Centocor had a role in the data analysis and the data interpretation, but not in the data collection or study design. Centocor supplied the study drug.
Comparison of ultrasonographic assessment of synovitis and joint vascularity with radiographic evaluation in a randomized, placebo-controlled study of infliximab therapy in early rheumatoid arthritis†
Article first published online: 5 APR 2004
Copyright © 2004 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 50, Issue 4, pages 1107–1116, April 2004
How to Cite
Taylor, P. C., Steuer, A., Gruber, J., Cosgrove, D. O., Blomley, M. J. K., Marsters, P. A., Wagner, C. L., McClinton, C. and Maini, R. N. (2004), Comparison of ultrasonographic assessment of synovitis and joint vascularity with radiographic evaluation in a randomized, placebo-controlled study of infliximab therapy in early rheumatoid arthritis. Arthritis & Rheumatism, 50: 1107–1116. doi: 10.1002/art.20123
- Issue published online: 5 APR 2004
- Article first published online: 5 APR 2004
- Manuscript Accepted: 11 DEC 2003
- Manuscript Received: 23 MAY 2003
- Arthritis Research Campaign
To investigate sensitive ultrasonographic imaging methods for detection of synovial thickness and vascularity to discriminate between patients with early rheumatoid arthritis (RA) receiving infliximab + methotrexate (MTX) versus placebo + MTX over 18 weeks, and to compare the relationship between synovial thickening and vascularity at baseline and radiologic damage to joints of the hands and feet at 54 weeks.
Patients with early RA (duration <3 years) receiving stable dosages of MTX were randomly assigned to receive blinded infusions of 5 mg/kg infliximab (n = 12) or placebo (n = 12) at weeks 0, 2, 6, and then every 8 weeks until week 46. At baseline and week 18, clinical assessments were performed, and metacarpophalangeal joints were assessed by high-frequency ultrasonography and power Doppler ultrasonography measurements. Radiographs of the hands and feet taken at baseline and at 54 weeks were evaluated using the van der Heijde modification of the Sharp method (vdH-Sharp score).
Using changes in the total vdH-Sharp score over 54 weeks and changes in synovial thickening and joint vascularity at 18 weeks, we were able to distinguish those patients receiving infusions of infliximab + MTX from those receiving placebo + MTX. Sonographic measurements of synovial thickening and vascularity at baseline in the placebo + MTX group demonstrated clear relationships with the magnitude of radiologic joint damage at week 54. Infliximab + MTX treatment abolished these relationships.
The delay or reversal of inflammatory and joint-destructive mechanisms in patients with early RA was already apparent following 18 weeks of treatment with infliximab + MTX and was reflected in radiologic changes at 54 weeks.
Recent advances in the therapy of rheumatoid arthritis (RA) include the introduction of biologic agents that neutralize the proinflammatory cytokine tumor necrosis factor α (TNFα). Treatment with TNFα inhibitors in the established phase of RA produces marked improvement in signs and symptoms of inflammatory joint activity in ∼60–70% of patients (1). Prior to the advent of biologic agents targeting TNFα, it was expected that up to 90% of RA patients with aggressive synovitis would have radiographic evidence of bone erosion within 2 years of diagnosis, despite pharmacologic intervention that included the use of available disease-modifying antirheumatic drugs (DMARDs) (2). More recently, in a cohort of patients (Anti–Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy [ATTRACT]) who had been treated unsuccessfully with multiple DMARDs and had active disease despite treatment with methotrexate (MTX) at a median dosage of 15 mg/week, significant arrest of structural joint damage, assessed radiographically, was observed after 54 weeks of treatment with combined infliximab and MTX as compared with treatment with placebo + MTX (3). This outcome was apparent both in clinical responders and in those failing to achieve this response level (response was defined as achieving the American College of Rheumatology [ACR, formerly, the American Rheumatism Association] composite criteria for 20% response to therapy at 54 weeks) (4).
Therefore, it might be postulated that infliximab + MTX treatment should be initiated early in the evolution of RA, before joint destruction becomes irreversible. In a cohort of patients treated with either MTX or etanercept in the early phase of RA, the rate of radiologic progression of bony erosions was significantly less in the group treated with etanercept after 6 months of therapy (5). Both of these studies (3, 5) used plain radiography to assess joint space narrowing and erosions, and assessed end points that represented relatively late measures of antecedent disease activity with resulting cartilage and bone destruction.
