Background Imaging has an essential role in the new spondyloarthritis (SpA) classification criteria for axial but not for peripheral manifestations. We evaluated the impact of imaging findings for identification and treatment decisions in patients with peripheral spondyloarthritis (pSpA) and controls (non-SpA).
Methods Patients with pSpA (Assessment of SpA international Society criteria, n=30) and non-SpA (n=30), aged <45 years, with painful heels or knees, were recruited. Conventional radiography, grey-scale ultrasound including power Doppler (US/PDUS) and MRI of symptomatic areas were performed to assess inflammatory and structural changes. Mann-Whitney U test was used for group comparisons.
Results In total, 105 painful entheses (71 heels, 34 knees) in 60 patients were examined. Differences between diagnoses were found for symptom duration (pSpA: 17.2±27.5 vs non-SpA: 4.4±4.3 months), human leucocyte antigen B27 prevalence (67% vs 13%) and gender distribution (53.3% vs 20% male, respectively), all P<0.05. Logistic regression analysis for baseline differences showed that chronic changes (erosions and calcification) in the heel were more frequent in pSpA versus non-SpA by US/PDUS (62.5% vs 28.6% patients and 59.5% vs 26.5% entheses, P<0.05). Inflammatory changes in heel or knee by US/PDUS and MRI could not differentiate between non-SpA and pSpA.
Conclusions Differentiation between pSpA and non-SpA was only possible based on structural but not inflammatory changes in the heels and knees of symptomatic patients. US/PDUS was superior to MRI for this purpose. These findings imply that pSpA is associated with erosive changes at enthesitic sites, while inflammation and susceptibility are of minor influence for the development of erosions and calcification to pSpA.
- magnetic resonance imaging
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Statistics from Altmetric.com
What is already known about this subject?
The concept of peripheral spondyloarthritis (pSpA) relies mainly on peripheral symptoms, but imaging characteristics are not well studied.
What does thi study add?
This is the first study to prospectively compare imaging findings in pSpA and controls.
Identification of patients with pSpA in daily practice may not only rely on active but frequently also on chronic changes.
How might this impact on clinical practice?
Chronic changes are more relevant for the differentiation between spondyloarthritis (SpA) and non-SpA than active changes such as enthesitis.
The term spondyloarthritis (SpA) covers non-radiographic axial SpA (nr-axSpA), ankylosing spondylitis (AS) and peripheral spondyloarthritis (pSpA). The SpAs are characterised by inflammatory and structural changes in the axial skeleton and in peripheral joints and entheses. Enthesopathy of the lower limbs and especially of the heels is a typical feature of SpA, which is predominant in pSpA,1 2 while nr-axSpA and AS mainly affect the axial skeleton.3 Clinical assessment of peripheral symptoms, including peripheral enthesitis, is included in the core set of the Assessment of SpA international Society (ASAS)4 for evaluation of SpA disease controlling and treatment. The term enthesitis is commonly used for referring to inflammatory entheseal involvement and is included in the overall term of enthesopathies, together with the occurrence of tendon or ligament pathology detected as loss of normal echostructure and/or thickening, as described in a consensus definition of the Outcome Measures in Rheumatology Clinical Trials Ultrasound Group.5 In daily practice, patient’s complaints and clinical investigation are the features that most physicians will rely on for decision-making with respect to individual treatment of patients with enthesopathy. Due to the increasing availability and knowledge about the use of imaging procedures in the last decade, imaging techniques such as ultrasound (US)6–8 including power Doppler (PD) US9 and MRI10 11 are playing and increasing role in the assessment of axial and peripheral symptoms in patients with SpA.9 12 Scoring systems for the evaluation and quantification of enthesitic lesions especially in the lower limbs have been introduced.6 7 13
However, and despite these efforts, imaging has an essential role in the new SpA classification criteria for axial3 but not for peripheral14 manifestations. This is largely due to limited knowledge about the value of imaging to detect peripheral arthritis and enthesitis in SpA and also about its ability to differentiate SpA from other rheumatic or non-rheumatic conditions (non-SpA).15
The aim of this study was to evaluate the value of imaging procedures commonly used in daily practice for the differential diagnosis of patients with peripheral involvement of the lower limbs in pSpA versus non-SpA, with a primary hypothesis that the two conditions can be differentiated. In addition, the influence of imaging on treatment decision in those patients was evaluated.
All patients gave informed consent for participation in the study (study number: 3988–11). All patients gave written consent to participate in the study and for publication of their data as results of the study.
