Abstract
Objective. To assess structural and inflammatory ultrasound (US) lesions of entheses in ankylosing spondylitis (AS) patients with active disease and to evaluate inflammatory lesions after 6 months of tumor necrosis factor (TNF-α) blocking therapy, in daily clinical practice.
Methods. Consecutive patients with AS were clinically evaluated and underwent US examination of 9 bilateral entheses before and after 6 months of TNF-α blocking therapy. US examination included the following as inflammatory lesions: bone erosions/cortical irregularities, enthesophytes, calcifications as structural lesions; adjacent bursitis, effusion, increased tendon hypoechogenicity or thickness; and positive power Doppler (PD) signal.
Results. At baseline, 105 (95%) of 111 included patients showed US abnormalities. Structural lesions were seen in 74 patients (67%) and inflammatory lesions in 88 (79%). Enthesophytes and positive PD signal were the most prevalent structural and inflammatory lesions, respectively. Most lesions were found at the lower extremities. Additionally, inflammatory lesions occurred at the lateral epicondyle of the elbow. Patients with structural lesions at baseline were significantly older, had longer disease duration, higher modified Stoke AS Spine score, and higher C-reactive protein. Individually, there was a great diversity in changes of inflammatory entheseal lesions during treatment, but on the group level no significant decrease was found.
Conclusion. This prospective observational cohort study in daily clinical practice shows a high prevalence of structural and inflammatory US lesions in AS patients with longstanding and active disease. Positive PD signal was the most common inflammatory feature. No significant change in inflammatory US lesions was found after 6 months of TNF-α blocking therapy.
- ANKYLOSING SPONDYLITIS
- SPONDYLOARTHROPATHIES
- ULTRASONOGRAPHY
- ENTHESITIS
- TUMOR NECROSIS FACTOR-α
Within the family of spondyloarthropathies (SpA), ankylosing spondylitis (AS) is a chronic inflammatory rheumatic disease that mainly affects the spine and sacroiliac joints. Inflammatory involvement of entheses, so-called enthesitis, is one of the characteristic extraspinal manifestations of AS. Clinical symptoms of enthesitis are pain, stiffness, and tenderness, with or without local soft tissue swelling at the entheseal site. Reported prevalence rates of enthesitis are high in axial SpA (axSpA) and vary from 40% to more than 70%1,2.
Clinical examination of enthesitis may reveal local tenderness by palpation, sometimes accompanied by swelling of the entheseal site. To evaluate entheseal involvement in clinical SpA studies, different enthesitis indexes have been developed such as the Maastricht AS Enthesitis Score (MASES)3 and the Spondyloarthritis Research Consortium of Canada Enthesitis Index4. Although these indexes are easy to perform, there may be discrepancies in “true” enthesitis and enthesitis assessed with these clinical enthesitis indices. It can be difficult to distinguish enthesitis from entheseal pain without inflammation by clinical examination. Currently, no gold standard is available to calibrate the presence of enthesitis. Probably, the most reliable method to demonstrate “true” enthesitis is histopathological examination of the entheses at the insertion. However, obtaining entheseal biopsies in clinical practice is hampered by practical and ethical problems due to the burden of this intervention for the patient. Therefore, alternative methods to assess enthesitis are investigated, including musculoskeletal ultrasound (US).
Musculoskeletal US is a reliable and easy-to-perform dynamic imaging technique that can visualize pathological changes such as enthesophytes, calcifications, or bone erosions in greyscale. These structural lesions can be present in more advanced disease. Applying the power Doppler mode (PD), an US technology to visualize blood flow, active inflammation at the entheseal site can be detected. Hypervascularization shown with PD is the main feature of active inflammation and can be found in early as well as advanced stages of the disease. PDUS could be an appropriate technology to monitor and evaluate the effect of treatment on enthesitis.
Severe enthesitis may lead to disability, especially in case of resistance to therapy. It can be treated by reducing biomechanical stress in combination with nonsteroidal anti-inflammatory drugs (NSAID). In case of insufficient effect, additional local US-guided corticosteroid injections at the involved entheseal site can be considered. Further, tumor necrosis factor–α (TNF-α) blocking therapy is available for axSpA patients with persistent high disease activity despite NSAID and conventional treatment5. Clinical effectiveness of TNF-α blocking therapy on enthesitis has been investigated in axSpA and peripheral SpA. However, those studies were heterogeneous regarding the evaluated entheses, scoring methods used, and followup time6,7,8.
Therefore, the 2 main objectives of our present study were to determine the prevalence of structural and inflammatory US lesions of the entheses in AS patients with active disease and to evaluate changes in inflammatory US lesions after 6 months of TNF-α blocking therapy in daily clinical practice.
