Final Report Summary - ENDURE (Enhanced Durability Resurfacing Endoprosthesis)
Project context and objectives:
The treatment of younger patients with severe hip disease using a conventional total hip arthroplasty (THR) presents a challenge. The success rate of the THR is currently very low within the first 16 years of implantation. To overcome this Finsbury instruments, in conjunction with surgeons, Midland Medical Technologies Ltd (MMT) and Doncaster's Centaur Precision Castings, developed the metal on metal (MoM) Birmingham hip resurfacing (BHR) system, which was developed for MMT. However, the MMT was subsequently bought by Smith and Nephew (United States of America (USA) based company) during 2005. Evidence suggests that the MoM resurfacing provides an ideal solution for implantation into younger patients who have a more active lifestyle with the survival rate being 93 % after 3 to 5 years over a conventional THR. However, there is concern about metal wear debris and its systematic distribution throughout the body. Several studies have reported high concentrations of cobalt and chromium in the serum and/or urine of patients with MoM resurfacing implants. Potential toxicological effects of the elevated metal ions have increased concerns about safety of MoM implants. This is a particular concern in younger and active patients in whom life expectancy after implantation is longer. Because patients who receive a MoM hip arthroplasty are shown to be exposed to high concentrations of metallic ions, the Medical Advisory Secretariat performed literature searches for the adverse biological effects of cobalt and chromium. Cobalt and chromium make up the majority of the substrate material used in the manufacture of MoM implants.
Concerns have been raised recently due to high levels of cobalt and chromium measured in blood and serum samples of patients with MoM devices. These concerns relate to the possible systemic effect of metal ions released by MoM prosthesis in terms of genomics, as well as immulogy, and histology, and reviews have recently been published, into the ion levels and their biological effects were marked when wear was abnormally increased, with low carbon (< 0.2 %) implants and loose or poorly positioned implants.
In November 2006, a meeting was organised by Dr Case, University of Bristol, United Kingdom (UK), with a panel of world experts in various medical fields where metal ions were thought to have a potentially detrimental effect. It was concluded that a risk to benefit analysis should performed for individual patients wishing to have a MoM implant. The long-term exposure concerns were related more specifically to younger, more active patients wishing to have children and to maintain their active life style. It was concluded that a risk benefit analysis of the patient situation should be implemented. For those patients wishing to have children post joint replacement, it was decided that further follow up studies were required, with a particular focus on the potential passage of the genetic damage to offspring in younger patients.
MoM resurfacing technology is superior to that of conventional THR implant for younger patients; however, the MoM has a series of potentially long-term problems of its own. To overcome these problems, the partners in the ENDURE project developed a ceramic on polymer resurfacing hip prosthesis. Like that of the MoM prosthesis the ceramic on polymer resurfacing prosthesis will only cover the femoral head thus retaining all of the desirable properties associated with retaining an intact femur.
Using the following key innovations developed during the course of the Endure project the partners we able to produce a new resurfacing hip prosthesis that gave significant advances over current existing systems.
These key innovations include:
- fabrication of a resurfacing (BHR-type) femoral, hemispherical ball-head and cup from ceramic and polymeric materials;
- the use of new materials, and processing methods to achieve a non metallic low ware resurfacing hip system;
- development of a flexurally tough alumina matrix composite (AMC) ceramic femoral head and thin walled reinforced polymer acetabulum cup;
- an innovative coating process that will allow for full bone integration with the implant;
- a novel design that includes the bone as part of the construction;
- an advanced low impact surgical procedure to reduce trauma during the operation.
The ENDURE system resurfaces the femoral head similar to that of the MoM system but unlike this system the ceramic wear particle will remain inert within the body thus eliminating any concerns relating to the long term effects of ions. It is well known that the use of ceramic and polymeric materials will lower the coefficient of friction between the mating surfaces and as a result will reduce the amount of wear and thus debris generated during the normal operation of the joint.
Overall project objectives
Our main objective was to develop a high durability ceramic on polymer resurfacing hip prosthesis that has a long working lifetime ideally in excess of thirty years. This was be achieved by unique combination of the best, established technologies, plus the additional advantages of novel AMC bioceramics, new processing methods and polymeric compounds. In order to realise these objectives, we broke them down into the following categories.
Scientific objectives:
Our scientific objectives were those relating to issues of durability, via influences in physical wear phenomena and biocompatibility.
Technological Objectives for the device:
The technological objectives were to acquire the science to enable the development of the first durable, long-life hip resurfacing joint prosthesis that would:
- produce a low friction / wear coupling by accurate and precise grinding of the component faces to the specification and tolerance identified during the work packages (WPs);
- achieve volumetric wear rates = 0.8 mm3 per million cycles under simulated steady state in-vivo lubrication conditions, by accurate and precise machining to asphericity specifications and tolerance identified;
- reduce the size of the wear particles so that they do not exceed 0.1 µm, by application of design specification from WP1;
- avoid adverse sensitivity and have complete biocompatibility by choice of AMC as the bioceramic and polymeric / CF PEEK materials used for the construction;
- ensure complete and stable osseointegration within the first three months following surgery by a coating combination of titanium and hydroxyapatite, applied through the development of the vacuum plasma spray technique;
- improvement of operative technique and instrumentation to improve optimise alignment and reduce the risk of femoral neck notching to achieve optimal range of motion and avoid post-operative femoral neck fractures;
- increase of candidate patients including women of child bearing age.
Economic objectives:
As well as the technical objectives the consortium partners also addresses the economic need for the technology developed during the course of the project. The results of the economic study showed that by the end of year 5 after the project, and through a network of trans-national and cross-sectoral licensees, the consortium would be able to sell the new joint technology over the time period.
Enabling innovation related objectives:
To achieve the societal and economical objectives that come from the dissemination and exploitation of the research results the consortium partners have defined an enabling set of objectives.
To enable innovation through the project team and to benefit Europe the objectives are:
- collation and preparation of the results of the project into a suitable format and apply for patent protection of the results of the project covering the manufacture of the novel AMC ceramic / polymeric hip technology;
- the transfer knowledge from the research and technology development (RTD0 performers to the small and medium-sized enterprise (SME) participants has been achieved using technology transfer events and interactions;
- to disseminate the results and benefits of the knowledge and technology developed beyond the consortium to potential users such as healthcare / surgical, dental, veterinary, scientific.
Societal and policy objectives are to benefit society by:
- reducing the number of surgical revisions following implantation of a medical device;
- as a result of the reduction in the numbers of people requiring revision surgery, the ENDURE project will have the affect of allowing reallocation much-needed medical resource to other areas.
Project results:
In order to deliver the work within the ENDURE project each objective was broken down into a separate WP.
WP1 - Design foundation
The objective of WP1 was to define the design envelopes for the femoral and acetabular components.
Literature search for the material data
During this task, several aspects of the prosthesis component design were considered and our research showed that currently the most common bearing components used is metal on polyethylene. The polyethylene is housed in a metal backing and implanted on the acetabular side. Other common bearing types are metal on metal and ceramic on ceramic. Although some are more successful than others, all of them still experience problems. The common problems associated with each implant range from high wear rates, stress shielding, osteolysis to implant loosening. In order to improve the longevity of the prosthesis, these problems must be overcome. This could potentially be achieved by the use of a new material implant combination.
Ceramic materials have been used successfully for the manufacture of hip joint replacement. These are used as a bearing material, articulating against itself or against polyethylene. These devices includes a metal outer shell in order to allow for an in growth surface to be added to the outer surface of the component, which would not be achievable on ceramic or polyethylene materials alone. The in growth surface is critical to the short and long term fixation between the implant and the surrounding bone. However, problems with fixation on the acetabular side are often reported, and loose cups are a common reason for revision in hip replacement. These problems are attributed to the high stiffness of the acetabular replacement component, due to the presence of a stiff metal shell. Polymer materials offer a solution to the stiffness mismatch. Various polymer materials have been successfully used in other areas of orthopaedics (spinal) and in dental applications. This has led to a growth of interest in polymers for use on the acetabular side of hip replacements. A polyaryletherketone (PEEK) biomaterial launched by Victrex in 1998 and marketed for medical use has many advantages over metals including elimination of imaging artefacts and the ability to view tissue/bone growth and repair using x-rays. The addition of a filler material in the PEEK namely carbon fibre formulates a different grade called CF-PEEK. The materials properties can be adjusted by varying the percentage of carbon fibre inclusion; this is of particular interest to the design of a replacement construct where some compliance is required at the bone interface whilst bulk strength is a basic requirement for the component itself. The properties of both filled and unfilled PEEK materials must be fully understood and verified for their use in-vivo and to determine where CF-PEEK is more appropriate to use over unfilled PEEK.
Required range of motion (ROM)of THR implants
In normal THA prosthesis, the range of motion is maximised through the use of a large head diameter and a small neck diameter on the femoral stem component which enhances the hip performance by allowing greater range of motion before the components impinge (Chandler et al., 1982; Barrack, 2003; Bengs et al., 2008). Hip resurfacing decreases the head to neck ratio as the natural femoral neck is retained. Impingement can occur when the retained femoral neck abuts against the acetabular component or anterior acetabular bony wall. This can be painful and restrict motion. The ISO standard 21535 states that 100° flexion / extension; 60° abduction / adduction and 90° internal / external rotation is required. However, even the earliest literature states that a greater range of motion needs to be achieved in order to allow everyday tasks to be performed. Johnson and Smidt (1969) and Roach and Miles (1991) have shown flexion / extension values in the range of 125 - 135° which is higher than the value recommended in the standard. Clearly if THA implants are designed to this specification, the patient's post operative ROM will be limited.
