for testing the firstbDMARDs 20 years ago, despite current recom-mendations for a treat-to-target strategy. Fur-ther research is needed to determine the impactof patient characteristics on patients’ responseto RA treatments in clinical trials. It also appears

5 10 15 20 25

Mean count

MONARCH (2017)

BUILD (2017)

BEAM (2017)

MOBILITY (2015)

AMPLE (2014)

Li (2013)

GO−FORTH (2012)

SERENE (2010)

JESMR (2010)

GO−FORWARD (2009)

CNTO 148 (2008)

START (2006)

Abe (2006)

van (2004)

Keystone (2004)

Edwards (2004)

STAR (2003)

ARMADA (2003)

ATTRACT (1999)

18.05 [16.00, 20.10]

13.40 [11.57, 15.23]

15.30 [13.60, 17.00]

16.70 [14.97, 18.43]

15.85 [14.12, 17.58]

11.25 [ 9.40, 13.10]

11.60 [ 9.65, 13.55]

19.67 [17.36, 21.97]

12.55 [10.57, 14.53]

12.50 [10.26, 14.74]

13.50 [10.94, 16.06]

15.00 [13.06, 16.94]

14.30 [11.75, 16.85]

20.15 [17.89, 22.41]

19.15 [17.23, 21.07]

21.00 [17.54, 24.46]

21.10 [19.26, 22.94]

17.10 [14.62, 19.58]

21.50 [18.86, 24.14]

17.90 [15.23, 20.57]1999−2009

15.66 [13.39, 17.93]2010−2017

16.43 [14.48, 18.38]All Studies

I2 89 %Tau2 8.94p value: 0.213

Study Mean [95% CI]

(a) Swollen joint count (range 0−66)

10 15 20 25 30 35 40

Mean count

MONARCH (2017)

BUILD (2017)

BEAM (2017)

MOBILITY (2015)

AMPLE (2014)

Li (2013)

GO−FORTH (2012)

SERENE (2010)

JESMR (2010)

GO−FORWARD (2009)

CNTO 148 (2008)

START (2006)

Abe (2006)

van (2004)

Keystone (2004)

Edwards (2004)

STAR (2003)

ARMADA (2003)

ATTRACT (1999)

27.35 [25.04, 29.66]

24.03 [21.52, 26.55]

23.37 [21.28, 25.46]

26.83 [24.70, 28.96]

25.85 [23.68, 28.02]

22.70 [20.03, 25.37]

13.15 [10.99, 15.31]

28.67 [25.90, 31.43]

14.95 [12.61, 17.29]

23.50 [20.58, 26.42]

25.00 [21.13, 28.87]

22.00 [19.59, 24.41]

18.40 [15.55, 21.25]

34.60 [31.82, 37.38]

27.70 [25.45, 29.95]

32.67 [28.73, 36.60]

27.45 [25.43, 29.47]

28.35 [25.27, 31.43]

31.50 [28.26, 34.74]

26.48 [22.59, 30.37]1999−2009

24.74 [20.43, 29.05]2010−2017

25.34 [22.13, 28.54]All Studies

I2 94 %Tau2 25.54p value: 0.557

Study Mean [95% CI]

(b) Tender joint count (range 0−68)

Fig. 4 Forest plots showing means and 95% confidenceintervals (CI; box and whisker plots) for the differentstudies according to year of publication for studiesproviding data on a swollen joint count (score range0–66), and b tender joint count (score range 0–68).Diamond shapes indicate 95% confidence intervals aroundthe means summarised by timeframe and overall,

respectively. I2 and Tau2 values indicate a high degree ofheterogeneity between studies, while p values indicate nosignificant difference in mean scores between studiespublished from 1999 to 2009 (studies above dashed line)and those published from 2010 to 2017 (studies belowdashed line)

Adv Ther (2020) 37:3676–3691 3685

that patients currently being enrolled in RCTsof RA treatments are not being treated accord-ing to a treat-to-target strategy before the startof the study. Future RCTs of RA treatmentsshould include patients who have been treatedusing this strategy if RCTs are to remain infor-mative and impactful.

