Timeline of US enthesitis

In the last 15 years, the OMERACT US Task Force has taken considerable effort to improve the standardisation of US assessment of entheseal involvement in patients with SpA and PsA3–6 and several goals were achieved.

The first milestone dates back to 2005 with the publication of the preliminary definition for US enthesopathy, which included all the main abnormalities detectable by US and laid the foundations for further international research.3

In 2011, a systematic literature review undertaken by the OMERACT group showed the lack of consensus on US examination (ie, scanning methods to use and number of entheses to assess), and the need to determine an US definition of enthesitis.4

In 2014, Terslev et al reached an agreement on the individual elementary lesions to include in the definition of US enthesitis.5 Six elementary components of US enthesitis were identified, re-defined and divided in US findings indicative of ‘active’ entheseal inflammation (hypoechogenicity, thickening and power Doppler signal) or ‘structural damage’ (bone erosions, enthesophytes and calcifications).

A few years later, in 2018, the following final US definition of enthesitis in SpA and PsA was proposed by the OMERACT group: ‘hypoechoic and/or thickened insertion of the tendon close to the bone (within 2 mm from the bony cortex), which exhibits Doppler signal if active and which may show erosions and enthesophytes/calcifications as a sign of structural damage’.6

OMERACT definitions of US enthesitis: recent data on controls

In the last 2 years, three studies obtained data at the large entheses of healthy subjects using the 2014 OMERACT US definitions developed by Terslev et al.5 The results of these three studies converged in detecting a remarkably high prevalence of pathological findings at subject level, being entheseal thickening and enthesophytes the most frequent, and Doppler signal and bone erosions those with the lowest prevalence. Guldberg-Møller et al found an US elementary lesion in at least one enthesis in 47 (73.4%) out of the 64 studied subjects,7 Bakirci et al detected entheseal thickening and enthesophytes, respectively, in 69 (86.3%) and 70 (87.5%) out of the 80 participants8 and Di Matteo et al found the prevalence of entheseal thickening significantly higher than the one of hypoechogenicity both at subject and at entheseal level.9 Finally, in a recent study by Falsetti et al, conducted in a cohort of 50 dysmetabolic patients, entheseal thickening and hypoechogenicity were found in at least one enthesis in more than two-thirds of the patients.10 Conversely, the number of entheses with positive Doppler signal was low in all these studies, namely, 4 (1.25%) out of 320 entheses in Guldberg-Møller et al’s study, 11 (1.15%) out of 960 entheses in the Bakirci et al’s study, 11 (1.34%) out of 820 entheses in Di Matteo et al’s study, and 6 (1%) out of 600 entheses in the Falsetti et al’s study.

The high prevalence of US enthesitis in healthy subjects questions the diagnostic accuracy and discriminant power of the OMERACT US definitions of ‘active’ inflammation at the enthesis. Moreover, taking literally the 2018 OMERACT US Task Force definition of enthesitis may entail a number of potential misinterpretations.

First, either hypoechogenicity or thickening is necessary and sufficient to detect enthesitis. As such, this would lead to a dramatic overestimation of the prevalence US enthesitis (figure 1A).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Metabolic syndrome: longitudinal scan of patellar tendon at its distal insertion into the anterior tibial tuberosity. Note the presence of hypoechoic areas (asterisks), entheseal thickening (arrowheads) together with an enthesophyte (arrow). (B) Psoriatic arthritis: while no B-mode signs of enthesitis are detectable, the image shows evidence of intense power Doppler signal at the entheseal area. att, anterior tibial tuberosity; pt, patellar tendon.

Second, even though included among the ‘inflammatory findings’, Doppler signal is somehow weighted differently. In fact, only Doppler mode can assess whether an enthesitis is active or not, but the presence of Doppler signal without B-mode findings is apparently not sufficient to define enthesitis.

The third and last unclear aspect regards the structural damage, which appears to be exclusively a bone-related matter; in fact, calcifications (the only finding accepted among the structural components related to the tendon side of the enthesis) were merged with the enthesophytes. Indeed, tendon damage within 2 mm from the cortical bone insertion does not have an US equivalent in this definition.

Open issues and possible solutions

Recent data on controls and a literal interpretation of the 2018 OMERACT US definition of enthesitis prompt out the following considerations.

First, the high prevalence of the B-mode findings defining enthesitis (ie, entheseal thickening and hypoechogenicity) in healthy subjects undermines their specificity and raises the need for refining the 2018 OMERACT definition. This could be obtained by adding quantitative measures (ie, cut-off values for entheseal thickening and Doppler score higher than 1). The Glasgow Ultrasound Enthesitis Scoring System (GUESS) and MAdrid Sonographic Enthesitis Index (MASEI), the most widely used scores for US enthesitis, have introduced cut-offs for measuring entheseal thickening.11 12 However, the known variability of tendons and ligaments thickness among individuals is a potential limitation to the systematic application of these cut-offs. Therefore, further investigations are needed to identify potential ‘confounders’ of entheseal thickening (eg, age, sex, body mass index and physical activities) and adjust cut-off values accordingly. A possible alternative could be the combinations of qualitative findings forming patterns of entheseal involvement distinctive of SpA/PsA; whether a combination of hypoechogenicity or thickening and Doppler improves the accuracy of a correct diagnosis in comparison to Doppler alone will be a matter of further research.

Second, Doppler signal has shown the highest specificity among the entheseal US inflammatory findings as well as a high sensitivity in detecting even a small amount of abnormal blood flow. Thus, the presence of Doppler signal alone (ie, without hypoechoic areas or entheseal thickening) should also be regarded as sufficient to define enthesitis, and not only to state its ‘activity’ (figure 1B). Indeed, this would arguably improve the diagnostic potential of US in patients with suspected SpA/PsA, especially in the early phases of enthesitis or when the clinical examination of the entheses is not conclusive. Moreover, focusing the research of Doppler findings only at entheseal level may lead to discarding the precious information, because Doppler signal can also be detected outside the 2-mm area in patients with SpA.13–15 Thus, in the presence of Doppler signal close to tendon-to-bone (or ligament-to-bone) insertion, also Doppler signal at tendon, ligament and/or bursal level should be considered an expression of active enthesitis. Moreover, enthesophytes disrupt the tendon-to-bone insertion line and focally move it proximally. This aspect may affect inter-observer reliability in the identification of the "entheseal area of interest", and requires a clear statement to avoid individual interpretations. We believe that together with the bony edge also the area where to detect specific abnormalities due to enthesitis should move proximally, and Doppler signal should be measured accordingly (figure 2A,B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Psoriatic arthritis. Longitudinal scan of Achilles (A and B) and extensor finger (C and D) entheseal insertions, in B-mode and power Doppler ultrasound. (A and B) Achilles enthesitis. The dashed line represents 2-mm distance around the bony surface of the enthesis that considers the pathological presence of the enthesophyte (arrow). The solid line represents the border of the enthesis area without the enthesophyte. (C and D) Finger extensor enthesitis at the base of the middle phalanx (mp). The dashed line represents 2-mm distance around the bony surface of the enthesis. At, Achilles tendon; c, calcaneal bone; pp, proximal phalanx.

However, Doppler signal can be observed also after intense physical activity, or in patients with tendinous tears (‘reparative Doppler’).16 17 Therefore, these possible predisposing conditions should always be investigated in the presence of Doppler signal at the enthesis; such finding should prompt a thorough US examination of the entire enthesis and tendon to rule out possible concomitant tears.

Third, hypoechoic areas and entheseal thickening have been frequently found in association with US findings of structural damage, such as calcifications and enthesophytes.9 This suggests that hypoechoic areas and entheseal thickening do not necessarily reflect only inflammation, but they may also be the expression of a reparative process at the tendon side of the enthesis. Indeed, it is known that these two mechanisms (ie, inflammation and degeneration) are not mutually exclusive, but might contribute together to the pathogenesis of tendinopathies/enthesopathies.18 19

New technology and small entheses

Finally, we must take note that the recent availability of very high frequency probes (ie, up to 24 MHz) allowed the assessment of small entheses of the hands (ie, insertions of finger extensor and flexor tendons, and pulleys), which have been demonstrated to be potential important targets of SpA (figure 2C,D).20–22

Such entheses might provide the opportunity to investigate sites expected to be less frequently exposed to the effect of concomitant metabolic syndrome and/or biomechanical forces in runners.

However, since the work of the OMERACT group referred on experience and data acquired in large entheses, whether the US findings and methods described and validated by that work can be applied sic et simpliciter to assess small entheses remains an open question.