Background Osteoarthritis (OA) is a heterogeneous disorder, which has been described in several different population studies. However there is a significant difference between patient populations recruited in OA interventional clinical trials and epidemiological studies. While epidemiological studies focus on incidence, prevalence, risk and cause of OA, clinical trials aim at enrolling patients with certain characteristics based on defined guidelines on clinical outcome measures. There is therefore a need for in-depth, post-hoc analysis of clinical studies in OA.
Objectives The aim of the analysis was to investigate the associations between JSW, KL-score, pain and JSN (joint space narrowing), as well as BMI (Body mass index), by combining data from two phase III studies (N=2,206).
Methods This is a post-hoc analysis of two randomized, double-blind, multi-center, placebo-controlled trials (CSMC021C2301 and CSMC021C2302), evaluating the efficacy and safety of oral salmon calcitonin in subject with painful knee OA, enrolling 1,176 and 1,030 patients, respectively. The analysis includes baseline data on KL-score, pain and function metrics from the WOMAC questionnaire, as well as demographics. The main inclusion criteria for the target knee were: ACR OA criteria, KL-score 2 or 3 and JSW of ≥2.0 mm, and significant WOMAC pain. Both knees were analyzed. Moreover, analysis of the two-year data on JSN was included for the placebo arm
Results The mean JSW was comparable in knees of KL-0 and -1 and significantly decreased with increasing KL-2, -3 and -4 (p<0.01) when including all knees in the analysis. JSW was significantly lower in the non-target knees as compared to the target knees (2.48±0.08mm vs. 3.11±0.06mm, p<0.0001). When analyzing the two-year progression data only knees from the placebo group was included. The mean JSN was 0.318±0.018mm, where the mean JSN for the non-target knee was 0.279±0.025mm and 0.356±0.026 for the target knee (p<0.05). While the KL score of the target knee was a more important clinical descriptor for progression than pain, pain was the more important descriptor for the contra-lateral knee. Clearly different levels of progression were observed in relation to KL score and pain, in which the third pain quartile (Q3), but not the fourth quartile (Q4), progressed significantly faster than the first and second pain quartiles. Similar observations were made for BMI and Pain, in relation to JSN, namely that patients with Q3 pain were progressing faster than those in Q4.
Conclusions These data from potentially the largest clinical trial dataset in OA to date clearly describe significant associations between KL-score, JSW, pain and BMI in patients with symptomatic knee OA. Consequently, this dataset is ideally suited for identification of different phenotypes of OA, and biomarkers associated with those. Patients with symptomatic OA at baseline progressed significantly faster than patients with asymptomatic disease, however with important variations that need accounting for when designing clinical trials, such as relations to pain, BMI and JSN. These results suggest that disease phenotypes rather than disease status are responsible for disease progression and may act as background material for designing future clinical trials in OA.
Disclosure of Interest M. Karsdal Employee of: Nordic Bioscience, A.-C. Bay-Jensen: None declared, A. Bihlet Employee of: Nordic Bioscience, I. Byrjalsen Employee of: Nordic Bioscience, P. Alexandersen Employee of: Nordic Bioscience, C. Christiansen Employee of: Nordic Bioscience