You are here

Article Text

Article menu

Download PDFPDF

Clinical and epidemiological research
Extended report
Progression of radiographic joint damage in rheumatoid arthritis: independence of erosions and joint space narrowing
  1. J S Smolen1,2,
  2. D M van der Heijde3,
  3. D Aletaha1,
  4. S Xu4,
  5. J Han4,
  6. D Baker4,
  7. E W St Clair5
  1. 1
    Division of Rheumatology, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria
  2. 2
    Second Department of Medicine, Center for Rheumatic Diseases, Hietzing Hospital, Vienna, Austria
  3. 3
    Leiden University Medical Center, Leiden, The Netherlands
  4. 4
    Centocor Incorporated, Malvern, Pennsylvania, USA
  5. 5
    Duke University Medical Center, Durham, North Carolina, USA
  1. Correspondence to J S Smolen, Division of Rheumatology, Department of Internal Medicine III, Medical University of Vienna, Waehringer Guertel 18–20, A-1090, Vienna, Austria; josef.smolen{at}wienkav.at

Abstract

Objective: To compare the progression of erosions and joint space narrowing (JSN) in patients with early active rheumatoid arthritis (RA) using data obtained in the “Active-controlled Study of Patients receiving Infliximab for the treatment of Rheumatoid arthritis of Early onset” (ASPIRE) study.

Methods: This was a post hoc analysis of patients in ASPIRE who received placebo plus methotrexate (MTX) or infliximab (3 or 6 mg/kg) plus MTX. Radiographs of the hands (870 patients) and feet (871 patients) were obtained at baseline and week 54 and scored using the van der Heijde/Sharp method. In total, 7160 joints in the placebo plus MTX group and 18 908 joints in the combined infliximab plus MTX group were included in this analysis.

Results: At baseline, 83.4% of joints in the placebo plus MTX group had no radiographic damage, 8.5% had only erosions, 4.4% had only JSN and 3.7% had both. The distribution was similar in the infliximab plus MTX group. In the placebo plus MTX group, the majority of joints did not have development or progression of radiographic damage from baseline to week 54; among joints that did have development or progression of damage at week 54, erosions occurred more often than JSN. The same pattern was observed in the infliximab plus MTX group, although the proportions of joints with damage at week 54 were generally larger in the placebo plus MTX group. There was a tendency for joints with existing erosions or JSN to have progression of damage, rather than development of new damage.

Conclusions: Erosions were the predominant type of damage observed in both treatment groups. Erosions and JSN are related but partly independent processes.

https://doi.org/10.1136/ard.2008.094128

Statistics from Altmetric.com

    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.

    Rheumatoid arthritis (RA) is characterised by persistent inflammation of the synovial membrane and often results in joint destruction within the first few years after disease onset.1 2 3 This joint damage, in the form of bone destruction and cartilage changes, can be significant and is a major contributor to disability in patients with RA.4 In clinical trials, this damage is routinely assessed using validated radiographic scores.5 6 7 In the Sharp score and its modifications,7 8 9 bone destruction is reflected in the erosion score and cartilage damage is reflected in the joint space narrowing (JSN) score; these are evaluated separately and the total score sums the erosion and JSN scores.

    The two components of joint damage appear to be a consequence of discrete, but inter-related, pathogenic processes. Bone erosion in patients with RA is mediated primarily by osteoclasts, which are activated by proinflammatory cytokines and other factors, and their destructive action is not counterbalanced by bone formation.10 11 12 13 14 Cytokines are also involved in inducing the breakdown of cartilage matrix via activation of chondrocytes and direct action of matrix metalloproteinases.15 16 17 Tumour necrosis factor α (TNFα), a proinflammatory cytokine, is involved in osteoclast activation and cartilage breakdown in patients with RA.18 19 20

