Imaging techniques such as magnetic resonance imaging (MRI) and ultrasound (US) have been increasingly used in psoriatic arthritis (PsA) providing additional clues to the pathogenesis of this peripheral, axial and dermatologic disease. This has improved our understanding of the disease and can be used to aid diagnosis and then to follow outcomes of treatment. Both imaging modalities have highlighted the differing involvement of PsA when compared with rheumatoid arthritis (RA) with a significant burden of entheseal disease, flexor tenosynovitis (occurring alone or as part of dactylitis) and other extra-capsular inflammatory changes. MRI scanning has also highlighted the link between the nail and the distal interphalangeal (DIP) joint confirming previous clinical observations. Imaging studies in psoriasis patients have discovered a high level of subclinical inflammatory change but the clinical importance of such findings has not yet been defined.
The potential use of MRI and US to monitor treatment outcomes has encouraged research in this field. In MRI, the PsA MRI Score (PsAMRIS) has been developed with promising initial validation. In US, work is ongoing with the OMERACT group to define key pathologies and to develop scoring systems. A few scoring systems are available for enthesitis scoring using US which are further being developed and refined.
Further improvements in technologies in both of these fields offer exciting possibilities for future research. New MRI techniques offer the chance to image previously ‘dark’ structures such as tendons which is key in spondyloarthritides (SpA). Sonoelastography may also improve our understanding of tendon involvement in SpA. Whole-body multi-joint MRI allows a ‘snapshot’ of inflammation in PsA including joints, entheses and spinal involvement. Three-dimensional US should improve reliability and comparability of US scoring reducing inter-operator variability. The latest machines offer real-time fusion imaging employing US machines with an in-built virtual navigator system linked to previous MRI acquisitions. All of these new techniques should aid our understanding of PsA and our ability to objectively measure response to therapy.
Psoriatic arthritis (PsA) is considered as part of the umbrella group of the seronegative spondyloarthritides (SpA), including PsA, ankylosing spondylitis (AS), reactive arthritis, inflammatory bowel disease-related arthritis and undifferentiated SpA. AS is considered as the ‘prototype’ SpA with typical features such as sacroiliitis, a high prevalence of human leucocyte antigen B27 (HLA-B27) and only minimal clinical variation. A proportion of patients with AS do have peripheral arthritis but this is much lower than that seen in PsA . However, PsA shows significant clinical heterogeneity with potential involvement of both the peripheral and the axial skeleton. In addition to arthritis, inflammatory changes are seen in many other tissues resulting in enthesitis and dactylitis which are considered hallmarks of PsA. The phenotype of PsA varies significantly and includes ‘RA-like’ polyarticular disease, oligo or monoarthritis typically of large joints and those with predominant axial or entheseal disease. This heterogeneity must be taken into account when reviewing imaging literature in this field.
Magnetic resonance imaging (MRI) and ultrasound (US) have helped in the identification of the different pathologies found within PsA joints and can be used in diagnosis to help differentiate between PsA and other inflammatory arthropathies. Both techniques are also increasingly being used to monitor disease both in clinical practice and in research studies with the use of specific scoring techniques. This chapter aims to summarise recent MRI and US research in PsA, discussing the use of these modalities in identifying pathology, aiding diagnosis and following disease progression.
Magnetic resonance imaging
The potential for imaging studies to contribute to the understanding of the pathogenesis of PsA is well recognised . However, there has been very little work done using modern imaging such as MRI and ultrasound in PsA, especially in early disease, to improve our insight into the disease. High-field MRI with contrast enhancement is arguably the ‘gold standard’ for simultaneously imaging soft tissue and bony pathology in PsA. MRI is the most sensitive imaging modality for the assessment of structures critical in the evolution of inflammatory disease . MRI allows visualisation of the area of interest in three planes and provides detail of both the bone and surrounding soft tissue. Studies in other inflammatory arthritides such as rheumatoid arthritis (RA) have confirmed the superiority of MRI when compared with plain film radiography for assessing structural damage. Unlike radiography or computed tomography (CT), there is also no ionising radiation exposure for the patient. The sensitivity of MRI as an outcome measure in RA allows the detection of differences between treatment groups earlier than conventional radiography, and smaller degrees of change may also be observable . As a result, MRI is being increasingly used in clinical studies, in terms of both identifying features for diagnosis and monitoring disease progression over time .
