Magnetic Resonance Imaging of Cartilage Contact and Bound Water in ACL-Deficient and ACL Reconstructed Knees

Abstract

Objectives:Osteoarthritis (OA) is common following ACL-reconstructive (ACLR) surgery (6). The cause of early OA is not understood, but theories have focused on osteochondral damage at the time of injury (2) and abnormal joint mechanics following surgical repair (7). In this study, we investigate the inter-relationship of cartilage mechanics and biomarkers of OA in both ACL-deficient (ACLD) and ACLR knees. Our approach employs a novel dynamic MR sequence to measure joint mechanics (3) and the recently developed mcDESPOT to assess regional variations in water bound to proteoglycan (PG) (5). We hypothesize that bound water will be diminished in the cartilage of ACLD knees and, after surgery, will continue to adapt in a manner that reflects altered cartilage loading. This abstract presents initial observations on a cross-section of healthy, ACLD and ACLR knees.Methods:The dominant knees of 8 healthy controls, ACLD knees of 5 patients and ACLR knees of 8 patients were imaged in a 3 T MRI scanner (Table). Controls had no history of pain, injury, or surgery to their knee. Patients had no additional ligament injury and no meniscal damage. ACLD subjects were imaged prior to reconstructive surgery. Femoral and tibial cartilage were segmented from MR images and cartilage thickness was calculated. The mcDESPOT sequence provided a fraction map of water bound to PG (Fpg). Subjects flexed their knee against an inertial load at 0.5 Hz, while a SPGR-VIPR sequence continuously acquired volumetric data. Kinematics were obtained using model tracking of the dynamic images (3). Cartilage was registered to the bone segments for all frames, and contact patterns were characterized by the proximity between surfaces. Spatial representations of tibial cartilage contact, thickness and Fpg were co-registered for each subject.Results:Our initial images suggest lower Fpg values in ACLD knees, primarily on the posterior-lateral tibia. This is also observed in ACLR knees, with additional evidence of diminished Fpg on the weight-bearing medial tibia. Contact patterns were altered in both groups. ACLD tended to exhibit increased contact on the posterior lateral tibia and anterior contact in the medial tibia. Contact differences in the ACLR knees were more subtle, but tended to show a posterior-lateral shift on the medial tibia when compared to control knees (Figure). These initial observations support our hypotheses that cartilage composition may be altered in ACLD knees and continues to adapt following ACLR. While contact in ACLR knees appears to be restored close to the healthy condition, we observed a residual shift in the medial plateau. Interestingly, this shift corresponds with a decrease in PG content not observed in ACLD knees. Loss of PG occurs early in OA, prior to any morphological changes (1, 4). Decreased PG content was also observed in ACLD and ACLR knees in the posterio-lateral tibia, consistent with observations of edema and cartilage damage following an ACL injury (2).Conclusion:Initial observations of our novel dynamic and quantitative MR images suggests altered cartilage composition due to both injury and abnormal mechanics following surgical repair.