Article Text


Concise report
Combination etanercept and methotrexate provides better disease control in very early (≤4 months) versus early rheumatoid arthritis (>4 months and <2 years): post hoc analyses from the COMET study
  1. Paul Emery1,2,
  2. Tore K Kvien3,
  3. Bernard Combe4,
  4. Bruce Freundlich5,
  5. Deborah Robertson6,
  6. Tahmina Ferdousi6,
  7. Eustratios Bananis6,
  8. Ronald Pedersen6,
  9. Andrew S Koenig6
  1. 1Division of Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK
  2. 2NIHR Leeds Musculoskeletal Biomedical Research Unit, Leeds Teaching Hospital NHS Trust, Leeds, UK
  3. 3Department of Rheumatology, Diakonhjemmet Hospital, Oslo, Norway
  4. 4Department of Rheumatology, Hospital Lapeyronie, Montpellier I University, Montpellier, France
  5. 5Division of Rheumatology, University of Pennsylvania, Philadelphia, Pennsylvania, USA
  6. 6Inflammation and Immunology, Pfizer Inc, Collegeville, Pennsylvania, USA
  1. Correspondence to Paul Emery, Arthritis Research UK Professor of Rheumatology, Director of Musculoskeletal Division. Division of Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Chapeltown Road, Leeds LS7 4SA, UK; p.emery{at}


Objective The objective of this post hoc analysis was to test the benefits of treating very early rheumatoid arthritis (VERA; ≤4 months) using COMET trial data. Treatment response in VERA and early rheumatoid arthritis (ERA; >4 months to 2 years) with combination etanercept+methotrexate (ETN+MTX) or MTX monotherapy was compared.

Methods Data assessed at week 52 for baseline disease duration effect included remission (disease activity score (DAS)28 <2.6, SDAI ≤3.3, Boolean), low disease activity (LDA; DAS28 <3.2), Boolean components of remission and radiographic non-progression. Subjects who discontinued because of lack of efficacy were included as non-responders.

Results Higher proportions of VERA subjects achieved LDA (79%) and DAS28 remission (70%) than ERA (62%, 48%, respectively, p<0.05) with ETN+MTX. Such high responses with MTX monotherapy were not observed (VERA, LDA=47%, DAS28 remission=35%; ERA, 47% and 32% respectively, p>0.70 for each). Regardless of disease duration, no radiographic progression was seen in 80% of subjects with ETN+MTX. In contrast, a higher proportion of VERA subjects showed no radiographic progression compared with ERA subjects treated with MTX (73.9% vs 50%, p=0.01).

Conclusions Treatment of VERA with ETN+MTX provides qualitatively improved clinical outcomes not seen with MTX monotherapy, supporting the pivotal role of TNF inhibition in early disease.

Statistics from


New treatment strategies, particularly tumour necrosis factor alpha (TNFα) antagonists, have provided historically unprecedented outcomes for rheumatoid arthritis (RA) and made clinical remission an achievable target. The window of opportunity for optimal outcomes in RA treatment is hypothesised to be very early in the disease (<2 years or possibly shorter durations)1,,3 when inhibition of TNFα may have a clinically significant impact on disease progression.4 The window-of-opportunity hypothesis is supported by clinical and preclinical evidence that has shown the strong influence of TNFα on pre-RA and early RA (ERA).4,,6 Disease duration has been shown to be an important predictor of response to treatment.6 7 Recent data indicate that subjects treated for ERA, who achieve remission, have better radiographic and quality-of-life outcomes over the long-term,6 8 9 and subjects treated with TNFα antagonists have higher rates of remission relative to subjects treated with other standard therapies.4 10 Furthermore, recent recommendations address the importance of treating RA patients to a target and establish that clinical remission or a minimum of low disease activity (LDA) should be the primary goal of treatment in order to optimise outcomes,11 12 which may in turn lower the long-term costs of healthcare and decrease work loss.13

The COMET (COmbination of Methotrexate (MTX) and ETanercept (ETN) in active early rheumatoid arthritis) trial demonstrated that treatment of active ERA (<2 years) with ETN+MTX was superior to MTX alone regarding disease activity reduction and remission rates.4 The COMET study population included subjects who had very early RA (VERA) (disease duration of ≤4 months) and ERA (>4 months and <2 years).

