Abstract
Introduction:
For 30 years, uncemented anatomic hip stems have been implanted with documented clinical results[1,2]. Their geometry can be linked back to the geometry of the PCA and ABG stems. Modifications to date include stem length, body geometry, material, and reduction in distal geometry. New tools have been developed allowing anatomical measurements and analysis of three-dimensional digital femora geometry through CT scans[3]. The purpose of this study is to analyze three-dimensional contact of various anatomic hip stem designs using this technique.
Methods:
Six femora (57–87 yrs, 72–88 kg), were selected from a CT scan database (SOMA™) of 604 Caucasian bones. They were selected based on femoral anteversion (average +/−1.5 * std. dev.) with three measuring[4] 8–10° and three 31–33° of anteversion. The CT scans were segmented into cancellous/cortical bone and converted into CAD models in PRO/Engineer Wildfire (v.5). A/P views of the bones were scaled to a 120% magnification to allow three surgeons to surgically template and choose the stem size and location (maximizing fill (n = 1); restoring the head center (n = 2)) with two implant designs (1-Citation TMZF and 2-ABG II Monolithic, Stryker Orthopaedics, Mahwah). Measurements from templating were used to virtually implant CAD models of the implants into the bones (n = 36 bone/stem assemblies). The assemblies were imported into Geomagic Qualify 2012 for 3D deviation analysis comparing the coated region of the implant to the cortical-cancellous boundary. The analysis generated color map profiles based on the following categories: Contact (−2.0 to 0.5 mm), Conformity (0.5 to 2.0 mm), Proximity (2.0 to 5.0 mm), and Gap (5.0 to 12 mm) and the percent of the surface that was within each of these categories. These results were compared for patterns within and across the anatomic families.
Results:
Similar patterns of fit were observed within and across both families. The same size implants were not always used together across both systems. The strongest commonality was found regarding the percentage of the implant adjacent to more than 5 mm of cancellous bone (Gap, shown in red in Figure 1b) and the pattern of contact on the medial curvature of the implants. On average 61% and 56% of the metaphyseal region of Implants 1 and 2, respectively, is adjacent to 5–12 mm of cancellous bone between the implant and cortical bone. Implants 1 and 2 also demonstrated 30% and 37% between 0.5 and 5 mm of cancellous bone to the cortical boundary. Contact (< .5 mm) was only achieved in areas where bone would have been removed through femoral preparation. When maximizing fill, it was found that the percent Gap was reduced and distributed between conformity and proximity. There was also less variability between both systems when the goal was to maximize fill, however there was no statistical difference given the sample size between both stems regardless of method.
Discussion:
Proper load transfer is essential for positive bone remodeling for short/long term fixation. As anatomic stems load femurs circumferentially, it is important to note that common characteristics transfer load to bone potentially contributing to their success. Previously, technology has not permitted circumferential analysis of implant fit on a wide scale, reproducible basis.
