Diagnosis of chondral lesions of the knee joint:: can MRI replace arthroscopy?: A prospective study

The role of magnetic resonance tomography (MRI) for the diagnosis of chondral lesions of the knee joint is still unclear. The sensitivity of the method ranges from 15% to 96%. The scope of our daily experiences showed that there were considerable deviations between the tomographical and arthoscopical results, which vary from the results of experimental studies. Therefore we have conducted a prospective study to investigate the question of how MRI can replace arthroscopy (ASC) in the diagnosis of cartilage damages in the scope of daily routine. All 195 patients included in this study received a magnetic resonance tomography followed by an arthroscopy. A clear diagnosis of supposition had to be determined before the magnetic resonance tomography. The patients were divided into 3 Groups. Group A ( n =86) received a standard Military Hospital Ulm (MH) MRI - sagittal STIR TSE and PD TSE, coronal and transversal T2 FFE (TR=660 ms, TE=18 ms, FA=30degrees, 512 matrix). In addition, one sub-Group, AK (n =21) was examined with a special cartilage sequence of the cartilage fs T1 W FFE. Neither patients in Group AK nor in Group A as a whole received any contrast medium. Group B (n =88) was examined with an alternate MRI protocol (Radiological Joint Practice, Neu-Ulm - sagittal T1 SE, T2 SE and T2 FLASH (TR=608 ms, TE=18 ms, FA=20degrees, 256 matrix), coronal PD fs), employing gadolinium as a contrast medium. 156 cartilage lesions were found arthroscopically. In Group A the sensitivity was 33%, the specificity 99%, and the positive and negative prediction values 75% and 98% respectively. Group B reached a sensitivity of 53% and a specificity of 98%. The positive prediction value was 48% and the negative was 98%. Group AK showed a sensitivity of 38% and specificity of 98%; the positive and negative prediction values came to 50% and 97% respectively. In conclusion, our results indicate that the MRI examination techniques recommended in the literature at present are not able to replace the ASC for the diagnosis of cartilage damages of the knee joint. In view of the high specificity (97%-99%) and the high negative prediction value (97%-98%), MRI is suitable for the exclusion of cartilage lesions. For a negative MRI associated with a cartilage injury, a cautious attitude towards an operative cartilage treatment is therefore justified. Because the MRI can not replace the ASC for diagnostic of cartilage damage, the ASC still has to be seen as the method of choice for the evaluation of cartilage damage.