Treatment of early rheumatoid arthritis: A randomized magnetic resonance imaging study comparing the effects of methotrexate alone, methotrexate in combination with infliximab, and methotrexate in combination with intravenous pulse methylprednisolone†
To compare the effects of methotrexate (MTX), alone or in combination with intravenous (IV) methylprednisolone (MP) or infliximab, on magnetic resonance imaging (MRI)–detected synovitis, bone edema, and erosive changes in patients with early rheumatoid arthritis (RA).
Forty-four patients with early RA were randomized to receive MTX alone (MTX group), MTX plus IV MP (IV MP group), or MTX plus infliximab (infliximab group), infused on day 0 and weeks 2, 6, 14, 22, 30, 38, and 46. Gadolinium-enhanced MRI scans of the metacarpophalangeal joints, wrists, and metatarsophalangeal joints were performed at baseline, week 18, and week 52.
Scores for MRI-detected synovitis and bone edema improved over time in the 3 groups, with significantly lower synovitis scores in the infliximab group compared with the MTX group and significantly lower bone edema scores in the infliximab group compared with the MTX and the IV MP groups. Scores for MRI-detected erosion significantly increased over time in all groups. There were no differences in erosion scores between the MTX group and the other groups. It is of note that patients treated with IV MP showed more significant progression in MRI-detected erosions compared with patients treated with infliximab. At week 22, response rates according to the American College of Rheumatology 20% improvement criteria (ACR20), the ACR50, and the ACR70 were significantly higher in both the IV MP group and the infliximab group compared with the MTX group. At week 52, remission was achieved in 40% of patients in the MTX group and in 70% of patients in the IV MP and infliximab groups. Health Assessment Questionnaire scores improved significantly over time in all groups, with patients receiving IV MP experiencing significantly more improvement compared with patients treated with MTX alone. No severe side effects were observed, except 1 case of MTX-related pneumonitis.
The combination of MTX and infliximab is superior to MTX alone for reducing MRI-detected signs of synovitis and bone edema in patients with early RA. Progression of MRI-detected erosion was greater in patients treated with MTX plus IV MP compared with that in patients who received MTX plus infliximab.
Over the last decade, tumor necrosis factor (TNF) blockers have been successfully used to treat patients with rheumatoid arthritis (RA) in whom disease remained active despite therapy with methotrexate (MTX), not only by suppressing the symptoms and signs of arthritis but also by slowing structural damage (1). In addition, several studies have demonstrated a benefit of TNF blockade on radiographic progression in early RA (2–4). The place of glucocorticoid therapy in early RA is still a matter of debate. In a pivotal trial performed in patients with early RA, the combination of high-dose oral glucocorticoids, MTX, and sulfasalazine (SSZ) was found to be superior to SSZ alone in improving clinical outcome and the structural damage (5). Although these results suggest a role for intensive therapy including oral glucocorticoids in early RA, the advantages of this approach must be balanced with the potential side effects of glucocorticoids (6).
Inflammatory changes and early erosive damage can now be more readily detected by magnetic resonance imaging (MRI) (7). MRI evidence of synovitis and damage was recently shown to be significantly reduced in patients with early RA who were treated with a combination of MTX and infliximab compared with MTX alone (8). The effects of glucocorticoids and infliximab have never been compared in early RA using MRI as the primary outcome measure. In the current investigator-initiated study, we compared the effects of MTX alone or in combination with intravenous (IV) methylprednisolone (MP) pulse therapy or infliximab on MRI-detected synovitis, bone edema, and erosive changes in patients with early RA.
PATIENTS AND METHODS
Inclusion and exclusion criteria.
Patients were eligible for this study if they met the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) classification criteria for RA (9), had a disease duration of <1 year, had active disease (defined as a swollen joint count [SJC] ≥6 and a tender joint count [TJC] ≥8), and had not previously been treated with MTX. Exclusion criteria included past or current tuberculosis, congestive heart disease, previous treatment with glucocorticoids for >3 months (and during the 4 weeks before inclusion), and previous treatment with >2 disease-modifying antirheumatic drugs (DMARDs), MTX, or IV MP pulses. Patients receiving nonsteroidal antiinflammatory drugs (NSAIDs) could be included provided that they were receiving stable dosages.
