Diagnostic imaging of psoriatic arthritis. Part II: magnetic resonance imaging and ultrasonography

Abstract

Plain radiography reveals specific, yet late changes of advanced psoriatic arthritis. Early inflammatory changes are seen both on magnetic resonance imaging and ultrasound within peripheral joints (arthritis, synovitis), tendons sheaths (tenosynovitis, tendovaginitis) and entheses (enthesitis, enthesopathy). In addition, magnetic resonance imaging enables the assessment of inflammatory features in the sacroiliac joints (sacroiliitis), and the spine (spondylitis). In this article, we review current opinions on the diagnostics of some selective, and distinctive features of psoriatic arthritis concerning magnetic resonance imaging and ultrasound and present some hypotheses on psoriatic arthritis etiopathogenesis, which have been studied with the use of magnetic resonance imaging. The following elements of the psoriatic arthritis are discussed: enthesitis, extracapsular inflammation, dactylitis, distal interphalangeal joint and nail disease, and the ability of magnetic resonance imaging to differentiate undifferentiated arthritis, the value of whole-body magnetic resonance imaging and dynamic contrast-enhanced magnetic resonance imaging.

Introduction

One of the challenges of current rheumatology is to discriminate undifferentiated arthritis as early as possible, especially to identify early rheumatoid arthritis (RA) and other entities, including psoriatic arthritis (PsA). Despite the fact that histopathological differences of the synovium in RA and PsA have been reported, static magnetic resonance imaging (MRI) does not enable differential diagnosis of these two entities⁽¹⁾. However, it becomes possible when taking into consideration the other clinical and radiological features of PsA, such as dactylitis, enthesitis, sacroiliitis and spondylitis⁽²⁾.

Magnetic resonance imaging

Synovitis

The results of dynamic contrast-enhanced MRI (DCE-MRI) studies are still inconsistent (Fig. 1). Schraml et al.⁽³⁾ reported high synovial uptake in both RA and PsA groups of patients, with similar curves of enhancement (including maximal enhancement, ME and initial rate of enhancement, IRE) in the early phase. In the late phase (i.e. 15 minutes after administration of contrast medium), there was a statistically significant difference between these two groups. PsA patients had had lower levels of signal intensity than RA group. Quite distinct results were presented by Cimmino et al.⁽⁴⁾ In the 0.2T extremity MRI of the wrist joints they showed that RA patients had higher volume of inflammation at nearly all sites investigated, however PsA patients showed higher IRE and ME values. This discrepancy between these two studies could result from different region of interest (ROI) positioning⁽²⁾. The strength of the magnet could also have influenced the results. Nevertheless, Coates et al. showed that the quality of the assessment of synovitis and erosions in low-field MRI scanners is comparable to those in high-field MRI scanners, however, it appears less sensitive in the evaluation of bone marrow edema (BME).

Fig. 1.

MRI of the hand of a 43 y.o. male with PsA: A. TIRM T2, coronal plane; B. T1 FS CM, coronal plane; C. T1 FS CM, axial plane; D, E. color map of the maximum enhancement and 3D reconstruction (Image Analysis Ltd., UK): synovitis in the 2^nd metacarpophalangeal joint, tenosynovitis of the 3^rd digit's flexor digitorum tendons, to the lesser extent of the 2^nd digit, extracapsular inflammation around the extensor complex of the 2^nd finger

DCE-MRI studies can also assess vascularity of a bone in the course of osteitis, thus also enable early diagnosis and quantification of inflammatory activity in the bone(3). Poggenborg et al.(5) monitored the efficacy of biological treatment for PsA by means of DCE-MRI and showed reduction of inflammation within the bone, synovium and tendon sheaths, although not its complete disappearance.

BME and bone erosions

Bone marrow edema in RA involves typically subchondral layer of the bone. In PsA it is reported to be more extensive and expanding to diaphysis⁽⁵⁾. Narvaez et al.⁽⁶⁾ compared MRI images of the wrists and hands of the patients with PsA and those with RA. Enthesitis, extensive BME in the bones’ diaphyses and subcutaneous tissue involvement were specific for PsA. The authors did not report any differences in synovitis between RA and PsA. However, as for tenosynovitis, tendon sheaths of the flexors were more often involved in PsA, whilst, on the contrary, tendon sheaths of the extensors in RA⁽⁶⁾. Additionally, Tan et al.⁽⁷⁾ demonstrated that BME is more severe in patients with arthritis mutilans PsA phenotype than in other clinical manifestations of this entity.

As far as bone erosions are considered, there is no MR detectable disparities between PsA and RA⁽²⁾. A high-resolution micro-computed tomography provides differential diagnosis: erosions in PsA are smaller and deeper than in RA, and they become more irregular and indistinct when bone repair process comes to its voice.