Recent studies addressing the use of ultrasonographic imaging technologies in the evaluation of rheumatoid synovial inflammation and joint damage indicated that clinical joint examination and plain radiography are relatively insensitive tools (6–8), and thus may be inadequate measures for use in clinical trials designed to assess reduction in signs and symptoms of RA as well as retardation of structural damage to joints. These observations led to the present study. In this placebo-controlled, double-blind, randomized study of infliximab therapy added to preexisting MTX treatment in patients in the early phase of RA, we used high-frequency ultrasonography and power Doppler (PD) technology to evaluate changes in synovial thickness and vascularity within the first 18 weeks of therapy and compared these findings with the rate of radiologic damage to joints of the hands and feet over 54 weeks. Because the PD signal is related to the fractional blood volume at a given depth, we reasoned that quantitative PD might be a useful method for monitoring responses to therapy. The results presented here demonstrate unequivocally the value of ultrasonographic technology as a powerful tool in the evaluation of inflammatory joint disease.
This is the first placebo-controlled study to confirm the clinical and structural benefits of infliximab + MTX therapy in patients with early RA. Furthermore, this is the first time that ultrasonographic measures of synovial thickness and vascularity have been successfully used to assess outcome in the context of a placebo-controlled, double-blind, randomized study.
PATIENTS AND METHODS
Patients were eligible to participate in this study if they had a diagnosis of RA according to the 1987 ACR criteria (9), symptoms for 6 months to 3 years, a minimum of 2 swollen metacarpophalangeal (MCP) joints despite treatment with MTX, and seropositivity for IgM rheumatoid factor (as measured by agglutination assay). In addition, eligible patients were required to have either (a) erosion of at least 1 MCP joint as demonstrated on plain radiography or as a cortical break with irregular margins (or contour) on gray-scale ultrasound in both the longitudinal and transverse scanning planes or (b) erosions of at least 2 MCP joints defined by cortical breaks with irregular margins/contour on gray-scale ultrasound in either the transverse or the longitudinal plane associated with a strong vascular signal in PD mode at the site of the cortical break. Cortical breaks without irregular margins were not defined as erosions because they are often artifactual (7).
Patients must also have been receiving oral MTX for a minimum of 8 weeks at a minimum stable dosage of 12.5 mg/week, but not exceeding 17.5 mg/week for at least 4 weeks prior to screening, and must have been receiving a stable dosage of folic acid. Patients taking oral corticosteroids must have been receiving a stable dosage not exceeding 10 mg prednisolone per day for at least 4 weeks before screening. Patients not taking corticosteroids must not have received corticosteroids for at least 4 weeks prior to screening. The screening laboratory test results must have met the following criteria: hemoglobin value ≥8.5 gm/dl, total white blood cell count ≥3.5 × 109/liter, neutrophil count ≥1.5 × 109/liter, platelet count ≥100 × 109/liter, serum transaminase value not exceeding twice the upper limit of normal, and serum creatinine value not exceeding 150 μmoles/liter.
The protocol was approved by the Riverside Research Ethics Committee, and, after patients gave their informed consent, enrollment took place at Charing Cross Hospital, London.
Twenty-four consenting patients were randomized within 14 days of their screening visit. A pharmacist who did not participate in evaluating patient response randomly assigned patients to 1 of 2 treatment groups. Half of the patients received infliximab at 5 mg/kg, and the other half received placebo infusions (normal saline) at weeks 0, 2, and 6, and then every 8 weeks until week 46. All infusions were administered over 2 hours. All physicians, patients, nurses, and other nonclinical members of the study team were blinded with respect to the nature of the study agent being administered. Baseline dosages of MTX or corticosteroid were maintained throughout the first 18 weeks of the study. After week 18, if any patient failed to achieve a 50% reduction from baseline in the number of swollen hand and wrist joints, the weekly dosage of MTX was increased by 2.5 mg once every 4 weeks until a 50% reduction from baseline was achieved, until the dosage of oral MTX reached 25 mg/week, or until the escalation in dosage was limited by toxicity. Thereafter, irrespective of response status, patients continued with the maximum tolerated MTX dosage up to 54 weeks. Clinical monitoring of the study was conducted independently (Centocor, Malvern, PA).
