Skip to main content
  • Research article
  • Open access
  • Published:

Changes in clinical disease activity are weakly linked to changes in MRI inflammation on treat-to-target escalation of therapy in rheumatoid arthritis



Rheumatoid arthritis (RA) treat-to-target (T2T) regimens often use the disease activity score (28 joints) incorporating C-reactive protein (DAS28CRP) as an outcome measure. We compared changes in the DAS28CRP with changes in magnetic resonance imaging (MRI) inflammation on treatment escalation.


Eighty seropositive RA patients with active disease were enrolled. Group A (N = 57) escalated to another conventional disease-modifying therapy (cDMARD) combination, and Group B (N = 23) to anti-TNF therapy/cDMARDs. Contrast-enhanced 3T-MRI wrist scans were obtained before and 4 months after regimen change. Scan pairs were scored for inflammation (MRI(i)) and damage. Disease activity was assessed using the DAS28CRP.


Eighty patients were enrolled and 66 MRI scan pairs were available for analysis. Intra-reader reliability was high: intraclass correlation coefficient (average) 0.89 (0.56–0.97). ΔDAS28CRP did not differ between groups: Group A, −0.94 (−3.30, 1.61); Group B, −1.53 (−3.59, 0.56) (p = 0.45). ΔMRI(i) also did not differ: Group A, 0 (−25, 10); Group B, −1 (−15, 28) (p = 0.12). Combining groups, ΔMRI(i) correlated weakly with ΔDAS28CRP (Spearman’s 0.36, p = 0.003). Using multiple linear regression analysis adjusting for confounders, ΔDAS28CRP was associated with ΔMRI(i) (p = 0.056). Of the individual MRI measures, only Δtenosynovitis correlated with ΔDAS28CRP (Spearman’s 0.33, p = 0.007). ΔMRI(i) was negatively associated with the MRI erosion score at entry (p = 0.0052).


We report the first study investigating the link between changes in clinical and imaging inflammation in a real-world RA cohort escalating to conventional and biologic DMARDs. The association was significant but relatively weak, suggesting that MRI targets cannot yet be advocated as outcomes for T2T escalation.

Trial registration

ANZCTR 12614000895684. Registered 22 August 2014.


Rheumatologists routinely manage rheumatoid arthritis (RA) using a treat-to-target (T2T) regimens [1], escalating therapy in patients who fail to reach remission or a low disease activity state [2]. Management is often initiated with methotrexate alone and then escalation proceeds to combination therapy with conventional disease-modifying anti-rheumatic agents (cDMARDs) and biologic DMARDs (bDMARDs). In New Zealand, cDMARD combinations include triple therapy (methotrexate, sulphasalazine, hydroxychloroquine) and methotrexate/leflunomide, although other options are available. The New Zealand Pharmaceutical Management Agency (PHARMAC), responsible for buying and funding pharmaceuticals, has defined criteria for funding bDMARDs including a trial of three different cDMARD combinations [3]. For those with an inadequate response to cDMARDs, escalation to methotrexate/anti-tumour necrosis factor (anti-TNF) agents follows and subsequently other bDMARDs. Adequacy of response is usually measured by the DAS28CRP [4, 5].

MRI is a sensitive imaging modality that can reveal articular inflammation and joint damage [6, 7]. The OMERACT RA MRI score (RAMRIS) [8] and the Harvaardsholm tenosynovitis score [9] can quantify inflammation in a relatively reproducible manner. Studies show that MRI can detect more inflammation than physical examination. Up to 96% of asymptomatic RA patients with normal joints have MRI synovitis and 46% have MRI bone oedema/osteitis [10]. Krabben et al. [11] found MRI inflammation in 66% of non-swollen wrist joints in patients with early RA. Severe grades of bone oedema (osteitis) were common in non-swollen joints. We and others have found that osteitis is a predictor of joint erosion [12, 13]. Although MRI is more sensitive than clinical assessment for detection of joint inflammation and damage, MRI outcomes are not yet part of accepted clinical practice because of feasibility issues including cost and accessibility. It is also not clear whether using the change in MRI inflammation scores to dictate escalation of therapy would lead to better outcomes than using easily administered clinical scores such as the DAS28CRP. While clinical trials of bDMARDs including golimumab [14] and tocilizumab [15] indicated that clinical responses were associated with improvements in MRI inflammation scores, there is little information comparing clinical and MRI responses to cDMARDs in the T2T “real-world” setting.

