Objective: The purpose of this study was to assess the accuracy of fast spin-echo MR imaging for depicting the severity of articular cartilage abnormalities in patients with osteoarthritis.
Subjects and methods: Twenty-three subjects (10 volunteers less than 35 years old and 13 patients with proved, symptomatic, idiopathic osteoarthritis of the knee of 6 months' to 10 years' clinical duration) underwent fast spin-echo MR imaging of the knee. Two observers graded each articular surface using a five-category scale that took into account abnormalities in the signal intensity of cartilage as well as thickness and contour. The 13 patients also underwent arthroscopic evaluation (as part of a separate protocol) in which cartilage abnormalities were graded by using a similar five-category grading scale, without the graders knowing the results of MR imaging. Articular cartilage was assumed to be normal in the volunteers.
Results: One hundred thirty-seven joint surfaces were graded; one surface was obscured by artifact and was excluded. The Spearman rank linear correlation between arthroscopic and MR grading was highly significant (p < .002) for each of the six articular regions evaluated. The MR and arthroscopic grades were the same in 93 (68%) of 137 joint surfaces, they were the same or differed by one grade in 123 surfaces (90%), and they were the same or differed by one or two grades in 129 surfaces (94%).
Conclusion: Our results suggest that fat-presaturated fast spin-echo MR imaging depicts the severity of articular cartilage abnormalities in osteoarthritis with reasonable accuracy, as compared with arthroscopic evaluation as the standard of reference.