Kluess et al. (2008) performed a study on how the implant designs affect the flexion / extension and internal and external rotation angles that are achieved post-operatively. The results of the study highlight that the current implant designs do not achieve the required ROM in all patients. Using the data gather from the literature searches, it was possible to generate a simple finite element (FE) model that considered various potential loading scenario, simulating activities more demanding than the conventionally modelled walking gait, and a more accurate prediction of worst cases stresses generated within the construct during service was obtained.
Design aspects specific to ceramic material use
FE analysis studies have particular use in the pre-clinical analysis of new prosthesis designs and are used to make safe design decisions with regards to performance. FE models also provide technical evidence during the regulatory submission process, a necessary step in most new product development projects. Whilst it is relatively easy to simulate ISO tests of implant components mounted in simple elastic supports, it can be argued that this does not represent the whole range of typical loading conditions experienced by an implant in-vivo. Even to begin to recreate these conditions, it is necessary to mount the implant in realistic supporting bone with geometry with materials properties derived from computer tomography (CT) data. The FE model developed to assess the stresses the component may be subjected to during its life in vivo was developed using state of the art methods in order to improve its accuracy. In particular, surgical data (CT scanned bones) were used to generate the bone models. The establishment of design requirements, including pass criteria for the output of the FE analysis, is critical to the safety of the design.
FE models of resurfacing femur
The following most fundamental design criteria were considered in this study:
- The volume of damaged bone under stumbling and falling loading conditions must be lower for the new designs than for the ADEPT or BHR prostheses, hereby referred to as 'the traditional design',
- The percentage volume of bone with sufficient remodelling stimulus to generate stress shielding and hypertrophy around the prosthesis will be lower than for the traditional design.
- The peak stresses in the new design must be considerably below the fatigued fracture strength for the material. Absolute strength tests are beyond the scope of this report, but extreme stumbling loads must be sustainable safely.
Details for ceramic manufacturing WP
Ceramic heads were produced by the following process:
1. powder batch preparation, e.g. wet milling and spray-drying powder granulation,
2. pressing into a cylindrical billet, of low theoretical density and strength,
3. green machining, into an approximate, over-sized shape,
4. thermal 'presintering' to achieve approximately 96 % theoretical density,
5. hot isostatic pressing (HIP) in oil or gas to achieve approximately 100 % theoretical density and strength,
6. hard machining of the bearing surface, to precise, tightly toleranced dimensions,
7. polishing the bearing surface to the required surface roughness,
8. laser marking with part and serial numbers,
9. proof testing, and
10. final inspection.
The manufacturing process has been developed to produce heads of adequate strength to avoid fracture in-vivo, through careful control of grain size (in powder processing), volumetric flaw content (in pressing, sintering and HIP) and surface flaw content (in progressive machining and polishing). The diameter, roundness and sphericity of the final bearing surface are critical to a bearing component's wear performance, so the manufacturing process was also designed to allow target bearing dimensions to be achieved within tight tolerances.
The particular difficulty with the production of ceramic resurfacing heads is the lower wall thickness than modular heads. This is relevant to the bearing surface dimensions, as gripping the piece for machining in a chuck may distort it, so that it deviates from the required roundness when it springs back upon release. Although specific details of manufacturing processes, fixtures and target dimensions and tolerances cannot be released, this work package involved:
- further development of pressing of larger sized billets for green machining of resurfacing heads and other CT products with large diameters based on existing and established technology,
- development of machine mounting fixtures to allow the thin-walled resurfacing component to be gripped and machined with sufficiently low distortion to give adequate final roundness and sphericity.
This permitted a series of resurfacing head shaped prototype prostheses to be produced, allowing mechanical testing to commence. The approach involved developing the existing manufacturing technique with as few changes as possible, as past clinical results of other manufacturers have demonstrated the risk associated with departures from known techniques.
Both the design and the manufacturing techniques were verified with mechanical testing.
Design aspects specific to CF-PEEK material use
There are many design features that are required in order to make a new resurfacing hip system successful. The areas that required extensive research were the geometry and tolerances between surfaces as well as the overall design. It was important to also include the design principles of the dedicated surgical instruments.
From the material data report it was confirmed that the femoral head would be manufactured from Biolox ceramic, whilst the acetabular cup would be made of carbon fibre reinforced polyetheretherketone (CF-PEEK) or a combination of CF-PEEK and unfilled PEEK.
A literature review was performed on performance of the current acetabular cup designs that employ CF-PEEK as their bearing surface which include the Cambridge and MITCH cups. Fixation methods in the form of a coatings and porous layers to encourage bony in-growth and improve initial fixation of the cup were also explored.
The investigation also looked at patents on PEEK acetabular cups, coatings and porous outer surface.
Designs ideas discussed during the ENDURE kick-off meeting were reviewed to determine the optimum option and each design was scored against the following criteria: complexity, manufacturability, bony in-growth, surface finish, novel idea, initial fixation, coating ability and difficulties to grade each design. The optimum design includes a multiple layered design consisting of CF-PEEK MOTIS as the bearing surface and a porous layer of PEEK Optima.
WP2 - Implicit and explicit computational analysis
The aim of WP2 was to produce computational analysis tools:
1) to inform the design of the developing implant concepts,
2) for their initial theoretical structural verification, and
3) to enable physical tests to be developed that would reproduce clinically representative stress distributions in the implant components, for their more concrete physical test verification.
The goals of the structural analysis were to predict:
- the bulk stresses in the bone supporting the implants,
- the stimulus and progression of bone remodelling around the implants and
- the fixation interface conditions, indicating their long-term stability,
- the bulk stresses in the implants under a range of in-vivo loading conditions (normal and traumatic scenarios), and
- the bulk stresses in the implants under surgical impact loading conditions.
First, a matrix of load cases was produced by conducting a review of the scientific literature and published biomechanical data of live patients with instrumented joint replacement implants, in order to capture a wide range of normal activity load cases and traumatic loading events. As a second modelling input, intact and generic implantable pelvis and femur FE models were generated using live patient CT scan data. These models were implanted with several iterations of the implant designs, allowing progressive informed design development.
Analyses were conducted to verify the ENDURE cup design. First, monoblock CFR PEEK, CoCr and Delta cups were implanted in the replica hemipelvis model, with ideal implant positioning, for three (smallest, largest and mid-range) sizes. The CFR PEEK cup was predicted to generate a reduced remodelling stimulus in the supporting bone than the stiffer metal and ceramic cups, indicating that it would preserve the density of the supporting bone for longer. The worst-case (smallest) size CFR PEEK cup was predicted to have a peak stress of 35.5 MPa under walking loads, giving a safety factor of 2.7 compared to the material's quoted 95 MPa fatigue strength. All cup materials were predicted to produce comparable levels of implant-bone micromotion, indicating that biological fixation through bone in growth would be achievable. Second, FE analysis of incorrect implantation was conducted to verify the cup's strength with excessive inclination causing edge loading. Even at worst-case steep inclination, the peak stress in the worst-case cup did not exceed 50 MPa under walking loads. A third FE analysis modelling the same conditions but including incomplete cup rim support produced a peak stress of 66 MPa, still a safety factor of 1.4 compared to the fatigue strength and 2.3 compared to the tensile strength. Finally, FE analyses of the cup deflection under implantation 'pinch' loading and in-vivo gait loading were conducted, which indicated the likely change in implant shape and therefore the required bearing design clearance to produce the required net clearance for correct bearing performance when implanted and in service.
Second, analyses were conducted to verify the ENDURE head design. The modelling results supported the fundamental design philosophy in terms of the stress in the prosthesis and the strain in the supporting bone. Specifically, it was predicted that incorporation of the stiff ceramic material would not lead to additional adverse bone adaptation in comparison to existing metal implants, thanks to the proposed geometric design developments over the precedent technology. The model indicated that with incorporation of a shortened metaphyseal stem design in comparison to currently clinically used metal resurfacing head designs, the net increase in stress shielded bone volume could be minimised, although the analysis supported the use of conventional cemented implant fixation methods, over less forgiving cementless fixation. The analysis indicated peak ceramic tensile stresses between 143 and 305 MPa, which it should withstand safely, indicating a safety factor of 3.8 compared to the characteristic flexural strength of the material. The models also analysed the implant fixation surface conditions, and indicated a maximum tangential micromotion (sliding) displacement of 35.9 µm and maximum normal (opening) micromotion of 20.2 µm. The scientific literature suggests that bone in-growth can be achieved with small micromotions (in the range 20 - 56 µm), so this supports the generation of a stable implant-bone interface, in theory.
On the basis of the in-vivo predicted stress distribution, mechanical tests were designed with load cases and pass criteria to represent the extremes of in-vivo loading. Three tests were developed to reproduce the three main stress concentrations in the structure. First, a stem test evaluated the bending stress generated at the root of the stem. Second, a crush test evaluated the strength of the head's shell under worst-case implant malpositioning. Third, an ultimate strength test evaluated the burst strength of the whole head with correct positioning. Additionally, an interface torque test determined the strength of the prosthesis-bone interface under worst case loading conditions, where the torque axis aligns with the head axis. All three tests were designed with the FE modelling procedure, and the design was verified in theory when these tests were simulated. The subsequent WP6 conducted these tests physically, for more concrete implant design verification.
