Objective Subclinical enthesopathy is recognised in both psoriasis and psoriatic arthritis (PsA). This study used ultrasonography with power Doppler (PD) to test the hypothesis that subclinical enthesopathy in PsA was associated with an ‘inflammatory’ or vascular phenotype compared to that seen in psoriasis.
Methods 100 patients with a mean age of 46.3 years (SD 15) (42 with psoriasis and 58 with PsA) and 23 matched healthy controls (HC) from two centres were included. 1230 lower limb entheses were scanned by ultrasonographers blinded to clinical details. Both inflammatory and chronic features of enthesopathy were scored.
Results Psoriasis patients (with or without arthritis) were more likely to express a vascular phenotype, with higher inflammation-related enthesopathy scores than HC (for inflammation p<0.0001, for chronicity p=0.02, for total ultrasound scores p<0.0001). The PsA patients had higher ultrasound enthesopathy scores than psoriasis patients (inflammation p=0.04, chronicity p=0.02) and HC (inflammation p<0.0001, chronicity p=0.003). When symptomatic entheses were excluded, PsA patients still had higher PD scores than psoriasis patients (p=0.003). Doppler positivity in at least one entheseal site was observed more frequently in PsA (21/58, 36.2%) versus psoriasis (4/42, 9.5%; p=0.002).
Conclusions This study shows that the ultrasound appearances of subclinical enthesitis in psoriasis differ from the subclinical enthesitis in PsA, with PsA patients having more PD. This is suggestive of a more inflammatory or vascular process in PsA, and offers potentially novel insights into the progression from skin to joint disease in psoriasis.
- Psoriatic Arthritis
Statistics from Altmetric.com
During the past decade it has emerged that enthesitis and associated osteitis are the common denominators underlying the multifaceted skeletal manifestations of psoriatic arthritis (PsA) that include axial and appendicular disease.1 In keeping with the importance of enthesitis as the key pathological lesion in PsA, some studies have shown enthesopathy in asymptomatic large insertions of the lower limbs in patients with spondyloarthropathies including PsA.2 This is reminiscent of studies in inflammatory oligoarthritis, in which ultrasound detected synovitis in joints that were clinically uninvolved.3 Of even greater importance is that several studies have shown subclinical enthesopathy or osteitis in up to 50% of psoriasis patients with no skeletal symptoms.4
Ultrasonography is well suited to the assessment of entheses and is able to depict soft tissue thickening and oedema in addition to new bone formation and erosions.5 In recent years, power Doppler (PD) ultrasonography has been increasingly used in rheumatology as it identifies vascular abnormalities known to be associated with inflammation. PD ultrasound to some extent provides a reflection of the degree of angiogenesis, which is critically related to joint damage and therapeutic responses to drugs.6–9 An extra-articular PD signal has been demonstrated in both PsA and psoriasis patients.10–13 However, there is limited work directly comparing PD changes at the insertions between patients with PsA and psoriasis.
The purpose of this study was to undertake ultrasonography of symptomatic and asymptomatic insertions in cases with PsA and cases with psoriasis, the latter having no clinical arthritis. The hypothesis tested was that the imaging phenotypes might differ between PsA patients and psoriasis cases without clinical arthritis. In particular, we hypothesised that the enthesopathy associated with PsA would have a greater degree of vascularisation compared to that seen in psoriasis without arthritis. We also postulated that subclinical vascular changes at the entheses might be associated with more widespread disease activity in PsA. Such an imaging biomarker could be helpful in understanding and predicting the disease evolution from psoriasis to PsA, which at the present time is not fully understood.
The study was carried out at two European centres—at the Leeds Teaching Hospitals (UK) and the University Hospital Department of Verona (Italy). Ethics approval for the study was obtained at both sites.
Patient groups and clinical assessment
One hundred consecutive patients (42 with psoriasis and 58 with PsA) and 23 healthy controls (HC) were recruited. PsA patients were included if they fulfilled the CASPAR criteria.14 Psoriasis patients were excluded if they had a current or previous history of arthralgia, arthritis or enthesitis. The clinical assessment was performed by one rheumatologist at each centre, who was blinded to the ultrasound data. This assessment included the psoriasis area and severity index (PASI), tender and swollen joint counts (TJC, SJC) and the Leeds and SPARCC enthesitis indices in the patients with PsA. Patients receiving glucocorticoids and anti-tumour necrosis factor therapies were excluded from the study, those receiving disease-modifying antirheumatic drug therapy were recruited if they had active skin or joint disease (inadequate responders).
Ultrasonography was performed by three rheumatologists fully trained in musculoskeletal ultrasound and with a special interest in scanning enthesitis (SZA and CCG in UK, IT in Italy). Ultrasound was performed using a Logiq E9 machine (General Electric, Wauwatosa, Wisconsin, USA) in the UK and a Logiq 5 machine (General Electric) in Italy, both with a linear probe at 6–15 MHz.
