Background Cartilage and synovium are two central tissues remodeled in osteoarthritis (OA). Recent data has shown that type II collagen degradation (CTX-II) is a biomarker of cartilage and subchondral turnover. CRPM and C3M are biomarkers of synovial inflammation and turnover.
Objectives To identify different subtypes of OA related to progression by measuring biomarkers of cartilage degradation and synovial inflammation.
Methods Fasting serum (n=767) and urine (n=620) samples from knee OA patients, of the placebo arms of two phase III RCTs (NCT00486434 and NCT00704847) were analyzed for Serum C3M, CRPM, (type III collagen and C-reactive protein neo-epitope) and creatinine-corrected urinary CTX-II (type II collagen C-terminal telopeptide). The relationship between the biomarkers and the 5 baseline-reported WOMAC pain subscale questions, using the sum of target and non-target knee, were analyzed: A; during walking on a flat surface, B; using stairs (up or down), C; at night while in bed, D; sitting or lying and E; while standing. Joint space width was analyzed at baseline and at 24 months. Log-transformed data were analyzed by Spearman's rho and by multiple regression analysis, where WOMAC or radiographic scores were the dependent variable and biomarkers, age, gender and BMI as covariates. Cut-offs was based on reference value and difference between the high/low groups was analyzed by Mann-Whitney.
Results CTX-II was significantly correlated with WOMAC with all 5 sub-questions (p<0.01), while level of C3M was correlated with question B and E (p<0.05). CTX-II remained significantly associated with all sub-questions after adjustment: A) r=0.20, p<0.0001; B) r=0.16, p=0.0005; C) r=0.11, p=0.017; D) r=0.13, p=0.0046; and E) r=0.17, p=0.0002. But not C3M: B) r=0.08, p=0.063 and E) r=0.08, p=0.069. CTX-II was significantly correlated JSW in both knees (r=-0.17, p<0.0001), whereas C3M was correlated to joint space narrowing of the target knee (r=0.11, p=0.013). CRPM was not associated with the clinical measures. Patients were separated into 2 groups; 1) low CRPM and low C3M, and 2) either high CRPM or high C3M, or with both high. 16% of the patients were found group 2. WOMAC subscale question E was significantly higher in this group (p=0.019). Group 1 patients (low in both CRPM and C3M) progressed structurally more group 1 patients (p=0.046).
Conclusions CTX-II was associated with all types of pain, especially with pain associated with walking on a flat surface, indicating that CTX-II is related to pain from continuous skeletal load, supported by the strong correlation with JSW. C3M, from the synovium, was weakly correlated pain by using stairs and standing still. C3M was correlated to progression indicating that patients with certain levels of C3M had a higher likelihood for progression. We found that the two inflammatory markers could facilitate patient segregation. These biomarkers may assist in identification of the inflammation driven OA phenotype, which needs to be treated different than non-inflammatory OA.
Acknowledgements The CSMC021C2301/2 investigators
Disclosure of Interest A. Bay-Jensen Shareholder of: Nordic Bioscience, Grant/research support from: EC FP7 programs, Employee of: Nordic Bioscience, I. Byrjalsen Employee of: Nordic Bioscience, K. Musa Employee of: Nordic Bioscience, A. Bihlet Shareholder of: Nordic Bioscience, Employee of: Nordic Bioscience, P. Alexandersen: None declared, J. Andersen: None declared, B. Riis: None declared, C. Christiansen Shareholder of: Nordic Bioscience, M. Karsdal Shareholder of: Nordic Bioscience, Grant/research support from: EC FP7 programs, Employee of: Nordic Bioscience