Purpose: This prospective study was performed to investigate whether 3-Tesla magnetic resonance imaging (MRI) provides an accurate assessment of the articular cartilage in clinical practice.
Methods: Forty patients with persistent knee pain and suspected cartilage lesions underwent 3-T MRI shortly before arthroscopy with the following sequences: axial/coronal/sagittal proton density-weighted turbo spin echo with spectral fat suppression, axial/sagittal 3-dimensional T1-weighted gradient echo with selective water excitation, and axial T2-weighted gradient echo (Intera 3.0T; Philips Medical Systems, Best, The Netherlands). Knee cartilage surfaces were divided into 6 regions; lesions detected on MRI were classified into stages I to IV and compared with the arthroscopic grading.
Results: For the 240 cartilage surfaces evaluated, the sensitivities, specificities, positive predictive values, and negative predictive values of 3-T MRI were 74%, 95%, 74%, and 95%, respectively, for the detection of grade IV lesions; 63%, 90%, 60%, and 91%, respectively, for grade III lesions; 62%, 90%, 57%, and 92%, respectively, for grade II lesions; and 29%, 95%, 39%, and 92%, respectively, for grade I lesions.
Conclusions: In these preliminary clinical studies 3-T MRI provided convincing visualization of the hyaline cartilage with comparatively good diagnostic values. Nonetheless, it must be pointed out that the positive predictive values were low for all grades of lesions. Thus, when 3-T MRI suggests a cartilage defect, the probability that the arthroscopic finding corresponds exactly to the MRI result is between 39% and 74%. Therefore, the value of arthroscopy for a detailed assessment and grading of a cartilage disorder with regard to definitive planning of a therapeutic procedure cannot be replaced by 3-T MRI.
Level of evidence: Level I, testing of previously developed diagnostic criteria in a series of consecutive patients with universally applied gold standard.