The lower limb entheses that have previously been described in the GUESS scoring system (including Achilles, plantar fascia, quadriceps, patellar tendon origins and insertions) were scanned.2 To enable blinding, patients were asked not to communicate with the sonographer about their disease during the ultrasound assessment. The sonographer did not perform a physical examination, therefore sites other than those involved in the ultrasound scan were not seen by the sonographer. The room was completely darkened starting from the beginning of the ultrasound assessment. The light from the ultrasound machine alone is not enough for the sonographer to see the skin clearly. The patients were placed in a supine position to assess the entheses around the knee. The knee was extended to assess the presence of Doppler signal and semiflexed to 30° to assess the grey-scale (GS) changes. They were then asked to take a prone position with the feet over the end of the examination Table in a neutral position for visualisation of the entheses around the heel.
The PD settings were standardised with a pulse repetition frequency of 800 Hz, a colour-mode frequency of 9.1 MHz and low wall filters. The colour gain was increased to the maximum level not generating PD signals under the bony cortex. All ultrasound assessments were performed using a multiplanar scanning technique.
Ultrasound images interpretation
The GUESS scoring system was modified and the outcome measures in rheumatology clinical trials definition of enthesopathy was used to interpret ultrasound images: ‘abnormally 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’.5 Bursal enlargement was also scored. The thickness of the enthesis was measured at the level of insertions and was evaluated on a quantitative basis. Normal values for each insertion were accepted as reported in the literature2 and less than 1 mm of increase exceeding the threshold was scored as grade 1, 1 mm or greater but less than 2 mm of increase was scored as grade 2 and 2 mm or greater was scored as grade 3. Erosions were also scored quantitatively by the maximum diameter of the erosion (grade 1, >0 mm but <2 mm; grade 2, ≥2 mm and <3 mm; grade 3, ≥3 mm). The rest of the assessments were scored on a semiquantitative basis (mild changes grade 1, moderate grade 2 and severe grade 3).
It is now recognised that the enthesis represents an integrated organ comprising the insertion (which is avascular in health) and adjacent tendon, fibrocartilages and the immediately adjacent synovium when present.15 For the PD assessment we therefore looked for PD signal in all the components of the enthesis organ, including signals at the insertion and inside the adjacent bursa, wherever these were present. It is not possible to define the margins of the enthesis by ultrasound; however, the entheseal vascularity was assessed when the Doppler signal was close to the cortical bone.
Ultrasound findings were categorised according to the following: GS changes related to inflammation (entheseal hypoechogenicity, thickening and bursal enlargement) were added up to calculate a ‘GS inflammation score’. PD scores were added up to calculate a ‘Doppler inflammation score’. These two scores created a ‘total inflammation score’.
The GS changes related to chronic findings (calcifications, erosions and enthesophytes) were added up to calculate a ‘chronicity score’. These scores were summated to give a total ultrasound enthesopathy score.
Data are expressed either as frequencies, means (SD) or medians (range) according to the variability. The prevalence of each individual lesion by ultrasound in patients with or without arthritis was compared by using a χ2 test. The Mann–Whitney U test was used to compare ultrasound scores between groups.
Correlations between clinical parameters (TJC, SJC, PASI and disease duration) and ultrasound scores (separately for ultrasound scores related to inflammation, damage and total) were analysed using the Pearson correlation test. Statistical analysis was performed using SPSS V.11.5.
To provide agreement between sonographers, all investigators agreed on definitions and the scoring system before the study both on saved images and while acquiring sample images. Stored images from the first 21 patients included in the study (total 1020 images) were scored by all three sonographers for GS and PD for all entheses. Intraclass correlation coefficient values were calculated for each pair of investigators and for all types of ultrasound scores used for the purpose of this study.
Interobserver agreement on sonography scoring
A moderate to excellent agreement between both investigator pairs was found for different scores: for GS inflammation the intraclass correlation coefficient values were in the range 0.91–0.93 (95% CI 0.79 to 0.97), for PD inflammation 0.74–0.95 (95% CI 0.45 to 0.98), for chronicity scores 0.89–0.93 (95% CI 0.76 to 0.98) and for total ultrasound scores 0.92–0.95 (95% CI 0.81 to 0.98).
Patients and HC were of similar age (46.3 years (15) vs 52.2 years (11), respectively; p=NS) and body mass index (26.7 (19.3–42.4) vs 24.2 (21–31.3); p=0.08). The gender distribution was also similar in both groups (43% female in the psoriasis group and 52% in the HC). None of the patients with psoriasis had clinical enthesitis compared to 51.7% of the patients with PsA (30/58). In the PsA patients, the median TJC was 4 (0–20) and the SJC was 4 (0–19). PASI scores were slightly lower in the patients with PsA (3.3 (0–18)) compared to those with psoriasis (4.3 (0–22); p=0.02). The groups had similar disease durations (19.1 years (12.5) for PsA and 17.5 years (11) for psoriasis). In general, hypoechogenicity, thickening and enthesophytes were the most frequent ultrasound findings for each site, whereas PD and erosions were relatively rare (table 1) (figure 1). These findings are described in more detail below.
