Revision knee prostheses are often augmented with intramedullary stems to provide stability following bone loss. However, there are concerns with the use of such stems, including loosening caused by strain-shielding, end-of-stem pain, and removal of healthy bone surrounding the medullary canal. Extracortical fixation plates may present an alternative. The aim of the study was to quantitatively evaluate and compare strain-shielding in the tibia following implantation of a knee replacement component augmented with either a conventional intramedullary stem (design1), or extracortical plates (design2) on the medial and lateral surfaces.

Eight composite synthetic tibiae were implanted with one of the two designs, painted with a speckle pattern, loaded in axial compression (peak 2.5 kN) using a materials test machine, and imaged with a 5-megapixel digital image correlation (DIC) system throughout loading. Bone loss was simulated in all models by removing a volume of metaphyseal bone. For four tibiae, the tibial tray was augmented with a cemented stem (∼150 mm). The others were augmented by extracortical plates (maximum 90 mm long) along the medial and lateral surfaces (Fig. 1). Strains were computed using an ARAMIS 5M software system between loaded and unloaded states in the longitudinal direction, for the medial, posterior and lateral surfaces of the tibiae. Strains were checked locally by use of strain gauge rosettes at three levels on medial, lateral and posterior aspects.

The bone strains measured on the posterior surfaces were reported in three regions; proximal (0–70 mm, where the medial extracortical plate lies), middle (70–130 mm, the stem is present but not the extracortical plates), and distal (130–200 mm, beyond the stem). Mean longitudinal strains for both implant types were comparable in the distal region, and were greater than in the other regions (Fig 2). The mean strains differed considerably in the middle region: 565–715 μstrain with stemmed components 1050–1155 μstrain with plated components. Strains followed a similar pattern in the proximal region, particularly very close (20 mm) to the tibial tray component, where the stemmed component bones (775 ± 160 μstrain) displayed less surface strain than the plated component bones (1210 ± 180 μstrain).

Strain-shielding was observed for both designs. The side plates were shorter than the intramedullary rods, so the region of the bone distal to the plates was not strain-shielded, while the same region was strain-shielded when a stemmed component was implanted. It was also shown that in the region of bone just distal of the tibial tray component, design1 shielded the bone from strain 56% more on average than design2. From these results, it can be speculated that the use of extracortical plate rather than intramedullary stems may lead to improved long-term results of revision TKA, assuming the plates and screws provide adequate stability. The extramedullary fixation system preserves more bone than IM fixation, and has the advantage of allowing use of primary TKA components, cemented over the subframe. Similar components have been developed for the femur.