We prospectively reviewed 2440 Cemented Anantomical (side specific) AGC total knee replacements performed on 2074 patients between 2002 and 2008 in our unit. The surgery was performed by a number of surgeons, both Consultant and Trainee grades. 1357 underwent PCL sacrificing surgery with implantation of Posterior Stabilised Femoral Implant (Cam and Grove) and 1083 underwent PCL retaining surgery with implantation of a Non-Stabilised Femoral Implant. The mean age at date of surgery for Posterior Stabilised prosthesis was 69 years (range 23-94) and Non-Posterior Stabilised prosthesis was 70 (range 33-97) with no significant difference p< 0.05 (C.I 95%). Follow up at present mean 4 years (range 1-6 years).

11 patients underwent secondary patella resurfaced for anterior knee pain at a mean of 1.5 years following the index procedure (range 0.5-4.1 years) with 6 patients reporting a marked improvement in anterior knee pain post operatively. 3 patients underwent removal of patella buttons for symptomatic loosening at a mean of 2.7 years follow index procedure.

A total of 17 patients underwent a revision of either tibial or femoral components (0.7%). Indications for revision were infection in 11 patients, gross instability in 1 patient, periprosthetic fracture in 1 patient, persistent stiffness in 1 patient. 3 patients underwent revision for symptomatic loosening of the tibial component at a mean of 2.9 years (range 1.1-4.0).

These short to medium term results for the Anatomic AGC total knee replacement agree with published results from other centres.

Before proceeding to longer-term studies, we have studied the early clinical results of a new mobile-bearing total knee prosthesis in comparison with an established fixed-bearing device. Patients requiring bilateral knee replacement consented to have their operations under one anaesthetic using one of each prosthesis. They also agreed to accept the random choice of knee (right or left) and to remain ignorant as to which side had which implant. Outcomes were measured using the American Knee Society Score (AKSS), the Oxford Knee Score (OKS), and determination of the range of movement and pain scores before and at one year after operation.

Preoperatively, there was no systematic difference between the right and left knees. One patient died in the perioperative period and one mobile-bearing prosthesis required early revision for dislocation of the meniscal component.

At one year the mean AKSS, OKS and pain scores for the new device were slightly better (p < 0.025) than those for the fixed-bearing device. There was no difference in the range of movement.

We believe that this is the first controlled, blinded trial to compare early function of a new knee prosthesis with that of a standard implant. It demonstrates a small but significant clinical advantage for the mobile-bearing design.