Finally, explicit FE analyses were conducted of the proposed ceramic head's structural strength under dynamic impact loading, representing surgical impaction during implantation. The worst-case, highest load impacts were modelled, when the head is finally seated on the bone and its support is therefore stiffest. A variety of potential loading conditions were incorporated considering variability in the angle of applied force from the impactor. An off-axis impactor angle was found to modify the stress distribution in the implant adversely, by introducing bending stresses in its stem. However, the peak stresses were predicted to be 67.5 MPa in the worst (smallest size implant) case, considerably below the strength of the material. Finally, physical impaction tests were conducted to assess the validity of this novel analysis methodology. The characteristic impact force-time profile predicted by the FE model was observed in the physical tests. The FE model was found to over-predict the peak stresses and under-predict the impact duration, probably due to necessary simplifications in the modelling of the test support structure. As such, the models represented a slightly worse than realistic case, so were validated for the verification process.
WP3 - Primary fabrication development
All tasks in this WP were associated with the manufacture of the prosthesis components and the development of the tooling / manufacturing processes needed to produce the components. The work performed focused on the prototype manufacture of the femoral and acetabulum components addressing the hard machining of the ceramic head and the surface finish process to obtain the correct clearances and geometry. During this WP, the potential manufacturing route for the acetabulum cup and the required fixation of the cup into the pelvis along with the surface geometry of the articulating surface was established.
Ceramic manufacturing
In order to manufacture the femoral head a pre form tool of the head was manufacture, this perform tool allowed cylindrical billets to be made from which the green state femoral head where machined. The green state femoral head were generated as over-sized components that would reduce to their near final shape during the following operations.
The green state components were thermally pre-sintered to achieve an approximate 95 % theoretical density. This was followed by a hot isostatic pressing process which gave the final components a 100 % theoretical density and its final strength. The femoral components then went through a hard machining process which formed the outer and inner surfaces to the tight dimensions required. The outer surface was then polished to give the correct RA required and the inner surface was laser marked prior to proof testing and final inspection.
Prototype CF-PEEK manufacturing
Injection mould tooling for the manufacture of the CF-PEEK cups was designed and manufactured specifically for the very exacting demands of the materials. The first set of tooling produces oversized cups with no surface feature on either the inner face or the back of the cup. The idea behind the manufacture of the oversized cups was that these could be machined to form the final cup geometry for testing purposes. At the same time as these were being manufactured, rapid prototypes of the cup were also produced for assessment from the point of ease of manufacture and the benefits of the interlocking with the bone once implanted. Several tests were performed with the novel outer surface geometry to identify the optimum shape to gain the best fixation.
Following a review of all of the parameters required by the cup with respect to the materials properties, mechanical stability and range of movement a series of cup designs were assessed following the review a final design was derived this was the Horseshoe cup were the cup design mimicked the actual design of the acetabulum. The design incorporated the acetabulum notch in the Lunate (articular) surface. The design allows the cup to flex with the adjacent bone giving a more physiological bone strain pattern, to reduce stress shielding, but not to the extent that there is excessive cup-bone interface micromotion, which would result in poor initial stability. The raised area also allowed for loading of the horseshoe segment of the acetabular socket mimicking what usually occurs on the articular cartilage. The thickness of the cup was reduced to conserve bone, and allow a larger bearing diameter to be used whilst minimising dislocation risk. At the bottom of the recess is a hole which is designed to simulate the actual acetabulum allowing for the collection of synovial fluid. It also acts as a location point for the impaction instrumentation.
Having defined the internal geometry of the cup the research focused on the external surface which had to integrate with the acetabulum. Several proposed ideas were generated for an interlocking mechanism with the acetabulum. These included the incorporation of a location pin augmented with fins which once pressed into the acetabulum trabecular bone would give an initial good fixation.
The single pin location was developed together with the alignment instrumentation needed to implant the cup in the correct position. The instrumentation for the alignment of the cup was based on the Peter Ring implant procedure where the location pin runs along the ilium.
Results
As a result of the testing the final design of the femoral head and acetabular cup was established. The femoral head was made from Biolox Delta ceramic using predefined manufacturing protocols, e.g. hot isostatic pressing followed by hard machining and polishing to achieve an Ra of 0.02 µm. The femoral component consisted of the short location stem capable of withstanding a 1.27 kN stem tip load and a shell that will with stand a tensile stress load of 5.6 kN. The inner surface of the head was coated with a Ti and HA coating with thicknesses of 30 µm and 70 µm, respectively.
The acetabulum cup was manufactured from CF-PEEK by advanced injection moulding techniques; once moulded the hole in the bottom of the horseshoe was drilled. The cup has been designed with the single pin location which has been offset to allow for the maximum range of movement. The inner surface has been designed to have a similar Ra value of 0.02µm to that of the femoral component. The back of the cup has been designed to integrate with the dense trabecular tissue within the reamed acetabulum the design carries a series of fins that will mesh with the bone tissue. The final design mimics the actual acetabulum and the material characteristics have been developed to minimise the stress shielding under normal gait conditions.
The outer surface of the cup has a dual coating consisting of a 30 µm Ti layer over the whole surface with an increase in the coating around the circumference of 70 µm for a width of 10 mm to show the surgeon the position of the cup following implantation. This will then be over coated with a 70 µm coating of HA to gain good secondary fixation with the surrounding tissue.
WP4 - Development of secondary coating
Under WP4, an investigation in to the potential coating methods for the integration of both the femoral and acetabulum components was performed. This investigation consisted of vacuum plasma coating trials on the selected materials i.e. a ceramic substrate and samples of CF-PEEK. These samples were coating with hydroxyapatite HA, titanium (Ti) and a combination of Ti and HA. Several trails were made using the different combinations of the coating at different thicknesses and with different substrate preparations such as shot blasting with silica media and with HA powder. The preparation with the HA was performed to reduce the potential of contamination from the alumina blasting media, however the cost of blasting with HA proved significantly more than the cost associated with cleaning off the alumina media residue. Following coating, mechanical testing was performed on the samples to determine the mechanical bond strength of the coating.
Development of dual coating process parameter options
During this task, a study of coating a Ti foam backing bonded to the back of the CF-PEEK acetabulum cup was analysed. Following this extensive research, it was decided not to generate a foam material for the back of the acetabulum cup as this would prove too flexible for good early fixation of the cup. Therefore, our investigations focused on the coating of the CF-PEEK directly with a HA coating and a Ti with HA coating. In order to ensure that a good adhesion was achieved a series of grit blasting tests was performed on the substrate material. Two medias were used for the grit blasting process, these were conventional Alumina and the other media was HA. It was decided to try the HA to prevent the risk of cross contamination from the Alumina media, the results of this test showed that both materials provided a suitable surface finish for the adhesion of the plasma sprayed coatings. However the cost of using the HA was considerable more than that of the alumina media and from a cost exercise this will probably be the media used during production.
For the coating process, it was necessary to reduce the heat transfer into the cup, in ordered to achieve this we had to increase the velocity of the robot arm while at the same time reduce the powder size as this reduced the amount of energy required to melt the powder. The plasma gas used for the trial was approximately 25 000°C and the velocity at which the medial left the gun was in the order of supersonic to twice this speed. For the HA coating, it was decided to use a standard powder that conformed to ISO 13779-1,-2,-3 and -4, the results of the trials showed that it was possible to generate a coating of Ti and HA or just HA on the surface of the cup. It was also possible to generate a coating thickness the same as can be found on conventional cups of around 75 µm and surface roughness of approximately Ra 59.6 µm.
Selection of dual coating process parameters
Several test samples were manufactured these consisted of tensile test samples (standard dog bone shaped) which were coated with both HA and Ti HA following the grit blasting process. The result of the preliminary trails showed that it was possible to coat the samples with both HA and Ti HA and that the adhesion to the samples was good however the process distorted the samples and following the process the samples had bent by up to 2 to 3 mm. Modifications were made to the process and further trails were performed, the results of these trials were successful and produced samples with no distortion.
Following several coating experiments involving Ti and HA in different combination to determine if a Ti coating was needed or if it was possible to just have a HA coating on the back of the cup. It was determined that in order to obtain an adhesion of the coating to the cup and for the surgeon to be able to identify the cup position (post operation) it was decided to have both a Ti and HA coating on the cup. In the first instance a thin Ti coating (30 µm) was applied with a 70 µm HA coating, however, following CT scans of the samples it was determined that this could not be seen by the surgeon especially when implanted next to bone tissue
To make the cup more prominent against both soft tissue and bone the Ti coating was increased around the equator for approximately 10 mm from 30 µm to 70 µm. It was decided to only increase the coating in this rejoin so as to prevent the fins from becoming clogged with the coating media.
WP5 - Development of clinical instrumentation for new generation implant materials
Work comprised of designing and developing the alignment instruments and impaction cap needed during the insertion procedure. Following the design of the acetabulum cup a series of designs for the alignment instrumentation were generated. However, following feedback from the surgeon group, the final design of the alignment instrument was determined and consisted of a device that located in the reamed acetabulum with a location arm that hooked on the ischia spine of the pelvis. Once in position this would allow the location hole for the acetabulum cup to be drilled.
