ABSTRACT
PURPOSE OF THE STUDY: The knee has little capacity for spontaneous regeneration of deep cartilage defects. In 1999, the French Society of Arthroscopy initiated a multicentric clinical trial on autologous chondrocyte transplantation using the technique described by Brittberg and Peterson. The protocol of this prospective study was validated by the ethics committee and all patients provided the informed consent for participation. MATERIAL AND METHODS: Patients underwent surgery in seven hospitals: 28 patients (7 female, 21 male, mean age 28 years, age range 18-45 years). The underlying condition was: osteochondritis (n=14), isolated posttraumatic chondropathy (n=8), chondropathy plus ACL tear (n=6). All patients presented deep condylar cartilage defects (ICRS grades 3 and 4). Mean surface area involved after debridement was 490 mm2 (range 150-1050 mm2). Patients were reviewed two years at least after transplantation for functional assessment and an MRI performed 2 to 3 years after transplantation. Control arthroscopy was also performed in 13 patients with biopsy for histology and immunohistochemistry for 10. RESULTS: Twenty-six patients were reviewed with more than two years follow-up (mean 2 years 9 months). There were no general complications; three patients presented a partial avulsion of the autograft treated arthroscopically and one arthrolysis was performed at six months. Function was improved in all patients but four, but pain persisted in one patient. The mean ICRS score improved from 41 points (19-55 points) to 74 points (54-86 points), for an 80% gain. Follow-up MRI was available for 16 knees: the graft was hypertrophied in 11, at level in 3 and insufficient in 2; marginal integration was good in 10 knees and fair in 6. Items of marginal and subchondral integration had a very high positive predictive value for good clinical outcome. The arthroscopic score was nearly normal (range 8-11) in 8 knees and abnormal (range 4-7) in 5. The Knutsen histological groups according to richness of hyaline cartilage were: group 1 (>60%) (n=1); group 2 (>40%) (n=3), group 3 (<40%) n(=4) and group 4 (bone or fibrous tissue) (n=1). The function scores (r=0.80) and the MRI scores (r=0.76) were correlated with the arthroscopy scores. There was no correlation between the histological findings but the sample size was too small for meaningful analysis. DISCUSSION: The clinical results demonstrate an improvement in more than 80% of knees, findings similar to earlier reports. The arthroscopic and histological results were equivalent to those reported by Knutsen, but less satisfactory than those reported by Bentley or Peterson. Cell injections under a periosteal patch constitute the first generation of autologous chondrocyte grafts. Resorbable matrices loaded with chondrocytes before implantation are under development and have provided promising early results.
