Fixation of Unicondylar Knee Prostheses

There is increasing use of Unicondylar or Unicompartmental Knee Replacements (UKR), especially following publication of good survival data and a trend towards ‘minimally invasive surgery’. The UKR preserves one of the femoral condyles and its meniscus, plus both of the cruciate ligaments. Therefore, the knee functions more normally following UKR than after Total Knee Replacement (TKR). However, the odds for failure of the UKR are higher than the TKR, and a principal reason is loosening of the tibial and femoral components. There is a need for the development of more reliable UKR fixation designs. The overall aim of this research was to understand fixation of UKR and make recommendations for improvement to designers and surgeons. Since the Oxford mobile-bearing UKR is most widely used in the UK, it was used as the benchmark in this study. To assess initial fixation, in-vitro bone-constructs were prepared from ten cadavers implanted with the Oxford mobile-bearing UKR and tested for bone strain and bone-implant interface motion with the implants fixed using first cementless and then cemented methods. Cementless fixation produced higher proximal tibia strain and bone-implant displacement than cemented fixation. Peak bone strain increased with reduced bone density, such that the lowest density specimen fractured when implanted with the cementless UKR. To assess long-term fixation, an in-vivo prospective follow-up study of 11 Oxford UKR patients was developed and conducted for one-year, taking measurements of bone density using Dual X-Ray Absorptiometry (DXA) scanning. The average bone resorption under the tibial implant was found to be low; while it was higher under the femoral component and very high under the tibial intercondylar eminence. The fixation of the Oxford UKR implant was considered to be adequate at 1-year. Finite Element (FE) simulation techniques were reviewed and developed to simulate the UKR knee for investigation of bone strain, bone-implant interface micromotion and bone remodelling to assess initial and long-term fixation performance. Computer simulations of the tibiae and femora of 2 patients and 4 cadaveric specimens (obtained from the in-vivo and in-vitro studies) were developed and validated for bone strain, bone-implant interface micromotion and bone remodelling. Comparative multi-specimen computational studies were conducted to understand how particular design features affected fixation. Good fixation was indicated for cementless UKRs when implanted in dense bone, but bone strains were very high in low density tibia. Cementation of the implants spread the loads more evenly and reduced bone strains. The cementless tibial implant caused less bone resorption (compared to the cemented equivalent) but the difference in the femur was small. Bone resorption was highest at the anterior tibia and posterior to the femoral peg. Bone density was an important factor in the fixation performance of implant design features. Less bulky fixation features reduced bone resorption, provided that the underlying bone was sufficiently dense to maintain bone strains below the failure limit of bone. For patients with dense bone, fixation could be improved with shorter tibial keels and less stiff femoral implants. For patients with low density bone, fixation could be improved with cementation and bone resection that avoids creating stress-raisers