In order to impact the cup in the acetabulum, a series of designs for the impaction cap were generated. The main criterion for the impaction cap was that it fully supported the cup during the implantation procedure thus reducing the possibility of distortion to the CF-PEEK cup. A final design for the impaction cap was generated which gave full support to the cup while locating in the holes at the bottom of the horseshoe. A collet design with a central locking pin holds the cap to the cup. The collet fits into the hole which has the reverse taper to that of the collet and once the locking pin is inserted which open the collet locking the two components together.
Trials with both the alignment instrumentation and the impaction cap were performed both in the laboratory and by a surgeon during the cadaver trials. The results of these test showed that the alignment instrument was able to position the cup in the optimum position to allow for the best range of movement. The result of the testing performed with the impaction cap showed that it held the cup firmly during the impaction procedure and allowed the cup to be fitted into the acetabulum without distortion. The impaction cap also permitted the removal of the cup from the acetabulum.
Clinical instrument development
Following an in-depth study of the instrumentation used for the current implantation of the ADEPT resurfacing implant, it was decided that new instrumentation would not be required for the implantation of the new femoral component as this was similar in design. Only slight modifications to the existing instrumentation were needed in order to fit the femoral components. The main problem was with the development of the acetabulum cup instrumentation; this required the development of a location device that would allow the surgeon to fit the cup in precisely the correct position first time. This was based on the knowledge that once impacted in place it would not be possible to realign the cup without compromising the initial fixation. Therefore, it was decided to incorporate location pins on the back of the cup which would fit into template pre-drilled holes in the pelvis. The second technical challenge was that of location and support of the cup during the implantation process. To prevent damage to the relatively compliant cup during the implantation process the cup requires a large area of contact with the impaction cap. The impaction cap was therefore designed to conform to the same geometry as the internal surface of the cup which not only prevented damage but also ensured that no distortion would occur during impaction. The final obstacle to overcome was that of location of the cup and cap to allow the surgeon to manipulate the cup during the insertion process, following several design concept brainstorms and prototype modelling it was decided to use the hole in the bottom of the acetabulum cup for this purpose. Following the design of this new cap several rapid prototypes were made for analysis and with some final minor adjustments the finished design was determined and prototypes were manufactured. Following consultations with a surgeon on the design of the cup it was decided to remove the three location pins from the back of the cup as there were concerns that one or more of these could protrude through the back of the pelvis if the cup was positioned poorly during surgery. Therefore, a new location feature was designed for the back of the cup which would eliminate this possibility. The new location feature consisted of a single longer stem (pin) that protrudes down the centre of the iliac spine.
The new cup location design meant that a new design for the instrumentation was needed, therefore a design was produced where the alignment instrument locates in the reamed acetabulum but also fixes around the back of the pelvis onto the ishia spine.
Design of the impaction cap
In order to impact the new cup in the acetabulum a new impaction cap was needed as methods used for metallic cups were unsuitable for the PEEK. Therefore, a series of designs for fixing the impaction cap to the cup were generated. These looked at attaching the impaction cap to the cup through the hole in the bottom of the cup.
Following the development of the impaction cap, prototype caps were manufactured for test and analysis. Following the manufacture of the prototype components, injection moulded cup were attached to the impaction caps and implanted into cadavers to test the system.
WP6: Mechanical testing and surface analysis
During this WP, a number of physical tests and simulations were developed for the verification of the design and material integrity of the femoral and acetabular components. In order to conduct the work, a number of key factors were examined:
- the mechanical strength of the ceramic components under in-vivo loading
- the geometric conformity of the acetabular components after implantation,
- the dynamic friction in the bearing between femoral and acetabular components,
- the predicted tribological conditions in the bearing couple,
- the loads experienced by femoral and acetabular components under surgical loading,
- the loads required to remove the acetabular cup when implanted, and
- the resistance to lever-out loads of the acetabular cup when implanted.
Ceramic component testing
Physical testing of representative prototype ceramic heads was conducted using worst case support and implant positioning for three main stress concentrations identified under in-vivo loading, as reported and simulated in WP2. The heads were loaded in servo-hydraulic test machines, and the fracture load recorded.
Results
The heads were required to sustain in excess of 16 kN axial burst loading. Two heads sustained 50 kN without fracture, and the third fractured above 47kN. The stems were required to sustain in excess of 1.27 kN, and sustained a minimum of 3.38 kN. The heads were required to sustain in excess of 5.6 kN axial load, and sustained a minimum of 13.73 kN.
Geometric evaluation of the femoral components
Uncoated and coated cups were implanted into a range of under-reamed hemispherical bores in representative foam biomechanical test material.
The cup deformation was measured, and the resulting frictional torque generated in internal-external rotation articulation was recorded.
Further tests were conducted for the bearings using full gait and stair climbing / descent loading cycles, incorporating flexion-extension and internal-external rotation articulations. This was conducted with a novel test platform incorporating a multi-axis robot to apply the required load magnitude and direction cycles, taken from the literature.
Side loading tests to evaluate the effect of incorrect fitting
Tests and computational analyses were conducted to consider the influence of cup malpositioning upon the integrity of the implant-bone interface and the bearing performance.
First, a test was conducted where cups were pressed-in to representative foam blocks. The cups were then levered out by applying force on the cup ri, representing an impingement load in-vivo, as a result of cup-femoral neck impingement contact, which would be a possibility if the cup were malpositioned.
Second, a finite element analysis was conducted to predict the tribological conditions in the cup and the proximity of the contact patch in the bearing to the edge of the cup's bearing surface. Edge loading on the hemispherical part of the bearing surface would be no different than for currently used sub-hemispherical resurfacing cups, but the novel aspect of this cup, without clinical precedence, was the bearing surface cut-out, so this was the focus of the investigation. The cup was oriented with relatively steep abduction according to current surgical recommendations, and extreme anterior axial rotation. Then the contact patch location was predicted simply with computer-aided design (CAD) geometric analysis, and the contact pressure was predicted in more detail with a FE analysis model. This was conducted for load cases representing normal gait, stair ascent and descent, and deep-flexion rising from seated. The peak joint contact force instant of each cycle was applied in quasi-static models. Analyses were run for the same range of bearing clearances tested previously (250 µm, 500 µm and 750 µm). The peak stress in the cup was also predicted in these scenarios, to support WP2's prior cup strength analysis.
Results
In physical testing, average torques of 16 Nm to 34 Nm were required to lever the cup out of its supporting foam block, with the lever-out torque increasing with a greater interference fit between the bone and the cup. The lever out resistance was found to be considerably lower when uncoated cups were tested.
In the CAD analysis, the contact patch was predicted to lie within the bearing surface for gait and stair ascent / descent activities. However, in the deep flexion 'rising from seated' load case, the contact patch was predicted to reach the edge of the acetabular notch in the bearing surface.
The peak contact pressure predicted was 21.6 MPa in stair ascent, when the bearing clearance was largest (750 µm). The contact patch was predicted to reach the edge of the bearing surface during rising from seated when the patch was largest (250 µm clearance), but owing to the lower magnitude of the joint contact force in this case, the pressure was not expected to exceed 9.3 MPa, which was lower than the peak contact pressure for the stair descent load case.
A combined computational modelling study and physical test programme was conducted to assess the strength of the femoral head implant under surgical impaction loading. A prototype head prosthesis was cemented onto a representative shape and stiffness foam biomechanical test material cylinder, and attached to a load cell. A head impactor instrument used in current resurfacing surgery was obtained, and modified to feature a slide hammer.
The force-time characteristic of the impact, and the calculated cumulative impulse over time, were analysed in comparison to the FE model results, for model validation. Two models had been produced, including and excluding the mass and stiffness influence of the surgeon's hand, holding the impactor. Ten impact loads were required to verify the head's strength against impaction loads.
Results
During the tests, the head was impacted ten times without fracture. The force-time characteristic recorded in the physical tests and the calculated cumulative impulse both showed good correlation with both basic and hand-damped FE models in terms of peak force and duration. The basic (undamped) model over-predicted the peak force by 49 %, but agreed closely with the cumulative impulse. The hand-damped model only over-estimated the peak force by 23 % but underestimated the cumulative impulse by 16 %, so the basic model was selected as a worst-case.
WP7: Tribological and geometry study
WP7 aimed to 'optimise the wear and friction performance' of the prosthesis components. Deliverables were:
- to provide details for the wear and friction systems and methods ready for tribological testing, and
- to derive the optimal geometry for the ceramic on polymer bearing surfaces with respect to their tribological performance.
Two main approaches were used to assess the design's tribological characteristics:
- FE analysis of the contacting bodies, coupled with theoretical contact calculations, and
- wear testing using a 10-station ProSim hip wear simulator.
The theoretical calculations were used to predict the performance of three bearing clearance options for the proposed Biolox delta ceramic on carbon fibre reinforced PEEK (CFRPEEK) prosthesis. Then, the wear simulator testing was conducted to assess the performance of the preferred design relative to two traditional bearing couples: ceramic on ultra-high molecular weight polyethylene (UHMWPE) and ceramic on highly cross linked polyethylene (HXLPE).
The theoretical predictions indicated that the ENDURE bearing is likely to operate in a boundary lubrication regime, and literature review showed that this is consistent with typically employed bearing clearances with this material combination, which are insufficient to achieve fluid film lubrication. The literature also indicates that friction factor in similar ceramic-CF-PEEK bearings is independent of clearance, provided the clearance is above 250 µm. Therefore the design clearance must be sufficient that implantation and service deformations do not reduce the in-vivo clearance to below this level.
