Osteoarthritis and Cartilage Vol. 14, Supplement B S27 Fig. 1. A: normal, B: pre-arthritic, C: early-arthritic hip. Conclusions: This preliminary study showed that T2 time as- sessment using 3.0T MR imaging is promising for detailed qual- itative assessment for both acetabular and femoral cartilages in vivo. The findings of increased T2 time at the weight-bearing zones in the early-arthritic hips were consistent with previous studies concerning knee osteoarthritis. However, pre-arthritic hips showed adverse tendency of T2 time distribution in the aceatabular cartilage, with low T2 values in the superficial carti- lage area. This finding may be related with intrinsic difference of long-term stress distribution and resultant cartilage matrix in the aceatabular cartilage between the normal and dysplastic hips, and further extensive studies are needed. A20 PRECISION OF 3 TESLA MR IMAGING OF CARTILAGE MORPHOLOGY IN A MULTICENTER CLINICAL TRIAL F. Eckstein 1 , C.H. Charles 2 , M. Hudelmaier 1 , R. Buck 3 , B. Wyman 3 , M.-P. Hellio Le Graverand 3 1 Institute of Anatomy, PMU, Salzburg, Austria, 2 DIAL, Duke University, Durham, NC, 3 Pfizer GRD, Ann Arbor, MI Purpose: Quantitative MR imaging of cartilage morphology has become an important and powerful tool in OA research. Recent studies (at single sites) have indicated that measurements at 3.0 Tesla (T) are more reproducible than at 1.5 T, but the precision errors in large, multicenter clinical trials with equipment from multiple vendors have not yet been evaluated. Methods: 157 female participants, aged ≥ 40 years, were re- cruited at 7 clinical centers, using Siemens Magnetom Trio and GE Signa Excite (short and long bore) magnets. Conventional standing ap knee radiographs were obtained to determine the Kellgren Lawrence grade (KLG). 79 subjects had a BMI ≤ 28, no symptoms, and a KLG = 0 bilaterally. 78 subjects had a BMI ≥ 30, symptoms in one knee, and mild to moderate ra- diographic OA (16 = KLG 1, 35 = KLG 2, 27 = KLG 3). Two coronal MR image acquisitions were obtained in each participant with a water excitation T1-weighted spoiled gradient recalled sequence at 3T with a 1.0 x 0.31 x 0.31 mm 3 resolution. One scan was acquired at baseline and one 3 months later. Car- tilage volume (VC), mean cartilage thickness (ThC.Me), area of cartilage surface (AC), and total area of subchondral bone (tAB) were quantified in the medial tibia (MT), lateral tibia (LT), medial femoral condyle (cMF) and lateral femoral condyle (cLF) using proprietary software (Chondrometrics GmbH, Germany). Segmentation of paired data sets was performed by the same (of 7) technicians in the same session, with blinding to time point. The segmentation was quality controlled by one expert. From the two consecutive measurements the root mean square (RMS) coefficient of variation (CV%) was computed. Results: The RMS CV% values for measurements of cartilage volume (VC) were 2.4% in MT, 2.6% in LT, 3.0% in cMF, 3.4% in cLF, and 2.8% when averaging the CV% values in the 4 cartilage plates. The average CV% was 2.5% in the non-OA subjects (KLG 0), 3.3% in KLG 1, 2.6% in KLG 2, and 3.4% in KLG 3 participants. When only considering healthy participants (KLG 0), the RMS CV% was 2.7% on the Siemens Trio (n = 36), 2.4% on the GE short bore (n = 22), and 2.2% on the GE long bore magnet (n = 21). Precision errors for cartilage thickness were < 3% in single plates across all subjects (average across the 4 plates = 2.5%), and < 1.5% for surfaces (average across all plates = 1.4% for AC and = 1.2% for tAB). The trends in the subgroups were similar to those for VC. Systematic differences in cartilage thickness (3 months versus baseline) ranged from -0.5% (MT) to + 0.9% in LT in healthy participants, and from -0.3% (cMF) to + 1.4% (cLF) in KLG 2 and 3 subjects. Conclusions: The findings show that 3.0 Tesla MR imaging pro- vides highly reproducible measurements of cartilage morphology in multicenter clinical trials with equipment from different vendors, and a high degree of stability over short- to intermediate-term time intervals. If used appropriately, the methodology can thus be efficiently applied to large-scale, multicenter clinical trials. A21 SAFETY AND EFFICACY OF NEAR INFRARED LIGHT FOR CARTILAGE RE-GROWTH OF DEEP OSTEO- CHONDRAL DEFETC IN SHEEP AS ANIMAL MODEL D. Fortuna 1 , G. Rossi 2 , B. Grigolo 3 , R. Buda 4 , A. Zati 5 , S. Giannini 4 , T. Bilotta 5 , P. Mondardini 6 , A. Crovace 7 , L. Masotti 1 1 El.En. S.p.A., Calenzano (Florence), Italy, 2 University of Camerino, Veterinary Science Department, Matelica, Italy, 3 Rizzoli Orthopedic Institute, Laboratory of Immunology and Genetics, Bologna, Italy, 4 Rizzoli Orthopedic Institute, VI Divisione di Chirurgia ortopedico-traumatologica, Bologna, Italy, 5 Rizzoli Orthopedic Institute, Servizio di Recupero e rieducazione funzionale, Bologna, Italy, 6 Sport Medicine Institute - CONI, Bologna, Italy, 7 University of Bari, Dept. of Emergency and Organ Transplantation, Valenzano (Bari), Italy Purpose: To assess the safety and efficacy of near-infrared light from a power laser (HILT) to induce light-activated articular chondrocytes proliferation in vivo. High Intensity Laser Therapy (HILT) is a new non-invasive method to stimulate tissue repair in vivo. Methods: The study was approved by the Italian institutional animal use and care committee at the Italian Ministry of Research and Education. At time zero (T0) monolateral cartilage full thickness surgical defects reaching the subchondral bone was created in ten female adult sheep, weighing 60±5 kg, by drilling (Ø = 14 mm) of the femoral trochlear sulcus of patella. All subjects (n=10) received antibiotics prophylaxis for 6 days postoperatively. One week later surgery all subjects were divided into two groups: the HILT group received 15 treatments (Tx) of laser while un- treated group didn’t received laser. Each subject of the HILT group received 15 Tx in 3 weeks (5 Tx/week) up to T1 (30 d. after T0). Each treated subject received the same amount of energy (2,500 J) with the following setting: 19 kW/cm 2 of peak intensity and 2.5 J/cm 2 of fluence with a spot size of Ø = 10 mm. We used a prototype power laser designed by El.En. S.p.A. (Italy). To asses the effects of HILT we compared histological and immunohistochemical (IHC) findings of the samples collected from the lesion at 30 (T1), 45 (T2) 90 (T3), 120 (T4) and 180 (T5) days later the induction of surgical defect. In particular the samples were stained with H&E, Safranin-O, Alcian Pas and Herovici’s stain. Mono and polyclonal antibodies specific to IL-1β, MMP-9, TIMP-2, COMP, ILGF-1 and collagen types I-II were employed for IHC evaluations. Results: The macroscopic observation of the defect areas has showed a progressive re-growth of a new tissue from the edges to the central area of the lesions in Treated group.