    Treatment with disease modifying antirheumatic drugs, such as sulfasalazine, leflunomide and methotrexate (MTX), can slow the progression of radiographic damage and improve physical function.21 22 23 The advent of the anti-TNFα agents infliximab, etanercept and adalimumab has further increased the therapeutic potential for preventing the progression of joint destruction. Anti-TNFα agents, especially in combination with MTX, can significantly slow, and possibly halt, the progression of joint damage in patients with early RA24 25 and those with established RA.26 27 In the “Active-controlled Study of Patients receiving Infliximab for the treatment of Rheumatoid arthritis of Early onset” (ASPIRE) population, patients with early RA received either placebo plus MTX or infliximab plus MTX. Treatment with infliximab plus MTX effectively inhibited radiographic progression in these patients.25

    While there has been extensive research on the pathology of joint damage in RA, it is still largely unknown if erosions and JSN develop concomitantly or independently. These two components can be best studied as they are assessed, namely separately and in individual joints, particularly in patients with early RA who are more likely to have many joints with little or no destruction. We therefore performed a post hoc analysis of the ASPIRE population, evaluating the development and progression of erosions and JSN separately in patients with early RA.

    Patients and methods

    Patients and study design

    The ASPIRE study design and patient population have been described previously.25 Briefly, patients were between 18 and 75 years of age, had a diagnosis of active RA, had persistent synovitis for ⩾3 months and ⩽3 years and had not been previously treated with MTX. Patients were randomly assigned to receive infliximab 3 mg/kg, infliximab 6 mg/kg, or placebo infusions at 0, 2 and 6 weeks, and every 8 weeks thereafter to the end of week 46. All patients received MTX, with a rapid dose escalation to 20 mg/week. The ASPIRE protocol was approved by an institutional review board at each study centre and was carried out in accordance with the Helsinki Declaration and all subsequent revisions.

    Joint evaluations

    As previously reported, single radiographs of the hands (posteroanterior) and feet (anteroposterior) were taken at baseline and week 54.25 Radiographic damage (erosions and JSN) was evaluated using the van der Heijde modification of the Sharp score.9 Films were scored by two readers independently, without knowledge of the order of the radiographs and the treatment assignment. The intraclass correlation coefficients were 0.87 at baseline and 0.88 at week 54. The average score of the two readers was used for all analyses. Only joints that were evaluated for the presence of erosions and JSN were included in the present analysis. Scores for proximal interphalangeal (PIP), metacarpophalangeal (MCP), interphalangeal (IP) and metatarsophalangeal (MTP) joints were included, while scores of the wrists were excluded because different joints are scored for erosions and JSN. Based on this scoring method, the total score per patient could range from 0 to 340. The erosion scores for individual joints in the placebo plus MTX group and the infliximab plus MTX group, respectively, ranged from 0 to 7.5 and 0 to 8.5 at baseline and from 0 to 8.0 and 0 to 10.0 at week 54. JSN scores ranged from 0 to 4.0 at baseline and week 54 in both treatment groups.

    Data for all patients who had evaluable radiographs at baseline and week 54 were included in this post hoc analysis. Specifically, 870 patients had evaluable radiographs of the hands and 871 patients had evaluable radiographs of the feet. It should be noted that the population in this analysis differed from that reported in ASPIRE,25 which implemented different criteria for data analysis. Joints that did not have erosions or JSN at baseline were evaluated to examine the development of new radiographic damage at week 54. Worsening of existing damage was determined by evaluating joints that had only erosions, only JSN, or both at baseline.

    Mean erosion and JSN scores per joint were determined at baseline and at week 54, and the mean change from baseline per joint was also determined. The presence of erosions and JSN was defined by a score greater than zero, and worsening or new development of radiographic damage was defined as any increase from baseline greater than zero. Because the changes in total van der Heijde/Sharp scores were similar for the two infliximab plus MTX groups in the main ASPIRE study,25 the data for these two groups were combined for comparison with the placebo plus MTX group in this analysis.