Patients fulfilling the inclusion criteria were randomly allocated to the MTX, IV MP, and infliximab groups. All patient received MTX at a dosage ranging from 7.5 mg/week at baseline to 20 mg/week at week 14. Patients randomized to the IV MP and infliximab groups received an IV infusion of MP (1 gm) or infliximab (3 mg/kg) at weeks 0, 2, and 6 and then every 8 weeks until week 46. Treatment with oral NSAIDs remained unchanged throughout the study. Intraarticular steroid therapy was not allowed, nor was the introduction of oral glucocorticoids or other DMARDs. After the first year, IV MP and infliximab infusions were discontinued, and all patients continued treatment with MTX only. In case of a flare, the treating physician was allowed to add another DMARD and, if necessary, to (re)start infliximab.
The trial was conducted at the Université Catholique de Louvain and the Katholieke Universiteit Leuven hospitals. The ethics committees of both hospitals approved the study, and informed consent was obtained from all patients prior to the screening procedure. Evaluation of MRI scores over time was the primary end point of the trial. Secondary end points included clinical and functional assessments.
The following items were evaluated at every visit: the SJC (based on assessment of 66 joints and 28 joints), the TJC (based on assessment of 68 joints and 28 joints), patient's and physician's global assessment of disease activity as scored using a 100-mm visual analog scale (VAS), patient's assessment of pain as scored using a 100-mm VAS, morning stiffness, the Health Assessment Questionnaire (HAQ) score based on a scale of 0–3 (10), and serum titers of C-reactive protein (CRP). From baseline until the last visit, patients were always evaluated by the same physician, who was blinded to treatment allocation. The Disease Activity Score in 28 joints (11) using the CRP level (DAS28-CRP) and response rates according to both the ACR preliminary criteria for improvement in RA scores (12) and the European League Against Rheumatism (EULAR) response criteria (13) were computed at every visit.
MRI was performed at baseline, week 18, and week 52 using an Intera 1.5T superconducting magnet (Philips, Best, The Netherlands). Both hands and both feet were imaged successively within a circular surface coil (“knee coil” and “head coil”). The patients were positioned in a prone position for the hand examinations, with the arms above the head and the hands in a “prayer position” in the center of the coil, and in a supine position for the feet examinations. Transverse images were obtained perpendicular to the wrists, the metacarpophalangeal (MCP) joints, and the metatarsophalangeal (MTP) joints.
After intravenous bolus injection of 0.1 mmole/kg of body weight of gadopentetate meglumine (Dotarem; Guerbet, Paris, France), fat-suppressed transverse T1-weighted spin-echo and fat-suppressed 3-dimensional (3-D) transverse gradient-echo (fast field-echo) images were obtained. The sections were obtained, respectively, through the wrists, the MCP joints (except the MCP joints of the thumbs), and the MTP joints. The first series covered the wrist from the distal radioulnar joint to the distal carpal bones. For fat-suppressed gadolinium-enhanced T1-weighted imaging, 15, 13, and 15 sections were obtained through the wrists, MCP joints, and MTP joints, respectively. For fat-suppressed 3-D transverse gradient-echo gadolinium-enhanced T1-weighted imaging, 60, 60, and 80 sections with a thickness of 0.65 mm were obtained through the wrists, MCP joints, and MTP joints, respectively. The overall examination time was ∼30 minutes.
MR images were reviewed by a radiologist (JM) with >15 years of experience in musculoskeletal MRI, who was blinded to the clinical details and the treatment received. All examinations were reviewed on an EasyVision workstation (Philips), using multiplanar reconstruction software to obtain coronal and sagittal images from the 3-D sequences. Synovitis and bone erosions were evaluated according to the Outcome Measures in Rheumatology Clinical Trials RA MRI scoring system (14). For each patient, the first examination was evaluated separately, and the 2 followup examinations were evaluated by comparison with the first, in order to better evaluate the modifications during treatment.
Synovitis was assessed semiquantitatively on the basis of examination of the fat-suppressed gadolinium-enhanced T1-weighted images. A global score scale of 1–3 was used. A score of 0 indicated absence of significant synovial area enhancement, and scores of 1, 2, and 3 indicated discrete, mild, and severe synovitis, respectively. In the wrists, the score was assigned at 2 sites, the radioulnar and radiocarpal joints, yielding an aggregated score of 0–12 for both wrists. A score was also assigned to each of the 8 MCP joints, yielding an aggregated score of 0–24 for both sides, and to each of the 10 MTP joints, yielding an aggregated score of 0–30 for both sides. The global synovitis score for all examined joints ranged from 0 to 66.
