Quantitative Measurements of Cartilage, Osteophyte, and Bone Marrow Lesion volume in Knee Osteoarthritis

Schaefer L.F., Yin M., Sury, M., Jamieson S., Collins J., Smith S., Duryea J.

Brigham and Women’s Hospital, Harvard Medical School, Boston, MA

Background/Purpose: Clinical trials and other large studies of knee osteoarthritis (OA) require efficient and objective methods to assess disease changes. Magnetic resonance imaging (MRI) offers superb soft tissue contrast and is ideal for visualizing many structures associated with knee OA. We have developed the MRI Osteoarthritis Software Score (MOSS), a software-based method that assesses volume change of osteophytes, cartilage, and bone marrow lesions (BMLs) for knee OA. The goal of our study is to assess all 600 subjects of the OA Biomarkers Consortium FNIH Study. This will further validate the method and demonstrate its suitability for use in highly-powered studies and clinical trials of knee OA.

Methods: The OA Biomarkers Consortium FNIH Study is a case-control study of knee OA progression nested within the Osteoarthritis Initiative (OAI). The study cohort was divided into four subgroups based on radiographic and pain progression: Group 1: radiographic and pain progression (n=194) Group 2: radiographic-only progression (n=103) Group 3: pain-only progression (n=103) Group 4: no radiographic or pain progression (n=200). Double echo steady state (DESS) pulse sequences were used for cartilage and osteophytes, and turbo spin echo (TSE) for BMLs. Baseline to 24 month volume change was analyzed as a predictor of case-control status in univariate models using logistic regression. This highly-powered longitudinal study represents the result of 3,600 individual measurements of OA structures on MRI scans.

Results:   Change in volume of cartilage and osteophytes were strongly associated with radiographic progression: decreasing cartilage volume was associated with an increased odds of being a case, while increasing osteophyte and BML volume were associated with an increased odds of being a case. The association of cartilage and osteophyte volume change was lower but significant for pain progression. BMLs showed marginally significant associations with radiographic and pain progression. All three measurements (cartilage, osteophytes, and BMLs) can be performed in a total time of less than 30 minutes per knee, split roughly evenly between a research assistant and an experienced reader. Table 1: Results  

Conclusion: We have demonstrated that measurements of cartilage, BML, and osteophyte volume using the MOSS method have clinical validity in a case-control study. The method is efficient and ideal for current and future large studies and clinical trials of knee OA requiring assessment of many thousands of MRI scans.