The objective of this post hoc analysis was to test the benefits of treating VERA, and establish if treatment of VERA with ETN+MTX or with MTX alone results in improved outcomes relative to ERA.


The methodology of the COMET study has been published elsewhere.4 In these post hoc analyses, observed data were analysed by baseline disease duration (VERA, ≤4 months or ERA, 4 months to 2 years) and by treatment group to determine the proportion of subjects who achieved remission after 52 weeks of treatment (the 52-week data were used because of the limited study population and subdivision of treatment groups at later time points).14 Remission was based on disease activity score (DAS28) <2.6, simplified disease activity index (SDAI) ≤3.3 and ACR/EULAR Boolean-based remission15 (defined as the subject must simultaneously achieve total swollen joint count ≤1, total tender joint count ≤1, C reactive protein (CRP) ≤1 mg/dl, and patient global assessment (PtGA) ≤1 on a 1–10 scale). The analyses also included the proportions of subjects achieving LDA (DAS28 <3.2), the change in individual Boolean components, SDAI components, physician global assessment and radiographic non-progression (mean Δmodified total Sharp score (mTSS) ≤0.5).

Statistical methods

Fisher's exact tests were performed to compare the treatment responses (LDA, DAS28 remission, SDAI remission, SDAI/Boolean components of remission and radiographic non-progression) at week 52 of subjects in the VERA and ERA subgroups within each treatment group and between treatment groups. These analyses included subjects who completed the study and subjects who discontinued due to lack of treatment efficacy (included as clinical non-responders).


Of the 417 subjects included in these analyses, 112 (26.9%) had VERA and 305 (73.1%) had ERA. Overall, demographics and baseline disease characteristics were similar between the VERA and ERA subjects for both treatment groups (table 1). Subjects with VERA had a slightly higher number of tender and swollen joints at baseline compared with ERA subjects, but also had slightly lower radiographic damage.

Table 1

Baseline demographics and disease state

Treatment of VERA subjects with combination ETN+MTX resulted in significantly better disease control (LDA and DAS28 remission) relative to treatment of ERA subjects (p=0.014 and 0.004, respectively; figure 1). The differences between VERA subjects and ERA subjects were insignificant when the groups were treated with MTX alone. VERA and ERA subjects fared similarly well when SDAI and Boolean remission rates were examined, although combination treatment led to consistently higher percentages of subjects achieving remission than MTX monotherapy (figure 1 and table 2).

Figure 1

Proportions of subjects achieving endpoints at week 52 by treatment group and baseline disease duration. *p<0.05 for treatment effect. DAS, disease activity score; ERA, early rheumatoid arthritis; ETN, etanercept; LDA, low disease activity; MTX, methotrexate; SDAI, simplified disease activity index; VERA, very early rheumatoid arthritis

Table 2

Percent of subjects with improvement in the Boolean* and SDAI remission components, radiographic non-progression and remission at week 52 by patient group and treatment

A greater proportion of subjects with VERA who received ETN+MTX achieved ≤1 swollen joint, ≤1 tender joint, CRP ≤1 mg/dl and physician global assessment ≤1 than subjects with ERA receiving the same treatment (table 2). Similarly, greater percentages of subjects in the VERA subgroup who received MTX alone met the criteria of ≤1 swollen joint, CRP ≤1 mg/dl and physician global assessment ≤1 than in the ERA subgroup. The percentages of subjects who achieved PtGA ≤1 were lower than the other Boolean components for both disease duration groups and both treatment regimens. Among those treated with MTX alone, a significantly greater proportion of subjects in the VERA group showed no radiographic progression relative to the ERA group (p=0.01, table 2). In both disease duration groups, between 80% and 81% of subjects treated with ETN+MTX had no radiographic progression at week 52 which was higher than MTX monotherapy (50%–74%, p=0.41 for treatment effect within VERA and p<0.01 for treatment effect within ERA). A significantly higher proportion of subjects treated with combination therapy (ETN+MTX) achieved DAS28, SDAI and Boolean remission in addition to ≤1 swollen joint compared with subjects treated with MTX alone, regardless of disease duration (p<0.01 for all, table 2).


