Article Text

PDF
SAT0175 Dynamic Automated Synovial Imaging (DASI) for Differentiating between Rheumatoid Arthritis and Simil-Rheumatoid Psoriatic Arthritis
  1. B. Raffeiner1,2,
  2. E. Grisan3,
  3. L. Bernardi1,
  4. C. Botsios1,
  5. F. Ometto1,
  6. A. Coran4,
  7. V. Beltrame4,
  8. L. Ciprian1,
  9. R. Stramare4,
  10. L. Punzi1
  1. 1Rheumatology Unit, Department of Medicine, University of Padova, Padova
  2. 2Rheumatology Unit, Department of Medicine, General Hospital of Bolzano, Bolzano
  3. 3Department of Information Engineering
  4. 4Radiology, Department of Medicine, University of Padova, Padova, Italy

Abstract

Background Although both diseases are characterized by specific features the differentiation between rheumatoid arthritis (RA) and simil-rheumatoid psoriatic arthritis (srPsA) is extremely difficult except by hard-to-gain biopsy specimens (1). On contrast enhanced magnetic resonance imaging (CE-MRI) some extra-articular manifestations may direct diagnosis, but synovitis could not be discriminated between RA and srPsA in these studies (2-4). Contrast-enhanced ultrasound (CEUS) using “real” intravascular agents is believed to allow more accurate study of synovial vascularization.

Objectives To determine the feasibility to discriminate between RA and srPsA using CEUS derived flow parameters by ad hoc developed software program for analysis of synovial vascularization.

Methods 64 outclinic patients with polyarthritis of hands, 32 with RA and 32 with srPsA, were recruited. The most active joint was chosen for CEUS examination using a US device (Mylab70, Esaote) equipped with Contrast tuned Imaging (CnTI, Esaote), and as contrast agent sulfur hexafluoride microbubbles (SonoVue; Bracco International). Both the anatomical B-mode image and the CnTI cineloop video were digitally stored for subsequent software analysis. Image analysis was performed firstly applying a semi-automatic detection of synovial boundaries (5). Then, the contrast time-activity curve of all pixels belonging to the synovial and perisynovial region was analyzed fitting a gamma curve f(t) = A(t − t0)a $times$ e(t −t0)/b on the data. The statistics summarizing the distribution of the estimated kinetics parameters in the synovial and in the perisynovial tissue were computed and their difference between the two groups (RA and srPsA) analyzed, so to study the existence of different vascularization patterns. Finally, a supervised classifier (random forest) was trained to classify each patient through its CEUS-derived parameters, validating the classifier diagnostic power using a leave-one-out strategy. To further increase diagnostic power data about DAS28, CRP, ESR and autoantibodies were added.

Results Vascularization pattern constituted of 40 flow parameters discriminated effectively RA from srPsA. Accuracy was 0.93 during training and 0.83 during test phase. Adding rheumatoid factor (RF) and anti-CCP increased diagnostic accuracy to 0.99 in training and 0.93 in test phase decreasing needed flow parameters to 28, whereas DAS28, CRP and ESR did not.

Conclusions The Dynamic Automated Synovial Imaging (DASI) is actually the only imaging method that accurately discriminates RA from srPsA, especially in the presence of RF and anti-CCP data.

References

  1. Kruithof E. Arthritis Res Ther 2005;7:569-80.

  2. Jevtic V. Handchir Mikrochir Plast Chir 2012;44:163-170.

  3. Cimmino MA. J Rheum 2012;39(89):43-8.

  4. Schoellnast H. AJR 2006;187:351-7.

  5. Veronese E. Med Eng Phys 2013; 35, 188–194.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.3779

Statistics from Altmetric.com

Request permissions

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.