The aim of this study was to compare the response in clinical disease activity (DAS28CRP) with response in MRI inflammation scores after T2T-dictated escalation of therapy in two groups of RA patients: those escalating from cDMARDs to a different cDMARD combination and those escalating to a combination that includes anti-TNF therapy.


Power calculation

Studies exploring the relation between ΔMRI synovitis scores and ΔDAS28CRP [7, 16, 17] indicated a simple correlation of 0.4. Using a two-sided test at the 5% significance level with 80% power, the required total sample size would be 46 patients. Allowing for 10% dropout, the estimated sample size required was 50. For comparing MRI(i) responses between patients escalating to cDMARDs versus cDMARD/anti-TNFs, sample size was calculated based on the one-sided hypothesis that triple therapy would not be inferior to MTX/anti-TNF. We estimated that with 50 patients per group the study would have 80% power at 5% significance to establish non-inferiority, assuming a between-group difference in change from baseline to 3 months of 2 (SD = 2). Although enrolment was continued for 3 years, only 80 patients with seropositive RA were recruited from rheumatology outpatients’ clinics from Auckland and Christchurch, New Zealand between 2014 and 2016. Two cohorts were recruited: Group A included patients on cDMARD therapy (either methotrexate alone or a cDMARD combination), escalating to a different cDMARD combination; and Group B included patients on a cDMARD combination who were escalating to bDMARDs including an anti-TNF agent. Patients were required to have a DAS28CRP ≥ 3.2 at the first clinical visit. Tender joint counts (28 and 68 joints) and swollen joint counts (28 and 66 joints) were performed by trained research nurses and patients filled out questionnaires to assess function. Serum was obtained for inflammatory markers and rheumatoid serology. Written informed consent was obtained from all patients. The study was approved by the Health and Disability Ethics Committees, New Zealand.

MRI scans and scoring

The dominant wrist was scanned using a 3 T scanner (Philips MR Systems Achieve 3 T; Koninklijke Philips Electronics NV, Eindhoven, the Netherlands). An eight-element Philips SENSE 3.0 T Wrist Coil 8 Channel (Invivo Corp, Gainsville, FL, USA) (receive only) was used. The dominant hand in the wrist coil fitted snugly by the patient’ s side, palm to body, thumb anteriorly. The field of view included the carpus, distal radioulnar joint (dRUJ) and metacarpal (MC) bases. The following were used: turbo spin echo sequences, T1 weighted (T1w) and T2 weighted (T2w) with fat saturation (FS) using spectral adiabatic inversion recovery (SPAIR) in axial and coronal planes and proton density (PD) coronals (without FS). T1wFS axial and coronal sequences were obtained post intravenous gadolinium diethylenetriamine pentaacetic acid (GdDTPA) at standard dosage (Omniscan (Gadodiamide; 5.0 mmol/10 ml or 2.87 g/10 ml); GE Healthcare, Inc., Princeton, NJ, USA). Synovitis, osteitis and erosions were scored according to RAMRIS [8] and tenosynovitis in the flexor and extensor compartments [9]. MRI scan pairs (baseline, 4 months) were scored in known chronological order by one experienced reader (AD), blinded to clinical and X-ray data. Ten scans were selected randomly and rescored 6 months later. The intra-reader reliability was high: ICC (av) = 0.89 (99% CI 0.56–0.97).


Plain radiographs of the hands and feet were obtained for all patients on enrolment. They were scored for damage using the Sharp van der Heijde (SvdH) score [18] by a rheumatologist (ND), blinded to clinical and MRI data.

Statistical analysis

Intra-reader reliability for MRI scores was determined using intraclass correlation coefficients (ICCs). Spearman’s correlations were used to examine the association between ΔDAS28CRP and ΔMRI scores. Clinical characteristics for Groups A and B were compared using the chi-square test and median two-sample test. The effect of ΔDAS28CRP and treatment group on ΔMRI(i) were analysed using multiple linear regression adjusting for age, gender, duration, anti-CCP titre and MRI erosion score at baseline. Four patients transited from Group A into Group B, providing eight scan pairs. Apart from analyses investigating baseline patient characteristics where the four patients were included only in their initial Group A, all analyses were performed including these patients in both groups and were conducted using SAS for Windows version 9.4 (SAS Institute Inc., Cary, NC, USA).


A total of eighty patients were enrolled, fifty-seven in Group A (42 female, 15 male) and 23 patients in Group B (5 female, 18 male) [5, 18]. Dropout throughout the protocol is summarized in Fig. 1. One patient began adalimumab for disease control 1 month after enrolment and remained on this bDMARD for 3 months before the second MRI scan and so was switched to Group B. Full clinical and contrast-enhanced MRI data from Visit 1 (baseline) and Visit 2 (4 months) were available for 62 patients. However, 66 scan pairs were available for analysis (including eight pairs from four patients who were sequentially enrolled into Group A then Group B).