The majority of MRI studies to date, particularly those comparing PsA to other inflammatory arthritides such as RA, have scanned the hand and wrist using conventional 1.5-T magnets with a surface coil. There are some studies in PsA which have looked specifically at large joints such as the knee or ankle. Conventional MRI can only be used to scan one joint region of interest at a time and this is a limitation when assessing PsA, which is very heterogeneous in terms of phenotype and joints involved. Conclusions that can be drawn from MRI studies about the pattern of disease in PsA are limited when only certain phenotypes, such as RA-like polyarticular disease, are included and when they are based on imaging of just one joint area.
Current role of MRI: pathology
MRI allows visualisation of the diverse pathological tissues in PsA. These differing features can be used to aid diagnosis of PsA.
Synovitis
Despite the fact that synovial tissue samples have suggested a difference in the histopathology of PsA and RA tissue , joint synovitis in PsA and RA is indistinguishable on static or dynamic MRI scanning when matched for disease activity . Other features such as enthesitis, dactylitis and spondylitis can be used to differentiate the two conditions . A small study comparing 10 patients with PsA and 10 with RA investigated the appearances of metacarpo-phalangeal (MCP) synovitis using a 1.5-T MRI scanner. Patients included in this study were selected for MCP synovitis although the majority had oligoarticular PsA ( n = 8). This highlighted similar post-contrast appearances of synovitis, although a little higher grade in RA patients . Predominant extracapsular enhancement was described as being ‘more striking’ in the PsA patients but was not a common enough feature (PsA patients n = 3 vs. RA patients n = 2, p = 0.252) to differentiate between diseases . A slightly larger study of post-contrast 1.5-T hand MRIs comparing PsA ( n = 18) and RA ( n = 21) also found similar proportions with synovitis but a different distribution. MRI-identified pathology including synovitis was more commonly seen in the proximal inter-phalangeal (PIP) joints in PsA, whereas synovitis in the wrist, mid-carpal and MCP joints was more common in RA . This study was a retrospective analysis of MRI scans performed as part of clinical practice and therefore is likely to have selected patients with hand involvement. Many patients with PsA, particularly those with large joint oligoarthritis, are not likely to have synovitis in the small joints of the hands. These findings therefore may not be generalisable to all PsA patients.
Bone erosions and bone marrow oedema
Bone erosions appear similar to those seen in RA and do not have disease-specific appearances on MRI . In both diseases, a break in the bone surface can be visualised on MRI . This may be filled with inflammatory tissue or with fibrous tissue. It has been shown in a variety of inflammatory arthritides, including PsA, that even large erosions seen on MRI may not be identified using conventional radiography . There is evidence that erosions progress at a slower rate in PsA than in early RA , but obviously this is dependent on the cohort of patients studied as radiographic progression in PsA is very variable.
The appearance of bone oedema across the inflammatory arthritides is similar on MRI. Images show an increased signal within the bone on short-tau inversion recovery (STIR) imaging, T2-weighted films with fat saturation and enhancement on post-contrast T1-weighted images . In RA, bone marrow oedema on MRI has been shown to correlate with osteitis on histology of peripheral joints However, in PsA, MRI bone marrow oedema has not been validated against histopathological change at the peripheral joints, although this has been examined at the sacroiliac joints of SpA patients .
Bone marrow oedema is commonly described as an MRI feature of PsA. Using non-contrast-enhanced MRI scanning, Giovagnoni and colleagues noted signal change in subchondral bone (bone oedema) in 43% of their PsA patients . Godfrin et al. showed that bone marrow oedema at entheses on MRI correlated with hot spots on radionuclide scanning . Bone oedema identified in the spine and peripheral joints (knee, hip and hands) has been shown to improve in PsA with anti-tumour necrosis factor (TNF) therapy also suggesting that bone oedema represents tissue inflammation. Similar to synovitis, the differences seen between RA and PsA for bone oedema and erosion relate to the sites of involvement rather than a different appearance of the pathology.
