Article Text

Download PDFPDF

Extended report
Short- and long-term efficacy of intra-articular injections with betamethasone as part of a treat-to-target strategy in early rheumatoid arthritis: impact of joint area, repeated injections, MRI findings, anti-CCP, IgM-RF and CRP
  1. Merete Lund Hetland1,
  2. Mikkel Østergaard1,
  3. Bo Ejbjerg2,
  4. Søren Jacobsen3,
  5. Kristian Stengaard-Pedersen4,
  6. Peter Junker5,
  7. Tine Lottenburger6,
  8. Ib Hansen4,
  9. Lis Smedegaard Andersen7,
  10. Ulrik Tarp4,
  11. Anders Svendsen6,
  12. Jens Kristian Pedersen7,
  13. Henrik Skjødt1,
  14. Torkell Ellingsen4,
  15. Hanne Lindegaard5,
  16. Jan Pødenphant8,
  17. Kim Hørslev-Petersen7,
  18. CIMESTRA study group
  1. 1Department of Rheumatology, Glostrup Hospital, Glostrup, Denmark
  2. 2Department of Rheumatology, Slagelse Hospital, Slagelse, Denmark
  3. 3Department of Rheumatology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark
  4. 4Department of Rheumatology, Aarhus University Hospital, Aarhus, Denmark
  5. 5Department of Rheumatology, Odense University Hospital, Odense, Denmark
  6. 6Department of Rheumatology, Vejle Hospital, Denmark
  7. 7Department of Rheumatology, Rheumatism Hospital, University of Southern Denmark, Gråsten, Denmark
  8. 8Department of Rheumatology, Copenhagen University Hospital, Gentofte, Denmark
  1. Correspondence to Kim Hørslev-Petersen, Rheumatisk Hospital, University of Southern Denmark, Toldbodgade 3, DK-6300 Gråsten, Denmark; horslev{at}


Objective To investigate the short-term and long-term efficacy of intra-articular betamethasone injections, and the impact of joint area, repeated injections, MRI pathology, anticyclic citrullinated peptide (CCP) and immunoglobulin M rheumatoid factor (IgM-RF) status in patients with early rheumatoid arthritis (RA).

Methods During 2 years of follow-up in the CIMESTRA trial, 160 patients received intra-articular betamethasone in up to four swollen joints/visit in combination with disease-modifying antirheumatic drugs. Short-term efficacy was assessed by EULAR good response. Long-term efficacy by Kaplan–Meier plots of the joint injection survival (ie, the time between injection and renewed flare). Potential predictors of joint injection survival were tested.

Results 1373 Unique joints (ankles, elbows, knees, metacarpophalangeal (MCP), metatarsophalangeal, proximal interphalangeal (PIP), shoulders, wrists) were injected during 2 years. 531 Joints received a second injection, and 262 a third. At baseline, the median numbers of injections (dose of betamethasone) was 4 (28 mg), declining to 0 (0 mg) at subsequent visits. At weeks 2, 4 and 6, 50.0%, 58.1% and 61.7% had achieved a EULAR good response. After 1 and 2 years, respectively, 62.3% (95% CI 58.1% to 66.9%) and 55.5% (51.1% to 60.3%) of the joints injected at baseline had not relapsed. All joint areas had good 2-year joint injection survival, longest for the PIP joints: 73.7% (79.4% to 95.3%). 2-Year joint injection survival was higher for first injections: 56.6% (53.7% to 59.8%) than for the second: 43.4% (38.4% to 49.0%) and the third: 31.3% (25.0% to 39.3%). Adverse events were mild and transient. A high MRI synovitis score of MCP joints and anti-CCP-negativity were associated with poorer joint injection survival, whereas IgM-RF and C-reactive protein were not.

Conclusion In early RA, intra-articular injections of betamethasone in small and large peripheral joints resulted in rapid, effective and longlasting inflammatory control. The cumulative dose of betamethasone was low, and the injections were well tolerated.

