Magnetic resonance imaging characterization of osteochondral defect repair in a goat model at 8 T

This study was performed to non-invasively visualize and characterize osteochondral (OC) repair in ex vivo goat stifles using an 8 T magnetic resonance imaging (MRI) scanner and to compare the MR morphology with images obtained from 1.5 T, gross morphology and histology.

Mature, neutered male goats were assigned to an 8-week (n = 4) or 16-week (n = 4) study period. Two cylindrical OC defects (7 mm diameter, full cartilage thickness and 1mm into subchondral bone) were surgically created in the right stifle: one in the medial femoral condyle (MFC) and the other in the trochlear groove (TG). The implant matrices (non-woven or foam) were secured in the defect using a bottom anchored fixation device (FD). The contralateral left stifles served as time zero controls. At the day of necropsy, implants were placed at both defect sites (MFC and TG) on the normal left stifles. Following necropsy, the ex vivo goat stifles (intact and encapsulated) were disarticulated. Within 24 h postnecropsy, MR scans of the stifles along the mid-sagittal plane of the OC defect were acquired at 8 T and 1.5 T. MR relaxation times, T1 and T2, were measured at the region of repair tissue (RT) and adjacent native cartilage. Immediately after MR imaging, the stifles were dissected, grossly examined, and a sagittal OC block corresponding to the MR region of interest was prepared for formalin fixation.

The high-resolution MR images enabled visualization of cartilage and bone integrity surrounding the implant as well as delineating the margins of RT/implant matrix and the FD. On spin echo sequence, the RT variably appeared as high, intermediate or low MR signal intensity; whereas, the FD always appeared as low signal intensity. In general, the MR signal intensity of 8-week RT was slightly higher compared to 16-week RT; however, there was no difference in RT morphology of stifles implanted with the non-woven matrix or foam matrix. Subchondral sclerosis appeared as low signal intensity. The 8 T MR images showed better delineation of the stifle tissues compared to the images acquired at 1.5 T. The T2 relaxation time of the RT appears to indicate (inconclusive due to small number of samples) a slight variation in the RT type between 8 weeks and 16 weeks. At both study times, the defects grossly appeared whitish to reddish but did not have the characteristic hyaline appearance typical of articular cartilage (AC). The gross appearance of the MFC and TG RT differed, which was predominantly mottled and recessed with fissuring of adjacent native AC in the MFC. Histologically, the RT of both 8-week and 16-week postsurgical defects predominantly comprised fibrovascular connective tissue with only few samples showing the presence of fibrocartilaginous and/or hypertrophic chondrocytes within the defect RT at 8 weeks. Also, compared to 8-week, the 16-week RT appeared to be more fibrotic.

Using 8 T scanner, high-resolution MR images of ex vivo encapsulated goat stifles confirmed the capability of high-field MR imaging to distinguish the defect RT from the FD and adjacent joint tissues. The extent of OC repair and adjacent bone lesions (at 8 weeks and 16 weeks) observed in the MR images compared well with those observed on the corresponding histological sections.