Anteroposterior radiographs of the hands and feet of all patients were taken at baseline and at week 54. Radiographs were scored as pairs in chronologic order (10) for both erosions and joint space narrowing according to the van der Heijde modification of the Sharp method (vdH-Sharp score) (11) by 2 independent observers (PCT, JG) who were unaware of the identity, treatment, and clinical status of the patients. The findings are reported both as the mean ± SD and the median with interquartile range (IQR) (12).
High-frequency ultrasound and PD imaging.
At baseline and at week 18, all patients underwent ultrasound assessment of all 10 MCP joints, which were scanned over the dorsal surface in the transverse and longitudinal planes. The same sonographer (AS) scanned the joints of each patient at both visits to ensure consistency, and was unaware of the patient's study group assignment. With a view to standardization of data acquisition, the hand was maintained in a position of rest by a splint. High-frequency ultrasound and PD imaging were performed using a 15L8 transducer (Acuson Sequoia; Siemens Medical Systems, Ultrasound Group, Issaquah, WA) with constant settings in both the B-mode and the power mode. Each MCP joint was scanned by high-frequency ultrasound at 13 MHz, and gray-scale images were stored digitally. Synovium was defined as an anechoic or hypoechoic region over the dorsum of the joint, visible in longitudinal and transverse planes. The images were evaluated for synovial thickness in the transverse plane and assigned a score of 0–5 by a single assessor (AS) who was unaware of the identity of the patients. A total thickness score was calculated as the sum of individual joint scores.
All MCP joints were also scanned in the PD mode (14 MHz), and the images demonstrating maximal synovial vascularity were stored for subsequent analysis using a computerized image analysis system (CQ; Kinetic Imaging, Nottingham, UK). The number of color Doppler pixels was determined in a defined region of interest for each joint, and a total vascularity score was calculated as the sum of the individual joint scores. The data were independently analyzed.
Scanning was performed in one room in which temperatures were maintained at a constant level all year (∼20°C), with a delay of at least 10 minutes if patients arrived from outside. This procedure was followed to avoid confounding effects from differences in ambient temperatures. Care was taken when scanning to avoid undue pressure with the probe in case this altered blood flow in the joint.
An independent assessor, blinded to the patient's treatment assignment or clinical response indices, performed clinical assessments at baseline, at each subsequent visit for study agent infusion, and at week 18. The clinical assessments comprised the number of tender joints, the number of swollen joints, early morning stiffness (minutes), visual analog scale for pain scores (0–10), the patient's global assessment of disease activity (0–5), and the physician's global assessment of disease activity (0–5). These clinical response parameters were chosen to allow the data to be analyzed using the ACR composite criteria for 20%, 50%, and 70% response to therapy (4) and the European League Against Rheumatism (EULAR) Disease Activity Score in 28 joints (DAS28) (13). At the same time points, blood samples were collected for assessment of complete blood count, erythrocyte sedimentation rate (ESR), and liver function.
The sample size for the present study was chosen on the assumption that high-frequency ultrasound measures of synovial thickening and PD measures of synovial vascular signal would be more sensitive to change after 18 weeks of treatment than the change in plain radiographs of the hands and feet after 54 weeks as assessed by the total vdH-Sharp score. Based on the results of change in the total vdH-Sharp score after 46 weeks of treatment in the ATTRACT trial (3), the power to detect differences of magnitude observed in ATTRACT in groups of 12 patients would be ∼38% for a comparison of MTX alone versus MTX + 3 mg/kg infliximab every 8 weeks and ∼58% for a comparison of MTX alone versus MTX + 10 mg/kg infliximab every 8 weeks.
We examined the effect of the study agent on continuous end points by evaluating the difference between treatment groups using the Mann-Whitney U test and investigating relationships between end points using Spearman's rank correlation. For categorical data, Fisher's exact test was used to compare numbers of patients. Values are expressed as the mean ± SD. All statistical tests were 2-sided. P values less than 0.05 were considered significant.