View Full Text

Statistics from


Modern treatment strategies in patients with rheumatoid arthritis (RA) aim at early, rapid and sustained control of the inflammatory process. Treatment with disease-modifying antirheumatic drugs (DMARDs) is the cornerstone in treatment-naïve, early RA, but the onset of effect has a delay of up to 6–8 weeks. Oral glucocorticoids are used both as bridging treatment initially in the disease course and as an important supplement in periods with disease exacerbation.1 2 They relieve signs and symptoms of RA rapidly and also reduce long-term joint destruction.1 3 4

Theoretically, intra-articular administration has advantage over systemic treatment by ensuring a high concentration of glucocorticoids at the site of inflammation, and it has been used successfully in studies of early RA.3 5,,7 Intra-articular administration may result in a lower cumulative dose of glucocorticosteroid and, consequently, fewer side effects, because it is administered in refractory doses and only when swollen joints are present.

The use of intra-articular injections of glucocorticoids as a central component of a treat-to-target strategy in patients with RA has not been studied systematically. It is not known whether the short-term and long-term efficacies vary between joint areas and with repeated injections. It remains to be investigated whether the effect of intra-articular injections is associated with serological markers and imaging pathology—for example, anticyclic citrullinated peptide (anti-CCP) antibodies, rheumatoid factor (RF) of the immunoglobulin M (IgM) subclass (IgM-RF), C-reactive protein (CRP) or MRI pathology.

The aim of this study was to investigate this topic in patients with early RA. The patients participated in the randomised, double-blind, placebo-controlled CIMESTRA study that aimed at strict inflammatory control by the use of intra-articular glucocorticoid injections in combination with DMARD treatment.7 8


One hundred and sixty patients with active RA of <6 months' duration and two or more swollen joints at baseline participated in the CIMESTRA trial.7 8 They were DMARD-naïve and had not received glucocorticoid treatment in the previous 4 weeks. The patients were randomised to receive either methotrexate (MTX) (7.5–20 mg weekly) plus ciclosporin 2.5 mg/kg/day (combination therapy group) or MTX plus placebo-ciclosporin (monotherapy group).7 8 In both treatment arms, patients were given intra-articular injections of betamethasone (7 mg/ml) in all swollen joints (maximum four joints or 28 mg/visit) at the following time points: weeks 0, 2, 4, 6, 8 and every 4 weeks thereafter up to 2 years. The volumes injected were knee: 2 ml; ankle/elbow/shoulder/wrist: 1 ml; metacarpophalangeal (MCP)/proximal interphalangeal (PIP)/metatarsophalangeal (MTP) joints: 0.5 ml, and larger joints were given priority. Oral glucocorticoids were not allowed.

All patients received calcium and vitamin D supplementation. Patients with a Z score <0 in the femoral neck or lumbar spine received alendronate 10 mg/day.

Outcome measures

At all visits, tender and swollen joint counts of 40 joints were performed by an independent, blinded and trained doctor. Other disease activity variables were also measured (table 1).

Table 1

Short-term efficacy of intra-articular injections with betamethasone

It was assumed that the DMARD treatment was not clinically effective until after week 6. The short-term efficacy of the joint injections was assessed by the European League Against Rheumatism (EULAR) responses from week 0 to weeks 2, 4 and 6.9 The long-term efficacy was assessed by calculating the time between injection of a joint and the same joint becoming swollen again (‘the joint injection survival’). The clinical and radiographical outcomes, the number of injections and the cumulative betamethasone doses were similar in the two treatment arms after 5 years.10 Therefore, for this study, data from the two treatment arms (MTX plus ciclosporin and MTX plus placebo-ciclosporin) were pooled and analysed together.

Potential predictors of joint injection survival

Anti-CCP (IgG antibodies) were determined by a second-generation ELISA (Immunoscan RA-kit, Euro-Diagnostica AB, Malmö, Sweden) with 25 U/ml as cut-off point.11 Serum CRP (mg/l) and erythrocyte sedimentation rate (mm/h) were measured using standard laboratory methods. IgM-RF was detected by ELISA as previously described with a few modifications.12 Cut-off levels were >16 IU/ml (≈95th centile of healthy subjects).

Contrast-enhanced MRI of the non-dominant wrist was performed at baseline in 130 patients, and in 89 of these, the field of view also allowed scanning of the second–fifth MCP joint. MRI sequences comprised coronal and axial T1-weighted images before and after intravenous gadolinium-contrast injection, and a coronal short τ-inversion recovery sequence. For details, see Hetland et al.13 The MR image sets were assessed for bone erosions, synovitis and bone marrow oedema according to the Outcome Measures in Rheumatology RAMRI scoring system (RAMRIS)14 by an independent, trained rheumatologist (BE), who was blinded to all patient data.

