Aseptic loosening, osteolysis, and infection are the most commonly reported reasons for revision total knee arthroplasty (TKA). This study examined the role of implant design features (e.g. condylar box, pegs) and stems in resisting loosening, and also explored the sensitivity of the implants to a loose surgical fit due to saw blade oscillation.
Finite element models of the distal femur implanted with four different implant types: cruciate retaining (CR), posterior stabilising (PS), total stabilising (TS) with short stem (12 mm × 50 mm), and a total stabilising (TS) with long stem (19 mm × 150 mm) were developed and analysed in this study. Two different fit conditions were considered: a normal fit, where the resections on the bone exactly match the internal profile of the implant, and a loose fit due to saw blade oscillation, characterised by removal of one millimetre of bone from the anterior and posterior surfaces of the distal femur. Frictional interfaces were employed at the bone–implant interfaces to allow relative motions to be recorded.
The results showed that interface motions increased with increasing flexion angle and loose fit. Implant design features were found to greatly influence the surface area under increased motion, while only slightly influencing the values of peak motion. Short uncemented stems behaved similarly to PS implants, while long canal filling stems exhibited the least amount of motion at the interface under any fit condition.
In conclusion, long stemmed prostheses appeared less susceptible to surgical cut errors than short stemmed and stemless implants.