s / Osteoarthritis and Cartilage 22 (2014) S57–S489 S412 high sensitive C-reactive protein (hsCRP), matrix metalloproteinase (MMP) mediated degradation of CRP (CRPM) and MMP-mediated collagen type I, II, and III degradation (C1M, C2M, and C3M, respectively). Statistical analyses were performed in IBM SPSS Statistics, version 19. One-way ANOVA was used to compare serological biomarkers and PPT measures and Fisher’s Least Significant Difference (LSD) post hoc test was applied. Results: Group B (BLOKS 1⁄4 2,3) had decreased PPTs compared to controls (p< 0.05, figure 1) and increased levels of C1M (p1⁄4 0.001), C2M (p 1⁄4 0.001) and hsCRP (p 1⁄4 0.023), but a lower level of C3M (p 1⁄4 0.033) compared to controls (Figure 2). Group A had increased levels of C1M (p 1⁄4 0.039), C2M (p1⁄4 0.001) and CRPM (p1⁄4 0.028) and lower level of C3M (p 1⁄4 0.019), but no difference was found in PPTs compared to controls. Dividing Group A into BLOKS 1⁄4 0, (n 1⁄4 16) and BLOKS 1⁄4 1 (n 1⁄4 22) revealed that patients with BLOKS 1⁄4 0 had decreased PPTs compared to patients with BLOKS 1⁄4 1 (p < 0.05, Figure 1). Interestingly, the level of PPTs of the BLOKS1⁄4 0 group was at the level of the group with BLOKS 2. There was no significant difference in the biomarker level in the differentiated Group A in any of the investigated markers (Figure 2). Conclusions: Patients with widespread synovitis have decreased PPTs and significantly increased levels of bone and cartilage degradation and inflammation. However, these patients had a decrease in synovial membrane degradation compared to controls. Dividing patients with localized synovitis into two groups based on the synovitis score identified subgroups with significantly different PPTs, but the same level of the biomarkers studied. This study illustrate that patients with different levels of sensitization have different profiles of tissue turnover and each constitute a specific group of severe KOA patients. 741 A CLINICAL TOOL FOR THE PREDICTION OF PATIENT-REPORTED OUTCOMES AFTER KNEE REPLACEMENT SURGERY M.T. Sanchez-Santos y, A. Judge yz, R.N. Batra y, D. Murray y, A. Price y, A.D. Liddle y, M.K. Javaid yz, C. Cooper yz, N.K. Arden yz. yNIHR Musculoskeletal BioMed. Res. Unit, Nuffield Dept. of Orthopaedics, Rheumatology and Musculoskeletal Sci., Univ. of Oxford, Oxford, United Kingdom; zMRC Lifecourse Epidemiology Unit, Univ. of Southampton, Southampton Gen. Hosp., Southampton, United Kingdom Purpose: To develop a clinical risk prediction tool to predict patientreported outcomes after undergoing primary total knee replacement (TKR) surgery. Methods:We used data from 1,649 patients from the Knee Arthroplasty Trial (KAT) who received primary TKR across 34 centres in the United Kingdom between July 1999 and January 2003. The main outcome was Oxford Knee Score (OKS) 12-months after surgery. A wide range of preoperative predictor variables including patient characteristics, clinical and surgical factors, were considered. Bootstrap backward linear regression analysis was used to identify risk factors of outcome. Results: Presence of a fixed flexion deformity, presence of varus/valgus deformity (compared with no deformity), and an absent preoperative ACL (compared with intact ACL) were associated with better OKS at 12 months. The clinical factors associated with worse post-operative OKS were worse ASA grade, presence of other conditions affecting mobility and previous knee surgery. Lower pre-operative OKS, living in poor areas, high BMI and worse SF-12 mental health score were also associated with worse outcome. Discrimination and calibration statistics were good. Patients’ characteristics explained 16.6% of the variability in outcome, when clinical variables were included 18.9% of the variance of outcome was explained, reaching 20.2% when surgical variables were added. Conclusions: The clinical risk tool will help to inform patients and clinicians regarding expectations of the outcome after knee replacement surgery, and may also provide useful information to support shared decision-making. External validation of the tool, to confirm its performance and validity, will be undertaken in a new prospective cohort of patients. 742 ASSOCIATION BETWEEN SYNOVITIS AND DEPRESSION IN PATIENTS WITH KNEE OSTEOARTHRITIS: A CROSS-SECTIONAL STUDY Y. Shimura y, H. Kurosawa y, M. Tsuchiya y, H. Kaneko z, L. Liu z, R. Sadatsuki z, A. Yusup z, S. Hada z, M. Kinoshita z, Y. Makino y, M. Sawa y, Y. Iwase y, K. Kaneko z, M. Ishijima z. y Juntendo Tokyo Koto Geriatric Med. Ctr., Tokyo, Japan; z Juntendo Univ. Graduate Sch. of Med.,