Ethical considerations

All patients gave their written informed consent. The protocol was approved by the national health authorities and ethics committees in all five participating counties. The trial was performed in accordance with the Declaration of Helsinki and the International Conference on Harmonisation 1966 revised Guidelines for Good Clinical Practice in the European Community. The trial was registered at (reference NCT00209859).

Statistical analysis

Analysis was by intention-to-treat and included all available data at the selected time points. Unless otherwise specified, the data presented are on the joint injections given at the baseline visit (week 0). Data are presented as the median (IQR). The Cochran–Armitage test for trend (dichotomous variables), and Jonckheere–Terpstra test for trend (continuous variables) were used.

For each joint area (knees/shoulders/elbows/wrists/ankles/MCP/PIP/MTP joints) the average joint injection survival was calculated and evaluated by Kaplan–Meier plots and log-rank statistics. Patients were censored at the time of the last visit or if they withdrew from the study.

Potential predictors of joint injection survival were tested in univariate Cox regression analyses. The covariates were included as continuous variables and the analyses were also performed with them as dichotomous variables (synovitis score as tertiles owing to the distribution of data) as a sensitivity analysis. Variables with p<0.20 in univariate analysis were then tested in a multiple Cox regression analysis. When including adjustment for treatment arm (MTX plus ciclosporin vs MTX plus placebo-ciclosporin) the results were similar (data not shown).

A p value <0.05 was considered significant. The R software package15 was used for the statistical analyses.


Number of joint injections and doses of betamethasone

In the 160 participating patients with early RA, a total of 2166 joint injections were given in 1373 different joints (in number order: wrists>knees>MCP>shoulders>ankles>PIP>elbows>MTP) during 2 years. The median dose of betamethasone at each visit decreased from 28 mg at baseline to 0 mg at the following visits (figure 1A and table 1). The cumulative dose after 2 years was (median (IQR)) 77 (IQR 49–119) mg, corresponding to < 1 mg prednisolone a day, and the cumulative number of injections for each patient was (median (IQR)) 13 (8–19). After the baseline visit, 57 (36%) of patients had more than four swollen joints at one or several visits. A total of 1373 injections were first-time injections, 531 were second-time injections in a previously injected joint and 262 were third-time injections (figure 1C). At the baseline visit, 522 injections were given (see figure 1B for details). At the following visits, the majority of patients received no joint injection (figure 1A and table 1).

Figure 1

(A) The amount of betamethasone (mg) injected at each visit during the 2-year study. Shown are medians, 25 and 75 centiles (box), and 5 and 95 centiles (whiskers). Note that the median is the broad band, decreasing from 28 mg at week 0 to 0 mg at all subsequent visits. The number of patients with available data at each visit is shown at the bottom. (B) The 2-year joint injection survival for injections given at the baseline visit by joint area (MCP, metacarpophalangeal joint; MTP, metatarsophalangeal joint; PIP, proximal interphalangeal joint). Below is shown the number of joints still at risk at the corresponding time points. (C) Kaplan–Meier plot of the joint injection survival of first, second and third time injections. Injections from all visits are included here. Below is shown the number of joints still at risk at the corresponding time points.

Short-term efficacy of joint injections

Two weeks after the initial joint injections, the 28-joint count Disease Activity Score (DAS28) score had decreased from 5.5 to 3.0, with further improvement to 2.6 at 6 weeks, p<0.0001 (table 1). Thus, the median number of swollen and tender joints had decreased sharply to 0 and 3, respectively, and remained low. By weeks 2, 4 and 6, respectively, 50.0%, 58.1% and 61.7% of patients had achieved a good EULAR response, and 39%, 42% and 47% were in DAS28 remission. Health Assessment Questionnaire scores and other patient-reported measures (Visual Analogue Scale for pain and global assessments) decreased rapidly (table 1).