Enthesitis

The are plenty of studies indicating that enthesitis does exist in spondyloarthritis, including PsA^(8–10). However, are these data satisfactory and really confirm that enthesitis exists, and imaging studies are able to differentiate enthesitis from the healing process of enthesis microinjury? In both MRI and ultrasonography (US), entheseal thickening, altered signal (or echogenicity), posttraumatic scars on various levels of mineralization, irregularity of the bone, bony erosions, cysts and additionally BME of the bony part of the enthesis detected on MRI are regarded as features specific for enthesitis. These are the same features as those seen in posttraumatic, degenerated entheses, thus in the majority of cases there is no possibility of differentiation between primary enthesitis and inflammatory changes seen in the healing phase of microtrauma (Fig. 2).

Fig. 2.

T1 FS sagittal plane MRI of the Achilles tendon in a 19 y.o. female with PsA: increased signal of the soft tissues adjacent to the plantar aponeurosis with bone marrow edema of the neighbouring part of the tuber calcanei, achillobursitis calcanei

Poggenborg's et al.⁽¹¹⁾ whole-body MRI (WB-MRI) study confirms that statement. The authors did not report statistically significant differences in entheseal involvement in patients with axial SpA (according to EULAR Synovitis Study Group guidelines), with PsA (according to Moll and Wright's criteria) and in healthy controls. The most frequently involved entheseal sites were ischial tuberosity, humeral attachment of the supraspinatus tendon and heel enthesis of the Achilles tendon. Other authors reported pelvic ligaments to be more frequently involved⁽¹²⁾. Poggenborg et al.⁽¹¹⁾ showed that patients with PsA had more often clinically diagnosed enthesitis than patients with axial SpA, however they did not differ in frequency of enthesitis on WB-MRI. On the contrary, in healthy controls, the frequency of enthesitis on WB-MRI was higher than in clinical exam⁽¹¹⁾. The Authors conclude that future studies on enthesitis should include proper evaluation of a lifestyle (activity level) and body weight for microtrauma-enthesopathy differentiation, and to truly diagnose enthesitis.

Coates et al.⁽¹³⁾ showed that in most cases clinical examination was not able to discriminate between PsA patients and the healthy, and a great percentage of controls (both healthy patients and with RA) had clinical features of enthesitis. Weckbach et al.⁽¹⁴⁾ reported more cases of enthesitis in MRI than in clinical examination, including 80% of the patients with PsA, due to the MRI protocol that covered greater number of entheses (1st and 7th costochondral joints, Achilles tendon and additionally pelvic entheses). However, the readability of WB-MRIs varied significantly from very good for centrally located joints to quite low for distal peripheral joints, because of evaluation of thicker layers than on conventional MRI, the assessment based on one plane only and the off-centre artefact, i.e. poor image quality in the area located in the periphery of the scanner, etc.).

Dactylitis

Dactylitis (i.e. sausage finger) is one of the hallmarks of PsA with the incidence of 16–48%. Despite unclear pathogenesis, the inflammatory process may affect surrounding tissue⁽²⁾. It can originate from digital flexor tenosynovitis, interphalangeal joint inflammation (synovitis) or subcutaneous tissue oedema of the fingers. According to researchers⁽²⁾, the inflammatory process is secondary to tenosynovitis. Yet our own imaging studies (unpublished data) do not enable us to endorse this connection because quite frequently we observe subcutaneous tissue involvement without any signs of ongoing or past tenosynovitis or synovitis (Fig. 1).

Extracapsular inflammation

One of the features regarding as specific for PsA is extracapsular inflammation. In several studies, contrast enhancement of the extracapsular soft tissues of hands with coexistence of the collateral ligaments thickening was observed, especially in early stages of the disease, and in some cases, surprisingly, without synovitis⁽¹⁵⁾. This feature in the Authors own experience (unpublished data), although actually seen in PsA patients (Fig. 1), seems not to be specific for PsA and may also be present in RA, frequently at the dorsal aspect of the hand, over proximal interphalangeal, and also metacarpophalangeal joints, without signs of extensors’ bursae inflammation, which occasionally can be seen at the level of metacarpophalangeal joints.

