The influence of the strength of bone on the deformation of acetabular shells

Concerns have been raised that deformation of acetabular shells may disrupt the assembly process of modular prostheses. In this study we aimed to examine the effect that the strength of bone has on the amount of deformation of the acetabular shell. The hypothesis was that stronger bone would result in greater deformation. A total of 17 acetabular shells were inserted into the acetabula of eight cadavers, and deformation was measured using an optical measuring system. Cores of bone from the femoral head were taken from each cadaver and compressed using a materials testing machine. The highest peak modulus and yield stress for each cadaver were used to represent the strength of the bone and compared with the values for the deformation and the surgeon’s subjective assessment of the hardness of the bone. The mean deformation of the shell was 129 μm (3 to 340). No correlation was found between deformation and either the maximum peak modulus (r2 = 0.011, t = 0.426, p = 0.676) or the yield stress (r2 = 0.024, t = 0.614, p = 0.549) of the bone. Although no correlation was found between the strength of the bone and deformation, the values for the deformation observed could be sufficient to disrupt the assembly process of modular acetabular components. During the 1970s uncemented acetabular components became popular, despite more than a decade of success using cemented components.[[1]] Uncemented components remain popular [Jameson et al. 2013]; data from the 2013 National Joint Registry of England, Wales and Northern Ireland showed that between 2004 and 2012 the use of cemented acetabular components declined.[[2]] Cementless acetabular components require initial stability through a press-fit within the acetabulum.[[3,4]] Although uncemented components offer advantages such as preservation of bone stock,[[5]] there are concerns that the forces required to achieve initial fixation may result in significant deformation of the acetabular shell.[[6-8]] Deformation is thought to be linked to several variables. Previous in-vitro, cadaveric and finite element studies have examined deformation of the acetabular component with regard to different variables such as the diameter of the component,[[4,9]] underreaming,[[6]] the thickness of the component[[9,10]] and positioning.[[11]] We are not aware of studies that have related deformation the strength of the bone. The aim of this study in cadavers was to determine whether deformation of the acetabular shell correlates with the stiffness or strength of the underlying bone. The hypothesis was that stiffer and/or stronger bone would result in greater deformation. Materials and Methods Cadaver and implant details A series of 17 titanium acetabular shells (each used only once) were inserted into eight cadavers in a temperature-controlled laboratory. The posterior surface of the shells was grit blasted with an estimated roughness average of 2 to 5 μm. There were three female and five male cadavers. Their mean age at the time of death was 74.1 years (58 to 87); their mean weight was 72.4 kg (32 to 113) and their mean height was 170 cms (152 to 193). The shells were custom manufactured for these experiments by CeramTec (CeramTec GmbH, Plochingen, Germany) from a titanium alloy (TiAl6V4) with a thickness of 3 mm. As the sizes of the cadaver acetabula could not be determined prior to the start of the experiment, a range of external diameters of the shell between 44 mm and 62 mm were available. Table 1. Details of the cadavers Cadaver Gender Age (yrs) Height (cm) Weight (kg) 1 Female 72 170 43 2 Male 70 168 73 3 Male 80 193 80 4 Female 82 152 32 5 Male 58 175 113 6 Male 85 163 84 7 Female 59 163 68 8 Male 87 178 86 Surgical method A large incision was created using a posterior approach and up to 20 cm of the proximal femur was resected in order to provide the necessary access for the measurement system. Reaming was performed in 2 mm increments until the true floor of the acetabulum was exposed. Next, the reaming was expanded to remove the cartilage down to a hemispherical subchondral bed of bone suitable to securely press-fit a hemispherical shell without the need for screw fixation. After reaming, the surgeon graded the bone on a 3-point scale, 1 being the hardest and 3 the softest. The shells were then implanted as intended in a standard hip replacement. The amount of reaming was adjusted to achieve stable primary fixation that would be acceptable in live surgery in all cases. This involved reaming “to size” in eight shells, and under-reaming by 1 mm in nine shells. Fixation was defined as complete medial contact with the surgeon unable to pull the shell out or rotate it manually. In order to maximise the usefulness of the cadaver, in large patients it was sometimes possible to reream the acetabulum and insert a larger shell while still obtaining a secure press-fit. Shells were therefore inserted into 11 acetabula, with five acetabula having more than one implanted. Measurement of Deformation Measurements were performed immediately before and within five minutes after introduction of the shell using the GOM ATOS III Triple Scan optical system (GOM GmbH, Braunshweig, Germany). These two measurements were then compared to determine the deformation in a plane at 1 mm below the rim of the shell. This distance was chosen because previous measurements using this system have shown that although the shell is subject to asymmetrical deformation, the highest values are recorded at the rim.