The early diagnosis of rheumatoid arthritis can be very difficult radiologically, although it is somewhat easier for the rheumatologist when faced with a young to middle-aged female patient who has bilateral symmetric arthritis accompanied by pain, stiffness, and swelling in her wrists and metacarpophalangeal joints.

Often, the first radiographic findings may be discrete juxta-articular osteoporosis; however, even this can be difficult to substantiate. With digital imaging, the digitization process may correct for this finding–thus, juxta-articular osteopenia may be a radiological finding of the past.

The next radiographic finding is of discrete erosions and the location of the practice will depend on where they are first seen. Often, the first erosions are seen on the ulnar aspect of the base of the fifth metacarpal adjacent to the insertion of the ulnar collateral ligament. However, early erosions also can be seen on the ulnar styloid either at its tip, on the ulnar aspect of the bone, or at its base in the notch between the ulnar styloid and the distal metaphysis. Early erosions also may be seen on the ulnar aspect of the third, fourth, and fifth metacarpophalangeal joints, the distal aspect of the hamate adjacent to the base of the fifth metacarpal, and the radial aspect of the base of the first metacarpal.

Because it often takes 3 to 5 years for these erosions to appear, what other radiological procedures can help? Ultrasound has been used, but its use is controversial because it often will only pick up major changes in superficial joints.

Computed tomography scanning is inferior to both plain radiographs and magnetic resonance imaging (MRI). Bone scanning has been used but it is expensive, cumbersome, and not particularly accurate. Magnetic resonance imaging has been established as the method of choice for imaging in patients with early inflammatory arthritis. However even MRI, although very sensitive, is not very specific and the clinical picture as well as the biochemical results need to be considered.

In the 60-year-old patient who presented with a 5-year history of decreasing range of motion (ROM) and swelling in her knees, wrists, and metacarpophalangeal joints, the stiffness was worse in the morning and wore off during the day. It also was relieved to some extent by aspirin, ibuprofen, and other nonsteroidal anti-inflammatory drugs. The patient was otherwise healthy and her erythrocyte sedimentation rate (ESR), rheumatoid factor, and antinuclear antibody were and have remained negative. Initial radiographs were negative; however, later hand and wrist radiographs showed subtle juxta-articular osteoporosis and erosions (Figures 1 and 2). Magnetic resonance imaging of the hands and wrists was pathognomonic for an inflammatory arthritis.




Figure 1: AP radiograph of the left hand shows juxta-articular osteopenia and subtle erosions at the base of the fifth metacarpal. Figure 2: AP radiograph of the right hand shows similar changes.


MRI
Diagnosing Early Rheumatoid Arthritis

Although MRI is time-consuming and expensive, it has become the imaging modality of choice in early rheumatoid arthritis. Coronal PD and T1-fat saturation sequences should be used in conjunction with axial and sagittal T1- and T2-weighted sequences. One of the sequences should be a special sequence for cartilage evaluation.

Magnetic resonance imaging is capable of showing evidence of hyperemia, synovitis, and ultimately synovial thickening with erosions (Figures 3 and 4). Rheumatoid arthritis not only involves the articular synovium but the synovium surrounding tendon sheaths so that tenosynovitis is also common. Frequently MRI will show early subcortical cysts that are thought to be precursors of subcortical erosions and although well described in the pathological literature, they have only recently been described in the radiological literature.1

Another common finding in early rheumatoid arthritis are joint effusions. Bone marrow edema also can be seen before the erosions occur either as a result of hyperemia or as a result of synovial overgrowth. As the disease progresses, narrowing of the joint space also will occur but this is only seen in relatively advanced rheumatoid arthritis. Pannus formation also can be identified in more chronic disease

In the long-term follow-up of patients with rheumatoid arthritis receiving treatment either with the older drug regimens or with the new recombinant human IgG monoclonal antibody-derived drugs, it is possible to measure the actual volume of the synovium. This is an excellent way to assess disease activity; however, it is time consuming and not recommended for routine use.

Zikou et al2 used MRI to quantify hand synovitis in patients with rheumatoid arthritis treated with adalimumab. They followed 13 patients with refractory rheumatoid disease before and after one year of therapy. Eleven (85%) patients showed a decrease in synovitis, decreased C-reactive protein, and a decrease in ESR after therapy. The volume of inflammatory tissue decreased overall by 37%.





Figure 3: Two coronal images of the right hand show similar changes as in the left hand. Early T1 coronal (A). Delayed T2 fat saturation sequence (B).


Differentiating Rheumatoid Arthritis from Other Forms of Inflammatory Arthritis

Boutry et al3 studied 28 patients with early rheumatoid arthritis, 14 patients with systemic lupus erythematosus, and 5 patients with primary Sjögren’s syndrome. All of the patients had inflammatory polyarthritis of the hands clinically but had normal radiographs. They used a coronal T2-weighted short inversion time inversion recovery sequence, an axial T1-weighted fat-suppressed gadolinium-enhanced spin echo sequence, and a coronal gadolinium-enhanced 3-D gradient echo MR sequence. They assessed the radiocarpal joints, the intercarpal joints, the carpometacarpal joints, and the metacarpophalangeal joints in each patient. They found no statistically significant difference in patients with rheumatoid arthritis and other forms of polyarthritis in terms of synovitis, bone erosions, reactive edema, subchondral cyst formation, and tenosynovitis.

Boutry et al3 also discussed a newly developed scoring system for inflammatory arthritis. Cimmino et al4 reported that MRI does not differentiate rheumatoid arthritis from polymyalgia rheumatica and mixed connective tissue disease.