MATERIALS AND METHODS
Patients
Between November 2004 and October 2008, consecutive out-patients with AS who started TNF-α blocking therapy at the Medical Center Leeuwarden (MCL) were included in this study. All patients participated in the Groningen Leeuwarden AS (GLAS) cohort, a prospective longitudinal observational cohort study with followup visits according to a fixed protocol9. All patients were over 18 years of age, fulfilled the modified New York criteria for AS, and started TNF-α blocking therapy because of active disease [Bath AS Disease Activity Index (BASDAI) ≥ 4 and/or expert opinion] according to the Assessment of SpondyloArthritis international Society (ASAS) consensus statement10.
Patients were clinically evaluated and underwent complete US examination at baseline (before starting TNF-α blocking therapy) and after 6 months of treatment. Clinical evaluation of enthesitis was performed with the MASES (range 0–13; Supplementary Figure 1, available with the online version of this article)3. Disease activity was assessed with the BASDAI and the AS Disease Activity Score (ASDAS). Further, C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) were measured. Physical functioning was assessed with the Bath AS Functional Index (BASFI) and quality of life with the AS Quality of Life (ASQoL) questionnaire11,12. Spinal radiographic damage was scored by 2 independent and trained readers blinded to patient characteristics using the modified Stoke AS Spine Score (mSASSS)13.
The GLAS cohort was approved by the local ethics committees of the MCL and the University Medical Center Groningen (TPO 364). All patients provided written informed consent according to the Declaration of Helsinki.
US protocol
US examinations of the entheses in brightness (B) mode and PD mode were performed by 2 rheumatologists (GAWB, ENG) who are experts in the field of ultrasonography. The US examiners were blinded to clinical data such as the disease status of the patient and previous results of the US examination. US examinations at baseline and 6 months were performed by the same investigator. Patients were instructed to discontinue the use of NSAID a week before US examination, both at baseline and after 6 months, because of the potential effect on enthesitis.
An Esaote Technos MPX US machine (Esaote) was used, with 2 transducers including a linear array 7.5–15 MHz and a 3.5–5 MHz convex transducer. All entheses were scanned with the linear array transducer, except the greater trochanter of the femur because it is more deeply seated in most patients. US examination was done according to a specific scanning protocol easily applicable in daily clinical practice. The following 9 entheseal sites were scanned bilaterally in 2 orthogonal planes: plantar fascia, Achilles tendon, patellar ligament on the patellar apex and the tibial tuberosity, quadriceps femoris, pes anserine, greater trochanter of the femur, common extensor and flexor tendon on the lateral and medial epicondyle of the elbow (Supplementary Figure 1, available with the online version of this article). In B mode, the following abnormalities were scored: bone erosions/cortical irregularities, enthesophytes, calcifications, adjacent bursitis, effusion, increased hypoechogenicity, and increased thickness of tendon. In addition, the entheses were scanned for increased vascularization in PD mode. Special caution was taken for the recognition of normal nutrient vessels entering the entheseal bone. The settings for US were Doppler frequency of 7.5 MHz, low wall filter, and pulse repetition frequency of 750 KHz. Gain was adjusted until background noise was removed.
All abnormalities in B and PD mode were scored as absence (0) or presence (1). Bone erosions/cortical irregularities, enthesophytes, and calcifications were considered structural lesions. Adjacent bursitis, effusion, increased hypoechogenicity, increased thickness, and positive PD were considered inflammatory lesions.
Statistical analysis
Prevalence rates were expressed as number of patients or lesions (%). Further, normally distributed data were reported as mean ± SD and non-normally distributed data as median (range).
Independent samples t test, Mann-Whitney U test, chi-square test, and Fisher’s exact test were used to compare patient characteristics. Wilcoxon signed-rank test was used to evaluate the change in clinical and laboratory variables from baseline to 6 months. Generalized estimating equations (GEE) were used to evaluate the change in inflammatory US lesions from baseline to 6 months. This model takes into account the within-patient correlation of the 9 bilateral entheses. The exchangeable correlation matrix was used. Spearman’s correlation coefficient was used to investigate changes from baseline to 6 months in clinical, laboratory, and US variables. Statistical analysis was performed using PASW Statistics 22 (SPSS).
RESULTS
In total, 111 consecutive patients with AS underwent US examination before starting TNF-α blocking therapy. The mean age of all patients was 42.9 years (SD ± 10.9), 71% were male, median symptom duration was 15 years (range 2–49), 81% were HLA-B27–positive, mean BASDAI was 6.0 (± 1.57), and 77% had ≥ 1 tender enthesis according to clinical examination (Table 1).
Of these 111 patients, 85 (77%) had a second US examination after 6 months. Baseline characteristics, including disease activity, were comparable between patients with and without the 6-month US examination, except for symptom duration and time since diagnosis (17 vs 8 yrs, p < 0.001; and 9 vs 3 yrs, p < 0.005, respectively).