Bone erosions were assessed semiquantitatively by examination of the fat-suppressed 3-D gradient-echo images. Erosions were defined as sharply marginated bone lesions, with juxtaarticular location and high signal intensity, which are visible in at least 2 planes, with a cortical interruption seen in at least 1 plane. The scoring scale ranged from 0 to 10, based on the proportion of eroded bone volume compared with the total assessed bone volume (0 = no erosion, 1 = 1–10% of bone eroded, 2 = 11–20% of bone eroded, etc.). We did not consider the normal vascular channels (thin tubular interruption perpendicular to the bone surface) and synovial bone cysts (intraosseous bone lesions with a very thin communication with the bone surface). Multiplanar examination of the sections allows exclusion of pseudoerosions due to a simple sulcus seen in 1 plane. Bone erosions were evaluated at 12 sites for the wrists (distal portions of the radius and ulnas, scaphoids, lunates, triquetrums, and pisiforms), 8 sites for the MCP joints, and 10 sites for the MTP joints (metacarpal and metatarsal heads). Therefore, the aggregated bone erosion score for wrists, MCP joints, and MTP joints ranged from 0 to 120, from 0 to 80, and from 0 to 100, respectively, and the aggregated bone erosion score for all joints ranged from 0 to 300. Bone edema was evaluated at the same sites. The aggregated bone edema scores for the wrist, MCP joints, and MTP joints ranged from 0 to 36, from 0 to 24, and from 0 to 30, respectively.
In preparation for the final reading of the MR image, the interobserver reproducibility and the intraobserver reproducibility of scoring by 2 readers were evaluated. Sixteen complete MRI examinations were read by 2 observers, a trained radiologist with 2 years of experience in MRI (PD; see Acknowledgments) and the radiologist for the present study (JM). For interobserver reproducibility, the intraclass correlation coefficients (ICCs) were 0.88 (95% confidence interval [95% CI] 0.68–0.96), 0.94 (95% CI 0.83–0.98), and 0.82 (95% CI 0.55–0.94) for global synovitis, erosions, and edema scores, respectively. For intraobserver reproducibility, the ICCs were 0.92 (95% CI 0.78–0.97) and 0.95 (95% CI 0.86–0.98) for global synovitis scores, 0.99 (95% CI 0.96–1) and 0.96 (95% CI 0.88–0.97) for global erosion scores, and 0.91 (95% CI 0.75–0.97) and 0.89 (95% CI 0.71–0.96) for global edema scores from the first and second observers, respectively. Reproducibility was not as good for synovitis scores for the MTP joints and erosion scores for the wrists.
The sample size in this investigator-initiated pilot study was 15 patients per treatment arm, which is comparable with sample sizes in previous studies using MRI as an outcome measure in early RA (8). Changes from baseline to week 18 and from baseline to week 52 for the MRI-derived synovitis, edema, and erosion scores were compared between the 3 groups using the Kruskal-Wallis test, and differences between treatment arms were further explored using the Mann-Whitney U test. In this explorative study, a correction for multiple testing was not performed. Within-group changes over time in the MRI-derived scores for synovitis, edema, and erosion were analyzed using Friedman's test. Clinical data were analyzed using repeated-measures analysis of variance, paired and unpaired t-tests, Mann-Whitney U tests, Wilcoxon's signed rank tests, Friedman's test, Fisher's exact tests, or chi-square tests, as appropriate. Survival curves were derived using the Kaplan-Meier method and were statistically tested by log rank test. We calculated the hazard ratios and their 95% CIs with the univariate Cox proportional hazards model.
Forty-four MTX-naive patients with early RA and active disease were randomly allocated to 3 treatment groups: 14 patients were assigned to receive MTX alone, 15 patients were randomized to receive the combination of MTX and IV MP pulses, and 15 patients were randomized to receive combination therapy with MTX and infliximab. During followup, 2 patients withdrew from the MTX group, 1 patient withdrew from the IV MP group, and 1 patient withdrew from the infliximab group, as shown in Figure 1. As indicated in Table 1, baseline demographic, clinical, biologic, and radiographic characteristics did not differ between groups. Of note, rheumatoid factor (RF) and anti–cyclic citrullinated peptide (anti-CCP) antibodies were more prevalent among patients treated with IV MP. Gadolinium-enhanced MRI scans of the MCP joints, wrists, and MTP joints were performed at baseline, week 18, and week 52.