In PsA, evidence linking bone erosion to preceding joint inflammation is not as clear as in RA. A cross-sectional MRI study of 28 patients with erosive PsA did show that bone marrow oedema PsA MRI (PsAMRIS) scores were significantly higher in the arthritis mutilans type than in other non-mutilans PsA, and these patients also had high radiographic scores for joint damage using the modified Sharp–van der Heijde method . However, no longitudinal studies have confirmed a definite link between bone marrow oedema and bone erosion to date. This study also highlighted an unusual pattern of diaphyseal bone oedema in four of the 28 patients which may be specific to PsA. This was seen in one patient with arthritis mutilans and three patients with a non-mutilans pattern of PsA.
Enthesitis
Inflammation in tendons and ligaments can be viewed on MRI as swelling of an involved tendon and increased signal on T2-weighted images. In addition, enthesitis is represented by increased signal on STIR images at the tendon/ligament insertion and associated signal change in the underlying bone. McGonagle et al. studied enthesitis in knee arthritis associated with new-onset SpA (including PsA) and RA . Extracapsular oedema was seen adjacent to the enthesis at the patellar tendon, iliotibial band and the posterior capsule of the knee in all 10 of the SpA group but only 4/10 in the RA group. Six of the SpA patients also showed bone marrow oedema which was maximal at the site of entheseal insertions . The same group also imaged calcaneal enthesopathy showing a similar increased peri-entheseal signal and bone marrow oedema .
Extracapsular inflammation
When considering extra-capsular anatomy, the features of PsA on MRI imaging are significantly different from RA and more closely resemble changes seen in other SpAs. Jevtic and co-workers first described the extensive extra-capsular inflammation seen on contrast-enhanced MRI scans in PsA. In their series of 16 PsA patients, half demonstrated changes similar to those seen in RA with active synovitis and pannus where inflammatory changes were always located within the joint capsule. However, the other half of the patients showed inflammation also involving neighbouring structures including thickened collateral ligaments and periarticular soft tissue. In one joint, predominant extracapsular inflammation was seen without significant associated synovitis, thus raising the possibility that the joint capsule and synovium is not always the primary target in PsA . However, although these features were interesting, they were not uniform across the PsA cohort with around half having intracapsular disease and no extracapsular inflammation. This research suggested that there may be heterogeneity in PsA where some patients have a predominantly synovial disease as in RA and some show an extracapsular focus for their inflammation, consistent with the proposal that PsA, like SpA, is an entheseal-driven disease. It is also possible that some cases of sero-negative RA may have been included if patients had co-existent psoriasis and fulfilled the Moll and Wright criteria.
Giovagnoni et al. also showed extensive inflammation in the periarticular tissue beyond the joint capsule and involving the surrounding subcutaneous tissue in patients with PsA. In patients with RA, the inflammation was confined to the joint capsule in most cases and where periarticular oedema was seen in RA patients, it did not extend to the subcutaneous tissue . A review by Spira et al. confirmed the ‘typical’ pattern of PsA with early extracapsular enhancement associated with diffuse soft-tissue oedema spreading to the sub-cutis .
McGonagle et al. studied knee synovitis in patients with a recent onset (mean duration 10 weeks) of knee effusion . Ten patients with RA and 10 patients with SpA (including three with PsA) were included. Focal soft-tissue oedema and bone marrow oedema adjacent to the entheseal insertions were more common in SpA, despite little clinical evidence of enthesitis. This MRI pattern of inflammation is similar to that seen in peripheral enthesitis suggesting the possibility of a common pathogenesis.
Dactylitis
Dactylitis, one of the hallmark clinical features of PsA, occurs in 16%–48% of PsA and can be seen in many PsA phenotypes. It is often painful, but a chronic, non-tender dactylitic swelling can also occur. The pathogenesis of dactylitis is still not fully understood with abnormalities seen in many of the tissues on imaging.
