Subclinical inflammation on MRI of hand and foot of anticitrullinated peptide antibody-negative arthralgia patients at risk for rheumatoid arthritis

Abstract

Introduction: It is known that anticitrullinated peptide antibody (ACPA)-positive rheumatoid arthritis (RA) has a preclinical phase. Whether this phase is also present in ACPA-negative RA is unknown. To determine this, we studied ACPA-negative arthralgia patients who were considered prone to progress to RA for local subclinical inflammation observed on hand and foot magnetic resonance imaging (MRI) scans.

Methods: We studied a total of 64 ACPA-negative patients without clinically detectable arthritis and with arthralgia of the small joints within the previous 1 year. Because of the character of the patients' symptoms, the rheumatologists considered these patients to be prone to progress to RA. For comparisons, we evaluated 19 healthy, symptom-free controls and 20 ACPA-negative RA patients, who were identified according to the 1987 American Rheumatism Association criteria. All participants underwent MRI of unilateral wrist, metacarpophalangeal and metatarsophalangeal joints. Synovitis and bone marrow oedema (BME) were scored according to the OMERACT rheumatoid arthritis magnetic resonance imaging scoring system, and the scores were summed to yield the 'MRI inflammation score'. Scores were compared between groups. Among the ACPA-negative arthralgia patients, MRI inflammation scores were related to C-reactive protein (CRP) levels and the tenderness of scanned joints.

Results: MRI inflammation scores increased progressively among the groups of controls and ACPA-negative arthralgia and RA patients (median scores = 0, 1 and 10, respectively; P < 0.001). The MRI inflammation scores of ACPA-negative arthralgia patients were significantly higher than those of controls (P = 0.018). In particular, the synovitis scores were higher in ACPA-negative arthralgia patients (P = 0.046). Among the ACPA-negative arthralgia patients, inflammation was observed predominantly in the wrist (53%). The synovitis scores were associated with CRP levels (P = 0.007) and joint tenderness (P = 0.026). Despite the limited follow-up duration, five patients developed clinically detectable arthritis. These five patients had higher scores for MRI inflammation (P = 0.001), synovitis (P = 0.002) and BME (P = 0.003) compared to the other patients.

Conclusion: Subclinical synovitis was observed in the small joints of ACPA-negative arthralgia patients, and especially in patients whose conditions progressed to clinically detectable arthritis. This finding suggests the presence of a preclinical phase in ACPA-negative RA. Further longitudinal studies of these lesions and patients are required to confirm this hypothesis.

Figure 1

Magnetic resonance imaging-based inflammation scores shown separately for the three study groups. (A) Magnetic resonance imaging (MRI) inflammation scores (synovitis plus bone marrow edema (BME)). (B) Synovitis scores. (C) Bone marrow oedema scores. The three study groups are the symptom-free controls, the anticitrullinated peptide antibody (ACPA)–negative arthralgia patients and the ACPA-negative rheumatoid arthritis (RA) patients, based on the 1987 criteria for RA [16]. The scores presented are for all participants individually (dots) and the median scores per group (horizontal lines). The red dots indicate the ACPA-negative patients who developed clinically detectable arthritis during the median follow-up of 9 months. The y-axes are split because RA patients had higher scores than the symptom-free controls and ACPA-negative arthralgia patients. The presented P-values were obtained by comparing the scores of ACPA-negative arthralgia patients and symptom-free controls. All P < 0.001 for differences in MRI-based inflammation, synovitis and BME scores between the three groups.

Figure 2

Subclinical inflammation visualised by magnetic resonance imaging of two different anticitrullinated peptide antibody–negative arthralgia patients without clinically detectable arthritis. Images show the metacarpophalangeal (MCP) joints and wrists of anticitrullinated peptide antibody (ACPA)–negative arthralgia patients without clinically detectable arthritis. The white lines in the top coronal images reflect the localisation of the bottom axial images. (A) Post–contrast enhancement coronal (A1) and axial (A2) T1-weighted fast spin echo (FSE) images with fat saturation showing enhancement of the MCP2, MCP3 and MCP5 joints, which is consistent with active synovitis. Also, pronounced tenosynovitis in the third flexor tendon is present, although tenosynovitis is not included in the OMERACT rheumatoid arthritis magnetic resonance imaging scoring system score and was not evaluated in the present study. This patient developed clinically detectable arthritis during follow-up. (B) Post–contrast enhancement coronal (B1) and axial (B2) T1-weighted FSE images with fat saturation showing bone marrow oedema (BME) and erosions (confirmed on the pre–contrast enhancement T1-weighted FSE sequence) in the lunate. Also, there is active synovitis in the intercarpal joint.