Patient selection and examination procedures
Patients were recruited either by their treating rheumatologists working in a private setting or by the outpatient clinic of our tertiary rheumatology hospital (Rheumazentrum Ruhrgebiet). According to the study protocol, patients were asked to participate in the study if they either visited their rheumatologist because of a painful heel or knee or if there was a finding of heel or knee pain after taking a history or at the physical examination (the sites of interest are described in table 2A, where also the imaging examinations were performed in a standardised manner). In addition, the physician had to confirm based on the clinical examination that the pain was located in the enthesis of the same sites. However, the physicians were told to not proactively provoke tenderness at these sites in order not to include patients in the study without an appropriate reason. ‘Enthesitic’ pain was defined as pain precisely localised at the area of the tendon insertion at the bone, which was also present or worsened by local pressure at exactly the same area. Any signs of tendinopathy differing from this definition were not taken into account. In addition, patients with swelling in the respective painful joints were also not recruited, in order to avoid painful sites due to swelling and misinterpretation of the clinical and imaging results.
All patients had to be ≤45 years of age at the time point of the investigation to be included in the study. Furthermore, the recruiting rheumatologist had to set the diagnosis of pSpA or non-SpA. In addition, the study protocol required that all patients with the diagnosis of pSpA had also to fulfil the classification to pSpA as described in ASAS.14 Even though some patients recognised with pSpA also had axial involvement, they were all classified as pSpA because their predominant symptoms were peripheral (heel or knee). Non-SpA patients could have any other rheumatological diagnosis. Exclusion criteria were a contraindication for conventional radiographs and MRI examinations, the primary diagnosis of psoriatic arthritis (PsA) and any previous surgery in the heel or knee that was currently painful for enthesitis.
After fulfilling the inform consent, all patients were sent to the Rheumazentrum Ruhrgebiet where they were examined clinically by another experienced rheumatologist (‘examiner’) for assessment of pain in the sites that were previously reported to be painful. This examination was performed within 2 working days after the examination of the recruiting rheumatologist. The ‘examiner’ was blinded for the patient’s diagnosis throughout the study. In case of more than one painful heel or knee per patient, all painful heels or knees of the same patient were assessed by the examiner. Non-painful areas were not examined by the ‘examiner’ and not included in the study. At the time of presentation to the ‘examiner’, also patient’s pain level and physician’s judgement of impairment due to enthesitic pain on a 0–10 Numeric Rating Scale in the symptomatic sites were recorded.
For patients with pSpA, disease activity (Bath AS Disease Activity Index16) and function (Bath AS Functional index17), AS Disease Activity Score18 and the AS Quality of Life Questionnaire19 were assessed but were not shown to the ‘examiner’ and were entered in the electronic data analysis sheet at the end of the study. In addition, the non-steroidal anti-inflammatory drug (NSAID) intake of all patients was calculated according to recently proposed ASAS index20 and documented in the same manner as for the patient’s reported outcomes described above.
After clinical examination of the enthesitic pain (performed without knowledge of the patient’s diagnosis), the ‘examiner’ set a decision on the possible further treatment (eg, analgesics only, local corticosteroid injection, non-steroidal antirheumatic drugs, disease-modifying antirheumatic drugs (DMARDs)) of each patient based on his clinical findings and diagnostic assumption only and recorded this decision. Thereafter, on the same day, a conventional radiograph and US/PD of all painful sites was performed by the ‘examiner’. MRI examinations of the same sites were also performed either on the same day or during the following day (without change of medication prior to performance of MRI) and were evaluated also by the ‘examiner’. Documentation of the US/PD and MRI findings was performed by a standardised scoring sheet that included all possible imaging findings for inflammatory and chronic lesions (see the Evaluation of imaging section and also the description of the lesions in table 2). After performance and evaluation of all imaging procedures, the ‘examiner’ took the imaging results into consideration and re-evaluated his initial treatment decision. Changes in this decision were documented for more intensive versus equal versus less intensive treatment. The examiner also noted the imaging findings that lead to the decision of adjustment of treatment.
Conventional radiographs were performed using the sagittal view for the heel and both the sagittal and the anteroposterior view for the knee.
US/PD examinations were performed with the patients lying prone and their feet hanging relaxed over the edge of the examination table.
MRI examinations were performed in T1-sequences before and after injection of contrast agent (gadolinium), with a slice thickness of 3–4 mm and sagittal, coronal and axial slice orientation, according to the standard protocol of our hospital.