In these post hoc analyses of the data from the COMET trial, we attempted to examine the window-of-opportunity hypothesis by analysing the clinical effects of treating VERA relative to ERA with combination ETN+MTX and MTX monotherapy. The percentage of VERA subjects who achieved DAS28 remission (when treated with ETN+MTX) was disproportionately higher than the ERA subjects. The proportions of subjects who achieved SDAI and Boolean remission were lower than those who achieved DAS28 remission. The lower SDAI and Boolean remission rates may reflect the lower percentage of subjects who achieved PtGA scores ≤1. Regardless of disease duration, significantly more patients on ETN+MTX achieved ≤1 swollen joint, DAS28, SDAI and Boolean remission compared with MTX alone.

Early diagnosis and treatment may alter the natural destructive course of RA,9 16 a hypothesis that was discussed as ‘the window-of-opportunity.’2 With DAS28 remission rate of 70% (one of the highest in modern RA literature), these data suggest that combination therapy for VERA subjects provides the best outcomes currently achievable. These data support the need for further research on the window-of-opportunity hypothesis and the probable benefits of very early treatment.

One of the limitations of this analysis is that the original protocol for the COMET study did not define all measures of RA used in this report; available data from the COMET study were used to calculate the SDAI and Boolean remission outcomes. In addition, the study population was a highly select group that qualified for the trial and the conclusions drawn here may not be applicable to the overall RA population. Another limitation was introduced by the definition of disease onset in the COMET study, which was the time of diagnosis rather than the time when symptoms first began. This criterion does not exclude the possibility that some subjects may have had persistent unrecognised disease for some time, as indicated by an average mTSS of 5.5 and 6.7 with high SD values in the VERA subjects (table 1). Further, the structure of a clinical trial does not allow new or modified treatment regimens for non-responsive subjects and thus does not reflect clinical practice. Finally, this is a short-term analysis of 52 weeks and subjects may live with RA for decades. More research is needed to address these limitations.

Future data analyses will need to examine the long-term impact of early versus late intervention in the progression of RA. Some data have been studied in a post hoc manner,8 10 including the present study; however, prospective studies would provide substantial benefit to clinicians and subjects. Whereas the impact of early intervention on cessation of therapy has been addressed elsewhere,17 it remains an important question not considered in this analysis. Retention and maintenance of employment are known to be improved by effective treatment of ERA,18 but future analyses will also need to compare the effect of treatment in ERA versus VERA. Future studies also need to be conducted on the barriers that may prevent patients from seeking early diagnosis and treatment of RA.

In the absence of a cure for RA, clinical remission is the ultimate goal of treatment for preventing structural joint damage and improving long-term quality of life.12 13 With the development of modern TNFα antagonist therapies, remission has become an achievable target for treatment.12 The data presented here include some of the highest rates of remission reported to date and support both the window-of-opportunity hypothesis and the hypothesis that TNFα may mediate disease progression in the early stages of RA. The data also strongly support the need for early diagnosis and treatment of RA. In conclusion, treatment of RA with combination therapy as early as possible may reduce the lifetime burden of disease, consistent with unique pathological events at that time.


The COMET study was funded by Wyeth, which was acquired by Pfizer Inc in October 2009. We wish to thank all participants, investigators and staff in the trial. Medical writing support was provided by Patricia McChesney, PhD, of UBC Scientific Solutions and was funded by Pfizer Inc.


View Abstract


  • Handling editor Johannes WJ Bijlsma

  • Competing interests The sponsor designed the study and provided assistance in manuscript preparation; data were held and analysed by the sponsor. All authors contributed to data interpretation and prepared, reviewed and approved the manuscript for submission. The corresponding author had full access to all the data and had final responsibility for the decision to submit for publication.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.