Characteristics of randomized population.
The baseline characteristics of the 2 patient groups were generally well matched (Table 1). In the infliximab + MTX group, 10 of 12 patients were women and in the placebo + MTX group, 8 of 12 patients were women (P = 0.640). There was a moderate level of disease activity at baseline (mean DAS28 scores of 5.2 in the placebo + MTX group and 5.4 in the infliximab + MTX group) despite treatment with MTX at a mean weekly dosage of ∼15 mg for a mean duration of 0.91 years in both groups (Table 1). The 2 groups were well matched for ESR, but there were differences in baseline C-reactive protein (CRP) levels between the treatment groups (Table 1), although these differences did not reach statistical significance (P = 0.069).
|Parameter, treatment group||Mean ± SD||Median||IQR||P†|
|Placebo + MTX||8.8 ± 2.7||8.5||7.5, 11.0||0.581|
|Infliximab + MTX||9.5 ± 3.8||9.5||7.5, 11.5|
|Placebo + MTX||13.8 ± 9.2||14.0||7.0, 22.0||0.224|
|Infliximab + MTX||19.1 ± 9.1||23.5||14.0, 24.5|
|Placebo + MTX||35.7 ± 21.2||35.0||16.0, 51.0||0.340|
|Infliximab + MTX||27.4 ± 19.5||21.0||16.0, 34.0|
|Placebo + MTX||25.0 ± 25.4||16.0||4.0, 47.0||0.069|
|Infliximab + MTX||12.7 ± 20.7||4.0||1.0, 13.0|
|Placebo + MTX||5.2 ± 1.1||5.4||4.5, 6.1||0.665|
|Infliximab + MTX||5.4 ± 1.1||5.7||4.9, 6.2|
|Placebo + MTX||51.4 ± 14.0||51.7||45.2, 59.9||0.751|
|Infliximab + MTX||55.2 ± 11.8||57.5||44.6, 66.0|
|MTX dosage, mg/week|
|Placebo + MTX||14.8 ± 2.6||12.5||12.5, 17.5||0.793|
|Infliximab + MTX||15.0 ± 1.8||15.0||13.8, 16.3|
|MTX duration, years|
|Placebo + MTX||0.91 ± 0.65||0.84||0.27, 1.58||0.926|
|Infliximab + MTX||0.91 ± 0.38||0.88||0.63, 1.19|
|Placebo + MTX||71.1 ± 13.7||71.0||58.5, 83.5||0.644|
|Infliximab + MTX||67.9 ± 16.1||66.6||54.1, 83.4|
|Disease duration, years|
|Placebo + MTX||1.64 ± 0.63||1.72||1.09, 1.99||0.175|
|Infliximab + MTX||1.33 ± 0.64||1.17||0.75, 1.74|
No patient discontinued participation in the study during the first 18 weeks of therapy. One patient discontinued after week 18 because of lack of efficacy.
The Spearman's rank correlation coefficient for the changes from baseline to week 54 in total vdH-Sharp score noted by the 2 assessors was 0.79 (P < 0.001), indicating good agreement between the assessors. Linear regression analysis showed that ∼63% of the variability in 1 assessor's evaluations of the change from baseline was explained by a linear relationship with the evaluations of the second assessor.
The changes in total vdH-Sharp score from baseline to week 54 between treatment groups noted by both of the assessors showed that patients treated with infliximab + MTX had significantly less progression in joint damage than those treated with placebo + MTX (mean ± SD 2.8 ± 3.5 versus 11.3 ± 10.2 [P = 0.048] for assessor 1, and 3.8 ± 4.6 versus 13.2 ± 10.4 [P = 0.033] for assessor 2).
