COMPARISON OF CONCURRENT VALIDITY OF CHANGE IN 3D JSW ON WEIGHT-BEARING CT VS. 2D JSW ON RADIOGRAPHS FOR CARTILAGE STRUCTURAL WORSENING OVER 24 MONTHS

Abstract

Radiography is the current standard for assessing osseous structural knee OA, but it is a 2D image of a 3D structure and superimposition of bone impairs ability to sensitively detect changes. Weight-bearing CT (WBCT) allows improved visualization of OA features, shows greater correlation with central medial cartilage damage on MRI and has demonstrated excellent test-retest reliability, suggesting that WBCT may address shortcomings of radiographs. To determine the concurrent validity of change in 3D JSW on WBCT vs. change in 2D JSW on radiographs with radiographic progression defined by semiquantitative grading using of cartilage damage worsening on MRI (MOAKS) over 24 months. Radiographs, WBCT and MRI were obtained from 259 participants from the Multicenter Osteoarthritis Study (MOST) at baseline and 24-month follow-up (including one knee per participant). Radiographs were evaluated for change in 2D JSW at predetermined locations in the medial and lateral compartments. Change in 3D JSW on WBCT assessed JSW in each subregion (Figure). MRI scans were assessed using MRI Osteoarthritis of the Knee Score for cartilage morphology (MOAKS-CM). Concurrent validity was assessed by associations of JSW change on radiographs and WBCT with structural worsening of MOAKS-CM using logistic regression. Area under the receiver operating characteristic (AUROC) curve was calculated to compare the concurrent validity of 2D JSW vs. 3D JSW by compartment and subregion. A “no exposure” model was reported using only demographic variables. Exposure variables in each model contained either 2D JSWx or 3D JSW. Two separate parameter types were utilized for 3D JSW exposure, maximum change values and change in mean values. In addition to the main exposure JSW parameter, each model adjusted for demographic covariates (age, sex and BMI). JSWx=.225 was used when comparing to medial 3D JSW subregional change and JSWx=.750 was used when comparing to lateral 3D JSW subregional change. Mean age was 63.2 ± 9.0 years, BMI was 28.2±4.9 kg/m 2 , 57% were women and knees had KL Grade 2-3 at baseline for 60/259. For change in medial compartment MOAKS-CM, 3D Mean (AUROC=0.681) and Maximum (AUROC=0.682) were not superior to 2D JSW (AUROC=0.713). For change in lateral compartment MOAKS-CM, 3D Mean (AUROC=0.757) and Maximum (AUROC=0.730) showed no statistical difference with 2D JSW (AUROC=0.695). In subregional analyses, there was no significant difference between the association of change in MOAKS-CM and change in either 3D JSW (Mean, Maximum) or 2D JSW. In this study of the association of knee OA structural progression assessed by cartilage morphology on MRI, change in 3D JSW parameters measured on WBCT and 2D JSW on radiographs, WBCT did not demonstrate superior concurrent validity with worsening MOAKS-CM compared to change in 2D JSW assessed on knee radiographs. Thus, diagnostic value of this implementation of 3D JSW remained similar to that for 2D JSW. National Institutes of Health, University of Kansas (R01AR071648), University of Iowa (U01AG18832) and University of California-San Francisco (U01AG19069). AG and FWR are shareholders of BICL, LLC. AG is consultant to Pfizer, MerckSerono, TissueGene, Novartis, Regeneron and AstraZeneca. FWR is consultant to Grünenthal. NS is a consultant for Integra BioLife, Trice Medical and Pacira Biosciences. Others authors have no conflicts of interest to disclose. The authors would like to thank participants and staff of the MOST study. CORRESPONDENCE ADDRESS: nsegal@kumc.edu