    Statistical analysis

    We focused on the joints of patients who received placebo plus MTX and used the joints of patients treated with infliximab plus MTX for comparison. The Fisher exact test was used to compare the proportions of joints with erosions and/or JSN between the placebo plus MTX group and the infliximab plus MTX group. Continuous erosion and/or JSN scores per joint were compared between the two treatment groups using a two-sample t test on the van der Waerden normal scores. Nominal two-sided p values were reported. Logistic regression analysis was used to evaluate the influence of baseline joint damage and treatment group on the worsening of erosion and JSN scores from baseline to week 54. Generalised mixed linear modelling was used to model changes in erosion and JSN scores while adjusting for within-patient correlation. Statistical analyses were performed using the SAS V.8 software (SAS Institute, Cary, North Carolina, USA).

    Results

    Baseline characteristics

    The placebo plus MTX group included 7160 evaluable joints. Of these joints, 610 (8.5%) had only erosions at baseline, 315 (4.4%) had only JSN, 265 (3.7%) had both and 5970 (83.4%) had no radiographic damage (table 1). The infliximab plus MTX group included 18 908 evaluable joints; at baseline, 1794 (9.5%) had only erosions, 780 (4.1%) had only JSN, 574 (3.0%) had both and 15 760 (83.4%) had no radiographic damage. Overall, the proportion of eroded joints was significantly higher than the proportion of joints exhibiting JSN (p<0.001).

    View this table:
    Table 1

    Mean erosion and joint space narrowing scores per joint at baseline and week 54

    Assessment of radiographic damage at week 54 in the placebo plus MTX group

    Most (94.1%) of the 5970 joints that had no radiographic damage at baseline were also free of erosions and JSN at week 54; 4.4% developed only erosions, 0.8% developed only JSN and 0.8% developed both types of damage at week 54 (fig 1A). Among joints that had only erosions at baseline, 14.1% had worsening of erosions only, 1.3% developed only JSN and 0.8% had worsening of erosions and development of JSN (fig 1B). Thus, among these joints, worsening of erosions occurred more frequently (14.9%) than development of JSN (2.1%; p<0.001). Among joints that had only JSN at baseline, 3.5% had worsening of JSN only, 9.5% had formation of erosions only and 3.5% had worsening of JSN and the formation of erosions (fig 1C); therefore, a total of 13.0% of joints developed new erosions, while 7.0% had worsening of JSN at week 54 (p = 0.012).

    Figure 1
    Figure 1

    The percentage of joints with erosions, joint space narrowing, or both at week 54 among joints that had no radiographic damage at baseline (A), joints that had only erosions at baseline (B), joints that had only joint space narrowing at baseline (C) and joints that had erosions and joint space narrowing at baseline (D). ERN, erosion; JSN, joint space narrowing, MTX, methotrexate.

    Among joints that had both types of damage at baseline, 14.0% had worsening of erosions only, 8.3% had worsening of JSN only and 4.2% had worsening of both types of damage at week 54 (fig 1D). In total, 18.1% of joints had worsening of erosions and 12.5% had worsening of JSN at week 54 (p = 0.070). Thus, erosions were the dominant process at week 54, regardless of existing damage at baseline. Further, these findings pertained to all joint regions (PIP, MCP, IP and MTP) that were included in this analysis (data not shown).

    Assessment of radiographic damage at week 54 in the infliximab plus MTX group

    The results observed in the placebo plus MTX group were confirmed in the infliximab plus MTX group (fig 1A–D), revealing a similar pattern of radiographic damage at week 54. The proportion of joints affected, however, was significantly lower in the infliximab plus MTX group than in the placebo plus MTX group for several comparisons. Among joints that did not have radiographic damage at baseline, the proportion that developed new radiographic damage at week 54 was significantly lower in the infliximab plus MTX group (3.4%) than in the placebo plus MTX group (5.9%; p<0.001; data not shown). The proportions of joints in the infliximab plus MTX group that did not have radiographic damage at baseline and developed only erosions (2.5%) or erosions and JSN (0.2%) were also significantly lower than in the placebo plus MTX group (4.4% and 0.8%, respectively; p<0.001 for both comparisons; fig 1A).