The data were also analyzed based on the mean total vdH-Sharp scores assigned by the 2 observers. The values obtained by calculating the means of the scores from the 2 assessors corresponded to those obtained for each assessor individually. The mean value therefore has been used for all further analyses. The baseline total vdH-Sharp scores were similar in the 2 groups (Table 2). The annual progression rate for the entire group of 24 patients at baseline was a mean ± SD of 6.5 ± 11.8 and a median of 2.7 (IQR 1.5, 5.6), with no difference between the 2 treatment groups (P = 0.583). There was a greater progression of joint damage from baseline to week 54, as assessed by total vdH-Sharp score, in the placebo + MTX group as compared with the infliximab + MTX group (Table 2 and Figure 1). With respect to the components of the total vdH-Sharp score, there were numerically larger increases in erosion scores in the placebo + MTX group than in the infliximab + MTX group (median 1.5 [IQR 0.0, 5.0] versus median 0.5 [IQR −0.8, 1.8]; P = 0.193). There was also a greater progression in joint space narrowing in the placebo + MTX group than in the infliximab + MTX group (median 6.5 [IQR 2.0, 17.5] versus median 2.8 [IQR 1.6, 3.4]; P = 0.074).
|Parameter, treatment group||Mean ± SD||Median||IQR†||P‡|
|Baseline total vdH-Sharp score§|
|Placebo + MTX||7.1 ± 7.8||4.0||8.5||0.977|
|Infliximab + MTX||9.0 ± 15.9||3.5||6.0|
|Week 54 total vdH-Sharp score§|
|Placebo + MTX||17.7 ± 11.1||20.0||20.0||0.096|
|Infliximab + MTX||12.3 ± 17.4||5.8||7.5|
|Change from baseline to week 54 total vdH-Sharp score§|
|Placebo + MTX||12.2 ± 10.1||14.0||17.0||0.056|
|Infliximab + MTX||3.3 ± 3.6||3.3||3.3|
|Baseline total synovial thickness|
|Placebo + MTX||14.1 ± 5.5||13.5||7.5||0.931|
|Infliximab + MTX||14.4 ± 5.2||12.0||9.0|
|Week 18 total synovial thickness|
|Placebo + MTX||11.8 ± 6.8||9.5||8.5||0.034|
|Infliximab + MTX||6.8 ± 4.7||6.0||5.0|
|Percentage change from baseline to week 18 total synovial thickness|
|Placebo + MTX||−13.7 ± 34.4||1.2||58.9||0.014|
|Infliximab + MTX||−54.5 ± 21.6||−50.0||28.1|
|Baseline total CDA|
|Placebo + MTX||8,212 ± 6,479||6,158 ± 8,642||0.665|
|Infliximab + MTX||9,072 ± 6,718||9,664 ± 8,442|
|Week 18 total CDA|
|Placebo + MTX||5,130 ± 3,865||3,989 ± 5,792||0.005|
|Infliximab + MTX||1,228 ± 2,421||92 ± 1,610|
|Percentage change from baseline to week 18 total CDA|
|Placebo + MTX||−26.6 ± 50.3||−30.7 ± 34.4||0.017|
|Infliximab + MTX||−78.6 ± 39.4||−98.4 ± 16.8|
At baseline, the 2 treatment groups were well matched for total synovial thickening score as assessed by high-frequency ultrasonography (Table 2). At week 18, patients in the infliximab + MTX group achieved a median reduction in synovial thickness of 50% (IQR 71.4% reduction to 43.3% reduction) as compared with a median increase of 1.2% (IQR 46.4% reduction to 12.5% increase) in the placebo + MTX group (P = 0.014) (Table 2 and Figure 1). Examples of good and poor responders with respect to changes in synovial thickness are shown in Figure 2.
At baseline, the mean total color Doppler area (CDA) was 9,072 pixels in the infliximab + MTX group and 8,212 pixels in the placebo + MTX group (Table 2). At 18 weeks, the CDA had diminished by a median of 98.4% (IQR 100% reduction to 83% reduction) in the infliximab + MTX group as compared with a median reduction of 30.7% (IQR 48% reduction to 14% reduction) in the placebo + MTX group (P = 0.017) (Table 2 and Figure 1). Examples of good and poor responders with respect to changes in CDA are given in Figure 3.
MTX dosage escalation.
Between weeks 18 and 54, there was a significantly greater escalation in MTX dosage in the placebo + MTX group than in the infliximab + MTX group (median total increase of 6.3 mg/week [IQR 1.9, 10.6] versus 0 mg/week [IQR 0.0, 0.0]; P = 0.001) (Figure 4). Per the study design, MTX dosages were kept stable unless an increase was needed due to poor clinical response.
