PREDICTION OF TOTAL KNEE REPLACEMENT IN A 6-MONTH MULTICENTRE CLINICAL TRIAL WITH CHONDROITIN SULFATE IN KNEE OSTEOARTHRITIS: RESULTS FROM A 4-YEAR OBSERVATION

JP Raynauld1, J Martel-Pelletier1, M Dorais2, F Abram3, B Haraoui1, D Choquette1, F Morin4, L Bessette5, A Beaulieu6, JP Pelletier1. 1University of Montreal Hospital Research Centre (CRCHUM), Osteoarthritis Research Unit, Canada, 2StatSciences Inc., StatSciences Inc., Canada, 3ArthroLab Inc., Imaging Research & Development, Canada, 4Centre de recherche musculo-squelettique, Centre de recherche musculo-squelettique, Canada, 5Groupe de recherche en Rhumatologie et Maladies Osseuses, Groupe de recherche en Rhumatologie et Maladies Osseuses, Canada, 6Université Laval, Faculty of Medicine, Canada

Purpose: To identify predictive factors for the incidence of total knee replacement (TKR) during long-term follow-up of knee OA patients who formerly received treatment with chondroitin sulfate (CS) or placebo in a multicentre trial using clinical and quantitative magnetic resonance imaging (qMRI) data.

Methods: Knee OA patients participating in a previous 6-month randomized, double-blind controlled trial evaluating the impact of CS (400 mg b.i.d.; Condrosan®, Bioibérica, Spain,) vs. placebo who had serial MRI acquisitions of the symptomatic knee were recently contacted to evaluate retrospectively the incidence of TKR of the study knee. A sub-group of patients (n=70) who had taken all the study medication and had all clinical and MRI evaluations (according-to-protocol [ATP]) were selected for this post-hoc retrospective analysis. Of this cohort 51 patients were reachable for TKR incidence. The assessment was done blindly to treatment allocation with a standardized phone interview. Univariate logistic regressions were done to find baseline predictors of TKR. The Kaplan-Meier survival analysis was also used to compare the cumulative incidences of TKR over time between the two treatment groups. Cox regression analysis was used to find predictors using survival of not having a TKR over time as an outcome.

Results: The patients’ mean age was 62.9 years, 61% were female and the average BMI was 30.6 kg∕m2. A total of 7 (6 target knees and 1 contralateral) TKRs (13.7%) were performed on this sub-population in the timeframe of 3-4 years after completion of the original study. Interestingly, there were more TKRs performed within the placebo group (n=5) than the CS group (n=2) (71% vs. 29%, p=0.15, logistic regression). The predictors of long-term TKRs for the target knee were investigated by comparing the patients who had TKR (n=6) of the target knee to those who did not (n=45), using data at baseline or the change at 1 year. At baseline, the strongest predictors of TKR were WOMAC pain (p=0.02, logistic regression), stiffness (p=0.01) and function (p=0.04), bone marrow lesions of the medial tibial plateau (p=0.03), and C reactive protein level (p=0.03). Changes at 1 year in medial cartilage volume (p=0.05) and WOMAC stiffness (p=0.01) also predicted the occurrence of TKR.

Conclusion: These data demonstrate that, from a knee OA clinical trial, it is possible to predict a “hard” outcome such as TKR using clinical and qMRI data. Moreover, CS appeared to protect cartilage volume loss and improved clinical parameters.