Fig. 1
figure 1

Clinical and MRI data gathered as patients progressed through the study. IV intravenous, MRI magnetic resonance imaging, TNF tumour necrosis factor, Tx therapy

Clinical characteristics

Patients in Group A and Group B were of similar age and gender but those in Group B had a longer disease duration (Table 1). Consistent with their different places on the T2T escalation ladder, swollen and tender joint counts and disease duration (Visit 1) were lower in Group A than Group B (Table 2). The DAS28CRP was also lower for Group A than Group B: 4.24 vs 5.15 (p = 0.007). Interestingly, there was no difference between groups for MRI synovitis or tenosynovitis scores at Visit 1 but total MRI inflammation and osteitis scores were lower in Group A than Group B (p = 0.05 and 0.02 respectively) (Table 2). MRI erosion scores were also lower in Group A than Group B: 4 (0–50) and 9 (1–47) respectively (p = 0.03). The total SvdH X-ray damage score (Visit 1) was also lower in Group A than Group B: 20.5 (0–217) and 63 (0–209) respectively (p = 0.03).

Table 1 Patient characteristics on enrolment
Table 2 Clinical and imaging scores at Visit 1 and Visit 2 (4 months later)

Changes in clinical disease activity and MRI inflammation on escalation

Patients’ responses were measured by ΔDAS28CRP. Both groups improved from Visit 1 to Visit 2 so that ΔDAS28CRP for Group A was −0.94 (−3.30, 1.61) and for Group B was −1.53 (−3.59, 0.56). There was no difference between groups (p = 0.45). For Groups A and B, 83.3% and 87.5% respectively had a fall in the DAS28CRP from Visit 1 to Visit 2 indicating clinical improvement, while in 16.7% and 12.5% there was worsening. ΔMRI(i) from Visit 1 to Visit 2 varied between patients but median scores showed little change over 4 months, despite escalation of therapy, and there was no difference between groups. For Group A, ΔMRI(i) median (range) was 0 (−25, 10); for Group B, −1 (−15, 28) (p = 0.12). Table 3 shows how patients responded in terms of MRI inflammation (total and sub-scores) and MRI erosion score. Only a few showed MRI erosion progression (three patients (6.98%) in Group A and five patients (20.83%) in Group B). The majority had no change in MRI erosion scores (93.0% and 79.17% respectively).

Table 3 Direction of change for MRI scores from Visit 1 to Visit 2

Combining groups, there was a weak positive correlation between ΔMRI(i) and ΔDAS28CRP (Spearman’s 0.36, p = 0.003). Only ΔMRI tenosynovitis correlated with ΔDAS28CRP (Spearman’s 0.33, p = 0.007); MRI sub-scores for Δosteitis, Δsynovitis and Δerosions did not (Table 4). Multiple linear regression analysis was used to assess the effect of group and ΔDAS28CRP on ΔMRI(i) adjusting for age, gender, duration, anti-CCP titre and MRI erosion score at Visit 1. ΔDAS28CRP was associated with ΔMRI(i) so that for every unit increase in ΔDAS28CRP there was a 1.77-U increase in ΔMRI(i) (p = 0.056). Findings are summarized in Fig. 2.

Table 4 Correlations between ΔDAS28CRP and ΔMRI scores
Fig. 2
figure 2

Change in MRI inflammation score (over 4 months) plotted against change in DAS28CRP for each individual patient (labelled A or B according to group). Arrow points to an outlier in Group B (see text). DAS28CRP disease activity score (28 joints) incorporating C-reactive protein, MRI magnetic resonance imaging

Change in MRI score did not differ between groups

There was no difference between groups for ΔMRI(i) although the point estimate of the fall in score was numerically greater for Group A (−3.68 (95% CI −6.16 to − 1.20)) than Group B (−0.84 (95% CI −4.27 to 2.60), p = 0.18). We did note an outlier in Group B in whom there was a significant fall in the DAS28CRP (from 5.15 (V1) to 3.31 (V2)) despite a marked rise in MRI inflammation scores (from 25 to 53). Data for this patient were excluded in a sensitivity analysis but the difference between groups remained non-significant (p = 0.27).