ACKNOWLEDGEMENTS

Funding. The SLR was conducted by RTIHealth Solutions under the direction of Eli Lillyand Company, Indianapolis, IN, USA. Thisanalysis was also funded by Eli Lilly and Com-pany. Employees of Eli Lilly and Companycontributed to the study protocol design, datainterpretation, and report development andreview. The Rapid Service and Open Access Feeswere funded by Eli Lilly and Company.

Medical Writing Assistance. The authorswould like to acknowledge Dr Sue Chambersand Sue Williamson (Rx Communications,Mold, UK) for medical writing assistance withthe preparation of this manuscript, funded byEli Lilly and Company.

Authorship. All named authors meet theInternational Committee of Medical JournalEditors (ICMJE) criteria for authorship for thisarticle, take responsibility for the integrity ofthe work as a whole, and have given theirapproval for this version to be published.

Disclosures. Roberto Caporali has nothingto disclose. Walid Fakhouri, Claudia Nicolay,Serena Losi and Veronica Rogai are employeesand minor stockholders of Eli Lilly and Com-pany. Harriet Longley is a former employee ofEli Lilly and Company and is now affiliated withthe University of Bath.

Compliance with Ethics Guidelines. Theanalysis reported in this article is based on pre-viously conducted studies and does not involveany studies with human participants or animalsperformed by any of the authors.

Data Availability. Data are not publiclyavailable.

Open Access. This article is licensed under aCreative Commons Attribution-NonCommer-cial 4.0 International License, which permitsany non-commercial use, sharing, adaptation,distribution and reproduction in any mediumor format, as long as you give appropriate creditto the original author(s) and the source, providea link to the Creative Commons licence, andindicate if changes were made. The images orother third party material in this article areincluded in the article’s Creative Commonslicence, unless indicated otherwise in a creditline to the material. If material is not includedin the article’s Creative Commons licence andyour intended use is not permitted by statutoryregulation or exceeds the permitted use, youwill need to obtain permission directly from thecopyright holder. To view a copy of this licence,visit http://creativecommons.org/licenses/by-nc/4.0/.

REFERENCES

1. Verma MK, Sobha K. Understanding the major riskfactors in the beginning and the progression ofrheumatoid arthritis: current scenario and futureprospects. Inflamm Res. 2015;64(9):647–59.

2. Ogdie A, Maliha S, Sin D, et al. Cause-specificmortality in patients with psoriatic arthritis andrheumatoid arthritis. Rheumatology (Oxford).2017;56(6):907–11.

3. Ometto F, Fedeli U, Schievano E, Botsios C, Punzi L,Corti MC. Cause-specific mortality in a large pop-ulation-based cohort of patients with rheumatoidarthritis in Italy. Clin Exp Rheumatol. 2018;36(4):636–42.

4. Myasoedova E, Crowson CS, Kremers HM, Ther-neau TM, Gabriel SE. Is the incidence of rheuma-toid arthritis rising? Results from Olmsted County,Minnesota, 1955–2007. Arthritis Rheum.2010;62(6):1576–82.

5. Tobon GJ, Youinou P, Saraux A. The environment,geo-epidemiology, and autoimmune disease:rheumatoid arthritis. J Autoimmun. 2010;35(1):10–4.

3686 Adv Ther (2020) 37:3676–3691

6. Singh JA, Saag KG, Bridges SL, et al. 2015 AmericanCollege of Rheumatology guideline for the treat-ment of rheumatoid arthritis. Arthritis Rheumatol.2016;68(1):1–26.

7. Smolen JS, Landewe R, Bijlsma J, et al. EULAR rec-ommendations for the management of rheumatoidarthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update. AnnRheum Dis. 2017;76(9):960–77.

8. Smolen JS, Aletaha D, Bijlsma JWJ, et al. Treatingrheumatoid arthritis to target: recommendations ofan international task force. Ann Rheum Dis.2010;69(4):631–7.

9. Haugeberg G, Boyesen P, Helgetveit K, Proven A.Clinical and radiographic outcomes in patientsdiagnosed with early rheumatoid arthritis in thefirst years of the biologic treatment era: a 10-yearprospective observational study. J Rheumatol.2015;42(12):2279–87.

10. Nieuwenhuis WP, de Wit MPT, Boonen A, van derHelm-van Mil AHM. Changes in the clinical pre-sentation of patients with rheumatoid arthritisfrom the early 1990s to the year 2010: earlieridentification but more severe patient reportedoutcomes. Ann Rheum Dis. 1990s;75(11):2054–6.