Evaluation of imaging
All images were evaluated by the examiner, who is an experienced reader for interpretation of conventional radiographs and MRIs.
Conventional radiographs were evaluated for any pathological findings at the clinically painful sites.
Assessment of enthesitic pain by US/PD and MRI included the following sites, based on the features collected as proposed in different US scoring systems for enthesitis in SpA6 7 13 and using the definitions of5 (used similarly for both US/PD and MRI):
Intratendinous or peritendinous inflammatory signal.
Bone marrow oedema at the painful sites (MRI examinations only).
Thickness of the painful tendon at the area of clinical enthesitic symptoms. For patients with heel pain, the thickness of the Achilles’ tendon and plantar fascia was measured, while for patients with knee pain, the thickness of the quadriceps tendon at the upper and the lower patellar pole and at the tibial tuberosity was measured.
Normal versus pathological finding of fibrous structure including rupture of the painful tendon.
Bone erosion (defined as cortical disruption and loss of continuity) or calcification/enthesophyte formation at the site of examination.
In cases where not only sole chronic or only sole inflammatory findings but a combination of different chronic or different inflammatory findings was found (eg, combination of erosions and calcification and/or osteophytes in the calcaneus), this was considered as one chronic finding. Similarly, the combination of intratendinous and/or peritendinous inflammatory involvement and bone marrow oedema (on MRI) was considered as one inflammatory finding. In case of coexisting inflammatory and chronic changes, both changes were recorded separately for chronic and for inflammatory changes.
Comparisons of imaging findings were performed on patient basis but also based on single enthesitic sites. Continuous data are presented in a descriptive manner, and binary data (presence of a lesion yes/no) are presented as proportion of patients or sites being positive or negative. Comparisons of the baseline characteristics between both diagnoses were made by Mann-Whitney U test. Since from all baseline characteristics the duration of enthesitic symptoms, gender distribution and human leucocyte antigen (HLA) B27 status were found to be different between groups, comparison of the imaging findings was performed by univariate logistic regression analysis adjusted for these parameters. All analyses were performed using SPSS V.21.0.
A total of 60 patients were included, 30 with the diagnosis of pSpA and 30 with other diagnoses (non-SpA), and a total of 105 entheses, 71 heels and 34 knees were reported to be clinically painful and were evaluated.
In the SpA group, 16 patients also reported axial symptoms, and in the non-SpA group, 9 patients were diagnosed with rheumatoid arthritis, 8 with achillodynia, 4 with osteoarthritis of the hands, 2 with sarcoidosis (Löfgren’s syndrome) and 1 each with adult Still’s disease, systemic lupus, undifferentiated arthritis of the ankle, undifferentiated connective tissue disease and a degenerative heel spur.
The groups were similar for most demographic assessments but not for mean symptom duration of enthesitic pain (pSpA: 17.2±27.5 vs non-SpA: 4.4±4.3 months, P=0.005), HLA-B27-positive status (pSpA: 66.7% vs non-SpA: 13.3%, P=0.001) and gender distribution (pSpA: 53.3% male patients vs non-SpA: 20% male patients, P=0.008). At the time of examination, <66% of patients were taking NSAIDs, <35% DMARDs and <20% biologics in both groups. The detailed demographics and clinical characteristics of the patients in both groups are shown in table 1.
Of the 30 patients with pSpA, 17 (56.7%) reported heel symptoms only, 6 (20%) reported knee symptoms only and 7 (23.3%) reported symptoms in at least one heel and at least one knee. In comparison, of the 30 patients with non-SpA, 14 (46.7%) reported heel symptoms only, 8 (26.7%) reported knee symptoms only and another 8 (26.7%) reported symptoms in at least one heel and at least one knee.
Of the 105 entheses, 37 heels and 13 knees were assessed in patients with pSpA and 34 heels and 21 knees were assessed in patients with non-SpA.
No single findings in the clinical assessment were able to differentiate between pSpA and non-SpA, but there was a numerical trend towards lower scores in patients’ global assessment in patients with pSpA (5.6±1.6) versus non-SpA (6.5±2.5), which was similar also in the physicians’ global assessment (table 1).
After the clinical evaluation, a modification of the current treatment for enthesitic pain was proposed in 14/30 (46.7%) patients with pSpA and in 17/30 (56.7%) non-SpA patients (P=0.53).