FE indicated that the deformation under normal service loading is likely to be small in comparison to the bearing clearance. Deformation under press-fitted implantation is likely to be higher, and should be the focus of further research.
Finally, wear testing results for simple hemispherical total hip replacement cups of the correct material combination and clearance were obtained which showed no significant difference in wear between CFRPEEK cups and XLPE cups representative of those in clinical use. For this WP, we have combined the tasks:
7.1 Established testing systems
7.2 Literature search and review of similar designs and testing methods and task
7.3 Wear simulation testing in to one report that outlines all of the work performed as these task were so closely related.
Hard-on-soft hip replacement bearings (principally metal-on-ultra-high-molecular-weight-polyethylene (UHMWPE)) have been seen to function well up to 25 years in-vivo. However, the use of this bearing combination is unsuitable for younger patients and resurfacing. For patients below the age of 50 years the survivorship rate is only 79.3 % at 11 years. This is due to the patient's increased activity levels, resulting higher loads and greater number of cycles experienced by the implants adding to the chances of revision surgery. Osteolysis, the body's immune response to polyethylene wear formed through local mechanical damage on the contact surface has been suggested as the major contributing factor to the failure of total hip replacements. Wear particles are generated due to articulating mixed / boundary lubricated contact between the bearing surfaces. In order to reduce the incidence of osteolysis, the wear properties of the artificial joint need to be improved to reduce the volume and bioactivity of the wear debris generated.
Hard-on-hard (MoM or ceramic-on-ceramic (CoC)) bearings were introduced to encourage fluid film lubrication, in which the load is fully supported by a lubricating film, eliminating contact between the bearing surfaces to extend the lifetime of the prosthesis. Nevertheless, achieving fluid film lubrication in a MoM bearing is not instantaneous and involves an initial 'running-in' period in which wear rate is high. The film formation mechanism begins when the MoM components are in relative motion and are loaded. At first implantation, a full lubricant film cannot develop due to small polar bearing point contact between the new head and cup surfaces. Postoperative activities result in wear occurring at a high rate and the area of contact between the surfaces increases. This 'running in' phase ends when the clearance in the contact patch reduces towards zero and the geometry is most favourable for fluid-film lubrication. As migration into the 'steady state' phase occurs and the wear rate slows the contact patch continues to develop at a reduced rate. The effective clearance within the contact patch and within the surrounding entraining geometry is then slowly consumed by wear. Although the wear rate during the steady state is low there is concern over the initial high running in wear, releasing metal ions which are dispersed leading to systemic exposure and potentially have a negative effect on the patient's health.
A recent approach to improve the success of a joint prosthesis is to improve the wear resistance and mechanical properties of the acetabular cup material. CFRPEEK has been suggested as a potential bearing material due to its high wear resistance and mechanical properties which will allow the material to function alone in an acetabular component without the requirement for a metal backing shell for support. In addition, PEEK's proven biocompatibility indicates long term use without an adverse reaction in-vivo due to the material's excellent chemical stability which is a particular problem associated with the wear debris produced by UHMWPE and metal cups. Wang et al. (1999) carried out an investigation into the influence of carbon fibre content and wear resistance. The study concluded that PEEK resin blended with 30 w/w% of PITCH carbon fibre gives the optimal wear resistance, five times lower than 10 % reinforcement and three times lower if 50 % reinforcement was used.
Pin on plate studies have shown a combination of CFRPEEK and ceramic exhibits a five times reduction in wear factor compared to UHMWPE and CoCrMo combinations. Unlike, hip simulators pin on plate experiments do not replicate the conditions in-vivo. Instead the test is useful as a material screening technique to compare new potential materials.
To gain a greater understanding of wear rates of PEEK in a set up that more closely simulates the hip motion and loads through the components, hip simulator studies have been performed. The first hip simulator results were reported by researchers from Howmedica who performed an investigation on injection moulded 30 % discontinuous pitch carbon fibre PEEK acetabular cups articulating against zirconia heads. The CFRPEEK combination exhibited two orders of magnitude less wear than the ceramic-UHMWPE combination after 10 million cycles.
Materials and methods
Test 1: Theoretical analysis and modelling
Wear test evidence from the literature indicates that CF-PEEK cup prostheses will operate in mixed-boundary lubrication. The following is an analysis of the contact conditions in the proposed ENDURE cup, aiming to identify a target clearance, contributing to the detailed design. This study employed theoretical predictions and an FE analysis model. Three 48 mm diameter cups were analysed: the nominal cup as drawn (clearance 750 µm), plus variants with 500 µm and 250 µm. The contact pressure distribution and patch location was predicted under loads representing gait, stair ascent, stair descent and rising from seated.
The FE model from WP6 report 1 was re-used for this analysis, based on the ENDURE cup set in a cylindrical block of PMMA and loaded with a rigid spherical head.
Considering micro-elastohydrodynamic lubrication (uEHL), formulae from Hamrock and Dowson were employed to predict the lubricant film thickness 'hmin' in spherical-spherical bearings and to calculate the lambda ratio of the loaded bearing which indicates its likely lubrication regime. The lambda ratio compares the lubricant film thickness to the maximum surface roughness. A lambda ratio above three implies the achievement of fluid film lubrication, below 1 implies boundary lubrication and between 1 and 3 indicates a mixed lubrication regime. These formulae were input into an Excel spreadsheet and verified using example results given by Dowson.
The approach used in this study was to predict the minimum film thickness for several bearing clearances for the full size range of ENDURE cups, using the same service loading from the Dowson paper. The surface roughness of the CF-PEEK injection mouldings was not known at the time of writing. Therefore, with a surface roughness for the ceramic estimated as Ra=20 nm (range 7 nm - 50 nm for ceramic femoral heads in general), the equations were used to predict the threshold CF-PEEK surface roughness values to produce boundary or fluid film lubrication regimes.
With the uEHL calculation predictions, the results parameter was the maximum CF-PEEK roughness to permit mixed or fluid film lubrication under gait loading, across the full ENDURE size range. With the FE Model, the main output parameters were the peak contact pressure and the radial deformation of the cup rim, indicating the additional clearance required in the unimplanted prosthesis. This was calculated as the designed rim diameter minus the diameter of the maximum inscribed circle (MIC) fitted to the rim of the loaded, deformed cup. The MIC diameter was calculated using an algorithm written into a post--processing subroutine, which performed similar measurements to roundness testing machines.
Test 2: Physical wear testing
To assess the suitability of CFRPEEK as an acetabular component material and assess a few of the conditions described previously, long term wear tests were performed on pitch-based CFRPEEK cups articulating against Biolox Delta alumina femoral heads. Identical HXLPE and UHMWPE acetabular cups articulating against Biolox Delta alumina femoral heads were also tested for comparison.
The wear samples were nominal diameter 36 mm hip prostheses of Biolox Delta ceramic (CeramTec GmbH, Germany) heads against acetabular cups of three different materials - CFRPEEK (Invibio Ltd, UK), UHMWPE and HXLPE. Testing samples included four CFRPEEK, three UHMWPE and three HXLPE cups and were soaked in deionised water for a minimum six weeks before the wear test.
The test was carried out according to ISO14242 with the following conditions:
- Load: 50 - 3000 N single axis twin peak Paul type load curve
- Movement: ±10° internal / external and -15° to +30° extension / flexion
- Frequency: 1 Hz
- The cups were installed with an inclination angle of 35° in the medial-lateral plane and a single dynamic force was applied along the vertical axis
- Fluid: 25 % newborn calf serum with 0.1 % sodium azide
- The test was carried out for five million cycles
- The solution was changed at least every 0.5 million cycles
- Interruptions were made at 0.5 1, 2, 3, 4 and 5 million cycles for gravimetric determination of the wear and worn surface analysis
- Prior to the gravimetric measurement of wear, components were washed in detergent water, thoroughly rinsed in water and deionised water, cleaned in alcohol using an ultrasonic bath for at least 10 minutes to ensure removal of all traces of debris and lubricant, and then left in an atmosphere controlled room for at least 24 h to dry and thermally stabilise
- Wear was determined gravimetrically using an electronic balance (Sartorius ME235S, Germany) to a precision of 0.01 mg
- The gravimetrical wear was converted into geometrical wear using specific density of 1.41 for CFRPEEK, 0.934 for UHMWPE and HXLPE, and 4.36 for Biolox Delta ceramic
- The worn surfaces were analysed using the Alicona InfiniteFocus 3D optical microscope.
Results and discussion
Micro-elasto-hydrodynamic (µEHL) equation results
µEHL calculations led to the prediction of minimum CF-PEEK Ra roughness values to permit mixed and fluid film lubrication regimes to be achieved. These show that for the full size range, mixed lubrication should be achievable with Ra values of 32.9 nm, 48.3 nm and 86.6 nm for the 250 µm, 500 µm and 750 µm clearance cups respectively. Due to its smallest equivalent radius, the lowest film thickness was generated in the smallest size. Fluid film lubrication was predicted to be achievable with an Ra value of 21.8 nm for the 250 µm clearance cup. FFL would not be achievable for the smaller cups in the size range with clearances of 500 µm and 750 µm, because the 20 nm Ra roughness of the ceramic would take up a sufficient proportion of the lubricant film thickness that separation of the surfaces would not be possible.
