Introduction

Previous MRI classifications for meniscus degeneration were designed to identify tears within intra-meniscal high signal lesions, rather than determining the extent of tissue degeneration. The aim of this study was to devise a novel MRI classification for meniscus degeneration in osteoarthritic knees based on histological degeneration. We first aimed to characterize intra-meniscal high signal lesions seen in MRI. Secondly, we aimed to devise an MRI classification system by correlating degenerative MRI parameters with histology. Finally, we aimed to compare our classification with the previous classification in terms of histological degeneration grade and other osteoarthritis related parameters.

Methods

Human menisci were harvested during knee replacement surgery in osteoarthritic patients who had received knee MRIs within 3 months prior to surgery. For comparison of imaging and tissue samples, three axial tissue slices (anterior horn, mid-body, and posterior horn) were chosen and matched with respective MRI images from sagittal and coronal views. Characterization of intra-meniscal high signal lesions was done with microscopy, biochemical, and biomechanical tests. Meniscus degeneration was graded with our classification and Stoller’s classification on MRI, and with Pauli’s grading on histology, each by two separate readers. Both MRI classifications were correlated with histological degeneration grade. Both systems were further correlated with articular cartilage degeneration MRI grade and radiological medial joint space width in separate osteoarthritic subjects (n=100) without history of surgery. Intra-class correlation coefficients for intra and inter-observer agreement were analyzed for each grading system. Additional statistical analysis was done using Kruskal-Wallis test and Spearman’s rank correlation.

Results

Total of 202 menisci (119 medial, 83 lateral) samples from 139 patients were used, totaling 506 axial segments analyzed with both MRI and histology. Intra-meniscal high signals showed local decrease in collagen and increase in glycosaminoglycans, as well as biomechanical insufficiency according to the lesion size. A novel MRI classification was devised based on correlation between tissue degeneration and the following MRI parameters; intra-meniscal high signal intensity size, degree of surface integrity, and linear signals. Intra-class correlation coefficients for intra and inter-observer agreement for the novel classification were 0.916 and 0.859, respectively. Our novel classification showed better correlation with histological grade (r=0.819, p<0.001) compared to the previous classification (r=0.680, p<0.001). Significant differences in tissue degeneration scores were observed between mild (G1) and moderate grading (G2) in our system only (2.86±1.85 vs 7.12±2.74, p<0.01). Our novel classification also showed higher correlation coefficients than the previous classification with regards to articular cartilage degeneration (r=0.655 vs 0.418, p<0.001) and medial joint space width (r=0.584 vs 0.564, p<0.001).

Conclusion

A novel MRI classification for meniscus degeneration based on histology was devised, resulting in improved representation of tissue degeneration and degree of osteoarthritis. Our novel MRI classification may be a useful tool to determine and grade the tissue quality of meniscus prior to current therapies as well as future regenerative strategies targeting meniscus degeneration.