We performed a prospective randomized study to compare two fixation modes, with and without cement, for total knee arthroplasty.

The study series was composed of 96 cemented or noncemented total knee arthroplasties performed between May 1993 and October 1995. The only difference was the diamond interface used for cemented prostheses and the mesh interface used for uncemented prostheses. The operator was unaware of the type of fixation until the bone slices had been obtained. We assessed outcome in 73 cases with a mean follow-up of 27 months. The two populations were comparable for preoperative clinical status, bone tophicity and surgical procedure.

The mean duration of the operation was sgnificantly longer (> 10 min) for the cemented protheses. The complication rates were comparable but we did have two mobilizations of the tibial implant in the noncemented group. The total scores (127 ± 29 in the cemented group versus 135 ± 20 in the uncemented group) were significantly different. There were however more cases with degradation of the controlateral knee in the cemented group although the difference was not significant. When these cases were excluded from the analysis, the total scores for two groups were similar (143 and 140 respectively). Radiographic outcome was quite different with mobilization of the tibial implant in 2 cases and the rate of lucent borders was significantly higher in the noncemented group.

While the clinical outcome was comparable, the quality of the fixation was significantly better with cemented arthroplasty, which remains the gold standard.

Le but de cette étude était de comparer deux modes d'ancrage des prothèses de genou lors d'une série prospective randomisée. La série, incluait 96 arthroplasties cimentées et non cimentées avec traitement de surface sans hydroxyapatite, dont 73 cas revus au recul moyen de 27 mois. La durée d'intervention était significativement plus courte (10 minutes) en l'absence de ciment. Le score total IKS était significativement différent avec 127 ± 29 dans le groupe cimenté contre 135 ± 20 dans le groupe sans ciment, mais avec moins d'atteinte du genou controlatéral dans ce dernier (différence cependant non significative). Si nous excluons ces cas, le score total des deux groupes était voisin avec respectivement 143 et 140 points. Les résultats radiographiques différaient avec une mobilisation de l'implant tibial dans 2 cas du groupe sans ciment, aucune dans le groupe cimenté. Les taux de liserés radiographiques étaient significativement plus importants dans le groupe sans ciment. Si les résultats cliniques semblent comparables, la qualité de la fixation est significativement meilleure en cas d'arthroplastie cimentée qui reste pour nous la référence.