The value and limitation of gadopentetate-enhanced magnetic resonance imaging in detecting the condition of anterior cruciate ligament in rheumatoid knee: comparative study with histology.

Abstract

The aim of this study was to elucidate the utility and limitation of gadopentetate (Gd)-enhanced MRI as a method for evaluating the anterior cruciate ligament (ACL) in the rheumatoid arthritis (RA) knee, using both surgical macro findings and histological findings to ascertain the pathological condition of the affected knee. Thirty-six knees of 25 RA patients were studied in this study. Four imaging protocols were employed: protocol A, T1-weighted and T2-weighted sagittal images; protocol B, T1-weighted sagittal image, after infusion of Gd-DTPA (0.2 mmol/kg, i.v.); protocol C, T1-weighted angled coronal image, parallel to the ACL; and protocol D, T1-weighted angled coronal image, parallel to the ACL, after infusion of Gd-DTPA. Sagittal image was determined as previously described. Angle coronal image was newly determined as coronal image parallel to the ACL. Surgical and MRI findings of the ACL were classified into four types: Type I (normal group) indicated that the thickness of the ACL was almost normal, adequate tension was maintained (surgical findings),and the ACL had thick and a more complex appearance with a homogeneous signal intensity and well-defined borders (MRI findings). Type II (degenerated group): the ACL had degenerated and tension was reduced (surgical findings), and the ACL had thin and a more complex appearance with a less homogeneous signal intensity and less well-defined borders. This appearance was more evident on Type II than Type I (MRI findings). Type III (ruptured group): the parenchyma of the ACL remained but lacked continuity (surgical findings), and the ACL appeared as partial lack of low signal intensity (MRI findings). Type IV (absent group): the parenchyma of the ACL was practically absent (surgical findings), and the ACL appeared as complete lack of signal low signal intensity (MRI findings). The concordance rate between surgical and MRI findings was investigated. Moreover, we investigated the extent to which histological changes of the ACL could be discriminated using MRI. In RA knees, the overall concordance rate between surgical and MRI findings was 41.7% under imaging protocol A. The overall rate improved up to 69.4% under imaging protocol B. But the overall rate dropped to 36.1% under imaging protocol C. The overall rate improved up to 83.3% under imaging protocol D. Especially, significant differences between imaging protocols A and B ( p<0.05), and imaging protocols C and D ( p<0.01), with respect to ACL degenerated group, were recognized. But significant differences between imaging protocols A and C, and imaging protocols B and D, with respect to ACL degenerated group, were not recognized. The concordance rate between histological and MRI findings was 41.7% in ACL normal group, and 61.5% in ACL degenerated group. The concordance rate between surgical and MRI findings was 100% in ACL normal group, and 78.9% in ACL degenerated group. There was a significant difference in the concordance rates between histological, surgical, and MRI findings in normal group ( p<0.05). The results of this study suggested that with Gd-enhanced MRI, the degree of synovial proliferation around the ACL and the degree of degradation of the ACL in the RA knee can be evaluated more accurately than with conventional MRI; however, in RA knees with severe synovial proliferation, it may be difficult to discriminate between the invasive synovium going into the ligament from synovium surrounding the ligament. This may be a limitation of Gd-enhanced MRI at present. In the clinical setting, the present imaging technique does allow the ligament to be evaluated to a certain degree, and may prove useful in the evaluation of temporal changes in the RA knee.