Subchondral bone abnormalities in knee osteoarthritis

Osteoarthritis (OA) is the most common type of arthritis, with prevalence estimates expected to increase dramatically worldwide due to ageing and increasingly obese populations. The knee is the most commonly affected joint, leading to pain, loss of function and disability. Magnetic resonance imaging (MRI)-detected bone marrow lesions (BMLs) and osteophytes (OPs) are types of subchondral bone abnormalities whose aetiology and predictive value are uncertain. The aims of this thesis are to investigate associations between local cartilage morphology, systemic inflammatory cytokines and knee BMLs, and the predictive value of MRI-detected OPs on knee OA structural and symptomatic changes. Two data sources were used in this thesis. The first was a population-based study of older adults aged 50-80 years (mean age: 62 years; 51% female). Participants were randomly selected from the electoral roll in Southern Tasmania (population 229, 000) using sex-stratified random sampling. Follow-up measurements were performed at about 2.6 years later and again for questionnaire data at about 5.0 years later. MRI scans of the right knees were conducted at baseline and first follow-up. Knee cartilage defects, cartilage volume, tibial bone area, BMLs, effusion synovitis, infrapatellar fat pad and OPs were measured or scored based on MRI images. A standing anteroposterior semi-flexed view right knee with 15° of fixed knee flexion was performed at baseline with joint space narrowing (JSN) and radiographic OPs scored according to the Osteoarthritis Research Society International (OARSI) atlas. Knee pain was assessed using the Western Ontario McMaster Osteoarthritis Index (WOMAC) at all phases. The second was a randomized, multi-centre, placebo-controlled and double-blinded clinical trial that was designed to evaluate the effect of vitamin D supplementation on knee OA. Eligible participants were aged 50 to 79 years who had symptomatic knee OA (according to American College of Rheumatology criteria) for at least 6 months, and had pain of 20 to 80 mm on a 100-mm visual analog scale (VAS). Additionally, participants' serum 25OHD levels were >12.5 nmol/L and < 60 nmol/L. Knee cartilage defects, cartilage volume, BMLs and OPs were measured or scored based on MRI images at baseline and after 24 months. WOMAC knee pain, serum levels of inflammatory cytokines were assessed at baseline and after 24 months using enzyme-linked immunosorbent assay. This thesis encompasses five studies. In the first study, the natural history of patellofemoral joint (PFJ) BMLs over 2.6 years was described and associations between PFJ BMLs, knee pain and knee cartilage morphology were evaluated in a population-based sample of older adults. 109 (27%) of 406 participants who completed follow-up had PFJ BMLs at baseline. Of these participants, 49 (45%) of these participants' PFJ BMLs persisted in the same grade, 26 (24%) increased in grade, and 34 (31%) decreased in grade. Change in PFJ BMLs over 2.6 years was deleteriously associated with change of knee pain when going up/down stairs over 5 years. Baseline PFJ cartilage morphology predicted increases in PFJ BMLs over 2.6 years. In the second study, cross-sectional and longitudinal associations between serum high sensitivity C reactive protein (hs-CRP), knee BMLs and knee pain were investigated in a sample of knee OA patients. In these patients, serum hs-CRP is associated with knee BML scores and pain both cross-sectionally and longitudinally, suggesting inflammation is linked with BMLs and their associated pain. The third study described cross-sectional and longitudinal associations between serum levels of IL-17A, IL-17F, IL-23, IL-6 and knee BMLs in patients with knee OA. Baseline IL-6 were significantly associated with total knee BMLs as well as increased knee BML scores in both females and males. Baseline IL-17F and IL-23 predicted increased BML scores in females only. In the fourth study, MRI-detected OPs were measured and the cross-sectional and longitudinal association with knee structural abnormalities and knee pain were examined in older adults. Baseline MRI-detected OPs were significantly, independently and site-specifically associated with increases in cartilage defects, BMLs and loss of cartilage volume over 2.6 years. Medial tibiofemoral and total OP scores were dose-dependently associated with total knee pain change over 2.6 and 5 years but these became non-significant after further adjustment for cartilage defects and BMLs. In the fifth study, the prevalence of MRI-OPs detected only by MRI but not by standard x-ray was described, and the longitudinal associations with knee pain and structural changes were investigated in a population-based older adult sample. The prevalence of MRI-OPs was about 75%. Compared with participants without any OPs, participants with MRI-OPs had greater cartilage volume loss and increased cartilage defects and BMLs. MRI-OPs and established-OPs both predicted progression of knee structural abnormalities, but the associations for MRI-OPs were not as prominent as those for established-OPs. This suggests MRI-OPs may have a role to play in knee early-stage osteoarthritic progression. In conclusion, this series of studies indicate that BMLs are not static and changes in BMLs are clinically relevant. Systemic inflammation is important in the aetiology of BMLs in OA. MRI-detected OPs can predict knee OA structural and symptomatic progressions. MRI-OP detected only by MRI but not by standard x-ray can also lead to OA progression, suggesting MRI-OP, which largely represent early OP formation, can also serve as a biomarker to predict knee structural progression over time.