Efficacy at 18 and 54 weeks.
As anticipated on the basis of prior studies of patients with advanced disease, the reduction from baseline in DAS28 at 18 weeks was greater in the infliximab + MTX group than in the placebo + MTX group (median −1.21 [IQR −2.3, −0.26] versus median −0.39 [IQR −0.94, −0.21]; P = 0.157). At week 54, the corresponding median reductions from baseline in DAS28 for the infliximab + MTX group and the placebo + MTX groups were −2.3 (IQR −0.6, −2.8) and −0.3 (IQR 0.2, −1.8), respectively. The week 18 and week 54 data were also analyzed according to EULAR criteria for good, moderate, or absent responses to therapy (13) and ACR 20% and 50% response criteria (Table 3). There was a trend, although not significant, for more patients in the infliximab + MTX group than in the placebo + MTX group to meet EULAR criteria for good and moderate responses to therapy at both time points (Table 3). Similarly, more patients in the infliximab + MTX group met both ACR 20% and ACR 50% response criteria (Table 3).
|Placebo + MTX group, no. (%)||Infliximab + MTX group, no. (%)||P†|
|DAS28 at week 18|
|No response||8 (66.7)||4 (33.3)|
|Moderate response||4 (33.3)||7 (58.3)|
|Good response||0 (0.0)||1 (8.3)||0.220|
|ACR response level at week 18|
|≥20%||2 (16.7)||6 (50.0)||0.193|
|≥50%||1 (8.3)||3 (25.0)||0.590|
|DAS28 at week 54|
|No response||6 (54.5)||3 (25.0)|
|Moderate response||3 (27.3)||6 (50.0)|
|Good response||2 (18.2)||3 (25.0)||0.214|
|ACR response level at week 54|
|≥20%||2 (20.0)||6 (50.0)||0.193|
|≥50%||0 (0.0)||3 (25.0)||0.217|
Relationship between baseline ultrasonographic measures, baseline CRP, and radiologic change at 54 weeks.
In the placebo + MTX group, there were striking positive correlations between baseline synovial thickness and progression in total vdH-Sharp score (r = 0.69, P = 0.020), and between baseline synovial vascularity and progression in total vdH-Sharp score (r = 0.78, P = 0.005) (Figure 5). Infliximab treatment abolished these relationships such that there were weakly negative but nonsignificant correlations between baseline synovial thickness and progression in total vdH-Sharp score (r = −0.23, P = 0.479) and between baseline synovial vascularity and progression in total vdH-Sharp score (r = −0.28, P = 0.372) (Figure 5).
In contrast to the ultrasonographic findings, there was a positive but nonsignificant correlation between baseline CRP level and change from baseline to week 54 in the vdH-Sharp score for the placebo + MTX group (r = 0.58, P = 0.077). The corresponding relationship for the infliximab + MTX group was −0.19 (P = 0.562).
In this study of patients in the early phase of RA (mean duration <2 years), the rate of radiologic progression of structural damage over 54 weeks was significantly attenuated in patients treated with infliximab + MTX as compared with that for patients receiving placebo + MTX. This is a remarkable observation in a small cohort, given that this study was not designed to detect such a difference.
A noteworthy feature of this study design was the requirement for patients to have had at least 1 MCP joint erosion in the early phase of RA, as demonstrated by either plain radiography or as a cortical break with irregular margins on gray-scale high-frequency ultrasound in both longitudinal and transverse scanning planes, as defined by Wakefield et al (7). Alternatively, patients were also eligible for study entry if they had erosions of at least 2 MCP joints, defined by cortical breaks with irregular margins on gray-scale ultrasound in either transverse or longitudinal planes if associated with a strong vascular signal in PD mode at the site of the break. This eligibility criterion was based on our own previous observations using combined high-frequency ultrasound and PD to image the MCP joint in a series of 39 patients with early RA (disease duration <3 years). In this earlier cohort, we observed a striking association between the presence of bone erosions on high-frequency ultrasound imaging of MCP joints and increased intraarticular blood flow in PD mode imaging at sites of bone damage (14). In contrast, in a cohort of patients with RA in the established phase (mean 7.6 years), Qvistgaard et al (15) reported that high-frequency ultrasound and PD evaluation of small hand joints demonstrated no preponderance of blood flow near joint erosions.