    Among joints that had only erosions at baseline, the proportion of joints that had worsening of erosions at week 54 was significantly lower in the infliximab plus MTX group than in the placebo plus MTX group (8.7% vs 14.1%; p<0.001; fig 1B). Overall, there were no significant differences between the two treatment groups in the proportions of joints that had development or worsening of only JSN at week 54. Of joints that had erosions and JSN at baseline, a greater proportion in the infliximab plus MTX group did not have any progression of damage from baseline to week 54 when compared with the placebo plus MTX group (82.2% vs 73.6%; p = 0.006; data not shown); however, there were no significant differences between the treatment groups in the progression of erosions and JSN.

    Radiographic scores at week 54 for joints that had no radiographic damage at baseline

    Among joints that did not exhibit any damage at baseline, the mean erosion, JSN and total scores per joint at week 54 were significantly lower in the infliximab plus MTX group than in the placebo plus MTX group (p<0.001 for all; fig 2A).

    Figure 2
    Figure 2

    The mean changes from baseline to week 54 in erosion scores, joint space narrowing scores and total scores per joint for joints that had no radiographic damage at baseline (A), joints that had only erosions at baseline (B), joints that had only joint space narrowing at baseline (C) and joints that had erosions and joint space narrowing at baseline (D). ERN, erosion; JSN, joint space narrowing; MTX, methotrexate.

    Radiographic scores at week 54 for joints that had only erosions at baseline

    In joints with only erosions at baseline, the mean JSN scores per joint at week 54 were similar in the two treatment groups. Mean erosion scores decreased in both treatment groups, with the infliximab plus MTX group having a significantly greater mean (SD) decrease (−0.14 (0.49)) when compared with the placebo plus MTX group (−0.01 (0.57); p<0.001; fig 2B). The mean (SD) changes in total scores were −0.11 (0.56) in the infliximab plus MTX group and 0.02 (0.66) in the placebo plus MTX group (p<0.001).

    Radiographic scores at week 54 for joints that had only joint space narrowing at baseline

    In joints that had only JSN at baseline, there was no significant difference between the two treatment groups in the mean changes in JSN scores (p = 0.124; fig 2C). However, the mean (SD) increase in erosion scores from baseline to week 54 was significantly lower in the infliximab plus MTX group than in the placebo plus MTX group (0.05 (0.21) vs 0.13 (0.55); p<0.001), leading to mean (SD) changes in the total scores of −0.03 (0.50) and 0.10 (0.78), respectively (p = 0.004).

    Radiographic scores at week 54 for joints that had erosions and joint space narrowing at baseline

    In joints that had erosions and JSN at baseline, the mean erosion, JSN and total scores decreased from baseline to week 54 in both treatment groups (fig 2D). The mean (SD) decreases in JSN and total scores were greater in the infliximab plus MTX group than in the placebo plus MTX group (−0.09 (0.49) vs −0.03 (0.46); p = 0.047 and −0.20 (0.91) vs −0.06 (0.75); p = 0.014, respectively). There was a similar trend in the mean changes in erosion scores for the two treatment groups; however, the difference was not statistically significant (p = 0.059).

    Logistic regression analysis

    Regression analysis showed that the presence of only erosions, only JSN, or both at baseline was significantly associated with an increase in erosion score from baseline to week 54 (table 2). The presence of JSN at baseline, alone and in combination with erosions, was significantly associated with an increase in JSN score from baseline to week 54. However, the presence of only erosions at baseline was not significantly associated with an increase in JSN score. Overall, the probability of having an increase in erosion or JSN scores from baseline to week 54 was significantly lower for joints in the infliximab plus MTX group than for those in the placebo plus MTX group, regardless of baseline joint damage. In all of the regression models, inclusion of the interaction term for the treatment group and baseline joint damage found that this was not significantly associated with increases in erosion or JSN scores (data not shown).