In the analysis on a basis per lesion, significant differences between patients with pSpA and those with no SpA were found for bone erosions at the insertion of the Achilles’ tendon as assessed by US, with 59.5% lesions found in patients with pSpA versus 26.5% lesions found in non-SpA patients (P=0.008). After adjusting for symptom duration, this difference remained significant (P=0.041). In contrast, other comparisons including erosions of the Achilles’ tendon as assessed by MRI or inflammatory changes in and around tendons as assessed by both imaging techniques did not show any difference between pSpA and non-SpA. Findings on conventional radiographs could not differentiate between diagnoses at any of the painful sites. Detailed data of all lesions as detected by the different imaging techniques are shown in table 2A,B. An overview about the prevalence of chronic and inflammatory lesions on both US/PD and MRI is shown in figures 1-3.
In the analysis on the prevalence of lesions on a per patient basis, similar results were found, with 62.5% patients with pSpA versus 28.6% non-SpA showing bone erosions at the insertion of the Achilles’ tendon (P=0.036) as assessed by US/PD, while all the other analyses revealed no statistically significant differences between the two groups.
Impact of imaging on the patient’s treatment
After evaluation of US/PD and MRI, a change of medication based on the imaging findings (while the examining physician was still blinded for diagnosis) was decided in an additional 4/30 patients (additional 13.3%) in the SpA group, as compared with the decision based on clinical examination only. In all of these cases, the treatment modification was made towards a more intense anti-inflammatory treatment. In contrast, in the non-SpA group, the initial clinical decision for intensification of treatment by the clinical investigation was confirmed in only 8/17 patients (47%) after imaging, while in the remaining 9/17 patients, the clinical decision was not supported by the imaging result. As noted by the examiner, treatment modification was based on the general impression of the image, without a specific lesion having more or less impact on treatment decisions. Findings on conventional radiographs did not contribute to any change of medication in either group.
This study is the first to examine and compare clinical findings of patients with peripheral SpA and controls who both had heel and/or knee pain with the results of established imaging techniques such as conventional radiographs, US/PD and MRI. In addition, the influence of imaging on clinical decision-making for treatment was analysed in a daily practice setting.
The most interesting finding of this study is that it was rather the erosive changes at the heel enthesis and not the active inflammatory changes that were more frequent in patients with pSpA. This is in contrast to the landmark paper of McGonagle et al21 who reported enthesitis to be the main differentiating factor in patients with symptomatic knee arthritis with rheumatoid arthritis versus SpA (mostly PsA). Since the patients participating in the present study were included only with apparent clinical symptoms, we think that it fits to the imaging finding that there was a significant difference in the mean age between the groups and this, although the patients were all included based on their current age <45 years. However, formal proof of that will have to include a reinvestigation of the patients after some time to confirm the higher prevalence of chronic, postinflammatory changes in patients with pSpA compared with other diseases with similar symptoms and pathology. Our observation and the associated hypothesis may imply that, in patients with pSpA with an entheseal affection of the lower limbs, structural changes at these sites occur more frequently and probably earlier than in other diseases, which would also explain the longer symptom duration. However, we think that, if confirmed in other studies, our findings may become clinically even more important, since they indicate that, in patients, the item ‘enthesitis’, which is important in the ASAS classification criteria, may have to be reversed to ‘enthesopathy confirmed by imaging’.
The high prevalence of HLA-B27 found in patients with pSpA was expected. From our data, however, it appears that the occurrence of HLA-B27 may not so much be associated with susceptibility but rather with chronicity of entheseal pathology. This would be in line with older data in reactive arthritis22 and recent data of limited spinal inflammation in non-SpA patients, which are unlikely to develop into syndesmophytes as has been shown to occur in axSpA.23 However, as already mentioned, these data should be confirmed in other studies and centres.
Interestingly and not really expected, we obtained the same rate of inflammatory changes in patients with non-SpA as compared with pSpA in both the heel and the knee examinations. One reason to explain this result might be that definition used for imaging abnormalities was obtained based on the publication that has been used in our hospital5 but is slightly different from the most recent definitions published by Terslev et al.24 A less stringent definition of (inflammatory) abnormalities in our dataset might explain the relatively high prevalence of inflammatory abnormalities in controls, as compared with pSpA. Comparable data have been previously published in examinations of the heel only in patients with predominantly axial SpA or non-inflammatory back pain independent of the presence of peripheral symptoms15 and also in an earlier study that investigated patients with SpA-associated clinical symptoms and mechanically induced enthesopathy of the plantar fascia.25 Limitations of MRI to depict inflammation in the axial skeleton of patients with AS have been recently reported in two studies, in which histological examinations had been performed for direct comparisons.26 27 Whether the same would apply in patients with peripheral involvement in pSpA for comparison with other non-SpA conditions that are not known to be primarily associated with involvement of the entheses of the lower limb is unknown.
