Purpose: Use of autologous chondrocyte grafts for the treatment of knee cartilage damage appears to be a promising solution. We report preliminary retrospective results in 15 autologous chondrocyte grafts with maximum 4.1 years follow-up.

Material and methods: From September 1996 to December 2000, 15 autologous chondrocyte grafts were performed in 15 patients (13 men and two women), mean age 29.1 years (14.2–46.5) using the Carticel procedure of the Genzyme Tissue Repair laboratory. There were ten trauma-induced chondral lesions and five cases of osteochondritis dessicans. The lesions were located in the femoral condyles in 14 cases (12 medial and two lateral) and the patella in one. Mean surface defect was 6 cm2 (1–15) and all lesions were ICRS grade 3 or 4. The technique described by Brittberg et al. was used. Harvested cartilage was sent to Cam-bridge (USA) for culture. Reimplantation, performed by arthrotomy in all cases, was achieved 12 weeks on average (3.5–29) after initiating culture. Clinical and laboratory results at last follow-up were compared with preoperative data using the ICRS evaluation chart and the Tegner, IKDC, modified Cincinnati, and Lysholm scores. Patients were followed regularly with x-rays, MRI or arthroscan.

Results: Mean follow-up was 2.5 years (1.2–4.1). The subjective IKDC, modified Cincinnati, and Lysholm scores progressed respectively from 38.3 (9–46) to 71.3 (24–98), from 31.6 (18–69) to 58.4 (26–97) and from 41.9 (13–61) to 81.3 (29–100). According to this classification, outcome was excellent or good in ten knees, fair in four and poor in one. For the activity level assessed by the ICRS and Tegner classifications, scores declined respectively from 2.2 (1–3) and 7.4 (5–10) preoperatively to 2.8 (2–4) and 5.2 (2–7) at last follow-up.

Discussion: Due to the small number of patients, it is difficult to compare our results with those reported in the literature. Clinical and functional improvement appeared to be significant but in our experience, there was a decline in the mean sports level at last follow-up.

Conclusion: Autologous chondrocyte grafting appears to be a promising technique for repairing cartilage damage. Use of second-generation grafts with chondrocytes embedded in a solid matrix should facilitate their surgical implantation and improve outcome.

Purpose: We report aseptic complications observed in a retrospective series of 130 total knee arthroplasties using first generation Miller-Galante implants at 6.6 months mean follow-up.

Material and methods: Mean age of the patients at surgery was 65.4 years (35–82). One third of the patients was considered to be obese (BI > 30). The principal cause was primary degenerative joint disease (85%). Most of the implants were implanted without cement except for the first 9 prostheses (hybrid). Mean IKS score was 47.3 (12–70) preoperatively and 74.3 (30–99.5) postoperatively with 70% excellent or good results.

Results: The principal aseptic complications in this series were related to the patella (17%): loosening, fracture, dislocation, necrosis and metallosis. These complications were sometimes associated. Use of a metal-backed patellar insert was correlated with the development of certain patellar complications. The rate of aseptic loosening was 3.4% for the tibia, 1.7% for the femur. There was no statistically significant factor predictive of loosening (alignment, laxity, wear were not significant). Wear of the tibial component polyethylene predominated in the medial femorotibial compartment and was strongly correlated with varus and/or frontal laxity (p = 0.01). Images of bone rarefaction in the distal femur were observed in 61% of the cases. Their intensity was variable; stress shielding or osteolytic mechanisms appeared to be involved. Three revision procedures were required to treated clinically invalidating instability: one frontal laxity corrected by implantation of a thicker polyethylene tibial insert and two important sagittal laxities treated by posterior stabilised prostheses. Other aseptic complications were: stiffness requiring arthrolysis (n=1), unexplained painful prosthesis (n=1) and reflex dystrophy (n=1). These complications led to revision in 15% of the cases. Patient characteristics and mode of fixation were similar in patients with and without complications.

Discussion: These observations point out the multifactorial nature of failure of total knee arthroplasty, most often related to a complication of technical errors and implant properties. Nevertheless, patella-related complications was just one of the principal causes of the problems encountered with the Miller-Galante first-generation prosthesis, particularly the metal-backed insert, leading several teams to abandon this implant.