    View this table:
    Table 2

    Logistic regression analysis of the associations between baseline joint damage and treatment group and increases in erosion and joint space narrowing scores from baseline to week 54

    Generalised mixed linear modelling

    Estimated marginal mean changes in erosion and JSN scores are shown in fig 3. These results show that among joints with erosions at baseline (alone or with JSN), those in the infliximab plus MTX group had greater improvements in erosion scores than joints in the placebo plus MTX group. Further, among joints that did not have erosions at baseline, erosion scores at week 54 were lower in joints in the infliximab plus MTX group when compared with those in the placebo plus MTX group. Among joints that had JSN at baseline (alone or with erosions), those in the infliximab plus MTX group had greater decreases in JSN scores compared with those in the placebo plus MTX group. However, among joints that did not have JSN at baseline, the estimated marginal mean changes in JSN scores from baseline to week 54 were very small and were not significantly different between the two treatment groups.

    Figure 3
    Figure 3

    Estimated marginal mean changes from baseline to week 54 in erosion and joint space narrowing scores. Values are the estimated marginal means and 95% CIs. ERN, erosion; JSN, joint space narrowing; MTX, methotrexate.

    Discussion

    Erosions and JSN are specific features of RA and can be reliably evaluated using the van der Heijde/Sharp method.9 28 While both subcomponents of joint damage are elicited by the underlying disease processes, it is not clear if these components generally run a parallel, common, or at least, a partially independent course. The analyses reported here allowed a comprehensive view of the changes in joint damage among patients with early RA.

    Most of the joints included in this analysis did not have any radiographic damage at baseline, which is consistent with the characteristics of a patient population with early RA. Of the joints that did have damage at baseline, erosions were more common than JSN. Relatively few joints had both types of damage at baseline. This is consistent with a predominance of the erosive process in bone over cartilage degradation and with observations in TNF transgenic mice that suffered from severe arthritis and experienced severe bone destruction rather than severe cartilage damage, at least for a long period of time.19 Alternatively, this may also reflect limitations in the sensitivity of the instrument used to assess these two components, the radiographic scoring system.

    While most joints that did not have radiographic damage at baseline still did not exhibit radiographic damage at week 54, some of these joints (5.9% in the placebo plus MTX group and 3.4% in the infliximab plus MTX group) developed new joint damage. This is an important observation because it reveals that among joints apparently unaffected by structural damage, a considerable proportion developed radiographic damage despite therapy. This could be due to a delay in the visibility of radiographic changes, but it might also reflect de novo joint damage. It would be of interest to learn from prospective, controlled trials if such damage further accrues over time, and if such an increase is linear or reaches a plateau.

    Overall, at week 54, erosions were more common than JSN, regardless of the type of baseline joint damage. Of note, the worsening of existing JSN occurred more frequently than the development of new JSN. The worsening of existing erosions occurred slightly more often than the development of new erosions, which is consistent with previous results of the Combinatietherapie Bij Reumatoide Artritis (COBRA) trial.29 Our results, however, showed that the erosive process did not progress in the majority of joints affected at baseline and even appeared to improve, as indicated by an overall reduction in erosion scores among joints with erosions at baseline. In addition, changes in JSN scores showed that joints with this type of damage also tended not to show progression, with some appearing to improve, leading to a reduction from baseline in the JSN score at week 54. While the improvements in scores were relatively small, previous research has shown that even small differences in scores could be clinically meaningful over time.30

    The mean decrease in erosion scores in joints that had only erosions at baseline seemed counterintuitive based on the substantial proportion of joints that had worsening of erosions from baseline to week 54 (14.9%, placebo plus MTX group; 9.2%, infliximab plus MTX group). Upon further investigation, we found that among joints in the placebo plus MTX group that had only erosions at baseline, 64.4% had no change in score at week 54 and 20.7% had a decrease in score (data not shown). Thus, the overall improvement in the erosion score is explained by the large proportion of joints that had an improvement in score. Similar trends were observed for other joints that experienced a mean decrease in score. Importantly, for most of these mean decreases in scores, the effect of infliximab plus MTX was significantly larger than that of MTX alone. Consistent with previous observations,24 25 27 31 this indicates that TNF inhibition, such as with infliximab, in combination with MTX is more effective on erosions and JSN than MTX monotherapy. These findings were consistent for the development of new radiographic damage and worsening of existing damage.

