MRI-based screening for structural definition of eligibility in clinical DMOAD trials: Rapid OsteoArthritis MRI Eligibility Score (ROAMES). Issue 1 (January 2020)

• Record Type: Journal Article
• Authors: Roemer, F.W.; Collins, J.; Kwoh, C.K.; Hannon, M.J.; Neogi, T.; Felson, D.T.; Hunter, D.J.; Lynch, J.A.; Guermazi, A.
• Is Part Of: Osteoarthritis and cartilage. Volume 28:Issue 1(2020)
• Page Start: 71
• Page End: 81
• Publication Date: 2020-01
• Subjects: MRI -- Osteoarthritis -- Eligibility -- Clinical trial; Osteoarthritis -- Periodicals; Cartilage -- Periodicals; Osteoarthritis -- Periodicals; Cartilage -- Periodicals; Arthrose -- Périodiques; Articulations -- Maladies -- Périodiques; 616.7223005
• DOI: 10.1016/j.joca.2019.08.005
• Languages: English
• ISSNs: 1063-4584

Abstract

Summary: Purpose: Our aim was to introduce a simplified MRI instrument, Rapid OsteoArthritis MRI Eligibility Score (ROAMES), for defining structural eligibility of patients for inclusion in disease-modifying osteoarthritis drug trials using a tri-compartmental anatomic approach that enables stratification of knees into different structural phenotypes and includes diagnoses of exclusion. We also aimed to define overlap between phenotypes and determine reliability. Methods: 50 knees from the Foundation for National Institutes of Health Osteoarthritis Biomarkers study, a nested case-control study within the Osteoarthritis Initiative, were selected within pre-defined definitions of phenotypes as either inflammatory, subchondral bone, meniscus/cartilage, atrophic or hypertrophic. A focused scoring instrument was developed covering cartilage, meniscal damage, inflammation and osteophytes. Diagnoses of exclusion were meniscal root tears, osteonecrosis, subchondral insufficiency fracture, tumors, malignant marrow infiltration and acute traumatic changes. Reliability was determined using weighted kappa statistics. Descriptive statistics were used for determining concordance between the a priori phenotypic definition and ROAMES and overlap between phenotypes. Results: ROAMES identified 43 of 50 (86%) pre-defined phenotypes correctly. Of the 50 participants, 27 (54%) had no additional phenotypes other than the pre-defined phenotype. 18 (36%) had one and 5 (10%) had two additional phenotypes. None had three or four additional phenotypes. All features of ROAMES showed almost perfect agreement. One case with osteonecrosis and one with a tumor were detected. Conclusions: ROAMES is able to screen and stratify potentially eligible knees into different structural phenotypes and record relevant diagnoses of exclusion. Reliability of the instrument showed almost perfect agreement.