Scholes et al measured the roughness of the CFR-PEEK MITCH cup during 25 million wear cycles. The initial Ra was approximately 3300 nm, and reduced to approximately 393 nm after the wear test. If these roughnesses are typical of what can be achieved in manufacturing of the ENDURE cup, boundary lubrication is to be expected.
Recent literature studies have reported that clinically used ceramic-on-CF-PEEK bearing couples (MITCH) typically operate in boundary or mixed lubrication regimes, so the failure to achieve FFL may not be of concern. The MITCH cup, which this literature data refers to, had a larger nominal diametric clearance (0.80 - 0.91 µm) but greater ability to deform with its full cut-out horseshoe shape. The wear simulator data reported would probably not include the deformation resulting from press-fitting, so represents a worse case (high clearance). Older literature studies from the MITCH development process tested lower clearances and saw high friction with zero clearance, but no influence of clearance on friction for clearances between 250 µm and 750 µm.
The implication is that in the fully hemispherical, less flexible ENDURE design, the selection of bearing design clearance should be informed by prevention of cup clamping, reducing in-vivo clearance to zero.
WP8 Sterilisation and biocompatibility
This WP addresses the serialisation, packaging and biocompatibility of the new hip components and the instrumentation needed for its insertion. During this WP, a review of the materials to be used in the construction of both the femoral and acetabulum components was performed. Following a literature review, it is clear that the materials selected for the implants have already history in the medical field in similar applications and therefore problems with the sterilisation and biocompatibility were considered to be low risk. As part of the work performed, the materials to be used and the possible configuration of the coatings to be applied was identified. Having identified the materials to be used, a list of tests that would be needed in order to ascertain the biocompatibility was generated. Another key part of this work package was the development of handling and packaging protocols to allow the final products to be assembled where necessary and packaged in a form the was both cost effective and easy to serialise.
For further details please refer to the related project report.
The treatment of younger patients with severe hip disease using a conventional total hip arthroplasty (THR) presents a challenge. The success rate of the THR is currently very low within the first 16 years of implantation. To overcome this Finsbury instruments, in conjunction with surgeons, Midland Medical Technologies Ltd (MMT) and Doncaster's Centaur Precision Castings, developed the metal on metal (MoM) Birmingham hip resurfacing (BHR) system, which was developed for MMT. However, the MMT was subsequently bought by Smith and Nephew (United States of America (USA) based company) during 2005. Evidence suggests that the MoM resurfacing provides an ideal solution for implantation into younger patients who have a more active lifestyle with the survival rate being 93 % after 3 to 5 years over a conventional THR. However, there is concern about metal wear debris and its systematic distribution throughout the body. Several studies have reported high concentrations of cobalt and chromium in the serum and/or urine of patients with MoM resurfacing implants. Potential toxicological effects of the elevated metal ions have increased concerns about safety of MoM implants. This is a particular concern in younger and active patients in whom life expectancy after implantation is longer. Because patients who receive a MoM hip arthroplasty are shown to be exposed to high concentrations of metallic ions, the Medical Advisory Secretariat performed literature searches for the adverse biological effects of cobalt and chromium. Cobalt and chromium make up the majority of the substrate material used in the manufacture of MoM implants.
Concerns have been raised recently due to high levels of cobalt and chromium measured in blood and serum samples of patients with MoM devices. These concerns relate to the possible systemic effect of metal ions released by MoM prosthesis in terms of genomics, as well as immulogy, and histology, and reviews have recently been published, into the ion levels and their biological effects were marked when wear was abnormally increased, with low carbon (< 0.2 %) implants and loose or poorly positioned implants.
In November 2006, a meeting was organised by Dr Case, University of Bristol, United Kingdom (UK), with a panel of world experts in various medical fields where metal ions were thought to have a potentially detrimental effect. It was concluded that a risk to benefit analysis should performed for individual patients wishing to have a MoM implant. The long-term exposure concerns were related more specifically to younger, more active patients wishing to have children and to maintain their active life style. It was concluded that a risk benefit analysis of the patient situation should be implemented. For those patients wishing to have children post joint replacement, it was decided that further follow up studies were required, with a particular focus on the potential passage of the genetic damage to offspring in younger patients.
MoM resurfacing technology is superior to that of conventional THR implant for younger patients; however, the MoM has a series of potentially long-term problems of its own. To overcome these problems, the partners in the ENDURE project developed a ceramic on polymer resurfacing hip prosthesis. Like that of the MoM prosthesis the ceramic on polymer resurfacing prosthesis will only cover the femoral head thus retaining all of the desirable properties associated with retaining an intact femur.
Using the following key innovations developed during the course of the Endure project the partners we able to produce a new resurfacing hip prosthesis that gave significant advances over current existing systems.
These key innovations include:
- fabrication of a resurfacing (BHR-type) femoral, hemispherical ball-head and cup from ceramic and polymeric materials;
- the use of new materials, and processing methods to achieve a non metallic low ware resurfacing hip system;
- development of a flexurally tough alumina matrix composite (AMC) ceramic femoral head and thin walled reinforced polymer acetabulum cup;
- an innovative coating process that will allow for full bone integration with the implant;
- a novel design that includes the bone as part of the construction;
- an advanced low impact surgical procedure to reduce trauma during the operation.
The ENDURE system resurfaces the femoral head similar to that of the MoM system but unlike this system the ceramic wear particle will remain inert within the body thus eliminating any concerns relating to the long term effects of ions. It is well known that the use of ceramic and polymeric materials will lower the coefficient of friction between the mating surfaces and as a result will reduce the amount of wear and thus debris generated during the normal operation of the joint.
Overall project objectives
Our main objective was to develop a high durability ceramic on polymer resurfacing hip prosthesis that has a long working lifetime ideally in excess of thirty years. This was be achieved by unique combination of the best, established technologies, plus the additional advantages of novel AMC bioceramics, new processing methods and polymeric compounds. In order to realise these objectives, we broke them down into the following categories.
Scientific objectives:
Our scientific objectives were those relating to issues of durability, via influences in physical wear phenomena and biocompatibility.
Technological Objectives for the device:
The technological objectives were to acquire the science to enable the development of the first durable, long-life hip resurfacing joint prosthesis that would:
- produce a low friction / wear coupling by accurate and precise grinding of the component faces to the specification and tolerance identified during the work packages (WPs);
- achieve volumetric wear rates = 0.8 mm3 per million cycles under simulated steady state in-vivo lubrication conditions, by accurate and precise machining to asphericity specifications and tolerance identified;
- reduce the size of the wear particles so that they do not exceed 0.1 µm, by application of design specification from WP1;
- avoid adverse sensitivity and have complete biocompatibility by choice of AMC as the bioceramic and polymeric / CF PEEK materials used for the construction;
- ensure complete and stable osseointegration within the first three months following surgery by a coating combination of titanium and hydroxyapatite, applied through the development of the vacuum plasma spray technique;
- improvement of operative technique and instrumentation to improve optimise alignment and reduce the risk of femoral neck notching to achieve optimal range of motion and avoid post-operative femoral neck fractures;
- increase of candidate patients including women of child bearing age.
Economic objectives:
As well as the technical objectives the consortium partners also addresses the economic need for the technology developed during the course of the project. The results of the economic study showed that by the end of year 5 after the project, and through a network of trans-national and cross-sectoral licensees, the consortium would be able to sell the new joint technology over the time period.
Enabling innovation related objectives:
To achieve the societal and economical objectives that come from the dissemination and exploitation of the research results the consortium partners have defined an enabling set of objectives.
To enable innovation through the project team and to benefit Europe the objectives are:
- collation and preparation of the results of the project into a suitable format and apply for patent protection of the results of the project covering the manufacture of the novel AMC ceramic / polymeric hip technology;
- the transfer knowledge from the research and technology development (RTD0 performers to the small and medium-sized enterprise (SME) participants has been achieved using technology transfer events and interactions;
- to disseminate the results and benefits of the knowledge and technology developed beyond the consortium to potential users such as healthcare / surgical, dental, veterinary, scientific.
Societal and policy objectives are to benefit society by:
- reducing the number of surgical revisions following implantation of a medical device;
- as a result of the reduction in the numbers of people requiring revision surgery, the ENDURE project will have the affect of allowing reallocation much-needed medical resource to other areas.
Project results:
In order to deliver the work within the ENDURE project each objective was broken down into a separate WP.
WP1 - Design foundation
The objective of WP1 was to define the design envelopes for the femoral and acetabular components.
Literature search for the material data
During this task, several aspects of the prosthesis component design were considered and our research showed that currently the most common bearing components used is metal on polyethylene. The polyethylene is housed in a metal backing and implanted on the acetabular side. Other common bearing types are metal on metal and ceramic on ceramic. Although some are more successful than others, all of them still experience problems. The common problems associated with each implant range from high wear rates, stress shielding, osteolysis to implant loosening. In order to improve the longevity of the prosthesis, these problems must be overcome. This could potentially be achieved by the use of a new material implant combination.