    While these analyses addressed an issue that has not been previously investigated, this study has several limitations. First, this was a post hoc exploratory analysis. ASPIRE was not designed for the purpose of this analysis; although, radiographs were read by readers who were blinded to patient identity and clinical outcomes. This study addressed patients with early RA; therefore, the joint characteristics would presumably vary from patients with established RA. Additionally, these data were obtained after a relatively short 1-year treatment period, and the long-term patterns of radiographic progression have to be addressed separately.

    The scoring method used in this analysis also presents some limitations as it is intended to be used for assessing total joint damage in patients, rather than comparisons of individual joints; however, to obtain the total scores, each joint is scored separately. Importantly, there was substantial agreement between the scores of the two readers; although the precision, as well as the sensitivity to change, in scoring of the two components of joint damage may have differed. The data presented here pertain to changes in erosion and JSN scores and are limited by the sensitivity of the scoring method. It is possible that some bone abnormalities and cartilage degradation were too small to be detected and may have occurred without being reflected in the radiographic scores. Additionally, in the van der Heijde/Sharp method, JSN scores may also depict subluxation and luxation; however, these are uncommon in early disease and did not play a significant role in the present analysis. Finally, the wrist is an important location for JSN, but because this analysis included only those joints that were scored for erosions and JSN, the wrist could not be included. Consequently, we might have missed significant additional information on the occurrence of JSN.

    In summary, in contrast to most previously published data regarding radiographic damage in RA, this investigation evaluated erosions and JSN of individual joints separately and found them to be partly discrete processes. While this does not constitute an attempt to alter current methods of assessing the radiographic damage of RA, these data provide a more detailed view into the progression and development of erosions and JSN as separate, potentially independent events. The data obtained from this analysis also suggest that specific therapies could be developed that would preferentially affect one type of joint damage and allow for more targeted treatment strategies. Another interesting question is the temporal relationship of the occurrence of erosions and JSN. To address this issue, however, more frequent evaluations of joint damage would be required during a short time period. Additional studies are needed to evaluate the development and progression of joint damage and their roles in disease progression in early and established RA.

    Acknowledgments

    The authors thank Rebecca E Clemente and Mary Whitman of Centocor for writing support.