MRI erosion score at V1 was associated with change in MRI(i)

The MRI erosion score (V1) was negatively associated with ΔMRI(i) (parameter estimate −0.28; p = 0.0052), so those with higher MRI erosion scores at V1 had a greater fall in MRI(i) from V1 to V2. For every 1-U increase in the V1 erosion score, the MRI(i) score would be estimated to fall from V1 to V2 by 0.28 U. The X-ray erosion score (V1) was not associated with ΔMRI(i) (p = 0.79).


This is the first real-world study to investigate the link between changes in clinical disease activity and imaging (MRI) joint inflammation, in a cohort of RA patients undergoing escalation of therapy according to T2T principles. In this setting, the clinical target is most often a low disease activity state or complete remission (DAS28CRP ≤ 1.9). It has been suggested that more stringent imaging targets should be used so that patients are treated aggressively, aiming to reach a state where there is no imaging evidence of inflammation at all [19]. MRI scanning can inform about inflammation affecting different tissues including synovium (synovitis), bone (osteitis) and tendon sheaths/tendons (tenosynovitis). We found a significant but relatively weak association between change in clinical disease activity as measured by the DAS28CRP over the 4-month period after escalation of therapy, and change in the total MRI inflammation score. The only MRI sub-score separately associated with ΔDAS28CRP was the ΔMRI tendonitis score, although there was a trend towards an association with the ΔMRI synovitis score. These findings are consistent with observations by Krabben et al. [11], who found in their analysis of 1790 joints in 179 patients that synovitis and tenosynovitis on MRI were independently associated with clinical swelling. In that study, MRI inflammation was most often observed at the wrist, validating our choice of site for MRI scanning.

Although we did observe an association between change in clinical disease activity and change in MRI inflammation scores, it was obvious from inspection of the data that many individual patients improved clinically over the 4 months after treatment escalation, with only minor changes in their MRI inflammation scores. In some instances MRI inflammation worsened despite clinical improvement. Could this be due to the much greater sensitivity of MRI for detecting joint inflammation? Clinicians are familiar with high-grade clinical joint inflammation manifesting as Virchow’s triad of redness, swelling and tenderness, while lesser degrees may only be detectable as pain on movement. MRI can detect inflammation below the clinical threshold by revealing thickened synovium and tenosynovium, which often enhance post contrast, reflecting increased vascularity. However, if MRI inflammation simply represents the low end of the inflammatory spectrum, one might expect a closer correlation between clinical and imaging observations than we have found. A ceiling effect for MRI inflammation scores could be part of the explanation. Thus, a severely inflamed wrist joint might achieve a very high MRI inflammation score, which remains high even when joint swelling and tenderness have subsided. A period of 4 months might also have been insufficient time for MRI(i) to fall in some patients [20]. Another explanation for the relative disparity between changes in clinical and MRI scores is that they capture slightly different information. Clinical disease activity scores incorporate (through joint tenderness) an assessment of pain and this is missing from the MRI inflammation score. The MRI inflammation score incorporates osteitis, which is not directly detectable clinically. Additionally, low-grade synovitis may occur in normal individuals, especially older people [21], and this could contribute to background “noise” within the MRI inflammation score. In this study we have used the DAS28CRP as a “gold standard” against which changes in MRI inflammation were compared, but this measure has its own shortcomings and in some studies the physician global assessment is a better predictor of clinical decision-making [22].

Another possible explanation for the relatively poor correlation between change in MRI inflammation scores and change in the DAS28CRP might be that MRI scans in this study imaged one region only, the dominant wrist, while the DAS28CRP reflected disease activity at 28 joints. We were able to retrieve joint count data and found evidence for tenderness and/or swelling at the scanned wrist in 71.2% of patients at Visit 1 and 47% at Visit 2. Thus, the wrist was quite frequently involved. How well do MRI scans of the wrist reflect, or in other words act as a surrogate marker for, overall rheumatoid disease activity? The evidence for this is indirect. In an early study by McQueen et al. [12], MRI wrist bone oedema (osteitis) scores were predictive of radiographic erosion progression at the hands and feet 6 years later. In 2014, Baker et al. [20] reported a very similar result for MRI osteitis and synovitis from one hand (wrist and MCPs), which were again highly predictive of radiographic progression in the joints of the hands and feet 1 and 2 years later. In 2017, Glinatsi et al. [23] published a post-hoc analysis of clinical trial data from the CIMESTRA/OPERA cohorts and concluded that MRI-assessed inflammation at the wrist (rather than at the MCP joints) was the most important parameter reflecting physical function, pain and global assessment of disease activity in early RA. Thus, although MRI scanning of one joint area may potentially suffer from “sampling error”, there is evidence that the wrist is a particularly informative region.

