The relationship between specific tissue lesions and pain severity in persons with knee osteoarthritis

Abstract

Pain is the most common symptom in knee osteoarthritis (OA), a leading cause of chronic disability, and a major source of the disability attributable to OA in general. Pain severity in knee OA is variable, ranging from barely perceptible to immobilizing. The knee lesions that contribute to pain severity have received little attention. To examine whether worse pathology of specific knee tissues - i.e. cartilage, bone (attrition, cysts, bone marrow lesions, and osteophytes), menisci (tears and subluxation), ligaments, and synovium (synovitis/effusion) - is associated with more severe knee pain. One hundred and forty-three individuals were recruited from the community with primary (idiopathic) knee OA, with definite tibiofemoral osteophytes in at least one knee, and at least some difficulty with knee-requiring activity. Knee magnetic resonance (MR) images were acquired using coronal T1-weighted spin-echo (SE), sagittal fat-suppressed dual-echo turbo SE, and axial and coronal fat-suppressed, T1-weighted 3D-fast low angle shot (FLASH) sequences. The whole-organ magnetic resonance imaging (MRI) scoring (WORMS) method was used to score knee tissue status. Since summing tissue scores across the entire joint, including regions free of disease, may dilute the ability to detect a true relationship between that tissue and pain severity, we used the score from the worst compartment (i.e. with the poorest cartilage morphology) as our primary approach. Knee pain severity was measured using knee-specific, 100 mm visual analogue scales. In analyses to evaluate the relationship between knee pain severity and lesion score, median quantile regression was used, adjusting for age and body mass index (BMI), in which a 95% CI excluding 0 is significant. The increase in median pain from median quantile regression, adjusting for age and BMI, was significant for bone attrition (1.91, 95% confidence interval (CI) 0.68, 3.13), bone marrow lesions (3.72, 95% CI 1.76, 5.68), meniscal tears (1.99, 95% CI 0.60, 3.38), and grade 2 or 3 synovitis/effusion vs grade 0 (9.82, 95% CI 0.38, 19.27). The relationship with pain severity was of borderline significance for osteophytes and cartilage morphology and was not significant for bone cysts or meniscal subluxation. Ligament tears were too infrequent for meaningful analysis. When compared to the pain severity in knees with high scores for both bone attrition and bone marrow lesions (median pain severity 40 mm), knees with high attrition alone (30 mm) were not significantly different, but knees with high bone marrow lesion without high attrition scores (15 mm) were significantly less painful. In persons with knee OA, knee pain severity was associated with subarticular bone attrition, bone marrow lesions, synovitis/effusion, and meniscal tears. The contribution of bone marrow lesions to pain severity appeared to require the presence of bone attrition.