Ceramic materials have been used successfully for the manufacture of hip joint replacement. These are used as a bearing material, articulating against itself or against polyethylene. These devices includes a metal outer shell in order to allow for an in growth surface to be added to the outer surface of the component, which would not be achievable on ceramic or polyethylene materials alone. The in growth surface is critical to the short and long term fixation between the implant and the surrounding bone. However, problems with fixation on the acetabular side are often reported, and loose cups are a common reason for revision in hip replacement. These problems are attributed to the high stiffness of the acetabular replacement component, due to the presence of a stiff metal shell. Polymer materials offer a solution to the stiffness mismatch. Various polymer materials have been successfully used in other areas of orthopaedics (spinal) and in dental applications. This has led to a growth of interest in polymers for use on the acetabular side of hip replacements. A polyaryletherketone (PEEK) biomaterial launched by Victrex in 1998 and marketed for medical use has many advantages over metals including elimination of imaging artefacts and the ability to view tissue/bone growth and repair using x-rays. The addition of a filler material in the PEEK namely carbon fibre formulates a different grade called CF-PEEK. The materials properties can be adjusted by varying the percentage of carbon fibre inclusion; this is of particular interest to the design of a replacement construct where some compliance is required at the bone interface whilst bulk strength is a basic requirement for the component itself. The properties of both filled and unfilled PEEK materials must be fully understood and verified for their use in-vivo and to determine where CF-PEEK is more appropriate to use over unfilled PEEK.
Required range of motion (ROM)of THR implants
In normal THA prosthesis, the range of motion is maximised through the use of a large head diameter and a small neck diameter on the femoral stem component which enhances the hip performance by allowing greater range of motion before the components impinge (Chandler et al., 1982; Barrack, 2003; Bengs et al., 2008). Hip resurfacing decreases the head to neck ratio as the natural femoral neck is retained. Impingement can occur when the retained femoral neck abuts against the acetabular component or anterior acetabular bony wall. This can be painful and restrict motion. The ISO standard 21535 states that 100° flexion / extension; 60° abduction / adduction and 90° internal / external rotation is required. However, even the earliest literature states that a greater range of motion needs to be achieved in order to allow everyday tasks to be performed. Johnson and Smidt (1969) and Roach and Miles (1991) have shown flexion / extension values in the range of 125 - 135° which is higher than the value recommended in the standard. Clearly if THA implants are designed to this specification, the patient's post operative ROM will be limited.
Kluess et al. (2008) performed a study on how the implant designs affect the flexion / extension and internal and external rotation angles that are achieved post-operatively. The results of the study highlight that the current implant designs do not achieve the required ROM in all patients. Using the data gather from the literature searches, it was possible to generate a simple finite element (FE) model that considered various potential loading scenario, simulating activities more demanding than the conventionally modelled walking gait, and a more accurate prediction of worst cases stresses generated within the construct during service was obtained.
Design aspects specific to ceramic material use
FE analysis studies have particular use in the pre-clinical analysis of new prosthesis designs and are used to make safe design decisions with regards to performance. FE models also provide technical evidence during the regulatory submission process, a necessary step in most new product development projects. Whilst it is relatively easy to simulate ISO tests of implant components mounted in simple elastic supports, it can be argued that this does not represent the whole range of typical loading conditions experienced by an implant in-vivo. Even to begin to recreate these conditions, it is necessary to mount the implant in realistic supporting bone with geometry with materials properties derived from computer tomography (CT) data. The FE model developed to assess the stresses the component may be subjected to during its life in vivo was developed using state of the art methods in order to improve its accuracy. In particular, surgical data (CT scanned bones) were used to generate the bone models. The establishment of design requirements, including pass criteria for the output of the FE analysis, is critical to the safety of the design.
FE models of resurfacing femur
The following most fundamental design criteria were considered in this study:
- The volume of damaged bone under stumbling and falling loading conditions must be lower for the new designs than for the ADEPT or BHR prostheses, hereby referred to as 'the traditional design',
- The percentage volume of bone with sufficient remodelling stimulus to generate stress shielding and hypertrophy around the prosthesis will be lower than for the traditional design.
- The peak stresses in the new design must be considerably below the fatigued fracture strength for the material. Absolute strength tests are beyond the scope of this report, but extreme stumbling loads must be sustainable safely.
Details for ceramic manufacturing WP
Ceramic heads were produced by the following process:
1. powder batch preparation, e.g. wet milling and spray-drying powder granulation,
2. pressing into a cylindrical billet, of low theoretical density and strength,
3. green machining, into an approximate, over-sized shape,
4. thermal 'presintering' to achieve approximately 96 % theoretical density,
5. hot isostatic pressing (HIP) in oil or gas to achieve approximately 100 % theoretical density and strength,
6. hard machining of the bearing surface, to precise, tightly toleranced dimensions,
7. polishing the bearing surface to the required surface roughness,
8. laser marking with part and serial numbers,
9. proof testing, and
10. final inspection.
The manufacturing process has been developed to produce heads of adequate strength to avoid fracture in-vivo, through careful control of grain size (in powder processing), volumetric flaw content (in pressing, sintering and HIP) and surface flaw content (in progressive machining and polishing). The diameter, roundness and sphericity of the final bearing surface are critical to a bearing component's wear performance, so the manufacturing process was also designed to allow target bearing dimensions to be achieved within tight tolerances.
The particular difficulty with the production of ceramic resurfacing heads is the lower wall thickness than modular heads. This is relevant to the bearing surface dimensions, as gripping the piece for machining in a chuck may distort it, so that it deviates from the required roundness when it springs back upon release. Although specific details of manufacturing processes, fixtures and target dimensions and tolerances cannot be released, this work package involved:
- further development of pressing of larger sized billets for green machining of resurfacing heads and other CT products with large diameters based on existing and established technology,
- development of machine mounting fixtures to allow the thin-walled resurfacing component to be gripped and machined with sufficiently low distortion to give adequate final roundness and sphericity.
This permitted a series of resurfacing head shaped prototype prostheses to be produced, allowing mechanical testing to commence. The approach involved developing the existing manufacturing technique with as few changes as possible, as past clinical results of other manufacturers have demonstrated the risk associated with departures from known techniques.
Both the design and the manufacturing techniques were verified with mechanical testing.
Design aspects specific to CF-PEEK material use
There are many design features that are required in order to make a new resurfacing hip system successful. The areas that required extensive research were the geometry and tolerances between surfaces as well as the overall design. It was important to also include the design principles of the dedicated surgical instruments.
From the material data report it was confirmed that the femoral head would be manufactured from Biolox ceramic, whilst the acetabular cup would be made of carbon fibre reinforced polyetheretherketone (CF-PEEK) or a combination of CF-PEEK and unfilled PEEK.
A literature review was performed on performance of the current acetabular cup designs that employ CF-PEEK as their bearing surface which include the Cambridge and MITCH cups. Fixation methods in the form of a coatings and porous layers to encourage bony in-growth and improve initial fixation of the cup were also explored.
The investigation also looked at patents on PEEK acetabular cups, coatings and porous outer surface.
Designs ideas discussed during the ENDURE kick-off meeting were reviewed to determine the optimum option and each design was scored against the following criteria: complexity, manufacturability, bony in-growth, surface finish, novel idea, initial fixation, coating ability and difficulties to grade each design. The optimum design includes a multiple layered design consisting of CF-PEEK MOTIS as the bearing surface and a porous layer of PEEK Optima.
WP2 - Implicit and explicit computational analysis
The aim of WP2 was to produce computational analysis tools:
1) to inform the design of the developing implant concepts,
2) for their initial theoretical structural verification, and
3) to enable physical tests to be developed that would reproduce clinically representative stress distributions in the implant components, for their more concrete physical test verification.
The goals of the structural analysis were to predict:
- the bulk stresses in the bone supporting the implants,
- the stimulus and progression of bone remodelling around the implants and
- the fixation interface conditions, indicating their long-term stability,
- the bulk stresses in the implants under a range of in-vivo loading conditions (normal and traumatic scenarios), and
- the bulk stresses in the implants under surgical impact loading conditions.
First, a matrix of load cases was produced by conducting a review of the scientific literature and published biomechanical data of live patients with instrumented joint replacement implants, in order to capture a wide range of normal activity load cases and traumatic loading events. As a second modelling input, intact and generic implantable pelvis and femur FE models were generated using live patient CT scan data. These models were implanted with several iterations of the implant designs, allowing progressive informed design development.
Analyses were conducted to verify the ENDURE cup design. First, monoblock CFR PEEK, CoCr and Delta cups were implanted in the replica hemipelvis model, with ideal implant positioning, for three (smallest, largest and mid-range) sizes. The CFR PEEK cup was predicted to generate a reduced remodelling stimulus in the supporting bone than the stiffer metal and ceramic cups, indicating that it would preserve the density of the supporting bone for longer. The worst-case (smallest) size CFR PEEK cup was predicted to have a peak stress of 35.5 MPa under walking loads, giving a safety factor of 2.7 compared to the material's quoted 95 MPa fatigue strength. All cup materials were predicted to produce comparable levels of implant-bone micromotion, indicating that biological fixation through bone in growth would be achievable. Second, FE analysis of incorrect implantation was conducted to verify the cup's strength with excessive inclination causing edge loading. Even at worst-case steep inclination, the peak stress in the worst-case cup did not exceed 50 MPa under walking loads. A third FE analysis modelling the same conditions but including incomplete cup rim support produced a peak stress of 66 MPa, still a safety factor of 1.4 compared to the fatigue strength and 2.3 compared to the tensile strength. Finally, FE analyses of the cup deflection under implantation 'pinch' loading and in-vivo gait loading were conducted, which indicated the likely change in implant shape and therefore the required bearing design clearance to produce the required net clearance for correct bearing performance when implanted and in service.