The data demonstrate that ultrasonographic measures of synovial thickening and vascularity were able to discriminate between the 2 treatment groups after 18 weeks of treatment, with greater sensitivity than conventionally used outcome measures of change in disease activity that depend on the use of clinical evaluation, such as the numbers of tender and swollen joints in combination with measurement of ESR. We observed a 50% median reduction in the synovial thickening score measured by high-frequency ultrasound in patients treated with infliximab + MTX compared with a median increase of 1.2% in patients treated with placebo + MTX. This is consistent with our previously reported findings of diminished synovial tissue expression of TNFα, chemokines, and adhesion molecules, together with reductions in synovial inflammatory cell infiltration and granulocyte trafficking to RA joints after administration of infliximab (16–18). However, gray-scale images obtained with high-frequency ultrasound scanning of an MCP joint represent an overview of the amount of swollen tissue around the joint but do not differentiate between fibrous tissue and active synovitis, and for this reason we also used ultrasonographic technology capable of detecting synovial vascularity.
Histologically, RA synovitis is characterized by a mononuclear cell infiltrate and abundant vasculature (19). Furthermore, the disease activity of a given joint is correlated with the synovial vascularization (20, 21). Accordingly, we have shown that serum concentrations of the proangiogenic cytokine vascular endothelial growth factor (VEGF), measured at the first presentation of RA, predict the rate of radiologic damage during the subsequent year (22). Large vessel blood flow is at a high velocity and can be readily detected by conventional color Doppler sonography, which encodes the mean Doppler frequency shift. However, blood flow at the microvascular level, which is of interest with respect to rheumatoid synovitis, is at a lower velocity and is less readily detectable by this means. PD sonography, on the other hand, encodes the amplitude of the power spectral density of the Doppler signal and is a sensitive method for demonstrating the presence of blood flow in small vessels. The PD signal is actually a measure of the density of moving reflectors at a particular level, and thus of fractional vascular volume (23, 24).
PD, as with most other ultrasound methods, is insensitive to flow in submillimeter vessels, and therefore serves only as an indirect surrogate for measurement of capillary flow. However, several recent studies have demonstrated that PD ultrasound is capable of detecting synovial hyperemia in the inflamed RA joint (25–30), and that signal intensity correlates well with histologic assessment of synovial membrane microvascular density (28). Quantitative PD assessment of vascularized synovium in MCP joints of patients with RA has been reported to correlate with ESR (15). Three small, uncontrolled studies have demonstrated a reduction in PD signal following therapeutic intervention in RA (25, 26, 31). Similar findings have also been shown in PD studies of malignant neovascularity in cancer imaging, indicating that quantitative PD shows promise as a noninvasive method of monitoring oncologic therapy (32).
A previously unreported finding of this study is that, for patients treated with MTX alone, sonographic measures of synovial thickening and vascularity at baseline are strikingly correlated with the magnitude of joint damage over the following year. In particular, there was a highly significant, strongly positive correlation between the baseline summed color Doppler area in pixels for all 10 MCP joints imaged and the progression in total vdH-Sharp score over 54 weeks (Spearman rank correlation 0.78, P = 0.005). Treatment with infliximab + MTX not only abolished the positive relationship between the baseline vascular signal and progression of joint damage, but also resulted in a weakly negative but nonsignificant correlation. This negative correlation following infliximab + MTX therapy suggests that those patients with the highest baseline disease activity, as assessed by PD, may derive the greatest benefit from anti-TNFα treatment with respect to the suspension of structural damage to joints.
