Associations between biomarkers of bone and cartilage turnover, gender, pain categories and radiographic severity in knee osteoarthritis

BACKGROUND: Excessive cartilage degradation is a known characteristic of osteoarthritis (OA). Biochemical markers, such as uCTX-II, have been shown to be associated with disease severity, yet the tissue origin of CTX-II has been disputed. This analysis investigates the association between OA knee joints at different radiographic stages and pain categories with levels of uCTX-II and biomarkers of bone resorption and formation. METHODS: Baseline data of two randomised clinical trials (NCT00486434 and NCT00704847) in patients with radiographic OA and presence of pain were analysed post hoc. A subgroup with available urine samples and evaluable radiographs for both knees (N = 1241) was analysed. Urine CTX-I, urine CTX-II and serum osteocalcin were analysed for associations with combined Kellgren-Lawrence (KL) scores, gender and pain for both knees to assess the contribution of joints at different stages. RESULTS: Pain, BMI, age, gender and KL grade were all significantly associated with uCTX-II. The association between pain and CTX-II appeared to be driven by weight-bearing pain. The level of uCTX-II incrementally increased with higher radiographic severity of each knee. Levels of bone markers CTX-I and osteocalcin were both significantly associated with BMI and gender, but neither were associated with radiographic severity. Biomarker levels between male or female groups of identical KL scores were found to be higher in females compared to males in some but not all KL score groups. CONCLUSIONS: These results indicate that levels of uCTX-II are independently associated with radiographic severity of OA and pain intensity. CTX-II was associated with weight-bearing pain, but not non-weight-bearing pain, independent of co-variates. Bilateral OA knee joints appear to contribute to uCTX-II levels in an incremental manner according to radiographic severity of single joints. The data suggest that biomarker differences between genders should be taken into account when evaluating these markers in the context of structural features of OA.