Table 1. Baseline characteristics of the patients according to treatment group*
MTX (n = 14)
IV MP (n = 15)
Infliximab (n = 15)
Except where indicated otherwise, values are the mean ± SD. Differences between the groups were not statistically significant by unpaired t-tests or chi-square tests, except for a higher rate of rheumatoid factor (RF) positivity in the intravenous methylprednisolone (IV MP) group compared with the methotrexate (MTX) group (P = 0.011) and the infliximab group (P = 0.014) and a higher rate of anti–cyclic citrullinated peptide (anti-CCP) antibody positivity in the IV MP group compared with the MTX group (P = 0.007). CRP = C-reactive protein; DAS28-CRP = Disease Activity Score in 28 joints using the CRP level.
53.8 ± 15.2
50.3 ± 14.2
50.0 ± 9.9
Disease duration, years
0.45 ± 0.29
0.25 ± 0.33
0.36 ± 0.31
RF positivity, %
Anti-CCP positivity, %
Swollen joint count
10.3 ± 5.5
12.4 ± 7.6
12.5 ± 5.4
Tender joint count
11.6 ± 7.5
13.2 ± 9.1
15.9 ± 8.0
Serum CRP, mg/dl
2.5 ± 3.5
4.7 ± 5.1
4.8 ± 5.2
Health Assessment Questionnaire score
1.3 ± 0.6
1.2 ± 0.7
1.5 ± 0.8
5.2 ± 0.8
5.3 ± 1.3
5.3 ± 1.1
Erosive changes detected by conventional radiography, %
MRI results were not available for 2 patients (1 in the MTX group and 1 in the IV MP group) due to technical problems or poor quality of the scans, making scoring impossible. In 3 other patients (2 in the MTX group and 1 in the IV MP group), results of the MRI evaluation at week 52 were not available because they did not complete the study (see Figure 1). In the IV MP group and the infliximab group, the MRI-derived synovitis score improved significantly over time, and patients in all treatment groups experienced a significant improvement over time in the MRI bone edema score and a significant worsening of the MRI erosion score (Figure 2).
The synovitis score decreased significantly more from baseline to week 18 and from baseline to week 52 in the infliximab group compared with the MTX group, and the edema score decreased significantly more from baseline to week 18 in the infliximab group compared with the MTX group and the IV MP group (Figure 2 and Table 2). In patients treated with IV MP, the MRI erosion score tended to increase more from baseline to week 52 in comparison with patients treated with infliximab (Figure 2 and Table 2).
Table 2. Between-group differences in magnetic resonance imaging scores*
P = 0.003, MTX versus IV MP and versus infliximab, by Kruskal-Wallis test; P = 0.000, MTX versus infliximab, by Mann-Whitney U test.
P = 0.039, MTX versus IV MP and versus infliximab, by Kruskal-Wallis test; P = 0.018, MTX versus infliximab, by Mann-Whitney U test.
P = 0.045, MTX versus IV MP and versus infliximab, by Kruskal-Wallis test; P = 0.030, MTX versus infliximab, by Mann-Whitney U test; P = 0.041, IV MP versus infliximab, by Mann-Whitney U test.
P = 0.055, IV MP versus infliximab, by Mann-Whitney U test.
IV MP group
We also calculated the MRI scores without including the feet, as suggested by Ejbjerg et al (15), and the results were comparable. Taking only the hands into account, the MRI-detected improvement from baseline to week 18 was significantly greater in the infliximab group compared with the MTX group for both synovitis scores and edema scores (P = 0.005 and P = 0.05, respectively, by Mann-Whitney U test), and erosion scores progressed more from baseline to week 18 in the IV MP group compared with the infliximab group (P = 0.048 by Mann-Whitney U test).
To determine whether the difference in baseline prognostic characteristics could account for these results, we categorized the total patient population in the different subgroups according to the presence of anti-CCP antibodies and RF. We did not observe any difference in the baseline MRI erosion score or in progression of the MRI erosion score over time between the population that was positive for both RF and anti-CCP antibodies and the population that was negative for both RF and anti-CCP antibodies (data not shown).
Next, we compared the erosion scores over time in the 3 treatment groups, taking into account only the joints without MRI-detected erosions at baseline, in order to appreciate the appearance of new erosive changes in previously undamaged joints. As shown in Figure 3, patients in the IV MP group (compared with those in the MTX group) showed significantly greater progression of the mean erosion score in previously unaffected joints at week 52.