    REFERENCES

    1. ↵
      1. Fex E,
      2. Jonsson K,
      3. Johnson U,
      4. et al
      . Development of radiographic damage during the first 5–6 yr of rheumatoid arthritis. A prospective follow-up study of a Swedish cohort. Br J Rheumatol 1996;35:1106–15.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Sharp JT,
      2. Wolfe F,
      3. Mitchell DM,
      4. et al
      . The progression of erosion and joint space narrowing scores in rheumatoid arthritis during the first twenty-five years of disease. Arthritis Rheum 1991;34:660–8.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. van der Heijde DM,
      2. van Leeuwen MA,
      3. van Riel PL,
      4. et al
      . Biannual radiographic assessments of hands and feet in a three-year prospective followup of patients with early rheumatoid arthritis. Arthritis Rheum 1992;35:26–34.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Aletaha D,
      2. Smolen J,
      3. Ward MM
      . Measuring function in rheumatoid arthritis: identifying reversible and irreversible components. Arthritis Rheum 2006;54:2784–92.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Aletaha D,
      2. Smolen JS
      . The definition and measurement of disease modification in inflammatory rheumatic diseases. Rheum Dis Clin North Am 2006;32:9–44.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Larsen A
      . How to apply Larsen score in evaluating radiographs of rheumatoid arthritis in longterm studies. J Rheumatol 1995;22:1974–5.
      OpenUrlPubMedWeb of Science
    7. ↵
      1. Sharp JT,
      2. Lidsky MD,
      3. Collins LC,
      4. et al
      . Methods of scoring the progression of radiologic changes in rheumatoid arthritis. Correlation of radiologic, clinical and laboratory abnormalities. Arthritis Rheum 1971;14:706–20.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Genant HK
      . Interleukin-1 receptor antagonist treatment of rheumatoid arthritis patients: radiologic progression and correlation of Genant/Sharp and Larsen scoring methods. Semin Arthritis Rheum 2001;30:26–32.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. van der Heijde DM
      . Plain x-rays in rheumatoid arthritis: overview of scoring methods, their reliability and applicability. Baillieres Clin Rheumatol 1996;10:435–53.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Diarra D,
      2. Stolina M,
      3. Polzer K,
      4. et al
      . Dickkopf-1 is a master regulator of joint remodeling. Nat Med 2007;13:156–63.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Kitaura H,
      2. Zhou P,
      3. Kim HJ,
      4. et al
      . M-CSF mediates TNF-induced inflammatory osteolysis. J Clin Invest 2005;115:3418–27.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Lam J,
      2. Takeshita S,
      3. Barker JE,
      4. et al
      . TNF-alpha induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand. J Clin Invest 2000;106:1481–8.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Pettit AR,
      2. Ji H,
      3. von Stechow D,
      4. et al
      . TRANCE/RANKL knockout mice are protected from bone erosion in a serum transfer model of arthritis. Am J Pathol 2001;159:1689–99.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Redlich K,
      2. Hayer S,
      3. Ricci R,
      4. et al
      . Osteoclasts are essential for TNF-alpha-mediated joint destruction. J Clin Invest 2002;110:1419–27.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Pap T,
      2. Shigeyama Y,
      3. Kuchen S,
      4. et al
      . Differential expression pattern of membrane-type matrix metalloproteinases in rheumatoid arthritis. Arthritis Rheum 2000;43:1226–32.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Rawdanowicz TJ,
      2. Hampton AL,
      3. Nagase H,
      4. et al
      . Matrix metalloproteinase production by cultured human endometrial stromal cells: identification of interstitial collagenase, gelatinase-A, gelatinase-B, and stromelysin-1 and their differential regulation by interleukin-1 alpha and tumor necrosis factor-alpha. J Clin Endocrinol Metab 1994;79:530–6.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Stephenson ML,
      2. Goldring MB,
      3. Birkhead JR,
      4. et al
      . Stimulation of procollagenase synthesis parallels increases in cellular procollagenase mRNA in human articular chondrocytes exposed to recombinant interleukin 1 beta or phorbol ester. Biochem Biophys Res Commun 1987;144:583–90.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Kobayashi K,
      2. Takahashi N,
      3. Jimi E,
      4. et al
      . Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL-RANK interaction. J Exp Med 2000;191:275–85.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Redlich K,
      2. Hayer S,
      3. Maier A,
      4. et al