Prevalence of US lesions before start of TNF-α blocking therapy
At baseline, 105 of 111 patients (95%) showed US abnormalities. Structural lesions were seen in 74 patients (67%), inflammatory lesions in 88 patients (79%), and both structural and inflammatory lesions were seen in 57 patients (51%).
Patients with structural lesions were significantly older than patients without structural lesions at baseline (44.4 vs 39.8 yrs, p = 0.036), had longer time since diagnosis (9 vs 3.5 yrs, p = 0.033), higher mSASSS (12.9 vs 6.9, p = 0.007), and higher CRP levels (16 vs 11, p = 0.022). No significant differences in patient characteristics were found between patients with and without inflammatory lesions at baseline (Supplementary Table 1, available with the online version of this article).
Patients with only structural lesions had significantly longer time since diagnosis than patients with only inflammatory lesions (15 vs 7 years, p < 0.05). No relationship was found between the presence of structural and inflammatory lesions at baseline (p = 0.408).
Additionally, no significant differences in structural and inflammatory lesions were found between male and female patients (data not shown).
Structural US lesions
In total, 202 structural lesions were found in 74 patients, with an average of 2.7 per patient (Table 2). Enthesophyte was the most common structural lesion (65%). Most structural lesions were found at the lower extremities: 85 (42%) at the Achilles tendon, 37 (18%) at the quadriceps tendon, 28 (14%) at the greater trochanter of the hip, 21 (10%) at the patellar tendon, and 18 (9%) at the plantar fascia (Table 2).
Inflammatory US lesions
In total, 254 inflammatory lesions were found in 88 patients with an average of 2.9 per patient (Table 3). Positive PD and bursitis were the most prevalent inflammatory lesions (55% and 34%, respectively). Most inflammatory lesions were seen at the following entheseal sites: 59 (23%) at the pes anserine, 43 (17%) at the quadriceps tendon, 42 (16%) at the patellar tendon, and 39 (15%) at the lateral epicondyle of the elbow (Table 3).
The effect of TNF-α blocking therapy
As expected, significant decreases in disease activity (BASDAI, ASDAS, CRP, ESR), physical function (BASFI), and quality of life (ASQoL) were found after 6 months of TNF-α blocking therapy (Table 4). The clinical enthesitis index (MASES) decreased significantly from 2 (range 0–12) to 1 (range 0–9; p < 0.001).
Evaluation of the total number of inflammatory entheseal lesions at the group level showed an insignificant decrease from 210 lesions at baseline to 180 lesions after 6 months in 85 patients with a first and second US examination (p = 0.20). Evaluation at the individual entheseal sites showed that positive PD signal disappeared in 100 entheses, but it appeared in 79 entheses. A persisting positive PD signal was found in 25 entheses (Table 5).
GEE analysis revealed no significant change over time in inflammatory US lesions at both patient and lesion level (p = 0.218 and p = 0.193, respectively). Table 6 shows the diversity of changes in inflammatory lesions at the different entheses during TNF-α blocking therapy.
No significant correlations were found between the change in total number of inflammatory entheseal lesions and the change in MASES, BASDAI, ASDAS, CRP, ESR, BASFI, or ASQoL.
DISCUSSION
In our prospective observational cohort study, US lesions of entheses were found in 95% of patients with AS who had active and longstanding disease. This high prevalence of entheseal involvement is in accordance with a small cross-sectional study of 36 AS patients with less active disease (median BASDAI 4.5) and also longstanding disease (> 10 yrs), in which a prevalence rate of 97% was reported14. A larger cross-sectional study of 197 patients with SpA, of which 135 were diagnosed with AS and similar disease activity and also longstanding disease (mean 10 yrs) showed 91% greyscale or PD lesions8. Additionally, they reported 47% intraentheseal and up to 58% in perientheseal US lesions.
As expected in longstanding disease, a large proportion of our patients had structural US lesions of the entheses (67%). A new finding is that structural lesions at baseline were associated with more advanced and active disease such as older age, longer disease duration, more spinal radiographic damage, and higher CRP. The prevalence of inflammatory lesions was also high (79%). Positive PD was the most prevalent inflammatory lesion (55%) and frequently seen at the pes anserine of the knee (35%). The knee joint has numerous entheseal sites and structures, which makes it harder to connect clinical symptoms to defined anatomical structures. This may lead to underestimation of pes anserine enthesitis. On the other hand, the presence of the inferior geniculate artery may be regarded as a pitfall, because the Doppler signal may be mistaken for inflammatory activity of the enthesis. The increase in positive PD at the pes anserine after 6 months of treatment may be explained by this pitfall.