[[12]] The GOM system uses two cameras to determine the position and size of the object using an interferometric technique. A previous validation study [[12]] found that the maximum difference between the GOM system and a coordinate measuring machine (CMM) (Carl Zeiss Industrielle Messtechnik GmbH, Oberkochen, Germany) for this application was 5 μm. A thin titanium oxide (TiO2) coating was applied to the internal surfaces of the shells to dull the reflective surface. In order to prevent damage to the coating, surgical swabs were applied to the area surrounding the shell to absorb any moisture. The method of measurement and validation have been described previously.[[13]] Testing the strength of bone Core samples of cancellous bone were taken from each cadaver following the method described by Li and Aspden[[14]], to examine the mechanical properties of femoral cancellous bone. A total of 75 samples were taken from five sites across each of the 15 femoral heads: the superior, inferior, anterior, posterior and central surfaces. No samples were taken from one femoral head as no shells had been implanted in the corresponding acetabulum. Cores were removed using a bit and drill approximately perpendicular to the surface, so that the axis of the cylinder was towards the centre of the femoral head. For each sample, the subchondral bone was removed and the ends were trimmed. The mean diameter of the core was 7.8 mm (7.2 to 8.2) and the mean length was 12.7 mm (6.7 to 21.0). Each sample was then subjected to an unconstrained compression test on an Instron 5564 materials testing machine (Instron, High Wycombe, Buckinghamshure, United Kingdom) using the technique described previously.[[14]] A strain rate of 20% per minute was used and load–displacement curves were plotted. In order to avoid damaging the samples unnecessarily, tests were stopped manually once the slope of the curve was seen to be reducing. Data processing Loads and displacements were converted to stresses and strains by dividing by the original cross-sectional area and length of each sample. A program was written in MathCad (Parametric Technology Corporation, Needham, MA, USA) to import and analyse the stress– strain data. A polynomial curve was fitted to a section of the rising part of the curve and differentiated to determine the gradient. The region of the curve and the degree of polynomial (most commonly five) were selected to ensure close fitting to the steepest part of the curve without over-fitting, as judged visually and by least-squares distance between the fitted curve and the data. Key points of the test are shown in Figure 1. Peak modulus and yield stress were then obtained from the data. Peak modulus, and the corresponding stress and strain values, was taken as the greatest value of the gradient. Yield stress was the stress at the yield point, which was defined as the point at which the modulus decreased by 3% from its peak value. This may be too stringent, but uses the same approach as Li and Aspden.[[14]] The peak modulus is defined as the resistance of the material to the applied stress. Yield stress is the point at which elastic deformation stops and the sample is left with permanent or plastic deformation, once the load has been removed.[[15]] The greatest peak modulus and yield stress for each femoral head were recorded. Results from the left and right femora of the same cadaver were combined and the values for the overall greatest peak modulus and yield stress were used for that cadaver. These values were then compared with both the values for the deformation and the surgeon’s grading of the bone. Although this provided information on how surgical ‘feel’ matches against mechanical data, it is acknowledged that the surgical grading is very subjective. It should be noted that femoral bone quality was taken as a surrogate marker for acetabular bone quality. Statistical analysis The statistical significance of the relationships between deformation and both peak modulus and yield stress was assessed using the statistical package R (R Foundation for statistical computing, Vienna, Austria) and a linear regression. Deformation was assessed against the diameter of the shell using generalised linear models with a gaussian error structure to ensure that it did not influence the results, and p < 0.05 was considered to be statistically significant.