One of the aims of the study was to compare MRI parameters between patients escalating from one cDMARD combination to another cDMARD combination (Group A) and patients escalating to anti-TNF therapy/cDMARDs (Group B). Group B patients, who were further up the T2T escalation ladder, differed from Group A patients in that they had higher scores for MRI erosion and osteitis at their first visit. Interestingly, there were no differences between groups in terms of synovitis or tendonitis. This is consistent with the known association between MRI osteitis and erosion progression in RA [12, 13], and emphasizes that bone inflammation (osteitis) may be qualitatively and prognostically different from soft tissue joint inflammation (synovitis and tenosynovitis). Molenaar et al. [24] observed that clinically relevant damage progression can occur in patients in prolonged remission. Thus, MRI may have an important role in the detection of osteitis and prediction of damage progression; a role that cannot be duplicated by clinical assessment alone. Very few of our patients demonstrated MRI erosion progression over 4 months but longer follow-up would be required to determine whether those with clinically normal joints but high levels of osteitis had a worse erosive outcome. We did not find evidence for a greater reduction in MRI inflammation scores in those receiving anti-TNF therapy than those on cDMARDs. Although Groups A and B cannot be directly compared because of differences in disease duration, this supports the observations of O’Dell et al. [25] who showed equivalent clinical improvement in patients treated with triple therapy and those receiving anti-TNF/methotrexate therapy. Our findings also go some way towards validating the NZ PHARMAC escalation protocol that requires patients to have first tried cDMARD therapy before escalating to more costly bDMARD options.

What are the implications of our findings for the treating clinician? T2T regimens using clinical targets have been profoundly effective in limiting the progression of joint damage and disability in RA patients on cDMARD and bDMARD regimens [26]. If the goal posts are shifted to more stringent imaging targets, what is the evidence for better outcomes? The OMERACT group have studied this by looking for evidence of a “safe level” of MRI inflammation below which damage progression did not to occur, using MRI data from several RA clinical trial cohorts [27]. Gandjbakhch et al. identified an “acceptable MRI inflammatory activity state” as a RAMRIS synovitis score below 6. In these individuals, radiographic progression rarely occurred. As yet there is no evidence that clinical or functional outcomes are better in such patients. Our study has indicated considerable disparity between clinical disease activity and MRI joint inflammation, and has emphasized that the goal of complete imaging resolution of inflammation is far off in the majority of real-world patients. Is it appropriate to treat these individuals more aggressively? Issues of feasibility, cost and patient acceptability mean that clinical goals are probably to be preferred over MRI goals at least in the short term. Whether ultrasound can step into the gap remains an open question [28]. Both the ARCTIC [29] and the TaSER [30] studies have recently investigated whether ultrasound imaging can improve patient outcomes if used by clinicians to inform decision-making. Both studies concluded that an ultrasound-driven T2T strategy was not associated with significantly better clinical or imaging outcomes than a DAS28-driven strategy. Further large prospective studies of this type are now needed for MRI scanning.

Our study has a number of limitations. Firstly, the study was observational. It was not possible to stipulate what disease-suppressing therapies could be given and the associations we observed could have been influenced by individual DMARDs. We considered stratifying patients according to therapy but there were many combinations of cDMARDs used and groups would have been too small for useful comparison. However, we did establish that a similar proportion of Group A and Group B patients were taking methotrexate at their first visit. Secondly, although we found no difference between groups in clinical or MRI responses to therapy, this comparison was flawed by insufficient statistical power and also by the fact that the groups differed in terms of disease duration and radiographic erosions at baseline [31]. Despite extending enrolment by a year we did not reach the numbers required to satisfy the original power calculation. Nevertheless, our comparative clinical and MRI data do shed light on what is actually happening in terms of joint inflammation in rheumatoid patients undergoing escalation of therapy in routine rheumatology practice, and should be of help to clinicians working towards the aim of improving quality of life in their patients.


Full clinical and MRI data were available at two time points, before and 4 months after a change of therapy, in 66 out of 80 RA patients undergoing real-world T2T escalation. There was no or minimal change in the median MRI inflammation score over 4 months, despite clinical improvement. There was no difference in the MRI inflammation response between those escalating to a different cDMARD regimen and those escalating to a cDMARD/anti-TNF regimen. For the entire group there was a significant correlation between the change in the DAS28CRP and the change in the MRI inflammation score (Spearman’s 0.36, p = 0.003), but this was relatively weak. These results suggest that MRI outcomes should not be adopted to direct T2T escalation as clinically driven strategies are already successful in improving patient outcomes. A randomized controlled trial comparing MRI-driven and DAS-driven strategies is needed.