The CRP level declined over time in all groups, without between-group differences. Compared with the median CRP level at baseline, the level at week 52 dropped, from 7 mg/liter (interquartile range [IQR] 3–121) to 2.5 mg/liter (IQR 1–31) in the MTX group (P = 0.019), from 32 mg/liter (IQR 3–213) to 7.5 mg/liter (IQR 1–27) in the IV MP group (P = 0.002), and from 19 mg/liter (IQR 3–164) to 3.5 mg/liter (IQR 1–29) in the infliximab group (P = 0.008 by Wilcoxon's signed rank test). Clinical responses assessed by the ACR 20% improvement criteria (ACR20), the ACR50, and the ACR70 at week 22 were significantly better in the IV MP and infliximab groups compared with the MTX group (Figure 4A). A similar trend was observed at week 52 (Figure 4B) but without statistically significant differences between groups (as determined by Fisher's exact test). HAQ scores improved significantly over time in the IV MP and infliximab groups (P < 0.001 by Friedman's test), with patients receiving IV MP experiencing significantly more improvement compared with patients receiving MTX, from baseline to week 22 and from baseline to week 52 (P = 0.006 and P = 0.019, respectively, by Mann-Whitney U test) (Figure 4C). In addition, the DAS28-CRP improved over time in all groups, without between-group differences. Thus, at week 52, the mean ± SD DAS28-CRP had dropped from 4.85 ± 0.96 to 3.26 ± 1.31 in the MTX group (P = 0.005), from 5.39 ± 1.22 to 2.77 ± 1.09 in the IV MP group (P < 0.0001), and from 5.57 ± 1.03 to 2.79 ± 0.77 in the infliximab group (P < 0.0001 by paired t-test).
The time to EULAR-defined remission was analyzed by Kaplan-Meier survival curves (Figure 5). There was a statistically significant difference between only the IV MP group and the MTX group, in favor of the former. By week 52, remission was achieved in 40% of MTX-treated patients and in 70% of IV MP– and infliximab-treated patients. Clinical efficacy was sustained during the second year of followup for a large majority of patients in our cohort; only 8 of the 44 patients needed combination treatment with infliximab (no differences between groups were observed). Interestingly, these 8 patients had significantly higher baseline MRI-derived erosion scores (data not shown).
Treatment was generally very well tolerated. One patient (who was randomized to the infliximab group) experienced MTX-related pneumonitis at week 30. He completely recovered but was excluded from further participation in the study. No other serious adverse events, in particular no serious infections requiring in-patient IV treatment, were reported. Other side effects were minimal and were similarly distributed between groups (14, 12, and 12 episodes of benign infection in the MTX, IV MP, and infliximab groups, respectively; 5, 3, and 2 cases of mild hepatotoxicity in the MTX, IV MP, and infliximab groups, respectively).
In this independent and exploratory study, we evaluated 2 intensive combination therapies (infliximab plus MTX or IV MP plus MTX) compared with MTX alone in patients with early active RA. We selected MRI as the primary end point measure, in order to study the effects of these 3 treatment regimens on synovitis, erosions, and bone edema scores. Patients who received infliximab in combination with MTX showed more improvement in the MRI synovitis score compared with patients who received only MTX. The potent effect of infliximab on the synovitis score was also confirmed by a rapid decline in the CRP level and the DAS28. These data suggest that rapid control of inflammation was achieved. Such findings are consistent with the results of a similar study of infliximab in early RA by Quinn et al (8), in which MRI was used as an outcome parameter. Of note, in the study by Quinn et al, no comparison between infliximab and steroid treatment was made. In our study, we failed to demonstrate a significant difference between the effect of IV MP and that of infliximab on the evolution of the MRI-derived synovitis score. A recent study confirmed that use of combination therapy with infliximab plus MTX significantly retarded the progression of radiographic changes, especially in patients with persistent disease activity and an elevated acute-phase response (16).
Previous studies have also demonstrated the potential benefit of glucocorticoids on structural damage (17, 18). In our study, we showed that the rate of progression of the MRI erosion score was higher in the IV MP group compared with the infliximab group. This absence of a favorable effect of IV MP pulse therapy on the MRI erosion score in comparison with the other treatment regimens could be interpreted in different ways. Intravenous MP pulse therapy may act differently on bone erosion than a regimen of continuous low-dose oral glucocorticoids. It has been suggested that such a high dose of glucocorticoids may blunt osteoblast function and have an inhibitory effect on the synthesis of factors involved in bone repair, such as alkaline phosphatase (19, 20).