The majority of imaging work on dactylitis has been performed by Olivieri et al. This group specifically imaged 12 dactylitic fingers using MRI and US, showing that all dactylitic fingers had a moderate to severe flexor tenosynovitis but no peritendinous oedema was identified. They initially concluded that dactylitis was due to flexor tenosynovitis and that the peritendinous soft tissue was not involved . However, subsequent work by the same group showed that peritendinous inflammation and oedema were present in a proportion of dactylitic digits in addition to the flexor tenosynovitis . They have suggested that peritendinous oedema could have been caused by increased capillary permeability secondary to flexor tenosynovitis, but this remains hypothetical.
Jevtic’s earlier MRI study included two cases with clinical dactylitis, which was associated with a much greater degree of extra-capsular inflammation in the soft tissues around dactylitic digits. In these cases, there was extensive soft-tissue involvement but no associated tenosynovitis . Only coronal sequences were performed in this study which could mean that tenosynovitis was missed. A more recent study by Healy et al., imaging 19 dactylitic digits, found soft-tissue oedema and synovitis to be the most frequent abnormalities in 69% of digits. However, in keeping with previous studies, a wide range of other abnormalities were frequently found including flexor tenosynovitis and bone oedema – the latter in several patterns ranging from discrete periarticular involvement to abnormalities in the metaphysis of the phalanx . Spira et al. reviewed the imaging findings of PsA and highlighted the key pathology of tenosynovitis with additional synovitis or soft-tissue oedema in dactylitis and that flexor tenosynovitis was far more common than extensor involvement .
Relationship between the DIP joint and nail disease
Distal interphalangeal (DIP) joint involvement, although not exclusive to PsA, is certainly one of the characteristic features of this disorder. Nail involvement is more common in PsA than uncomplicated psoriasis and DIP joint involvement is often seen with co-existent nail disease . Investigating the pathogenesis underlying this link between arthritis and nail disease is likely to help us gain a better understanding of the disease.
Histological work has shown that the extensor tendon attaches to the base of the terminal phalanx and then extends distally to connect with the nail root. This can now be visualised on MRI using high-resolution techniques. MRI studies have confirmed the intimate relationship between the nail bed, the distal phalanx, the DIP joint and the insertion of the extensor tendon .
Scarpa et al. imaged 23 patients with PsA using a 1.5-T MRI and surface coil, 12 with clinical onychopathy and 11 without. They showed that nail thickening on MRI was extremely common in patients with PsA, even when clinical evidence of onychopathy was lacking, and that all of these patients also showed MRI involvement at the distal phalanx. In the majority of patients without clinical onychopathy, the changes were confined to the nail and distal phalanx with little evidence of DIP joint disease. However, involvement of the DIP joint was much more common in those with clinical onychopathy (58% of cases) .
The Leeds group compared 20 patients with DIP joint involvement, half with osteoarthritis (OA) and half with PsA patients. Again, they used a 1.5-T MRI scanner, but this study used a new 23-mm high-resolution microscopy coil which produced superior image quality of just one joint. They demonstrated that PsA patients have significantly more entheseal and ligament enhancement, extracapsular changes and diffuse bone oedema . Although there was a difference seen at the group level, it was not possible to differentiate between OA and PsA in individual patients using the MRI findings. This inflammatory response seemed to be focussed around the ligament origins/insertions with virtual normality of ligaments up to the enthesis. Bone oedema was seen diffusely throughout the distal phalanx in 80% of PsA digits but was seen maximally at the insertion of the collateral ligament enthesis . The inflammation seen was extensive enough in most cases to involve the nail bed, distal phalanx and DIP joint providing an explanation for the common association of nail and DIP joint involvement. It seems likely that inflammation in this region is transmitted via the entheses, which extend up to the nail bed and envelop the DIP joint, but these studies did not identify where the primary site of pathology lies. Both Scarpa and the Leeds group suggested that inflammation was likely to start in the nail and spreads proximally to the distal phalanx and then to the DIP joint. This seemed credible given the usual scenario in this disease, where disease of the skin (and nails) usually precedes joint disease.