3-Tesla magnetic resonance imaging


Anti-cyclic citrullinated peptide antibodies


Biological disease-modifying therapy


Conventional disease-modifying therapy


C-reactive protein


Disease activity score (28 joints) incorporating CRP


Distal radioulnar joint


Fat saturation


Gadolinium diethylenetriamine pentaacetic acid

ICC (av):

Intraclass correlation coefficient (average)




Total MRI inflammation score at the wrist (sum of synovitis, osteitis and tendonitis scores)


Methotrexate combined with anti-tumour necrosis therapy


Outcome Measures in Rheumatology


Proton density


Pharmaceutical Management Agency (New Zealand)


Rheumatoid arthritis


Rheumatoid Arthritis Magnetic Resonance Imaging Score


Statistical Analysis System


Spectral adiabatic inversion recovery


Sharp van der Heijde


T1 weighted


Treat to target


T2 weighted


Visit one


Visit two


Change in the DAS28CRP


Change in the total MRI inflammation score (wrist)


  1. Grigor C, Capell H, Stirling A, McMahon AD, Lock P, Vallance R, et al. Effect of a treatment strategy of tight control for rheumatoid arthritis (the TICORA study): a single-blind randomised controlled trial. Lancet. 2004;364(9430):263–9.

    Article  PubMed  Google Scholar 

  2. Smolen JS, Landewe R, Bijlsma J, Burmester G, Chatzidionysiou K, Dougados M, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update. Annals of the Rheumatic Diseases. 2017;76(6):960–77.

    Article  PubMed  Google Scholar 

  3. Pharmac. Decision relating to the funding of TNF inhibitors (Humira and Enbrel) and gabapentin (Neurontin) 2015 [updated 19 June 201720.6.17]. Available from:

  4. van Riel PLCM, Fransen J. DAS28: a useful instrument to monitor infliximab treatment in patients with rheumatoid arthritis. Arthritis Research and Therapy. 2005;7(5):189–90.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Prevoo ML, Van ’t Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis.[see comment]. Arthritis & Rheumatism. 1995;38(1):44–8.

    Article  CAS  Google Scholar 

  6. McQueen FM, Ostergaard M. Established rheumatoid arthritis - new imaging modalities. Baillieres Best Pract Res Clin Rheumatol. 2007;21(5):841–56.

    Article  Google Scholar 

  7. McQueen FM, McHaffie A, Clarke A, Lee AC, Reeves Q, Curteis B, et al. MRI osteitis predicts cartilage damage at the wrist in RA: a three-year prospective 3T MRI study examining cartilage damage. Arthritis Research and Therapy. 2014;16:R33.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Ostergaard M, Peterfy C, Conaghan P, McQueen F, Bird P, Ejbjerg B, et al. OMERACT Rheumatoid Arthritis Magnetic Resonance Imaging Studies. Core set of MRI acquisitions, joint pathology definitions, and the OMERACT RA-MRI scoring system. J Rheumatol. 2003;30(6):1385–6.

    PubMed  Google Scholar 

  9. Haavardsholm EA, Ostergaard M, Ejbjerg BJ, Kvan NP, Kvien TK. Introduction of a novel magnetic resonance imaging tenosynovitis score for rheumatoid arthritis: reliability in a multireader longitudinal study. Annals of the Rheumatic Diseases. 2007;66(9):1216–20.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Brown AK, Quinn MA, Karim Z, Conaghan PG, Peterfy CG, Hensor E, et al. Presence of significant synovitis in rheumatoid arthritis patients with disease-modifying antirheumatic drug-induced clinical remission: evidence from an imaging study may explain structural progression. Arthritis. Rheumatism. 2006;54(12):3761–73.

    Article  CAS  Google Scholar 

  11. Krabben A, Stomp W, Huizinga TW, van der Heijde D, Bloem JL, Reijnierse M, et al. Concordance between inflammation at physical examination and on MRI in patients with early arthritis. Annals of the Rheumatic Diseases. 2015;74(3):506–12.

    Article  CAS  PubMed  Google Scholar 

  12. McQueen F, Benton N, Perry D, Crabbe J, Robinson E. Bone Edema Scored on Magnetic Resonance Imaging Scans of the Dominant Carpus at Presentation Predicts Radiographic Joint Damage of the Hands and Feet Six Years Later in Patients with Rheumatoid Arthritis. Arthritis Rheum. 2003;48(7):1814–27.