Rather than a real difference between regimens, a bias in patient selection could also account for our results, because there were between-group differences in the distribution of prognostic factors at baseline due to the small size of this MRI study. To check this, we reanalyzed the data after separating the total patient population into subgroups according to the presence or absence of either anti-CCP antibodies or RF. We did not observe any difference in either baseline MRI-detected characteristics or MRI outcome over time between these subgroups. We could not perform this analysis in each treatment group separately due to the small number of patients, but our data suggest that a difference in prognostic status does not seem to have had a significant influence on the evolution of MRI-detected erosions at 1 year.
Although it seems that patients requiring anti-TNF treatment during the second year of followup had higher MRI erosion scores at baseline, it is difficult to evaluate the clinical and radiographic prognostic value of the baseline MRI data from our small study, especially because discrepancies between clinical improvement and MRI findings have been reported (21). Interestingly, in our study, the bone edema score improved most in patients treated with infliximab. It has been shown that bone edema on baseline MRI scans was the most powerful predictor of a significant change in the erosion score over time (22, 23). However, edema is not a specific finding, and the score is difficult to interpret even by expert radiologists.
We confirm that MRI scoring of both the hands and feet does not have much additional value compared with scoring of only the hands, as previously reported (15). If we had taken only the hands into account, we would have missed the finding that the MRI synovitis score also decreased significantly more from baseline to week 52 in the infliximab group compared with the MTX group. However, we showed the advantage of infliximab over IV MP on the progression of MRI erosions from baseline to week 18 only by leaving the feet out of the analysis, and we showed the advantage of infliximab over IV MP on the progression of MRI erosions from baseline to week 52 only by keeping the feet in the analysis, which of course could simply be explained by the small sample size of this study.
By the end of the first year, marked clinical improvement was observed in all groups. Combination therapy with MTX plus IV MP or with MTX plus infliximab resulted in better clinical control of RA as assessed by the ACR and DAS28-CRP response rates. Our data suggest that high doses of IV MP have an effect equivalent to that of TNF blockers at week 22. These findings are consistent with results from the Behandelstrategieën voor Reumatoide Artritis (BeSt) study, which confirm that intensive treatment of early RA is better than treatment with MTX alone (24). In our study, the ACR20, ACR50, and ACR70 response rates in the MTX group at week 22 were relatively low, probably due to the low incremental MTX dosage scheme used (the maximal dosage of 20 mg MTX weekly was reached only at week 22) and the lag period before MTX is effective.
The major clinical outcome for patients with early RA is to achieve disease remission. In our study, we demonstrated that 70% of the patients treated with MTX plus MP or MTX plus infliximab fulfilled the DAS28 criteria for remission after 1 year of treatment. This rate of remission is higher than the 30% rate of remission observed with infliximab in the BeSt study (24). The different methodologies, the small number of patients included in our study, and the different definitions of remission limit the interpretation of these findings. Finally, we demonstrated that the combination of MTX plus MP was more effective in improving physical function compared with the other treatment regimens; this effect might be explained by the potent psychological effect of glucocorticoids.
In a previous study, our group demonstrated that infliximab combined with MTX was clinically more beneficial than IV MP in combination with MTX in severe and refractory RA (25). This benefit was not confirmed in the current study of early RA, suggesting that the responsiveness to glucocorticoids could change over time in patients with chronic RA (26). The 3 regimens were well tolerated, confirming that the balance between the benefits and side effects of intensive therapy in early RA is largely positive.
In conclusion, intensive treatment of early RA is effective and well tolerated. MRI data indicate that infliximab has distinct effects on the progression of erosions that can be distinguished from those of high-dose IV MP. Future work is needed to better understand the significance of erosions detected on MRI and the mechanisms of action of these regimens on the rheumatoid process.
Dr. Durez had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Durez, Westhovens, Luyten, Houssiau, Verschueren.
Schering-Plough was not involved in the study set-up, data collection, or analysis and interpretation of the data, and had no influence on the publishing of the results.
We thank F. Vanheuverbeke and N. Vastesaeger (Schering-Plough Belgium) for the data management, P. Schmitz for technical assistance with the MRI protocol and for facilitating all MR examinations, and P. Draho for participating in a preliminary MRI analysis evaluation study. Special thanks to Steffen Fiews for his advice regarding the statistical analysis.