Long-term efficacy of joint injections

After 1 and 2 years, 62.3% (95% CI 58.1% to 66.9%) and 55.5% (51.1% to 60.3%), respectively, of the swollen joints that were injected at the baseline visit, had not relapsed.

figure 1B shows the efficacy of the joint injections given at baseline according to joint area, illustrated as the joint injection survival. After 2 years, no relapse had been seen in 51.2% (34.1% to 76.9%) of elbows, 60.1% (47.3% to 76.3%) of ankles, 54.8% (46.8% to 64.1%) of wrists, 55.3% (44.7% to 68.3%) of knees, 52.3% (41.7% to 65.6%) of MCP joints, 53.6% (35.3% to 81.3%) of MTP-joints, 49.5% (38.6% to 63.5%) of shoulders and 73.7% (79.4% to 95.3%) of PIP-joints (PIP-joints versus other joint areas: p<0.01).

figure 1C shows the joint injection survival of the first, second and third injections. The joint injection survival was highest for first-time injections (1 year: 63.6% (60.9% to 66.4%); 2 years: 56.6% (53.7% to 59.8%) and lowest for second-time (1 year: 48.7% (44.1% to 53.8%); 2 years: 43.4% (38.4% to 49.0%) and third-time (1 year: 32.4% (26.1% to 40.3%); 2 years: 31.3% (25.0% to 39.3%) injections (p<0.0001).

MRI scores

Synovitis scores at baseline were (median (IQR)) 5 (4–7) in the wrist, and 10 (8–14) in the wrist+MCP joints, respectively, indicating high inflammatory activity. Bone marrow oedema scores were 0 (0–1) in wrist and wrist+MCP joints and erosion scores were 1 (0–2).13

Adverse events

During the 2-year study period, no cases of septic arthritis or aseptic necrosis were seen. One patient had a spontaneous spinal fracture. Other adverse events were generally mild, transient and seen only in a minority of patients.10

Predictors of joint injection survival

In univariate Cox regression analyses, the MRI synovitis score of the MCP joints was associated with decreased survival for joint injections in general (HR=1.08 (95% CI 1.017 to 1.153) per synovitis score unit, n=296). When the joint injection survival and the MRI synovitis scores were compared at the individual joint level (eg, the survival of an injection in the left MCP2 was compared with the MRI score of the same joint), HR was 1.14 (0.786–1.654), (n=16) per synovitis score unit, but did not reach statistical significance. The reason might be lack of power due to the small number of joints with corresponding clinical and MRI synovitis changes. For the MRI synovitis score of the wrist, no significant associations were found, HR=1.02 (0.956–1.083), n=437. MRI bone marrow oedema was not associated with joint survival, whereas the MRI erosion score in the wrist (but not in MCP joints) was associated with poorer survival (HR=1.07 (1.003–1.144) per erosion score unit, n=437). Analysis with binary explanatory variables gave similar results, except that the MRI erosion score (wrist) no longer reached statistical significance. Anti-CCP positivity was weakly associated with longer survival (HR=0.805 (0.611–1.061), n=520). CRP and IgM-RF were not predictors of joint injection survival.

Multiple Cox regression analysis with MRI synovitis score of MCP joints and anti-CCP positivity as covariates showed that both variables contributed significantly to the joint injection survival. For MRI synovitis score, HR=1.078 (1.013–1.146) per synovitis score unit, n=294, for anti-CCP positivity, HR=0.661 (0.454–0.963), n=294.


This study adds significant new knowledge to the effect and feasibility of intra-articular glucocorticoid treatment in RA. In patients with early, DMARD-naïve disease, a longlasting effect was seen in both large and small peripheral joints when intra-articular glucocorticoid was co-administered with DMARD in a treat-to-target strategy. The injections had a rapid-onset anti-inflammatory action (<2 weeks), and the cumulative dose was low and well tolerated. Factors associated with a longlasting effect were identified and there were no serious adverse events.

The joint injections rapidly and effectively brought the inflammatory process under control. Two weeks after inclusion in the project, 50% of patients had achieved a good EULAR response, 39% were in DAS28 remission and 50% had no swollen joints. The good response was also reflected in the patient-reported outcomes such as scores for pain, global assessments and Health Assessment Questionnaire. The effect was long lasting, longest in the PIP joints, and of similar duration in the other joints. Thus, small and large joints had similar joint injection survival.

This study does not allow a comparison of oral and intra-articular glucocorticoid treatment. In studies of early RA with initial bridging treatment with oral prednisolone,16 17 the doses have been several-folds higher than in this study.7 8 When intra-articular administration is used, a higher concentration of glucocorticoids is reached at the site of inflammation, and therefore lower doses are needed to control joint inflammation. Glucocorticoids have been identified as disease-modifying drugs that retard radiographic progression in patients with RA.1 3 In this study, radiographic progression after 1, 2 and 5 years was modest (<1 unit a year),10 and the intra-articular injections may have contributed to this, although our design does not allow a definite association.