    Article  PubMed  Google Scholar 

  13. Hetland ML, Ejbjerg B, Horslev-Petersen K, Jacobsen S, Vestergaard A, Jurik AG, et al. MRI bone oedema is the strongest predictor of subsequent radiographic progression in early rheumatoid arthritis. Results from a 2-year randomised controlled trial (CIMESTRA). Annals of the Rheumatic Diseases. 2009;68(3):384–90.

    Article  CAS  PubMed  Google Scholar 

  14. Ostergaard M, Emery P, Conaghan PG, Fleischmann R, Hsia EC, Xu W, et al. Significant improvement in synovitis, osteitis, and bone erosion following golimumab and methotrexate combination therapy as compared with methotrexate alone: a magnetic resonance imaging study of 318 methotrexate-naive rheumatoid arthritis patients. Arthritis. Rheumatism. 2011;63(12):3712–22.

    Article  CAS  Google Scholar 

  15. Conaghan PG, Peterfy C, Olech E, Kaine J, Ridley D, Dicarlo J, et al. The effects of tocilizumab on osteitis, synovitis and erosion progression in rheumatoid arthritis: results from the ACT-RAY MRI substudy. Annals of the Rheumatic Diseases. 2014;73(5):810–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Emery P, van der Heijde D, Ostergaard M, Conaghan PG, Genovese MC, Keystone EC, et al. Exploratory analyses of the association of MRI with clinical, laboratory and radiographic findings in patients with rheumatoid arthritis. Annals of the Rheumatic Diseases. 2011;70(12):2126–30.

    Article  PubMed  PubMed Central  Google Scholar 

  17. McQueen FM, Clarke A, McHaffie A, Reeves Q, Williams M, Robinson E, et al. Assessment of cartilage loss at the wrist in rheumatoid arthritis using a new MRI scoring system. Annals of the Rheumatic Diseases. 2010;69:1971–5.

    Article  PubMed  Google Scholar 

  18. van der Heijde DM, van Riel PL, Nuver-Zwart IH, Gribnau FW. vad de Putte LB. Effects of hydroxychloroquine and sulphasalazine on progression of joint damage in rheumatoid arthritis. Lancet. 1989;1(8646):1036–8.

    Article  CAS  PubMed  Google Scholar 

  19. Porter D, Dale J, Sattar N. How low to aim in rheumatoid arthritis? Learning from other disciplines. Annals of the Rheumatic Diseases. 2014;73(3):480–2.

    Article  PubMed  Google Scholar 

  20. Baker JF, Ostergaard M, Emery P, Hsia EC, Lu J, Baker DG, et al. Early MRI measures independently predict 1-year and 2-year radiographic progression in rheumatoid arthritis: secondary analysis from a large clinical trial. Annals of the Rheumatic Diseases. 2014;73(11):1968–74.

    Article  PubMed  Google Scholar 

  21. Nieuwenhuis WP, Mangnus L, van Steenbergen HW, Newsum EC, Huizinga TW, Reijnierse M, et al. Older age is associated with more MRI-detected inflammation in hand and foot joints. Rheumatology (Oxford). 2016;55(12):2212–9.

    Article  PubMed  Google Scholar 

  22. Lindsay K, Ibrahim G, Sokoll K, Tripathi M, Melsom RD, PS H. The composite DAS Score is impractical to use in daily practice: evidence that physicians use the objective component of the DAS in decision making. Journal of clinical rheumatology. 2009;15(5):223–5.

    Article  PubMed  Google Scholar 

  23. Glinatsi D, Baker JF, Hetland ML, Horslev-Petersen K, Ejbjerg BJ, Stengaard-Pedersen K, et al. Magnetic resonance imaging assessed inflammation in the wrist is associated with patient-reported physical impairment, global assessment of disease activity and pain in early rheumatoid arthritis: longitudinal results from two randomised controlled trials. Annals of the Rheumatic Diseases. 2017;76(10):1707–15.

    Article  PubMed  Google Scholar 

  24. Molenaar ET, Voskuyl AE, Dinant HJ, Bezemer PD, Boers M, Dijkmans BA. Progression of radiologic damage in patients with rheumatoid arthritis in clinical remission. Arthritis & Rheumatism. 2004;50(1):36–42.

    Article  Google Scholar 

  25. O’Dell JR, Mikuls TR, Taylor TH, Ahluwalia V, Brophy M, Warren SR, et al. Therapies for active rheumatoid arthritis after methotrexate failure. N Engl J Med. 2013;369(4):307–18.

