BASELINE MRI AND RADIOGRAPHS AS PREDICTORS OF THE CLINICAL EFFECTIVENESS OF WEIGHT LOSS IN OBESE PATIENTS WITH KNEE OSTEOARTHRITIS. A HALF-YEAR RANDOMIZED CONTROLLED TRIAL.

2. [THU0356]   BASELINE MRI AND RADIOGRAPHS AS PREDICTORS OF THE CLINICAL EFFECTIVENESS OF WEIGHT LOSS IN OBESE PATIENTS WITH KNEE OSTEOARTHRITIS. A HALF-YEAR RANDOMIZED CONTROLLED TRIAL.

H. R. Gudbergsen*¹, M. P. Boesen¹, H. Bliddal², R. Christensen³, A. Astrup4
¹Imaging Laboratory, ²Head of Research, ³Musculoskeletal Statistics Unit, The Parker Institute, Frederiksberg Hospital, ⁴Department of Human Nutrition, Faculty of Life Science University of Copenhagen, Frederiksberg, Denmark

Background: Obese people with symptoms of osteoarthritis (OA) tend to progress in regards of symptoms, and studies have revealed a relationship between weight loss and level of improvement in symptoms (1).

Objectives: The objective of the study was to examine the opportunity to predict the long-term outcome of weight loss in obese patients with knee pain by assessing baseline radiographs and low-field (0.2T) MRIs.

Methods: This study is a sequel to a previously published trial (1), including a new cohort of patients with an add-on of baseline low-field (0.2T) MRI. Patients were recruited from the Department of Rheumatology, Frederiksberg Hospital, Denmark. Eligibility criteria were: age >18 years; primary knee OA according to the ACR-guidelines - no suspicion of secondary OA; no other substantial illness; BMI>28 kg/m2; motivation for weight loss. Subjects were randomly assigned (using minimization based on sex, BMI, and age) to either the intervention- (low-energy diet; LED) or control- (assumed hypo-energetic) group for an initial 8 weeks. MRIs were scored regarding cartilage abnormalities; synovitis, effusion and bone marrow oedema and radiographs were scored by the Kellgren-Lawrence Score. These variables were examined as independent predictors of changes in body weight and OA symptoms assessed by the WOMAC index. Baseline DEXA-scan was obtained in nt=29 and follow-up WOMAC in nt=27. Caregivers and patients were not blinded. The outcome assessor and statistician were both blinded.

Results: We randomized and analyzed the ITT population based on 30 patients (15 patients in each group). Patients in the LED and control group did not differ in regards of baseline variables. The LED and control group lost a mean (SE) of 15.9 (1.2) kg and -0.4 (0.8) kg respectively with a mean difference (MD) being 16.3 kg (95%CI:13.4-19.2;P<0.0001).

The WOMAC pain index improved in the LED but not in the control group with a MD of - 44.5 mm (95%CI: -114.1 to 25.2 mm;P=0.2). The WOMAC function index showed a significant difference favoring LED, with a group MD of - 266.0 mm (95%CI:-468.9 to –63.1 mm; P=0.01). We were not able to show any significant correlation in the Spearman correlation analysis when examining imaging variables as predictors of changes in WOMAC indexes (r<0.20 and P>0.56). No significant adverse events were reported.

Conclusion: Obese patients with knee OA achieved a significant weight reduction upon diet intervention in this new group of patients using the same approach as presented previously (1). The improvement in symptoms was achieved disregarding baseline OA status assessed by low-field (0.2T) MRIs and radiographs.

References: (1) Christensen R., Osteoarthritis Cartilage. 2005 Jan; 13(1): 20-7

Disclosure of Interest: None Declared.