As mentioned above, we did not identify any significant differences in the prevalence of other US/PD and MRI findings and no major differences between these techniques. This is different from what has been reported in other studies, where both focal thickening and very early calcification foci in the tendon of the heel or knee area could be more frequently detected by grey-scale US than by MRI.28 29 With respect to conventional radiographs, it may have been expected that erosive changes in heels and knees should have been more frequently detected. However, probably also due to the young age of the patients, this was not the case here.
Another interesting finding of our study is the different impact of the clinical examination and the imaging results on the physician’s treatment decisions in patients presenting with entheseal pain of the lower limbs. Based on clinical examination, there was only a slight difference in the proportion of patients with pSpA (47%) versus non-SpA (57%) in whom the blinded physician failed the decision of treatment modification. However, after inclusion of imaging, this decision changed the result only in patients with pSpA (additional 13%), while in the non-SpA group, the clinical decision of treatment modification could not be confirmed based on imaging. Interestingly, none of the lesions (neither inflammatory nor structural) assessed was found to be primarily responsible for these decisions, but it was rather the overall impression of the MRI and US/PD examination that drove this decision, while conventional radiographs were not considered to be helpful. In contrast to the inability in differentiation between diagnoses, these findings confirm an important role of US/PD and MRI for decision-making in daily clinical rheumatology practice in patients with SpA in general and with pSpA in particular. Furthermore, and since imaging in pSpA leads to an increase in the proportion of patients needing more intensive treatment, our findings confirm the knowledge that clinical examination underestimates (the frequently silent) enthesitic involvement, as compared with findings obtained by imaging.6 30–32
This study has some limitations. First of all, we did not assess the asymptomatic sites of the patients included but concentrated only on the sites where patients claimed to be painful prior to examination and which were the reasons for them to visit a rheumatologist. Definitely, an analysis of the clinically not painful sites of the same patients would have been interesting for examination of so-called ‘silent’ lesions. However, due to the limited resources especially for MRI examinations for all tendons in all patients, this analysis was not possible in the present study, and therefore, this information was not collected. The fact that all examinations were performed by one physician could also be considered as a limitation. Nevertheless, the examining physician is an experienced person in the use and interpretation of imaging with the different techniques used here. In addition, this physician was blinded for the diagnosis of the patients during the entire period of the examinations, in order not to influence the interpretation of the images and treatment decision before and after imaging. In fact, we believe that this approach is rather an advantage, since this person could perform US/PD and evaluate imaging results also from radiographs and MRI based on the clinical complaints and his own findings after clinical examinations before the final treatment decision. This procedure excludes faults in interpretation, if secondary physicians who were not involved in clinical examination would have evaluated the images.
In conclusion, in symptomatic patients with peripheral knee or heel involvement, the prevalence of inflammatory lesions, including enthesitis, was similar in patients classified as peripheral SpA based on the ASAS classification criteria and in those with other diagnoses. Imaging modalities such as MRI and US/PD could not differentiate between diagnoses; however, erosive changes were more frequently prevalent in patients with pSpA. Together with the higher prevalence of HLA-B27 in patients with pSpA, it seems that it is rather the chronicity than the susceptibility of the symptoms that is associated with SpA. Furthermore, imaging (US/PD and MRI) of symptomatic sites was associated with an increase in anti-inflammatory treatment in patients with pSpA but not in those with non-SpA.
Overall, these findings add to the up to now only limited knowledge about the role of imaging in patients with pSpA according to the ASAS classification criteria.
We thank all participating patients for their willingness to participate in this study.
Part of this work was presented at the ACR meeting 2013.
Contributors XB: idea, study organisation, patient examination, statistical analysis, writing of manuscript. JB: idea, study organisation, writing of manuscript. UK, HA, FD, MI, LK, CK, DK, ES, ES-B and FR: recruitment of patients, comments on manuscript. FR: study organisation.
Funding This study was supported by an unrestricted research grant from Pfizer Germany after an independent external project review process. The grant was used for performance of the imaging techniques, laboratory examinations and data documentation.
Disclaimer Pfizer had no influence on study design, protocol, patient selection, treatment, interpretation of results or writing of the manuscript.
Competing interests None declared.
Ethics approval Ethical Committee of the Ruhr-University of Bochum, Germany.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
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.