11. Solomon DH, Ayanian JZ, Yelin E, Shaykevich T,Brookhart MA, Katz JN. Use of disease-modifyingmedications for rheumatoid arthritis by race andethnicity in the National Ambulatory Medical CareSurvey. Arthritis Care Res (Hoboken). 2012;64(2):184–9.

12. Schmajuk G, Trivedi AN, Solomon DH, et al.Receipt of disease-modifying antirheumatic drugsamong patients with rheumatoid arthritis inMedicare managed care plans. JAMA. 2011;305(5):480–6.

13. Kilcher G, Hummel N, Didden EM, Egger M,Reichenbach S, GetReal Work Package 4. Rheuma-toid arthritis patients treated in trial and real worldsettings: comparison of randomized trials withregistries. Rheumatology. 2018;57(2):354–69.

14. Aletaha D, Smolen JS. Diagnosis and managementof rheumatoid arthritis: a review. JAMA.2018;320(13):1360–72.

15. Hartley L, Ahdesmaki O, Bell J, Mitchell S. System-atic literature review of clinical evidence of barici-tinib and comparator treatment in rheumatoidarthritis. RTI Health Solutions, Manchester, UK.Data on file, Eli Lilly and Company, 9 May 2018.

16. National Institute for Health and Care Excellence(NICE). The guidelines manual. Chapter 6.Reviewing the evidence. Process and Methods

[PMG6]. November 2012. https://www.nice.org.uk/process/pmg6/chapter/reviewing-the-evidence.Accessed 14 May 2019.

17. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMAGroup. Preferred reporting items for systematicreviews and meta-analyses: the PRISMA statement.Ann Intern Med. 2009;151(4):264–9.

18. Wan X, Wang W, Liu J, Tong T. Estimating thesample mean and standard deviation from thesample size, median, range and/or interquartilerange. BMC Med Res Methodol. 2014;14:135.

19. Viechtbauer W. Conducting meta-analyses in Rwith the metafor package. J Stat Softw. 2010;36:1–48.

20. Maini R, St Clair EW, Breedveld F, et al. Infliximab(chimeric anti-tumour necrosis factor alpha mono-clonal antibody) versus placebo in rheumatoidarthritis patients receiving concomitant methotrex-ate: a randomised phase III trial. ATTRACT StudyGroup Lancet. 1999;354(9194):1932–9.

21. Moreland LW, Schiff MH, Baumgartner SW, et al.Etanercept therapy in rheumatoid arthritis. A ran-domized, controlled trial. Ann Intern Med.1999;130(6):478–86.

22. Weinblatt ME, Kremer JM, Bankhurst AD, et al. Atrial of etanercept, a recombinant tumor necrosisfactor receptor: Fc fusion protein, in patients withrheumatoid arthritis receiving methotrexate.N Engl J Med. 1999;340(4):253–9.

23. Weinblatt ME, Keystone EC, Furst DE, et al. Adali-mumab, a fully human anti-tumor necrosis factoralpha monoclonal antibody, for the treatment ofrheumatoid arthritis in patients taking concomi-tant methotrexate: the ARMADA trial. ArthritisRheum. 2003;48(1):35–45.

24. Furst DE, Schiff MH, Fleischmann RM, et al. Adali-mumab, a fully human anti tumor necrosis factor-alpha monoclonal antibody, and concomitantstandard antirheumatic therapy for the treatmentof rheumatoid arthritis: results of STAR (Safety Trialof Adalimumab in Rheumatoid Arthritis).J Rheumatol. 2003;30(12):2563–71.

25. Edwards JC, Szczepanski L, Szechinski J, et al. Effi-cacy of B-cell-targeted therapy with rituximab inpatients with rheumatoid arthritis. N Engl J Med.2004;350(25):2572–81.

26. Keystone EC, Kavanaugh AF, Sharp JT, et al.Radiographic, clinical, and functional outcomes oftreatment with adalimumab (a human anti-tumornecrosis factor monoclonal antibody) in patientswith active rheumatoid arthritis receiving con-comitant methotrexate therapy: a randomized,

Adv Ther (2020) 37:3676–3691 3687

placebo-controlled, 52-week trial. Arthritis Rheum.2004;50(5):1400–11.

