Tibiofemoral joint osteoarthritis one year after anterior cruciate ligament reconstruction: a magnetic resonance imaging evaluation

Abstract

s / Osteoarthritis and Cartilage 21 (2013) S63–S312 S196 371 TIBIOFEMORAL JOINT OSTEOARTHRITIS ONE YEAR AFTER ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION: A MAGNETIC RESONANCE IMAGING EVALUATION J. van Leeuwen y, N. Beck y, M. van Middelkoop y, A. Culvenor z, S. Bierma-Zeinstra y, E. Oei y, K. Crossley z. y ErasmusMC, Rotterdam, The Netherlands; zUniv. of Queensland, Brisbane, Australia Purpose: Radiographic tibiofemoral joint (TFJ) osteoarthritis (OA) is known to be common more than five years after anterior cruciate ligament injury and reconstruction (ACLR). Magnetic resonance (MR) imaging enables evaluation of early, pre-radiographic features of OA, which may be important for individuals at high risk of OA. Specifically, MR imaging may be useful to assess the onset of OA at an earlier stage after ACLR. A recently proposed tool to assess OA features and prevalence of TFJ OA from MR images is the MR Imaging Osteoarthritis Knee Score (MOAKS). This study aimed to (i) establish the reliability of the MOAKS in novice readers; (ii) describe the prevalence of OA features and TFJ OA one year after ACLR; and (iii) investigate the relationship between meniscal or cartilage damage at the time of ACLR and the presence of meniscal or cartilage damage one year later. Methods: One hundred and eighteen participants (65% male; age 31 9 years; BMI 26.1 4.0kg$m-2; median time from injury to surgery 3 months) who were approximately one year post ACLR with a single bundle hamstring tendon autograft were recruited from two orthopaedic surgery clinics. High field MR imaging at 3 Tesla was performed on all participants and the MOAKS was used by two novice readers (medical students), following an intensive training regimen by a musculoskeletal radiologist, to score specific features of OA and to evaluate the prevalence of TFJ OA based on published criteria. Interrater reliability was calculated using the kappa statistic. Binary logistic regression analysis was used to study the relationship between meniscal or cartilage damage at time of surgery and meniscal or cartilage damage on MR imaging one year later. Results: Most measures of inter-rater reliability from the MOAKS were substantial (78% of kappa values> 0.6). Osteoarthritis features appeared slightly more prevalent in the medial compartment than in the lateral compartment. Prevalence of bone marrow lesions in the medial and lateral compartment were 62% and 48%, respectively, with articular cartilage lesions (21% and 15%), osteophytes (30% and 26%) and meniscal damage (32% medial and 30% lateral) also frequently observed in both compartments. Of the 118 participants 17 (14%) met the criteria for TFJ OA, following the MOAKS scoring system. Cartilage damage one year post ACLR was predicted by cartilage damage present at time of surgery. Meniscal damage one year post ACLR observed from MR images was predicted by meniscal damage seen at time of surgery, age and gender. Conclusions: Inter-rater agreement between the two novice readers using the MOAKS instrument to evaluate OA features in the TFJ one year after ACLR was high. These results add to other data recently published on the MOAKS and indicate that the MOAKS may be used by different researchers with varying levels of experience and at the early stages of OA disease. Although TFJ OA and OA features identified on MR imaging are evident as early as one year after ACLR, the stability of these features, particularly BMLs, which may resolve or progress over time, is not well understood. Further research investigating whether these features are risk factors for disease progression is therefore warranted. Table 1 T1r, SNR, NRMS error values compared to 4 point full thickness standard.