Soft tissue changes at insertions
As expected, patients with psoriasis (with or without arthritis) had higher enthesitis scores related to inflammation than HC (for inflammation p<0.0001, for total ultrasound scores p<0.0001). The PsA patients had higher ultrasound scores than both psoriasis patients without arthritis (for inflammation p=0.04, for total ultrasound scores p=0.02) and HC (for inflammation p<0.0001, for total ultrasound scores p<0.0001) (table 2). Certain findings including hypoechogenicity and bursal enlargement were more frequent in PsA compared to psoriasis (table 1).
Doppler positivity in at least one entheseal site was observed more frequently in PsA (21/58, 36.2%) than in psoriasis (4/42, 9.5%; p=0.002). A Doppler signal was not seen in any HC. The presence of Doppler positivity had a sensitivity of 36% (95% CI 24% to 50%) and a specificity of 91% (95% CI 77% to 97%) with a positive likelihood ratio of 3.8 to discriminate PsA and psoriasis. The most discriminative site for PD positivity was the retrocalcaneal bursa (seven PsA vs one psoriasis patient). Furthermore, six patients had a PD signal at more than one site and five of these patients had PsA.
Bone erosion and spur formation
Psoriasis patients (with or without arthritis) had more chronic changes compared with HC (p=0.02). The PsA patients had higher chronicity scores than both patients with psoriasis (p=0.02) and HC (p=0.003) (table 2). This may reflect previous episodes of inflammation. Calcification and enthesophytes were also more frequent in the entheses of patients with PsA compared to the psoriasis group, which may be in keeping with previous inflammatory episodes with subsequent tissue remodelling (table 1).
Clinical assessments and their association with ultrasound findings
When patients with clinical enthesitis were excluded from the PsA group, patients with arthritis still had higher PD enthesopathy scores than patients with psoriasis (p=0.003), as well as higher chronicity scores (p=0.01) (table 2).
The TJC (r2=0.21; p=0.03) and SJC (r2=0.29; p=0.003) correlated to entheseal Doppler scores. Furthermore, the SJC correlated to total ultrasound scores (r2=0.21; p=0.03). No link between PASI and disease duration and ultrasound scores was evident (data not given).
This study showed that the frequency of enthesis-related PD change was significantly higher in PsA compared to psoriasis even when only sites of asymptomatic enthesopathy were evaluated in PsA. Moreover, the frequency of subclinical enthesopathy was significantly higher in PsA compared to psoriasis. The results of this cross-sectional study suggest that there may be a trend towards increased entheseal thickening with subsequent vascular changes representing a step in the progression towards PsA in psoriasis cases. This will require confirmation in a longitudinal study.
It is noteworthy that abnormal patterns of vascularity have been reported in psoriatic arthritis synovium both by arthroscopic inspection and by histological assessment.16 ,17 Likewise, prominent vascular changes have been reported in skin disease and at the nail matrix in psoriasis.18 ,19 Furthermore, normal entheseal insertions are sites prone to microdamage and in the course of physiological tissue repair at such sites prominent histological evidence of vascular changes is seen even in healthy individuals.20 In this study we found a high prevalence of enthesophytes in the HC. Enthesophytes are not specific to enthesitis and may be related to mechanical forces. This finding also reflects the high sensitivity of ultrasound to detect these changes.
In this study, we found no entheseal PD signals in the HC group. The majority of the literature supports the absence of ultrasound-detectable blood flow in normal individuals,21 PD changes have been reported in a few cases.22 This raises the possibility of using Doppler signals for early diagnosis, although this would need to be demonstrated in prospective trials.
This study was performed in two different European populations by different sonographers, which raises issues about its reliability and general applicability. However, both study populations exhibited identical trends in patterns of disease. Likewise, the sonographers met and agreed on standard sonographic criteria for enthesitis and the interobserver agreement was quite high.
Taken together, these findings now need to be applied in a large inception cohort of psoriasis patients to ascertain whether they define the subgroup that will be likely to evolve into PsA in the following months and years. In that regard, we have previously used ultrasound to show a high frequency of subclinical entheseal involvement in patients presenting with psoriasis but without clinically evident arthritis.23 After following this cohort for 3.5 years we have seen a link between subclinical enthesopathy in patients with psoriasis and the future development of PsA.24
In conclusion, this study shows that the degree of systemic subclinical enthesopathy is much greater in PsA compared to psoriasis. Of particular note, the difference in vascular changes in PsA compared to psoriasis might have implications for determining the likelihood of a given patient progressing to PsA. This will form the basis for future studies.
The first two authors contributed equally
Contributors DGM conceived the study, supervised its conduct and takes responsibility for the integrity of the work as a whole, from inception to finish. ZRA, IT and SZA were involved in the design of the study. Ultrasound scans were performed by SZA, CCG and IT. All authors contributed to data acquisition and were involved in the critical revision of the final manuscript.
Funding This study was part-funded by the NIHR and by an unrestricted educational grant from Merck, Sharp and Dohme. SZA received a grant from the Turkish Educational Foundation. CCG was supported by grants from EULAR and the Spanish Foundation of Rheumatology.
Competing interest None.
Ethics approval Ethics approval for the study was obtained at Leeds Teaching Hospitals (UK) and the University Hospital Department of Verona (Italy).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.