Background X-ray and magnetic resonance imaging (MRI) demonstrated associations between osteophyte severity and current pain in knee osteoarthritis (OA) patients. Persistent pain over 5 years was also associated with x-ray osteophyte severity. A biomarker to identify knee OA subjects with persistent pain may reduce placebo response rates in clinical trials.
Objectives To determine whether the total knee osteophyte score (TKOS), as measured by MRI Osteoarthritis Knee Score (MOAKS), is a biomarker of current and persistent pain severity in knee OA subjects.
Methods Knee OA subjects from the Foundation for the National Institutes of Health Biomarker Consortium were longitudinally assessed over 4 years and categorized as pain and x-ray progressors (n=194), x-ray–only progressors (n=103), pain-only progressors (n=103), and non-progressors (n=200). Knee osteophyte severity was scored at baseline by MOAKS at 12 positions across the knee and summed to obtain the TKOS. TKOS was summarized as means and SDs in the 4 progression groups; each progression group TKOS was compared with the non-progression group using Wilcoxon rank sum tests (Table 1). The longitudinal Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain scores were plotted and tested with a linear mixed model, by high and low TKOS status, defined as above and below the TKOS median of 6 (Figure 1). Subjects with persistent pain were defined based on WOMAC pain scores at 4 out of 5 time points higher than the thresholds listed in Table 2; at each threshold, mean differences in TKOS were tested using analysis of variance among those with and without persistent pain.
Results The distribution of baseline TKOS differed by progression group, with the highest TKOS among pain and x-ray progressors (Table 1). In all 4 progression groups, baseline WOMAC pain scores were higher in subjects with high (>6) versus low (≤6) baseline TKOS status (P=0.0001, 0.0383, 0.0035, and 0.0466, respectively). The difference in WOMAC pain scores between TKOS high and low subgroups was constant over time (Figure 1; solid curves above and parallel to dashed curves with TKOS main effect always significant [P<0.01] in all 4 progression groups), but the TKOS×Time interaction term in the longitudinal mixed model was not statistically significant, indicating that the pain difference between the TKOS high and low subgroups did not change over time. TKOS was highly associated with persistent pain, with TKOS higher at all thresholds (Table 2).
Conclusions TKOS is a candidate biomarker for current and persistent pain in knee OA patients and may be a predictive biomarker for reduced placebo response rates in clinical trials.
Acknowledgements AbbVie funded the analysis; participated in analysis, design and interpretation and in abstract writing, review, and approval; and funded writing support by M. Theisen of CPS.
Disclosure of Interest S. Feng Shareholder of: AbbVie, Employee of: AbbVie, M. Levesque Shareholder of: AbbVie, Employee of: AbbVie