    Article  PubMed  Google Scholar 

  26. Steunebrink LM, Vonkeman HE, ten Klooster PM, Hoekstra M, van Riel PL, van de Laar MA. Recently diagnosed rheumatoid arthritis patients benefit from a treat-to-target strategy: results from the DREAM registry. Clin Rheumatol. 2016;35(3):609–15.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Gandjbakhch F, Haavardsholm EA, Conaghan PG, Ejbjerg B, Foltz V, Brown AK, et al. Determining a magnetic resonance imaging inflammatory activity acceptable state without subsequent radiographic progression in rheumatoid arthritis: results from a followup MRI study of 254 patients in clinical remission or low disease activity. J Rheumatol. 2014;41(2):398–406.

    Article  PubMed  Google Scholar 

  28. Mandl P, Kurucz R, Niedermayer D, Balint PV, Smolen JS. Contributions of ultrasound beyond clinical data in assessing inflammatory disease activity in rheumatoid arthritis: current insights and future prospects. Rheumatology (Oxford). 2014;53(12):2136–42.

    Article  PubMed  Google Scholar 

  29. Haavardsholm EA, Aga AB, Olsen IC, Lillegraven S, Hammer HB, Uhlig T, et al. Ultrasound in management of rheumatoid arthritis: ARCTIC randomised controlled strategy trial. Bmj. 2016;i4205:354.

    Google Scholar 

  30. Dale J, Stirling A, Zhang R, Purves D, Foley J, Sambrook M, et al. Targeting ultrasound remission in early rheumatoid arthritis: the results of the TaSER study, a randomised clinical trial. Annals of the Rheumatic Diseases. 2016;75(6):1043–50.

    Article  PubMed  Google Scholar 

  31. Hyrich KL, Watson KD, Silman AJ, Symmons DP. British Society for Rheumatology Biologics R. Predictors of response to anti-TNF-alpha therapy among patients with rheumatoid arthritis: results from the British Society for Rheumatology Biologics Register. Rheumatology (Oxford). 2006;45(12):1558–65.

    Article  CAS  PubMed  Google Scholar 

Download references


The authors thank all patients who participated in this study. They thank Mrs Megan Johnston, Ms Niamh Fanning and Clinical Nurse Specialist Jan Ipenburg for their assistance with patient recruitment for the Christchurch centre. The authors also thank the staff of Specialist Radiology and MRI Greenlane (SRG, Auckland), especially Mrs Desiree Mulders, charge radiographer, for their technical expertise with MRI scanning and documentation. They thank the staff of Hagley Radiology, Christchurch, for their technical expertise in performing MRI scans. The authors thank Dr Quentin Reeves (radiologist SRG) for his assistance in developing the MRI scan protocol used in the study.


The study was supported by the Auckland Medical Research Foundation (grant reference 1114001), Arthritis New Zealand (R234) and Lotteries Health NZ (3707583).

Availability of data and materials

The datasets used and/or analysed during the present study are available from the corresponding author on reasonable request


FMM has no disclosures. ND reports personal fees from Abbvie, outside the submitted work. The remaining authors have no disclosures.

Author information

Authors and Affiliations



FMM created and designed the project, supervised collection, analysis and interpretation of the data and wrote the manuscript. PC contributed to the study design, patient enrolment, data collection and analysis plus review of the manuscript. TP contributed to MRI data management and analysis and review of the manuscript. DD supervised patient recruitment and data collection and management. AL advised on initial study design and power calculation, performed statistical analysis and advised on manuscript presentation. ND advised on study design, scored X-ray images and contributed to manuscript submission. LS assisted in patient recruitment, data collection and review and advised on analysis. KL participated in the initial study design, oversaw review by the regulatory authorities and reviewed the manuscript. AD participated in study creation, read all MRI scans and formatted MRI scan data and reviewed the manuscript. Approval of the final version of the manuscript was obtained from all co-authors.

Corresponding author

Correspondence to Fiona M. McQueen.

Ethics declarations

Authors’ information

Not applicable.

Ethics approval and consent to participate

The New Zealand Health and Disability Ethics Committees approved this study design (14/CEN/118/AM01. Written informed consent was obtained from all subjects prior to participation in study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

McQueen, F.M., Chapman, P., Pollock, T. et al. Changes in clinical disease activity are weakly linked to changes in MRI inflammation on treat-to-target escalation of therapy in rheumatoid arthritis. Arthritis Res Ther 19, 241 (2017).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: