AB0941 Discrimination of enthesic inflammatory involvement in patients with spondyloarthritis through the use of computer analysis of static ultrasound images

Abstract

BackgroundOne of the weaknesses of the ultrasound study of the entheses is the absence of a universal, recognized and replicable ordinal evaluation scale that allows quantifying or semi-quantifying the degree of involvement of these structures.ObjectivesTo analyze computer analysis of static images of patients with inflammatory enthesopathies differentiation capacity.To determine its ability to discriminate between an affected and a healthy contralateral enthesis in the same individual.MethodsStatic images of the Achilles and epicondylar entheses of patients who consulted regarding enthesis pain associated with spondyloarthritis (ASAS/CASPAR criteria) were analyzed. We picked patients between 2016 and 2020. Images were obtained by the same observer under clinical practice routine circumstances and with the same ultrasound equipment (Logiq 8e, General Electric, USA). The same settings for each of the two enthesic examinations were used. Image analysis was performed using the ImageJ 1.53e program (Wayne Rasband & contributors – NIH, USA). Imaging records with contralateral studies in patients with recently diagnosed asymmetric enthesopathies (before the start of treatment) were selected. A repository of images from healthy controls, matched by age and sex of the patients, was used as control.We compared the mean gray intensity (MGI), its dispersion (DGI) and the mode (MoGI) of the studied entheses. Figure 1 explains how to acquire the image for the study of both entheses.Figure 1.Scheme to obtain the analysis area of ​​the Achilles (A) and epicondylar enthesis (B). The enthesic area was considered from the most distal point of the enthesis to the point of emergence or separation of the tendon from the cortex. Case A corresponds to a healthy control. Case B corresponds to a patient with psoriatic arthritis and severe epicondylar pain.ResultsWe included image records of 9 patients with axial spondyloarthritis (34.1 ±9.5 years, 7 males) and 10 with psoriatic arthritis (34.0 ±5.7 years, 8 males) with asymmetric involvement of 15 Achilles and 10 epicondylar entheses (LEI/MASES/ SPARCC).The distribution of the MGI, DGI, and MoGI variables was normal for both enthesis, contralateral patient controls and healthy subjects controls.In the records with Achilles involvement, the MGI on the affected side was 76.69 ±25.14, the DGI 24 ±3.48, and the MoGI 76.6 ±30.1. In the non-painful Achilles enthesis, the measurements were 83.99 ±27.29, 22.72 ±3.14 and 78.7 ±32.5, respectively. Statistically significant differences were detected in all pairwise comparisons (t -9.522, P=.000; t=3.393, P=.004; and t=-2.247, P=.041, respectively).In the epicondylar involvement records, the MGI on the affected side was 75.3 ±14.71, the DGI 25.93 ±2.52, and the MoGI 72.7 ±15.4. In the non-painful Achilles enthesis the measurements were 82.57 ±16.06, 25.7 ±2.99 and 74.9 ±16.5, respectively. Statistically significant differences between MGI and MoGI were identified (t=-9.849, P=.000; and t=-2.752, P=.022, respectively).The MIN ratio of the affected side and healthy side among patients was 1.18 ±0.047 while the MIN ratio between the dominant and non-dominant sides of the healthy athletes was 0.91 ±0.022 (t=25.81, P=.000).ConclusionComputer analysis of entheses static images correctly discriminates symptomatic from non-symptomatic entheses in early ultrasound studies. Even in non-symptomatic entheses, it detects differences compared to healthy subjects.This discriminatory capacity could be useful in updating ultrasound enthesis count indices. It may also be interesting as a criterion for evaluating changes over time. As a weakness, it is worth highlighting the potential confounding effect that enthesophytoses could have in this type of analysis if specific demarcations of the area studied are not made.Disclosure of InterestsCarlos Guillén-Astete Speakers bureau: Novartis, Janssen, Abbvie, Grunenthal, UCB, Gebro, Paid instructor for: Roche, Novartis, Janssen, Esteve, Menarini, Consultant of: Janssen, Novartis, Roche, Grant/research support from: Pfizer, Grunenthal, Gebro, Novartis, África Andreu-Suárez: None declared, Marina Tortosa-Cabañas: None declared, Miguel Martinez Aznar: None declared