Statistics from Altmetric.com
We read with interest the report by Kamel et al who highlighted the use of ultrasound (US) and magnetic resonance imaging (MRI) for the detection of patellar tendon enthesitis in patients with seronegative arthropathies without typical radiographic evidence.1 Their work adds to the growing body of evidence supporting the clinical use of US in rheumatological practice. US has previously been shown to be better than clinical examination for the detection of enthesitis,2,3 but data on MRI are more limited. The authors make interesting observations about the position of the abnormalities in the patellar tendon, when compared with the Achilles tendon, possibly relating to joint biomechanics and lines of force. However, we would like to raise a few points on what we regard as important omissions from the paper.
Firstly, the authors do not include the frequencies of the described US or the technical details of the MRI findings and do not state how the modalities correlated with each other. The authors also do not comment on the presence of bone marrow oedema adjacent to the enthesis on MRI. With regard to the plantar fascia, it has been reported that adjacent bone marrow oedema changes are more prominent than soft tissue entheseal changes.4 No data were presented on the reproducibility of either imaging technique for the detection of enthesitis.
Secondly, frequent mention of “early” US findings such as calcification and fatty degeneration is made. However, no correlation with disease or symptom duration is recorded for either image modality. Similarly, no correlation between patient age and the findings was made—that is, was calcification an age related phenomenon. Control groups of normal subjects and patients without spondyloarthropathy would have strengthened the study.
Thirdly, it would also have been relevant to know if the patients had had any previous corticosteroid injections, as calcified foci are not uncommonly found around the sites of injection, sometimes lasting for many months or years. It is possible, therefore, to over diagnose enthesitis if this is based on the presence of calcium deposits alone.
Finally, the authors make no mention of power Doppler, which has recently been shown to increase specificity of the grey scale findings for the detection of enthesitis.5 It would have been interesting to correlate this with the MRI findings.
In conclusion, although the findings in this report are interesting and we agree that US is a useful tool in the diagnosis of enthesitis, care needs to be taken in interpreting such data when all the information has not been presented.
We thank Dr Paul Emery and colleagues for their letter and we are happy that our study has stimulated useful comments. Indeed, the published “Letter” provided only a limited opportunity to describe detailed data. We were not able to publish the full text paper of knee patellar enthesitis because of some overlap of the data with our published study on heel enthesitis.1 Table 1 summarises the findings of both ultrasound (US) and magnetic resonance imaging (MRI) examinations of enthesis in each individual case. It also shows the frequencies of these described US and MRI findings and how these imaging modalities correlated with each other. We identified bone oedema in two cases: the two patients had reactive bone oedema secondary to patellar tendon inflammation that was maximal at the entheseal insertion, indicating osteitis associated with enthesopathy.
As regard the comment on calcification, the US was sensitive and accurate in detecting the early development of calcific foci in the patellar tendon in 2/16 (12.5%) patients, while MRI failed to recognise their presence (please refer to the previous published figures).2 The detection of an early calcification process by US was found to be a clinically important sign because it did not correlate with the disease duration. Further, the development of calcific foci in the patellar tendon was less frequent than the calcification of Achilles tendon of heel enthesitis.1,3 The two patients were aged 26 and 34 years, neither of them had a history of local steroid injection into the knee and so the calcification was not an age related phenomenon, rather it was secondary to a previous local steroid injection. We wonder how our colleagues came to the conclusion that we used the presence of calcium deposition as a clue to making a diagnosis of enthesitis. This was not mentioned in our letter or in our previous reports dealing with the diagnosis of enthesitis.1–5
The interobserver variability of sonographic readings was assessed by video recording the US examination and comparing the images obtained sequentially by three independent observers (sonographer, radiologist, and rheumatologist), who were unaware of the patient’s name or clinical diagnosis. Agreement between readers’ interpretation was statistically assessed using the weighted κ ranges from 0 (no agreement beyond chance) to 1.0 (perfect agreement beyond chance). The interobserver variability was negligible and yielded an excellent coefficient of r = 0.89 (baseline), r = 0.82 US, and r = 0.74 MRI. Therefore, the presented US data were statistically significant and clinically reproducible.
We are currently combining US examination with power Doppler for some cases when we expect proliferation of synovial tissue and/or other related soft tissue components. It was not practical to include detailed data in a “Letter”. These data deserve to be published in a separate report.
As our colleagues mentioned in their letter, we presented interesting data that describe the anatomical and pathological variations of enthesitis in the heels and knees. We strongly believe that the US examination is a very useful and reliable procedure for the diagnosis of enthesitis of different joints.