Article Text

PDF
OP0030 Identifying radiographic phenotypes of early knee osteoarthritis using separate quantitative features might improve patient selection for more targeted treatment
  1. P.M. Welsing1,2,
  2. M.B. Kinds1,3,
  3. A. Marijnissen1,
  4. M. Viergever3,
  5. P.J. Emans4,
  6. F.P. Lafeber1
  1. 1Rheumatology & Clin. Immunology
  2. 2Julius Center for Health Sciences & Primary Care
  3. 3Image Sciences Institute, University Medical Center Utrecht, Utrecht
  4. 4Ortthopaedic Surgery, Maastricht University Medical Center, Maastricht, Netherlands

Abstract

Background Osteoarthritis (OA) is a degenerative joint disease characterized by pain and functional disability. The expression of OA varies significantly between individuals and over time, implying the existence of different phenotypes.

Objectives This study aims at identifying phenotypes of progression of radiographic knee OA and to describe their radiographic and clinical characteristics.

Methods In individuals with early knee OA from the Cohort Hip & Cohort Knee (CHECK), baseline, two-year, and five-year follow-up radiographs were evaluated. Separate radiographic OA parameters were quantitatively measured by Knee Images Digital Analysis (KIDA). To identify phenotypes of radiographic knee OA progression, hierarchical clustering was performed using the KIDA measurements of participants with complete data at T0, T2y, and T5y (n=417 of 1002). The phenotypes were compared for development of joint space width (JSW), varus angle, osteophyte area, eminence height, bone density, and for clinical characteristics at T0. Additionally, logistic regression analysis evaluated whether baseline radiographic features predicted to which phenotype an individual belonged.

Results Overall, the radiographic features showed OA progression during follow-up. Based on the development, five clusters were identified that were interpreted as “severe“ (n=17; most progression of all radiographic features) or “no“ (n=108) progression, “early“ (n=110; progression of all features specifically between T0 and T2y) or “late“ (n=69; progression of all features specifically between T2y and T5y) progression, or specific involvement of “bone density“ (n=113). Clinical characteristics at T0 were not evidently different between the clusters, and WOMAC scores were only slightly lower in the “no“ cluster than in the other clusters. In the evaluation of predictors for the different clusters, the area under the curve (AUC) improved when radiographic features were added to basic demographic and clinical variables. Kellgren & Lawrence grading was not a significant predictor for any of the phenotypes. The predictors for “early“, “late“, and “no“ progression phenotypes generally had an opposite effect than the predictors for the “severe“ and “bone density“ phenotypes. Larger medial JSW, varus angle, osteophyte area, eminence height, and bone density at T0 were associated with “severe“ and “bone density“ progression. The “bone density“ model had AUC=0.91. Smaller eminence height and bone density were associated with “early“ and “late“ progression (AUC=0.79, and 0.76 respectively). Larger varus angle and smaller osteophyte area were associated with “no“ progression (AUC=0.72).

Conclusions This is the first study to identify specific phenotypes of radiographic knee OA progression in individuals with early OA complaints. Phenotypes represented the level (severe vs. no) and phase of progression (early vs. late), and the involvement of a specific feature (bone density). Baseline radiographic features could predict the phenotypes. The phenotypes might represent relevant subgroups for the evaluation of treatment strategies in clinical trials, and with that drive the discovery of more targeted treatment.

Disclosure of Interest None Declared

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.