DIP joint and the nail

According to the histopathological research, there is a link between distal interphalangeal (DIP) joint inflammation and onychopathy. Some authors revealed enthesopathy of the lateral slips of the extensor tendon or collateral ligaments on the level of the DIP joint, others reported MRI changes in patients with PsA in the form of both nail thickening and BME in the distal phalanx⁽²⁾. With the clinical absence of onychopathy the involved areas included only the nail and distal phalanx, whereas with coexisting onychopathy, the DIP joint was also affected. The Leeds group, with the use of 1,5T MRI scanner and a new 23-mm high-resolution microscopy coil, showed that PsA patients have significantly more entheseal and ligament involvement, extracapsular changes and diffuse bone edema than patients with osteoarthritis (OA)⁽¹⁶⁾. The inflammatory response in PsA, on the contrary to OA, seemed to occupy the collateral ligament insertion as its bony part of the enthesis showed apparent BME⁽¹⁶⁾. Diffuse BME was also detected in the distal phalanx, nail bed and the DIP joint. It was hypothesized that entheses, which reach to the nail bed and envelop the DIP joint, transmit the inflammation in this region. Moreover, both Scarpa⁽¹⁷⁾ and the Leeds group suggested that the nail is the site were inflammation is the most likely to commence, then spread proximally to the distal phalanx and the DIP joint. Other MRI study carried out by Dalbeth et al.⁽¹⁸⁾ proposed that the inflammation may have originated in the bone and subsequently extend to the nail bed.

MRI follow-up of PsA

The OMERACT group (Outcome Measures in Rheumatology Clinical Trials) elaborated the scoring system for evaluation of inflammatory and destructive changes in PsA hands – the Psoriatic Arthritis Magnetic Resonance Imaging Score (PsAMRIS). It assesses metacarpophalangeal (MCP), proximal interphalangeal (PIP) and DIP joints of fingers 2–5, searching for synovitis (score 0–3), flexor tenosynovitis (score 0–3), periarticular inflammation (absent or present), bone edema (score 0–3), bone erosion (score 0–3) and bone proliferation/ enthesophytes (absent or present).

In the clinical practice, the assessment is time-consuming and not enough reliable. Strube et al.⁽¹⁹⁾, in their study on 65 consecutive patients with PsA, evaluated the usefulness and repeatability of this scoring method on 0,2T MRI scans. They revealed low to moderate interobserver compatibility in separate components of the scale assessment, especially for tenosynovitis and periarticular inflammation. This may be caused by the fact, that T1-weighted sequences can not be saturated in low-field scanners, which worsens the quality of the reading significantly. Østergaard et al.⁽²⁰⁾ in their low-field MRI scanner study confirmed the interobserver divergence and Ejbjerg et al.⁽²¹⁾ showed low sensitivity of low-field MRI in assessing pericapsular inflammation. All the above results limit significantly the use of low-field MRI.

Poggenborg et al.⁽⁵⁾ compared the hand image on MR, CT (computed tomography) and plain radiography in patients with PsA under anti-TNF-alfa therapy. MRI inflammatory features (BME, synovitis, tenosynovitis) diminished after treatment, but did not withdraw completely. Moreover, dynamic MRI parameters of inflammation decreased during the study, so as the Authors encouraged further studies of DCE-MRI as a potential disease activity marker in PsA. For the first time a connection between BME and bony erosion development was stated.

Axial disease

With the introduction of ASAS (Assessment of SpondyloArthritis International Society) criteria, the use of MRI in detecting BME in the sacroiliac joints and the spine, in patients with all clinical forms of SpA, has become more common (Fig. 3). Castillo-Gallego et al.⁽²²⁾ assessed the lumbar spine and the sacroiliac joints, using the semiquantative Leeds Score System, in which BME is graded from 0 to 3, according to the severity of the lesions. They showed relationship between the severity and the extent of the disease and a positive test for human leucocyte antigen B27 (HLA-B27). Namely, HLA-B27 negative PsA patients had lower BME scores than HLA-B27 positive⁽²²⁾.

Fig. 3.

MRI of the sacroiliac joints in a 25 y.o. female with suspicion of PsA: A. TIRM T2; B. PD; C. T1 FS; D. T1 FS CM: bilateral subchondral bone marrow edema in the sacroiliac joints, with signal enhancement after contrast media administration, to the greater extent on the left side, bilateral minor bony erosions – bilateral sacroiliitis

Whole-body MRI

PsA can affect numerous joints, thus WB-MRI could be an optimal exam to identify all sites affected. Weckbach et al.⁽¹⁴⁾ using WB-MRI revealed in ca 80% of PsA patients more inflammatory changes than in clinical examination, which in 74% had an influence on the therapy. The sensitivity of MRI in assessing active and chronic changes in PsA is higher than in CT or plain radiography. However, as for MRI, Poggenborg showed that the quality of the peripheral joints scans on WB-MRI is inferior to the proximal ones, especially of the elbow (only 5–8% readable, 10% in case of DIP joints)⁽²³⁾. Synovitis in the peripheral joints was reported more frequently than BME and erosions. In hands (PIP and DIP joints), this feature was more common in PsA than in SpA and healthy controls. The frequency of synovitis and BME in PsA was similar in hands and feet, whereas SpA patients had more inflammatory lesions in feet. Large joints, usually glenohumeral joints, were inflamed with the same incidence in PsA and in SpA. The most frequently involved sites in PsA were: wrists joints, glenohumeral joints, knees, MCP and metatarsophalangeal joints. The Authors did not report BME in healthy controls.