To place the high prevalence of US lesions in AS into perspective, it is necessary to observe prevalence rates of US lesions in healthy controls. The limited publications show prevalence rates of morphostructural lesions between 0% and 29%15,16,17,18. Interestingly, a positive PD signal was never reported in healthy controls15,16,17,18. Therefore, positive PD appears to be a distinctive US feature to assess inflammation at the entheseal site in patients with SpA.
We found a large individual diversity in inflammatory lesions over time, especially positive PD and bursitis, but no decrease in total numbers of inflammatory lesions after 6 months of TNF-α blocking therapy or any relationship with clinical outcome. In contrast, Naredo, et al did find a significant decrease in several predefined entheseal US scores after 6 months of TNF-α blocking therapy8. This inconsistency can be caused by differences in study populations (85 patients with AS vs 197 patients with SpA, of which 135 were AS) and US protocol (daily clinical practice vs standardized approach). Wang, et al mentioned a significant improvement of Achilles enthesitis in 75 patients with AS after only 3 months of TNF-α blocking therapy; unfortunately, the exact results were not reported. In contrast with our study population, those patients were younger and had higher disease activity19. Recently, Song, et al demonstrated that a followup period of 1 or 2 years may be necessary to evaluate the effect of TNF-α blocking therapy on inflammatory entheseal lesions evaluated with MRI20. Overall, the minimum followup time needed to show an effect of TNF-α blocking therapy on inflammatory signs of enthesitis is not yet clear.
Unfortunately, no clear consensus is available on the exact US definition of enthesitis, neither on the location nor on the number of entheses to be examined with US in patients with SpA8,21,22,23. Several standardized quantitative scoring methods have been proposed. However, these various scoring methods include different entheseal sites and US techniques21,22,23,24,25,26. This leads to heterogeneous results, making the direct comparison of the several studies difficult and results hard to interpret. Therefore the Outcome Measures in Rheumatology (OMERACT) US working group published the definition of SpA-related enthesitis and the elementary lesions that should be included in US examination, based on a Delphi process27. Excellent agreement (93%) was reached on separating structural lesions from inflammatory lesions, as was also performed in our study. The selected lesions of this Delphi process show high comparability with the included US lesions of our study27. Additionally, we incorporated effusion and bursitis, which were present in 10% and 48% of patients with AS, respectively. The high prevalence of bursitis suggests that this may be of supplementary value.
To date, US studies of entheses in SpA have focused mainly on the lower limbs6,21,22,28,29. In our study, the lateral epicondyle of the elbow was also frequently involved (17%). Including the lateral epicondyle of the elbow in the US evaluation of the entheses should be considered for future SpA studies.
Our US study was embedded in a larger observational cohort study of patients with active disease before the start of TNF blocking agents in daily clinical practice. Therefore no control group was included in this US study, which is a limitation. Unfortunately, it is difficult to perform a randomized controlled trial on this subject because it is unethical to deprive patients with SpA who have active disease of a proven effective treatment.
Although both ultrasonographists trained extensively together prior to our study, and baseline and 6-month US examinations were performed by the same ultrasonographist, no formal interobserver and intraobserver reliability was obtained, and interpretation bias might be present. The 26 patients who did not have a second US examination had a shorter symptom duration and time since diagnosis than did the 85 patients who underwent US examination at both timepoints. This could have led to selection bias. However, no significant differences were found in the prevalence of structural and inflammatory lesions at baseline between patients with or without a second US examination.
The European League Against Rheumatology recommendations for the use of imaging in the diagnosis and management of SpA in clinical practice have been published30. Although these recommendations state that US provides additional information on peripheral disease activity, no clear advice can be given based on the results of US research. Regarding the results of our present study, routinely monitoring the effect of TNF-α blocking therapy on enthesitis after 6 months with US does not seem useful in daily clinical practice.
Our present study showed that structural and inflammatory US lesions were highly prevalent in AS patients with longstanding and active disease. Enthesophyte was the most prevalent structural lesion, and positive PD was the most frequently found feature of inflammation. Overall, there was no significant change in inflammatory entheseal US lesions after 6 months of TNF-α blocking therapy, with large individual variety of changes.
Until now, the absence of a clear US definition of enthesitis and description of location and number of entheses to be examined have led to heterogeneous study results. It is hoped that the outcome of the OMERACT Delphi process, the development of a clear US enthesitis definition, multiple followup visits, and longterm followup will help us to provide more robust data on entheseal involvement in SpA.
ONLINE SUPPLEMENT
Supplementary material accompanies the online version of this article.
Acknowledgment
The authors thank all participants of the GLAS cohort. Further, the authors acknowledge Dr. P.M. Houtman and A. Krol for their contribution to the data collection.
Footnotes
The Groningen Leeuwarden AS cohort was supported by an unrestricted grant from Pfizer pharmaceuticals. Pfizer had no role in the design, conduct, interpretation, or publication of this study.
- Accepted for publication December 15, 2016.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.