Similarly, in a previous immunohistochemical investigation of serial synovial biopsy specimens obtained from a small group of patients with established RA before and after a single dose of 10 mg/kg infliximab, we found that the patients with highest baseline synovial TNFα expression derived the greatest clinical responses (16). It is of interest to note that there was a positive but nonsignificant relationship between baseline CRP level and progression of radiologic damage in the placebo + MTX group, which was also abolished by infliximab treatment. However, these data must be interpreted with some caution, because there were differences between the 2 treatment groups in CRP values at baseline. Further studies are required to accurately determine whether baseline ultrasonographic parameters provide the power above and beyond baseline CRP levels to predict structural damage to joints.
Although there was good interobserver agreement between the radiograph assessors in this study, we missed the opportunity to further reduce the variability afforded by having both assessors read a set of test films and agree on a consistent approach for reading the study films. Reading and discussing a test set of images before reading the study images can reduce interreader disagreements by providing a common frame of reference. Similarly, we did not assess intraobserver variation for high-frequency ultrasound assessment of synovial thickening, although we tried to ensure consistency in assessment by having a single sonographer scan all patients' joints.
Limitations of ultrasonography include the lack of standardization for image acquisition and storage procedures. Furthermore, the quality of the examination is highly dependent upon the skill of the operator and the use of optimal equipment. Finally, there are potential problems with reproducibility based on intra- and interobserver variability and the use of different machines. In the future, 3-dimensional ultrasonography could be useful in reducing the sampling problem caused by spatial heterogeneity, and may permit greater standardization.
The significant reduction in PD area in patients treated with infliximab + MTX, as compared with those receiving placebo + MTX, supports the validity of the methodology used in this study and is consistent with our previous observations in studies of synovial biopsy samples obtained before and after infliximab + MTX treatment, which show a reduction in tissue vascular density (33). Diminished vascular permeability, a consequence of reduced VEGF levels (34), is likely to be a contributing factor in the rapid decrease in swollen joint counts observed following TNFα blockade, and in the reductions in ultrasonographic measurements of synovial thickening and PD area observed in this study.
The proportions of patients in the infliximab + MTX and the placebo + MTX groups achieving a moderate or good EULAR DAS28 response (67% and 33%, respectively) and meeting ACR 20% response criteria (50% and 17%, respectively) by 18 weeks are similar to the published results from the ATTRACT study at 30 weeks, comparing the efficacy of infliximab therapy added to once-weekly MTX with placebo infusions and MTX therapy (35). Although not powered for DAS28 or ACR20 assessments, ultrasonographic measures of synovial inflammation and vascularity are able to discriminate between the 2 treatment groups in this cohort, indicating that they are more sensitive than traditional clinical measures of response to therapy.
The findings of this study implicate vascular pannus regulated by TNFα in joint destruction in early RA, and suggest that the reversal of mechanisms that cause inflammation and joint destruction is already apparent soon after initiation of treatment with infliximab. These observations are of considerable practical importance for the individual patient and the efforts to optimize pharmacologic intervention in the treatment of early RA. Indeed, it has long been appreciated that for many patients, joint destruction and subsequent disability continue to progress despite the apparent suppression of synovitis as assessed by clinical examination (36). Until recently, the primary goal of treatment in RA was amelioration of signs and symptoms of disease. Another realizable goal of therapy in early RA is the significant attenuation of structural damage to joints. A hypothesis arising from the present study is that ultrasonographic imaging can be more effective in identifying those patients at highest risk of accelerated joint damage and those most appropriate for intervention with biologic agents targeting TNFα.
The authors are grateful to Rob Eckersley for his assistance in analyzing PD images. The authors wish to acknowledge Patricia A. Geraghty and Mary H. Whitman for their assistance in the preparation of the manuscript.
- 14Combined power color Doppler and greyscale ultrasound demonstrate an association between vascularity and erosions in rheumatoid arthritis [abstract]. Arthritis Rheum 2000; 43 Suppl 9: S292., , , , .
- 21Magnetic resonance imaging–determined synovial membrane volume as a marker of disease activity and a predictor of progressive joint destruction in the wrists of patients with rheumatoid arthritis. Arthritis Rheum 1999; 42: 918–29., , , , , , et al.
- 24Power Doppler sonography. Radiology 1996; 2000: 21–3., .
- 33Reduced synovial vascularity following TNFα blockade in rheumatoid arthritis [abstract]. Arthritis Rheum 1998; 41 Suppl 9: S295., , , , , .