[Synovial healing in reconstructed cruciate ligaments. Our personal experience compared in single interventions and combined reconstructions].

The healing process of tendon grafts used in cruciate ligament reconstruction is called ligamentization. The tendon structural architecture changes progressively into the histological appearance of normal ligament. Amiel and Clancy have demonstrated that this process is time-dependent in anterior cruciate ligament (ACL) reconstruction, the tendon graft taking one year to become similar to a normal cruciate ligament. Three different maturation phases of anterior cruciate graft ligamentization can be seen at MRI: periligamentous proliferation, intraligamentous proliferation and definitive healing. We report our MR findings in the reconstruction of single ACL, posterior cruciate ligament (PCL), and of both ACL and PCL. January 1995 to January 1999 we selected 60 patients submitted to arthroscopic cruciate ligament reconstruction. The ACL was reconstructed with the patellar tendon in 22 cases and the PCL in 23; fifteen patients underwent double reconstruction of ACL, with hamstring tendons, and PCL, with patellar tendons. Fifty-five patients were followed-up with MRI: 45 with a dedicated permanent magnet (Artroscan, Esaote Biomedica, Genoa, Italy) and 10 with a permanent low-field unit (AIRIS, Hitachi, Japan); the same technical parameters were used in all cases. Forty of 55 patients were examined at 1, 3, 6, 9 and 12 postoperative months, while 15 were followed-up longer (12 to 36 months postoperatively). At the first follow-up MR examination, 5 patients (2 ACL and 3 PCL) were excluded for incorrect tunnel position (1 case), hypertrophic scarring (2 cases) and new trauma (2 cases). MR findings were compared with clinical data of joint stability. MRI showed the different stages of the healing process in 20 ACL patients: proliferating soft tissue around the graft with the low signal intensity typical of tendons (stage I, 1-3 months postoperatively), the graft becomes progressively hyperintense (stage II, 3-9 months postoperatively), and finally the low signal intensity indicating completed ligamentization (stage III, 12 months postoperatively). As for PCL reconstructions, MR findings were similar but the process took longer, even 24 months. Finally, in the 15 cases of double reconstruction, both grafts were difficult to distinguish on T1-weighted images for a very long time (24-36 months postoperatively). Relative to the ACL, ligamentization takes longer for PCL and combined ACL and PCL reconstructions, probably because the morphological changes in PCL and double grafts may be impaired by many factors, such as gravity, long bone tunnels, hemarthrosis, hyperplastic synovial reaction in the intercondylar notch; also, rehabilitation protocols differ in ACL from PCL patients. No signs of instability were found at physical or arthrometric examinations. MRI demonstrates the different stages of ligamentization and thus proves a useful tool for postoperative follow-up in cruciate ligament reconstructions. Contrast-enhanced (Gd) studies are reserved to the cases of graft impingement or if other abnormalities are suspected.