Purpose: Magnetic resonance (MR) imaging offers the highest sensitivity for detecting bone necrosis. We evaluated osteonecrosis in rabbit models by calculating the percentage of fat to (fat + water) [F/(F+W)] on MR spectroscopy (MRS) and compared MR spectroscopy and imaging findings with corresponding histological results. Methods: To model the natural course of articular osteonecrosis, we removed the fourth tarsal bone in 45 rabbits, froze it for 5 min in liquid nitrogen to produce complete cellular necrosis, and then replaced the bone into the knee joint. We performed Carr-Purcell- Meiboom-Gill proton spectroscopic imaging to assess necrotic bone at 3 days and one, 2, 3, 4, 8, 12, 16, and 20 weeks after osteonecrosis and calculated the percentage of F/ (F+W) of each bone. We also performed conventional T1- and T2-weighted imaging and compared all data to histological findings to analyze the natural course of necrosis. Results: T1-weighted MR imaging demonstrated obvious low signal intensity at 2 to 8 weeks and recovery at 12 to 20 weeks, whereas T2-weighted imaging demonstrated inconsistent intensities throughout the period. The postoperative percentage of F/(F+W) measured using line scan MRS decreased to 8.88% at 3 weeks, 6.22% at 8 weeks, and 34.40% at 20 weeks results that were mostly consistent with MR imaging findings. Histological findings demonstrated complete absence of osteocyte nuclei and loss of osteoid-osteogenesis at 3 to 8 weeks. Recovery of bone marrow was identified as an increase in the area of fat after 12 weeks. Conclusion: Osteonecrosis delineated by T1-weighted MR imaging demonstrated fat content in the bone marrow that correlated with histology. The present MRS modality can be used to calculate the percentage of F/(F+W) of osteonecrosis to enable objective assessment of recovery and quantification of osteonecrosis to provide a numerical value for osteonecrosis.
ASJC Scopus subject areas