More recently, a longitudinal study of PsA DIP joints has clarified the relationship further. Dalbeth and colleagues performed longitudinal scanning over a 1-year period. At baseline, 34 patients were assessed clinically using the psoriatic nail severity score (PNSS) and had a 1.5-T MRI of their dominant hand. There was an association between clinical nail disease and MRI evidence of both bone erosion and bone proliferation at the DIP joint. Twenty patients attended for a repeat PNSS score at 12 months, giving follow-up data on 80 nails. Nails that had developed new onycholysis or hyperkeratosis during the study period were more likely to have had distal phalanx bone marrow oedema at baseline, suggesting that the inflammation may have originated in the bone and spread to the nail bed.
Subclinical disease in psoriasis
The majority of studies investigating subclinical joint disease in psoriasis have used ultrasound, but a small study from Italy has investigated MRI in this setting. Patients with active skin psoriasis but with no signs or symptoms of PsA ( n = 25) had a non-contrast-enhanced scan of their hand using a 1-T MRI with surface coil. They found that 17 patients had one or more signs of arthritis on MRI with capsular distension ( n = 11), periarticular oedema ( n = 9) and erosions at the MCP and interphalangeal (IP) joints ( n = 7) . By comparison, only one control patient had a bone cyst and they had no other features of PsA. It seems that MRI can be used to visualise signs of subclinical arthropathy, but the clinical significance of these findings has not been investigated.
Axial disease
MRI has also improved our ability to detect axial disease in PsA. Traditionally, the diagnosis of sacroiliitis in all forms of SpA, including PsA, has relied on radiological evidence of disease. However, it can take between 1 and 9 years from the onset of inflammatory back pain for radiological sacroiliitis to develop . In the early 1990s, MRI began to be investigated as a tool to detect sacroiliitis . Since then, MRI bone oedema has been demonstrated in all forms of SpA, affecting the sacroiliac (SI) joints and the cervical, thoracic and lumbar spine. Bone oedema adjacent to the SI joints has also been proven to correlate with histopathological evidence of inflammatory disease . MRI is now accepted as a diagnostic tool for axial disease in SpA including PsA and has been used as an outcome measure to evaluate treatment with TNF blockers . Further information on the use of MRI in axial SpA can be found in Chapter 6.
Monitoring treatment outcomes
While MRI has often been applied as a research tool to investigate the pathogenesis of arthritis, it has less commonly been used as an outcome measure for disease activity and joint damage. However, increasingly studies are looking for objective methods of measuring disease activity and MRI is able to visualise all structures in and around the joint as well as inflammatory change involving synovium, entheses and bone.
The Outcome Measures in Rheumatology Clinical Trials (OMERACT) MRI in inflammatory arthritis group have developed a scoring system for PsA MRI (PsAMRIS) , following on from the development of the RA MRI score (RAMRIS) . The first step was a literature review of PsA MRI studies to identify typical pathologies that would need inclusion into a scoring system . The RA version of the scoring system scores the wrist joints and the metocarpophalangeal (MCP) joints of fingers 2–5. There was a concern about excluding the proximal interphalangeal (PIP) and DIP joints in a potential PsA scoring system, given the increased involvement typically in the DIP joints. Therefore, the joints scored in the PsAMRIS are MCP, PIP and DIP of fingers 2–5. These joint regions were divided by the midpoints of the phalangeal bones and were then subdivided at the joint space line to give three joint regions and six sub-regions . Given that synovitis, bone erosions and bone oedema appear similar to that of RA, the semi-quantitative scoring system used in the RAMRIS was adopted . In addition to this, other key features were defined and added to the scoring system. Tenosynovitis is assessed in each joint region in each of the four flexor tendons on a scale of 0–3 depending on the thickness of enhancing or bright signal within the tendon. Periarticular inflammation is scored as present or absent adjacent to each joint region on the dorsal and palmar aspect of the finger. Bone proliferation, defined as “abnormal bone formation in the periarticular region, such as at the entheses (enthesophytes) and across the joint (ankylosis)” , is also scored as present or absent in each joint region. The joint regions and scoring system are outlined in Fig. 1 .
Related posts:
Imaging in rheumatology: New tools for use in clinical practice in 2012
Should modern imaging be part of remission criteria in rheumatoid arthritis?
Nuclear imaging of rheumatic diseases
Imaging of gout – An overview
Controversies in conventional radiography in spondyloarthritis
Ultrasound applications for the practicing rheumatologist
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
