Quantitative bone marrow lesion size in osteoarthritic knees correlates with cartilage damage and predicts longitudinal cartilage loss

• Published in: BMC musculoskeletal disorders 2011-09, Vol.12 (1), p.217-217, Article 217
• Main Authors: Driban, Jeffrey B, Lo, Grace H, Lee, Ji Yeon, Ward, Robert J, Miller, Eric, Pang, Jincheng, Price, Lori Lyn, McAlindon, Timothy E
• Format: Article
• Language: English
• Subjects: Bone marrow; Bone Marrow - pathology; Bone Marrow Diseases - etiology; Bone Marrow Diseases - pathology; Care and treatment; Cartilage; Cartilage, Articular - pathology; Complications and side effects; Cross-Sectional Studies; Diagnosis; Disease Progression; Female; Femur - diagnostic imaging; Femur - pathology; Humans; Hyaline Cartilage - pathology; Injuries; Knee; Knee Joint - pathology; Knee Joint - physiopathology; Longitudinal Studies; Magnetic Resonance Imaging; Male; Middle Aged; Musculoskeletal diseases; Osteoarthritis; Osteoarthritis, Knee - complications; Osteoarthritis, Knee - pathology; Pain; Pain Measurement; Physiological aspects; Radiography; Rheumatology; Severity of Illness Index; Studies; Tibia - diagnostic imaging; Tibia - pathology
• ISSN: 1471-2474; 1471-2474
• DOI: 10.1186/1471-2474-12-217

Bone marrow lesions (BMLs), common osteoarthritis-related magnetic resonance imaging findings, are associated with osteoarthritis progression and pain. However, there are no articles describing the use of 3-dimensional quantitative assessments to explore the longitudinal relationship between BMLs and hyaline cartilage loss. The purpose of this study was to assess the cross-sectional and longitudinal descriptive characteristics of BMLs with a simple measurement of approximate BML volume, and describe the cross-sectional and longitudinal relationships between BML size and the extent of hyaline cartilage damage. 107 participants with baseline and 24-month follow-up magnetic resonance images from a clinical trial were included with symptomatic knee osteoarthritis. An 'index' compartment was identified for each knee defined as the tibiofemoral compartment with greater disease severity. Subsequently, each knee was evaluated in four regions: index femur, index tibia, non-index femur, and non-index tibia. Approximate BML volume, the product of three linear measurements, was calculated for each BML within a region. Cartilage parameters in the index tibia and femur were measured based on manual segmentation. BML volume changes by region were: index femur (median [95% confidence interval of the median]) 0.1 cm3 (-0.5 to 0.9 cm3), index tibia 0.5 cm3 (-0.3 to 1.7 cm3), non-index femur 0.4 cm3 (-0.2 to 1.6 cm3), and non-index tibia 0.2 cm3 (-0.1 to 1.2 cm3). Among 44 knees with full thickness cartilage loss, baseline tibia BML volume correlated with baseline tibia full thickness cartilage lesion area (r = 0.63, p< 0.002) and baseline femur BML volume with longitudinal change in femoral full thickness cartilage lesion area (r = 0.48 p< 0.002). Many regions had no or small longitudinal changes in approximate BML volume but some knees experienced large changes. Baseline BML size was associated to longitudinal changes in area of full thickness cartilage loss.