Introduction The ABG I acetabular insert is an ultra high molecular weight polyethylene (UHMWPE) component used in primary hip arthroplasty. Studies have shown early osteolysis and aseptic loosening of the ABG I uncemented cup compared with other implants. Theories advocate that loosening is initiated by the biological response to insert wear debris; wear volume and the distribution of particle size are considered to be important parameters. This study analysed explanted plastic inserts to identify any mechanical properties that may have contributed to early failure.

Materials and Methods 21 ABG I acetabular components were revised due to aseptic loosening over a 16 month period. Silicone casts of the insert sockets were made and volumetric analysis performed using a shadowgraphing technique and a coordinate measuring machine (CMM). The UHMWPE inserts were divided into uniform pieces with a diamond-tipped microsaw and analysed for hardness, wear, stress and strain properties using a microhardness tool, pin-on-plate analysis and small punch testing. We performed identical tests on explanted inserts from other manufacturers.

Results We present the findings of the above tests and provide suggestions as to why these particular implants are more prone to early failure when compared with other common implants. We also discuss the results of volumetric analysis by shadowgraphing compared with CMM.

We found the ABG cementless hip has excessive acetabular wear and premature failure due to osteolysis.

In 60 patients implanted at mean age 56 years, 66 hips (mean follow up 48 months), 7 were revised and 7 have severe acetabular osteolysis. In some this is entirely asymptomatic. There was significant association with osteolysis, length of follow up and wear but no correlation between wear and acetabular component position, age, liner thickness, and use of ceramic or CoCr heads.

We recommend regular lifelong radiological review of these hips and suspension of use of this prosthesis until a wider review is undertaken.

We wished to determine the most accurate and reliable technique for insertion of tibial prostheses, with tibial resection guided by either intramedullary (IM) or extramedullary (EM) alignment jigs.

135 consecutive AGC cemented total knee replacements in 126 patients in a single unit were performed by, or directly supervised by, four consultant surgeons. Ethical approval and patient consent was obtained. Intramedullary alignment was used for the femoral cuts and patients were randomised at the time of operation to have either IM or EM guides for resection of the proximal tibia, cut with a zero degree posterior slope in both. The protocol only entered patients into the trial if their knees were suitable for use with both IM and EM tibial alignment although, in the event, no patients were excluded. Long leg radiographs (standing hip to ankle) were taken by a standardised method three months after the surgery. A blinded assessor, unaware of the alignment method used, evaluated acceptable films and measured tibial component alignment. The proportion of tibial prostheses aligned within two degrees of 90 was the endpoint of the study.

Of the 135 knees 100 suitable x-rays were assessed. Correct tibial alignment was more likely in the IM group (85%) than the EM group (65%), p=0. 019. Though mean alignment was similar, variation (standard deviation) was less in the IM group (2. 0 vv 2. 2).

In the AGC knee, intramedullary alignment guides are superior to extramedullary guides for alignment of the tibial prosthesis. We recommend the routine use of intramedullary tibial alignment.