Follow-up of collagen meniscus implants by MRI

The purpose of our study was to evaluate the usefulness of magnetic resonance imaging (MRI) in the follow-up of patients treated with collagen meniscus implant (CMI) and to identify MRI patterns suitable for defining its evolution.

Between March 2001 and June 2003, CMI was performed on 40 patients (27 men and 13 women, age 23-58 years, median 41 years) affected by irreparable medial meniscal lesions. All patients underwent MRI follow-up at 6 months and 1 year and 16 patients 2 years after the operation; 12 patients underwent second-look arthroscopy with implant biopsy. All MRI examinations were performed with a 1.5-T unit using GE T2*, spin-echo (SE) T1, and FatSat fast spin-echo (FSE) DP and T2-weighted sequences, with different orientations. At 24 months, MR arthrography was also performed. Implant evolution was assessed on the basis of MRI direct and indirect criteria. Direct criteria were morphology and signal intensity of the collagen meniscus/residual meniscus complex. Based on these characteristics, three pattern were identified and classified from 1 to 3, where a higher score corresponded to characteristics approaching those of the normal meniscus. Indirect criteria were chondral surface and subchondral bone marrow oedema at implant site and associated synovial pathology.

MRI follow-up at 6 months showed CMI shape and size to be normal (type 3) in 35/40 patients and type 2 in 5/40 patients. CMI signal intensity was type 1 in 32/40 patients and type 2 in 8/40. An interface between prosthetic and native meniscus was identified in 27/40 patients. Chondral lesions were present in 3/40 cases and subchondral bone marrow oedema in 8/40 cases. Reactive synovial effusion was seen in 2/40 patients. MRI follow-up at 12 months showed CMI shape and size to be normal (type 3) in 33/40 patients and type 2 in 7/40. Signal intensity was type 1 in 14/40 patients and type 2 in 26/40 patients. The interface was seen in 19/40 patients. The associated chondral lesions were unchanged, whereas subchondral bone marrow oedema was present in 3/40 patients. No synovial reaction was detected. At 24 months, CMI size was type 3 in 9/16 patients, type 2 in 6/16, and type 1 in one patient in whom the implant could not be identified, as it had been totally resorbed. CMI signal intensity was type 2 in 11/15 and type 3 in 4/16. The interface was identified in seven patients. MR arthrography depicted two additional chondral lesions and enabled correct grading of all lesions. Subchondral bone marrow oedema was present in two patients only.

MRI enables morphological and structural changes of CMI to be monitored over time. Follow-up can be extended beyond 2 years, until the CMI has stabilised and subchondral bone marrow oedema has completely resolved. In the single case with a poor CMI outcome, no related direct or indirect signs were identified.