Only one randomised controlled trial has compared multiple joint injections of glucocorticoids with systemic administration.18 Sixty-nine patients with RA were randomised to receive triamcinolone hexacetonide either intra-articularly (253 swollen joints injected) or intramuscularly (267 swollen joints). Four weeks after unguided, intra-articular injections, 44% of the patients had achieved an American College of Rheumatology 50 response, compared with 20% of the patients who received systemic (intramuscular) treatment with the same dose. The number of side effects was lower in the intra-articularly treated group.18

This study supported the use of unguided intra-articular injections with betamethasone in combination with initiation of DMARD treatment, which had a longlasting effect, especially when a joint was injected for the first time. However, even joints that were injected a second or third time went into longlasting remission. Thus, almost two-thirds of joints injected for the first time, half of joints injected for the second time, and one-third of joints injected for the third time, had not relapsed 1 year after injection. The long-term efficacy of intra-articular betamethasone is difficult to assess owing to the concomitant DMARD treatment. However, by comparison with previous early RA reports4 16 17 that used oral prednisolone according to protocol, the on-demand use of intra-articular injections resulted in a lower cumulative dose of glucocorticoids over time, corresponding to <1 mg prednisolone/day.10

To our knowledge, there has been no attempt in the past to identify subsets of patients who would benefit in particular from intra-articular glucocorticoid injections. A high MRI synovitis score in the MCP joints was associated with poorer joint injection survival in general, probably because high inflammatory activity is harder to control. The association seemed to be stronger when analysing data at the individual joint level, although the study lacked power to reach statistical significance. The effect was better in anti-CCP-positive patients than in anti-CCP-negative patients. We have previously shown that joint pathology is differently expressed and associated with components of the innate immune system and joint matrix dependent on anti-CCP status.19 20 This may reflect the emerging pattern of different pathogenesis and treatment responses in anti-CCP-positive and -negative disease.21 The increased joint injection survival in our anti-CCP positive patients with RA may indicate that the local immunosuppressive effect of joint injections targets specific processes in these patients. To this end, it is of particular interest that antibodies against citrullinated proteins seem locally to enhance the inflammatory process in experimental murine autoimmune arthritis.22 In contrast, serum CRP and IgM-RF were not associated with joint injection survival.

One weakness of this study is that it does not allow for a direct comparison of oral and intra-articular glucocorticoid treatment. Such a design would require intra-articular placebo injections to satisfy the criteria for a double-blind trial, and this was considered unethical.

In conclusion, intra-articular unguided injections of betamethasone in small and large peripheral joints of patients with early RA resulted in rapid, effective and longlasting inflammatory control of the injected joints, and were a well-tolerated treatment in a strategy aiming at remission. The cumulative dose of betamethasone was low, and no adverse events related to the intra-articular route of administration were reported. We suggest that rheumatologists consider including intra-articular injections with glucocorticoids of both small and large joints in the treatment of patients with early RA in combination with aggressive DMARD treatment as part of a treat-to-target strategy.


The CIMESTRA study group investigators: Gråsten Rheumatism Hospital: Jensen SH, Lorenzen T; Herlev Hospital: Bendtsen H, Faarvang KL, Hansen MS, Hansen TM, Nielsen H; Hvidovre Hospital: Jacobsen S, Lauridsen UB, Majgaard O; Odense Hospital: Beier J, Ejstrup L, Knudsen JB, Lasutrup H. Statisticians: Krogh NS, Zitelab ApS. Study nurses: Århus Hospital: Gerdes L; Gråsten Rheumatism Hospital: Frederisken J, Rasmussen P, Theilgård K; Herlev Hospital: Bukh G; Hvidovre Hospital: Pedersen-Zbinden B; Odense Hospital: Holm H, Lorentzen KB.


View Abstract


  • Funding The study was supported by a grant from the Danish Rheumatism Association, Novartis Healthcare Denmark A/S provided ciclosporin and placebo-ciclosporin and sponsored an independent good clinical practice monitor. Nycomed provided methotrexate, folic acid and calcium/vitamin D. Schering-Plough provided injectable betamethasone. Merck Sharp & Dohme provided alendronate. The sponsors were not involved in the study set-up, data collection, analysis or interpretation, and had no influence on the publishing of data.

  • Competing interests None.

  • Ethics approval Ethics committees in all five participating counties.

  • 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.