27. Lan JL, Chou SJ, Chen DY, Chen YH, Hsieh TY,Young M. A comparative study of etanercept plusmethotrexate and methotrexate alone in Taiwanesepatients with active rheumatoid arthritis: a12-week, double-blind, randomized, placebo-con-trolled study. J Formos Med Assoc. 2004;103(8):618–23.

28. Nishimoto N, Yoshizaki K, Miyasaka N, et al.Treatment of rheumatoid arthritis with humanizedanti–interleukin-6 receptor antibody: a multicenter,double-blind, placebo-controlled trial. ArthritisRheum. 2004;50(6):1761–9.

29. Klareskog L, van der Heijde D, de Jager JP, et al.Therapeutic effect of the combination of etanerceptand methotrexate compared with each treatmentalone in patients with rheumatoid arthritis: double-blind randomised controlled trial. Lancet. 2004;363(9410):675–81.

30. van de Putte L, Atkins C, Malaise M, et al. Efficacyand safety of adalimumab as monotherapy inpatients with rheumatoid arthritis for whom pre-vious disease modifying antirheumatic drug treat-ment has failed. Ann Rheum Dis. 2004;63(5):508–16.

31. Abe T, Takeuchi T, Miyasaka N, et al. A multicenter,double-blind, randomized, placebo controlled trialof infliximab combined with low dose methotrex-ate in Japanese patients with rheumatoid arthritis.J Rheumatol. 2006;33(1):37–44.

32. Kremer JM, Genant HK, Moreland LW, et al. Effectsof abatacept in patients with methotrexate-resistantactive rheumatoid arthritis: a randomized trial. AnnIntern Med. 2006;144(12):865–76.

33. Combe B, Codreanu C, Fiocco U, et al. Etanerceptand sulfasalazine, alone and combined, in patientswith active rheumatoid arthritis despite receivingsulfasalazine: a double-blind comparison. AnnRheum Dis. 2006;65(10):1357–62.

34. De Filippis L, Caliri A, Anghelone S, Scibilia G, LoGullo R, Bagnato G. Improving outcomes intumour necrosis factor a treatment: comparison ofthe efficacy of the tumour necrosis factor a blockingagents etanercept and infliximab in patients withactive rheumatoid arthritis. Panminerva Med.2006;48(2):129–35.

35. Westhovens R, Yocum D, Han J, et al. The safety ofinfliximab, combined with background treatments,among patients with rheumatoid arthritis and var-ious comorbidities: a large, randomized, placebo-controlled trial. Arthritis Rheum. 2006;54(4):1075–86.

36. Kim HY, Lee SK, Song YW, et al. A randomized,double-blind, placebo-controlled, phase III study ofthe human anti-tumor necrosis factor antibodyadalimumab administered as subcutaneous injec-tions in Korean rheumatoid arthritis patients trea-ted with methotrexate. APLAR J Rheumatol.2007;10(1):9–16.

37. Schiff M, Keiserman M, Codding C, et al. Efficacyand safety of abatacept or infliximab vs placebo inATTEST: a phase III, multi-centre, randomised,double-blind, placebo-controlled study in patientswith rheumatoid arthritis and an inadequateresponse to methotrexate. Ann Rheum Dis.2008;67(8):1096–103.

38. Miyasaka N. Clinical investigation in highly dis-ease-affected rheumatoid arthritis patients in Japanwith adalimumab applying standard and generalevaluation: the CHANGE study. Mod Rheumatol.2008;18(3):252–62.

39. Kay J, Matteson EL, Dasgupta B, et al. Golimumabin patients with active rheumatoid arthritis despitetreatment with methotrexate: a randomized, dou-ble-blind, placebo-controlled, dose-ranging study.Arthritis Rheum. 2008;58(4):964–75.

40. Smolen J, Beaulieu A, Rubbert-Roth A, et al. Effectof interleukin-6 receptor inhibition with tocilizu-mab in patients with rheumatoid arthritis (OPTIONstudy): a double-blind, placebo-controlled, ran-domised trial. Lancet. 2008;371(9617):987–97.

41. Keystone E, van der Heijde D, Mason J, et al. Cer-tolizumab pegol plus methotrexate is significantlymore effective than placebo plus methotrexate inactive rheumatoid arthritis: findings of a 52-week,phase III, multicenter, randomized, double-blind,placebo-controlled, parallel-group study. ArthritisRheum. 2008;58(11):3319–29.

42. Genovese MC, McKay JD, Nasonov EL, et al. Inter-leukin-6 receptor inhibition with tocilizumabreduces disease activity in rheumatoid arthritis withinadequate response to disease-modifying anti-rheumatic drugs: the tocilizumab in combinationwith traditional disease-modifying antirheumaticdrug therapy study. Arthritis Rheum. 2008;58(10):2968–80.

43. Keystone EC, Genovese MC, Klareskog L, et al.Golimumab, a human antibody to tumour necrosisfactor alpha given by monthly subcutaneousinjections, in active rheumatoid arthritis despitemethotrexate therapy: the GO-FORWARD Study.Ann Rheum Dis. 2009;68(6):789–96.

44. Smolen J, Landewe R, Mease P, et al. Efficacy andsafety of certolizumab pegol plus methotrexate inactive rheumatoid arthritis: the RAPID 2 study. A

3688 Adv Ther (2020) 37:3676–3691

randomised controlled trial. Ann Rheum Dis.2009;68(6):797–804.

45. Nishimoto N, Miyasaka N, Yamamoto K, et al. Studyof active controlled tocilizumab monotherapy forrheumatoid arthritis patients with an inadequateresponse to methotrexate (SATORI): significantreduction in disease activity and serum vascularendothelial growth factor by IL-6 receptor inhibitiontherapy. Mod Rheumatol. 2009;19(1):12–9.

46. Jones G, Sebba A, Gu J, et al. Comparison of tocilizu-mab monotherapy versus methotrexate monother-apy in patients with moderate to severe rheumatoidarthritis: the AMBITION study. Ann Rheum Dis.2010;69(1):88–96.

47. Kameda H, Ueki Y, Saito K, et al. Etanercept (ETN)with methotrexate (MTX) is better than ETNmonotherapy in patients with active rheumatoidarthritis despite MTX therapy: a randomized trial.Mod Rheumatol. 2010;20(6):531–8.

48. Emery P, Deodhar A, Rigby WF, et al. Efficacy andsafety of different doses and retreatment of ritux-imab: a randomised, placebo-controlled trial inpatients who are biological naive with activerheumatoid arthritis and an inadequate response tomethotrexate (Study Evaluating Rituximab’s Effi-cacy in MTX iNadequate rEsponders [SERENE]).Ann Rheum Dis. 2010;69(9):1629–35.

49. Kremer JM, Blanco R, Brzosko M, et al. Tocilizumabinhibits structural joint damage in rheumatoidarthritis patients with inadequate responses tomethotrexate: results from the double-blind treat-ment phase of a randomized placebo-controlledtrial of tocilizumab safety and prevention of struc-tural joint damage at one year. Arthritis Rheum.2011;63(3):609–21.

50. Kim HY, Hsu PN, Barba M, et al. Randomizedcomparison of etanercept with usual therapy in anAsian population with active rheumatoid arthritis:the APPEAL trial. Int J Rheum Dis. 2012;15(2):188–96.

51. Tanaka Y, Harigai M, Takeuchi T, et al. Golimumabin combination with methotrexate in Japanesepatients with active rheumatoid arthritis: results ofthe GO-FORTH study. Ann Rheum Dis. 2012;71(6):817–24.

52. van Vollenhoven R, Fleischmann R, Cohen S, et al.Tofacitinib or adalimumab versus placebo inrheumatoid arthritis. N Engl J Med. 2012;367(6):508–19.

53. Weinblatt M, Fleischmann R, Huizinga T, et al.Efficacy and safety of certolizumab pegol in a broadpopulation of patients with active rheumatoid

arthritis: results from the REALISTIC phase IIIbstudy. Rheumatol (Oxford). 2012;51(12):2204–14.

54. Gabay C, Emery P, Van Vollenhoven R, et al. Tocili-zumab monotherapy versus adalimumab monother-apy for treatment of rheumatoid arthritis (ADACTA):a randomised, double-blind, controlled phase 4 trial.Lancet. 2013;381(9877):1541–50.

55. Kang YM, Park W, Park YE, et al. Efficacy and safetyof certolizumab pegol (CZP) with concomitantmethotrexate (MTX) in Korean rheumatoid arthritis(RA) patients (PTS) with an inadequate response toMTX. Ann Rheum Dis. 2013;71(Suppl 3):666.

56. Li Z, Zhang F, Kay J, et al. Safety and efficacy ofsubcutaneous golimumab in Chinese patients withactive rheumatoid arthritis despite MTX therapy:results from a randomized, placebo-controlled,phase 3 trial. Arthritis Rheum. 2013;65(Suppl 10):S598–S599599.

57. van der Heijde D, Tanaka Y, Fleischmann R, et al.Tofacitinib (CP-690,550) inpatientswith rheumatoidarthritis receiving methotrexate: twelve-month datafrom a twenty-four-month phase III randomizedradiographic study. Arthritis Rheum. 2013;65(3):559–70.

58. O’Dell JR, Mikuls TR, Taylor TH, et al. Therapies foractive rheumatoid arthritis after methotrexate fail-ure. N Engl J Med. 2013;369(4):307–18.

59. Dougados M, Kissel K, Conaghan PG, et al. Clinical,radiographic and immunogenic effects after 1 yearof tocilizumab-based treatment strategies inrheumatoid arthritis: the ACT-RAY study. AnnRheum Dis. 2014;73(5):803–9.

60. Schiff M, Weinblatt ME, Valente R, et al. Head-to-head comparison of subcutaneous abatacept versusadalimumab for rheumatoid arthritis: two-yearefficacy and safety findings from AMPLE trial. AnnRheum Dis. 2014;73(1):86–94.

61. Kivitz A, Olech E, Borofsky M, et al. Subcutaneoustocilizumab versus placebo in combination withdisease-modifying antirheumatic drugs in patientswith rheumatoid arthritis. Arthritis Care Res(Hoboken). 2014;66(11):1653–61.

62. Yamamoto K, Takeuchi T, YamanakaH, et al. Efficacyand safetyof certolizumabpegol plusmethotrexate inJapanese rheumatoid arthritis patients with an inad-equate response to methotrexate: the J-RAPID ran-domized, placebo-controlled trial. Mod Rheumatol.2014;24(5):715–24.

63. Machado DA, Guzman RM, Xavier RM, et al. Open-label observation of addition of etanercept versus aconventional disease-modifying antirheumaticdrug in subjects with active rheumatoid arthritis

Adv Ther (2020) 37:3676–3691 3689

despite methotrexate therapy in the Latin Americanregion. J Clin Rheumatol. 2014;20(1):25–33.

64. Burmester GR, Rubbert-Roth A, Cantagrel A, et al. Arandomised, double-blind, parallel-group study ofthe safety and efficacy of subcutaneous tocilizumabversus intravenous tocilizumab in combinationwith traditional disease-modifying antirheumaticdrugs in patients with moderate to severe rheuma-toid arthritis (SUMMACTA study). Ann Rheum Dis.2014;73(1):69–74.

65. Genovese MC, Fleischmann R, Kivitz AJ, et al. Sar-ilumab plus methotrexate in patients with activerheumatoid arthritis and inadequate response tomethotrexate: results of a phase III study. ArthritisRheumatol. 2015;67(6):424–37.

66. Smolen JS, Burmester GR, Combe B, et al. Head-to-head comparison of certolizumab pegol versusadalimumab in rheumatoid arthritis: 2-year efficacyand safety results from the randomised EXXELE-RATE study. Lancet. 2016;388(10061):2763–74.

67. Peterfy C, Emery P, Tak PP, et al. MRI assessment ofsuppression of structural damage in patients withrheumatoidarthritis receiving rituximab: results fromthe randomised, placebo-controlled, double-blindRA-SCORE study. Ann RheumDis. 2016;75(1):170–7.

68. Burmester GR, Lin Y, Patel R, et al. Efficacy andsafety of sarilumab monotherapy versus adali-mumab monotherapy for the treatment of patientswith active rheumatoid arthritis (MONARCH): arandomised, double-blind, parallel-group phase IIItrial. Ann Rheum Dis. 2017;76(5):840–7.

69. Fleischmann R, Mysler E, Hall S, et al. Efficacy andsafety of tofacitinib monotherapy, tofacitinib withmethotrexate, andadalimumabwithmethotrexate inpatients with rheumatoid arthritis (ORAL Strategy): aphase 3b/4, double-blind, head-to-head, randomisedcontrolled trial. Lancet. 2017;390(10093):457–68.

70. Taylor PC, Keystone EC, Van Der Heijde D, et al.Baricitinib versus placebo or adalimumab inrheumatoid arthritis. N Engl J Med. 2017;376(7):652–62.

71. Dougados M, van der Heijde D, Chen YC, et al.Baricitinib in patients with inadequate response orintolerance to conventional synthetic DMARDs:results from the RA-BUILD study. Ann Rheum Dis.2017;76(1):88–95.

72. Bi L, Li Y, He L, et al. Rapid onset of responseobserved with certolizumab pegol in rheumatoidarthritis patients with inadequate response tomethotrexate: efficacy and safety results of a

randomized, double-blind, placebo-controlledphase 3 study. Arthritis Rheumatol. 2017;69:Abstr.2451.

73. Daıen CI, Morel J. Predictive factors of response tobiological disease modifying antirheumatic drugs:towards personalized medicine. Mediators Inflamm.2014;2014:386148.

74. Yamanaka H, Tanaka Y, Inoue E, et al. Efficacy andtolerability of tocilizumab in rheumatoid arthritispatients seen in daily clinical practice in Japan:results from a retrospective study (REACTIONstudy). Mod Rheumatol. 2011;21(2):122–33.

75. Leffers HC, Østergaard M, Glintborg B, et al. Effi-cacy of abatacept and tocilizumab in patients withrheumatoid arthritis treated in clinical practice:results from the nationwide Danish DANBIO reg-istry. Ann Rheum Dis. 2011;70(7):1216–22.

76. Maneiro RJ, Salgado E, Carmona L, Gomez-Reino JJ.Rheumatoid factor as predictor of response toabatacept, rituximab and tocilizumab in rheuma-toid arthritis: systematic review and meta-analysis.Semin Arthritis Rheum. 2013;43(1):9–17.

77. Curtis JR, Jain A, Askling J, et al. A comparison ofpatient characteristics and outcomes in selectedEuropean and US rheumatoid arthritis registries.Semin Arthritis Rheum. 2010;40(1):2–14.e1.

78. Diffin JG, Lunt M, Marshall T, Chipping JR, Sym-mons DP, Verstappen SM. Has the severity ofrheumatoid arthritis at presentation diminishedover time? J Rheumatol. 2014;41(8):1590–9.

79. Sato E, Tanaka E, Ochiai M, et al. Chronologicalchanges in baseline disease activity of patients withrheumatoid arthritis who received biologicDMARDs between 2003 and 2012. Mod Rheumatol.2015;25(3):350–7.

80. Rahman MU, Buchanan J, Doyle MK, et al. Changesin patient characteristics in anti-tumour necrosisfactor clinical trials for rheumatoid arthritis: resultsof an analysis of the literature over the past 16years. Ann Rheum Dis. 2011;70(9):1631–40.

81. Aaltonen KJ, Ylikyla S, Tuulikki J, et al. Efficacy andeffectiveness of tumour necrosis factor inhibitors inthe treatment of rheumatoid arthritis in random-ized controlled trials and routine clinical practice.Rheumatology. 2017;56(5):725–35.

82. Caporali R, Conti F, Covelli M, et al. Treatingrheumatoid arthritis to target: an Italian rheuma-tologists’ survey on the acceptance of the treat-to-

3690 Adv Ther (2020) 37:3676–3691

target recommendations. Clin Exp Rheumatol.2014;32(4):471–6.

83. Akdemir G, Markusse IM, Goekoop-Ruiterman YP,et al. Rheumatologists’ adherence to a diseaseactivity score steered treatment protocol in early

arthritis patients is less if the target is remission.Clin Rheumatol. 2017;36(2):317–26.

84. Ford JA, Solomon DH. Challenges in implementingtreat-to-target strategies in rheumatology. RheumDis Clin North Am. 2019;45(1):101–12.

Adv Ther (2020) 37:3676–3691 3691