Ultrasonography

The ultrasound examination, like MRI, allows detection of early inflammatory changes of the soft tissue. They have the potential to diagnose possible preclinical PsA, for example the foot involvement, which can be asymptomatic in patients with dermal psoriasis⁽¹⁾.

Synovitis and tenosynovitis

Ultrasound shows a typical spectrum of pathologies, like in other inflammatory processes: exudate, synovial pathology, i.e. thickening and increased vascularisation, with the following destructive component. As far as sausage digit (dactylitis) is considered, US can differentiate subcutaneous edema with tenosynovitis and synovitis⁽²⁾. Fournié et al.⁽¹⁵⁾ demonstrated extrasynovial soft tissue inflammation in 84% of cases (pseudosynovitis), including 60% coexisting with synovitis of the finger joints. Fournié also compared ultrasonographic features of PsA and RA and showed that enthesitis, subcutaneous soft tissue thickening and a positive doppler signal at the base of the nail are suggestive of periungual psoriatic involvement, and may be indicative of PsA. These interesting observations require further investigation (Fig. 4, Fig. 5).

Fig. 4.

Bilateral enthesopathy of the flexor pollicis longus tendons (arrows), more pronounced on the left side, in a 64 y.o. male with clinical diagnosis of PsA: A. PA radiograph; B. US (right hand on the left, left hand on the right side); C. negative power doppler signal

Fig. 5.

64 y.o. male with PsA: A. clinical image: onychopathy (nail bed seperation); B. radiographs: fluffy periostitis at the base of the distal great toe's phalanx; C–E. US: thickening of the nail of the affected toe, without enthesopathy of the lateral slips of extensor tendon (C), highly increased vascularisation nail bed (D), avascular enthesopathy of the flexor hallucis longus tendon (E)

The Author did not state which insertions are prone to inflammation: capsule-ligamentous complexes of radial or ulnar site or deep digital flexor tendon or lateral slips of extensor entheses.

Enthesitis

The OMERACT definition of the entheseal pathology in the course of SpA uses the term enthesopathy as: “abnormal hypoechoic (loss of normal fibrillar architecture) and/or thickened tendon or ligament at its bony attachment (may occasionally contain hyperechoic foci consistent with calcification), seen in two perpendicular planes, that may exhibit doppler signal and/or bony changes including enthesophytes, erosions, or irregularity”⁽²⁴⁾. In recent years, the presence of only one of the above features is defined as enthesitis (Fig. 6). For example, in one of the studies⁽²⁵⁾ clinical examination detected pathology of entheses in 22% of patients suffering from long-lasting PsA, while US in 56%. D'Agostino et al.⁽²⁶⁾ found features of enthesitis in almost 98% of patients with SpA, in 44% with mechanical back pain, and in 60% with RA. We agree with the Authors, that differentiation between mechanical, overuse and inflammatory changes is difficult, however not because of entheseal artefacts in doppler imaging. Most of mistakes (beyond erroneous interpretation of inflammatory changes in bursa as enthesitis) are caused by overlapping of imaging features both of inflammatory and mechanical/ overused enthesopathy. In our opinion, the definite diagnosis of enthesitis might apply to pediatric patients due to the minimal risk of imposing mechanical damages on the image, and to adult patient, with increased vascularity of the entheses, without clear history of trauma, whose treatment was not successful.

Fig. 6.

Ultrasound examination of the elbow joint: increased vascularity and tears in the enthesis of the common extensor tendon

Conclusion

Considering the inflammatory changes of the joints in the course of psoriatc arthritis, plain radiography is helpful in late, advanced stages of the disease. Early changes may be diagnosed using US and MRI. These two methods will optimise early diagnosis and consequently treatment. However, there is still a need for more prospective studies, in order to better understand PsA pathogenesis, and for further research, to optimize early diagnosis of PsA, to establish the role of US and MR as the diagnostic tools for the assessment of the disease severity, and to define prognostic factors, the evaluation of treatment efficacy and disease activity.

Conflict of interest

Authors do not report any financial or personal connections with other persons or organizations, which might negatively affect the contents of this publication and/or claim authorship rights to this publication.