Figure 4: Two coronal MRI sequences of the left hand. Early T1 coronal shows multiple erosions and subchondral cysts as dark grey areas in the carpal bones and at the metacarpal bases (A). Delayed T2 fat saturation sequence shows all these erosions and cysts as high signal areas (B).


Diagnosing Osteoarthritis

Degenerative joint disease, commonly referred to in synovial joints as osteoarthritis or osteoarthrosis, manifests itself in a predictable cascade with two prevailing theories to explain the pathogenesis.

The first attributes the articular degeneration to initially induced cartilage damage. Either an abnormal force is applied to an intrinsically normal joint or normal forces have a cumulative degrading effect on a weakened susceptible joint. The resulting force-tissue imbalance disrupts the cartilage matrix and integral chondrocytes.

The competing theory identifies subchondral bone bruising as the initial cause of osteoarthritis. This leads to subchondral microfractures that are less plastic to loading forces and transfer a disproportionate force to the cartilage itself that will then fail mechanically. In the pressure regions of the joint, the repetitive insult to the cartilage and subchondral bone results in cartilage thinning, eburnation, and subchondral cyst formation. Interestingly, injury isolated to cartilage alone does not stimulate an active reparative response, in contrast to injury extending into the subchondral bone that initiates neovascularization of the cartilage and subchondral bone. In the weight-bearing areas, this hypervascularity induces new bone formation in osteophytosis along the articular margins. If the insult is allowed to heal, the damaged hyaline cartilage will be replaced with fibrocartilage, eburnation may resolve, and cysts will be remodeled into trabecular bone.

Sensitivity for the various imaging modalities varies based on the degenerative process evaluated. Cartilage itself is radiolucent, but is quantitatively assessed on radiographs based on the width of the joint space. Knee radiographs often are obtained in the neutral, flexed, and weight-bearing positions providing physiologic positioned and dynamic evaluation of the joint.

Magnetic resonance imaging has the advantage of point analysis for cartilage thickness, morphology, and radiographically occult chondral defects. Newer sequences that are variations of proton density fat saturation techniques have improved correlation with pathologic quantitative and histological sampling. Unfortunately, the tomographic method of imaging and lower image resolution is less reliable for the evaluation of the joint space.

Subchondral cysts are easily identified on radiographs as well-circumscribed lucencies surrounded by trabecular bone. Despite its lower image resolution, MRI has demonstrated higher sensitivity in detecting small subchondral cystic changes in all joints, most notably in the carpals and knee. Magnetic resonance imaging’s principle water-sensitive sequences–T2-weighted spin echo and inversion recovery sequences–are able to detect small osseous cysts based on their higher image contrast than the adjacent osseous signal intensity. Tiny cysts often are obscured on radiographs by superimposed bone margins and trabeculation that is not a factor in MRI.

Eburnation is the cumulative effect of subchondral microfractures, trabecular collapse, and healing callus. With hyperemic repair, subchondral bone becomes transiently weakened allowing for further central degradation and lateral bone repair. As the disease worsens, the ROM decreases in the joint, which concentrates the mechanical forces on a limited region that continues to deteriorate.

Radiographs are more sensitive in detecting eburnation due to their three-dimensional image summation compared to the individual tomographic slice images of MRI. Sclerotic bone with its dense bone mineral lacks any MR signal and is thus represented on MRI as black (lack of signal). Differentiating mild sclerosis from normal cortical bone on tomographic slices is more difficult than on radiographs. However, bone edema from reactive bone hyperemia that often is associated with eburnation, microfractures, and hemorrhage is only detectable by MRI.

Osteophytosis may produce focal excrescence or broad-based growths. Detection of osteophytes on MRI suffers from the same limitation as eburnation due to solid bone’s lack of MR signal. As the osteophytes enlarge, marrow extension will occur and the sensitivity of MRI improves due to the fat signal of the marrow. In comparison to these substantially larger abnormalities, radiographs are more sensitive in detecting small osteophytes that can be seen as subtle contour deformities.

When evaluating a patient with joint pain, radiographs remain the initial method of detecting cumulative cartilage loss, small osteophytes, and determining disease severity. If radiographs do not identify degenerative changes, MRI is beneficial to evaluate for focal chondral injury as well as morphological changes, non-calcified loose bodies, reactive bone edema, or small subchondral cysts that often are radiographically occult and indicative of early osteoarthritis.

References
Sommer OJ, Kladoesk A, Weiler V, Czembirek H, Boeck M, Stiskal M. Rheumatoid arthritis: a practical guide to state-of-the-art imaging, image interpretation, and clinical implications. Radiographics. 2005; 25:381-398.
Zikou AK, Argyropoulou MI, Voulgari PV, et al. Magnetic resonance imaging quantification of hand synovitis in patients with rheumatoid arthritis with adalimumab. J Rheumatol. 2006; 33:219-223.
Boutry N, Hachulla E, Flipo RM, Cortet B, Cotton A. MR imaging findings in hands in early rheumatoid arthritis: comparison with those in systemic lupus erythematosus and primary Sjögren syndrome. Radiology. 2005; 236:593-600.
Cimmino MA, Iozelli A, Garlaschi G, Silvestri E, Montecucco C. Magnetic resonance imaging of the hand in mixed connective tissue disease. Ann Rheum Dis. 2003; 62:380-381.
Authors
Drs Klassen, Kidwai, and Griffiths are from the Department of Radiology, University of Florida, Shands Jacksonville, Fla.

Reprint requests: Harry J. Griffiths, MD, Dept of Radiology, University of Florida, Shands Jacksonville, 655 W 8th St, Jacksonville, FL 32209.