      . Tumor necrosis factor alpha-mediated joint destruction is inhibited by targeting osteoclasts with osteoprotegerin. Arthritis Rheum 2002;46:785–92.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Saklatvala J
      . Tumour necrosis factor alpha stimulates resorption and inhibits synthesis of proteoglycan in cartilage. Nature 1986;322:547–9.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Smolen JS,
      2. Kalden JR,
      3. Scott DL,
      4. et al
      . Efficacy and safety of leflunomide compared with placebo and sulphasalazine in active rheumatoid arthritis: a double-blind, randomised, multicentre trial. Lancet 1999;353:259–66.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Strand V,
      2. Scott DL,
      3. Emery P,
      4. et al
      . Physical function and health related quality of life: analysis of 2-year data from randomized, controlled studies of leflunomide, sulfasalazine, or methotrexate in patients with active rheumatoid arthritis. J Rheumatol 2005;32:590–601.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. van der Heijde DM,
      2. van Riel PL,
      3. Nuver-Zwart IH,
      4. et al
      . Effects of hydroxychloroquine and sulphasalazine on progression of joint damage in rheumatoid arthritis. Lancet 1989;333:1036–8.
      OpenUrlCrossRef
    24. ↵
      1. Breedveld FC,
      2. Weisman MH,
      3. Kavanaugh AF,
      4. et al
      . The PREMIER study: a multicenter, randomized, double-blind clinical trial of combination therapy with adalimumab plus methotrexate versus methotrexate alone or adalimumab alone in patients with early, aggressive rheumatoid arthritis who had not had previous methotrexate treatment. Arthritis Rheum 2006;54:26–37.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. St Clair EW,
      2. van der Heijde DM,
      3. Smolen JS,
      4. et al
      . Combination of infliximab and methotrexate therapy for early rheumatoid arthritis: a randomized, controlled trial. Arthritis Rheum 2004;50:3432–43.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Klareskog L,
      2. van der Heijde D,
      3. de Jager JP,
      4. et al
      . Therapeutic effect of the combination of etanercept and methotrexate compared with each treatment alone in patients with rheumatoid arthritis: double-blind randomised controlled trial. Lancet 2004;363:675–81.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Lipsky PE,
      2. van der Heijde DM,
      3. St Clair EW,
      4. et al
      . Infliximab and methotrexate in the treatment of rheumatoid arthritis. N Engl J Med 2000;343:1594–602.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Boini S,
      2. Guillemin F
      . Radiographic scoring methods as outcome measures in rheumatoid arthritis: properties and advantages. Ann Rheum Dis 2001;60:817–27.
      OpenUrlAbstract/FREE Full Text
    29. ↵
      1. Bruynesteyn K,
      2. van der Heijde D,
      3. Boers M,
      4. et al
      . Contribution of progression of erosive damage in previously eroded joints in early rheumatoid arthritis trials: COBRA trial as an example. Arthritis Rheum 2002;47:532–6.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. van der Heijde D,
      2. Landewe R,
      3. van Vollenhoven R,
      4. et al
      . Level of radiographic damage and radiographic progression are determinants of physical function: a longitudinal analysis of the TEMPO trial. Ann Rheum Dis 2008;67:1267–70.
      OpenUrlAbstract/FREE Full Text
    31. ↵
      1. Smolen JS,
      2. Han C,
      3. Bala M,
      4. et al
      . Evidence of radiographic benefit of treatment with infliximab plus methotrexate in rheumatoid arthritis patients who had no clinical improvement: a detailed subanalysis of data from the anti-tumor necrosis factor trial in rheumatoid arthritis with concomitant therapy study. Arthritis Rheum 2005;52:1020–30.
      OpenUrlCrossRefPubMedWeb of Science

    Footnotes

    • Funding This study was funded by Centocor, Horsham, Pennsylvania, USA and Schering-Plough, Kenilworth, New Jersey, USA.

    • Competing interests SX, JH and DB are employed by the sponsors of this study (Centocor). JSS has received research grants and/or honoraria from Abbott, Amgen, Astra-Zeneca, BMS, Centocor, Chugai, Novartis, Pfizer, Roche, Sanofi-Aventis, Schering-Plough, UCB and Wyeth. DMvdH has received research grants and/or consulting fees from Abbott, Amgen, BMS, Centocor, Chugai, Novartis, Pfizer, Schering-Plough, UCB and Wyeth. DA has received honoraria from Abbott, Roche, Sanofi-Aventis, Schering-Plough and Wyeth. EWS has received research grants from Genentech, Amgen and Centocor and consulting fees from Genentech, Centocor and Bristol-Myers Squibb, as well as a fellowship training grant from Centocor. SX, JH and DB own stock in Johnson & Johnson, of which Centocor is a wholly owned subsidiary.

    • Ethics approval Ethics approval was obtained.