Abstract
Background
Cementless Total Knee Arthroplasty has been developed to reduce the incidence of failure secondary to aseptic loosening, osteolysis and stress-induced osteopenia, especially in younger and more active patients. However, failures are still more common compared to cemented components, especially those involving the tibia. It is hypothesized that this is caused by incomplete contact between the tibial tray and the underlying bony surface due to: (i) inadequate flatness of the tibial osteotomy, or (ii) failure of implantation to spread the area of contact over the exposed cancellous surface. In the present study we compare the contact area developed during implantation of a cementless tray as a function of the initial flatness of the tibial osteotomy.
Method
Eight joint replacement surgeons prepared 14 cadaveric knees for cementless TKR using a standard instrumentation set (ZimmerBiomet Inc). The tibial osteotomy was created using an oscillating bone saw and a 1.27mm blade (Stryker Inc) directed by a slotted cutting guide mounted on an extramedullary rod and fixed to the tibia with pins and screws. The topography of the exposed cancellous surface was captured with a commercial laser scanner (Faro Inc, Halifax, approx. 33,000 surface points). 3D computer models of each tibial surface were generated in a CAD environment (Rapidform, Inuus). After scanning, a cementless tibial tray was implanted on the prepared tibial surface using a manual impactor. The tray-tibia constructs were dissected free of soft tissue, embedded in mounting resin, and sectioned with a diamond wafering saw. Points of bone-tray contact and interface separation were identified by stereomicroscopy and incorporated in the 3D computer models. Maps were generated depicting contacting and non-contacting areas Each model was subdivided into 7 zones for characterizing the distribution of interface contact in terms of anatomic location.
Results
The flatness for the tibial osteotomies averaged 1.1±0.35 mm (range: 0.56–1.81mm). After impaction, 79.8±0.3% of the tibial surface had plastically deformed to establish a contacting interface with the implant. 15.1% of the bony surface was within 0.2mm of the tray and 17.6% was within 0.3mm. Gaps large enough to impede ingrowth only occupied 2.6% of the exposed tibial These non-contacting areas were typically located centrally at the ACL, PCL and canal zones. There was an inverse linear relationship between the initial flatness of the tibial osteotomy and the percentage of tray-bone contact.
Conclusions
The amount of direct contact between the bone and implant is critical for the development of stability in cementless fixation. We found a relationship between ultimate bony contact and initial flatness. However, we also found that during impaction of the implant, bony contact increased through deformation of the most prominent peaks of the cancellous surface. Interface gaps were consistently observed in central areas of the tibia surface located above the medullary canal which may be reduced through selection of trays with distal keels.
For any figures or tables, please contact authors directly.