Second, analyses were conducted to verify the ENDURE head design. The modelling results supported the fundamental design philosophy in terms of the stress in the prosthesis and the strain in the supporting bone. Specifically, it was predicted that incorporation of the stiff ceramic material would not lead to additional adverse bone adaptation in comparison to existing metal implants, thanks to the proposed geometric design developments over the precedent technology. The model indicated that with incorporation of a shortened metaphyseal stem design in comparison to currently clinically used metal resurfacing head designs, the net increase in stress shielded bone volume could be minimised, although the analysis supported the use of conventional cemented implant fixation methods, over less forgiving cementless fixation. The analysis indicated peak ceramic tensile stresses between 143 and 305 MPa, which it should withstand safely, indicating a safety factor of 3.8 compared to the characteristic flexural strength of the material. The models also analysed the implant fixation surface conditions, and indicated a maximum tangential micromotion (sliding) displacement of 35.9 µm and maximum normal (opening) micromotion of 20.2 µm. The scientific literature suggests that bone in-growth can be achieved with small micromotions (in the range 20 - 56 µm), so this supports the generation of a stable implant-bone interface, in theory.
On the basis of the in-vivo predicted stress distribution, mechanical tests were designed with load cases and pass criteria to represent the extremes of in-vivo loading. Three tests were developed to reproduce the three main stress concentrations in the structure. First, a stem test evaluated the bending stress generated at the root of the stem. Second, a crush test evaluated the strength of the head's shell under worst-case implant malpositioning. Third, an ultimate strength test evaluated the burst strength of the whole head with correct positioning. Additionally, an interface torque test determined the strength of the prosthesis-bone interface under worst case loading conditions, where the torque axis aligns with the head axis. All three tests were designed with the FE modelling procedure, and the design was verified in theory when these tests were simulated. The subsequent WP6 conducted these tests physically, for more concrete implant design verification.
Finally, explicit FE analyses were conducted of the proposed ceramic head's structural strength under dynamic impact loading, representing surgical impaction during implantation. The worst-case, highest load impacts were modelled, when the head is finally seated on the bone and its support is therefore stiffest. A variety of potential loading conditions were incorporated considering variability in the angle of applied force from the impactor. An off-axis impactor angle was found to modify the stress distribution in the implant adversely, by introducing bending stresses in its stem. However, the peak stresses were predicted to be 67.5 MPa in the worst (smallest size implant) case, considerably below the strength of the material. Finally, physical impaction tests were conducted to assess the validity of this novel analysis methodology. The characteristic impact force-time profile predicted by the FE model was observed in the physical tests. The FE model was found to over-predict the peak stresses and under-predict the impact duration, probably due to necessary simplifications in the modelling of the test support structure. As such, the models represented a slightly worse than realistic case, so were validated for the verification process.
WP3 - Primary fabrication development
All tasks in this WP were associated with the manufacture of the prosthesis components and the development of the tooling / manufacturing processes needed to produce the components. The work performed focused on the prototype manufacture of the femoral and acetabulum components addressing the hard machining of the ceramic head and the surface finish process to obtain the correct clearances and geometry. During this WP, the potential manufacturing route for the acetabulum cup and the required fixation of the cup into the pelvis along with the surface geometry of the articulating surface was established.
Ceramic manufacturing
In order to manufacture the femoral head a pre form tool of the head was manufacture, this perform tool allowed cylindrical billets to be made from which the green state femoral head where machined. The green state femoral head were generated as over-sized components that would reduce to their near final shape during the following operations.
The green state components were thermally pre-sintered to achieve an approximate 95 % theoretical density. This was followed by a hot isostatic pressing process which gave the final components a 100 % theoretical density and its final strength. The femoral components then went through a hard machining process which formed the outer and inner surfaces to the tight dimensions required. The outer surface was then polished to give the correct RA required and the inner surface was laser marked prior to proof testing and final inspection.
Prototype CF-PEEK manufacturing
Injection mould tooling for the manufacture of the CF-PEEK cups was designed and manufactured specifically for the very exacting demands of the materials. The first set of tooling produces oversized cups with no surface feature on either the inner face or the back of the cup. The idea behind the manufacture of the oversized cups was that these could be machined to form the final cup geometry for testing purposes. At the same time as these were being manufactured, rapid prototypes of the cup were also produced for assessment from the point of ease of manufacture and the benefits of the interlocking with the bone once implanted. Several tests were performed with the novel outer surface geometry to identify the optimum shape to gain the best fixation.
Following a review of all of the parameters required by the cup with respect to the materials properties, mechanical stability and range of movement a series of cup designs were assessed following the review a final design was derived this was the Horseshoe cup were the cup design mimicked the actual design of the acetabulum. The design incorporated the acetabulum notch in the Lunate (articular) surface. The design allows the cup to flex with the adjacent bone giving a more physiological bone strain pattern, to reduce stress shielding, but not to the extent that there is excessive cup-bone interface micromotion, which would result in poor initial stability. The raised area also allowed for loading of the horseshoe segment of the acetabular socket mimicking what usually occurs on the articular cartilage. The thickness of the cup was reduced to conserve bone, and allow a larger bearing diameter to be used whilst minimising dislocation risk. At the bottom of the recess is a hole which is designed to simulate the actual acetabulum allowing for the collection of synovial fluid. It also acts as a location point for the impaction instrumentation.
Having defined the internal geometry of the cup the research focused on the external surface which had to integrate with the acetabulum. Several proposed ideas were generated for an interlocking mechanism with the acetabulum. These included the incorporation of a location pin augmented with fins which once pressed into the acetabulum trabecular bone would give an initial good fixation.
The single pin location was developed together with the alignment instrumentation needed to implant the cup in the correct position. The instrumentation for the alignment of the cup was based on the Peter Ring implant procedure where the location pin runs along the ilium.
Results
As a result of the testing the final design of the femoral head and acetabular cup was established. The femoral head was made from Biolox Delta ceramic using predefined manufacturing protocols, e.g. hot isostatic pressing followed by hard machining and polishing to achieve an Ra of 0.02 µm. The femoral component consisted of the short location stem capable of withstanding a 1.27 kN stem tip load and a shell that will with stand a tensile stress load of 5.6 kN. The inner surface of the head was coated with a Ti and HA coating with thicknesses of 30 µm and 70 µm, respectively.
The acetabulum cup was manufactured from CF-PEEK by advanced injection moulding techniques; once moulded the hole in the bottom of the horseshoe was drilled. The cup has been designed with the single pin location which has been offset to allow for the maximum range of movement. The inner surface has been designed to have a similar Ra value of 0.02µm to that of the femoral component. The back of the cup has been designed to integrate with the dense trabecular tissue within the reamed acetabulum the design carries a series of fins that will mesh with the bone tissue. The final design mimics the actual acetabulum and the material characteristics have been developed to minimise the stress shielding under normal gait conditions.
The outer surface of the cup has a dual coating consisting of a 30 µm Ti layer over the whole surface with an increase in the coating around the circumference of 70 µm for a width of 10 mm to show the surgeon the position of the cup following implantation. This will then be over coated with a 70 µm coating of HA to gain good secondary fixation with the surrounding tissue.
WP4 - Development of secondary coating
Under WP4, an investigation in to the potential coating methods for the integration of both the femoral and acetabulum components was performed. This investigation consisted of vacuum plasma coating trials on the selected materials i.e. a ceramic substrate and samples of CF-PEEK. These samples were coating with hydroxyapatite HA, titanium (Ti) and a combination of Ti and HA. Several trails were made using the different combinations of the coating at different thicknesses and with different substrate preparations such as shot blasting with silica media and with HA powder. The preparation with the HA was performed to reduce the potential of contamination from the alumina blasting media, however the cost of blasting with HA proved significantly more than the cost associated with cleaning off the alumina media residue. Following coating, mechanical testing was performed on the samples to determine the mechanical bond strength of the coating.
Development of dual coating process parameter options
During this task, a study of coating a Ti foam backing bonded to the back of the CF-PEEK acetabulum cup was analysed. Following this extensive research, it was decided not to generate a foam material for the back of the acetabulum cup as this would prove too flexible for good early fixation of the cup. Therefore, our investigations focused on the coating of the CF-PEEK directly with a HA coating and a Ti with HA coating. In order to ensure that a good adhesion was achieved a series of grit blasting tests was performed on the substrate material. Two medias were used for the grit blasting process, these were conventional Alumina and the other media was HA. It was decided to try the HA to prevent the risk of cross contamination from the Alumina media, the results of this test showed that both materials provided a suitable surface finish for the adhesion of the plasma sprayed coatings. However the cost of using the HA was considerable more than that of the alumina media and from a cost exercise this will probably be the media used during production.
For the coating process, it was necessary to reduce the heat transfer into the cup, in ordered to achieve this we had to increase the velocity of the robot arm while at the same time reduce the powder size as this reduced the amount of energy required to melt the powder. The plasma gas used for the trial was approximately 25 000°C and the velocity at which the medial left the gun was in the order of supersonic to twice this speed. For the HA coating, it was decided to use a standard powder that conformed to ISO 13779-1,-2,-3 and -4, the results of the trials showed that it was possible to generate a coating of Ti and HA or just HA on the surface of the cup. It was also possible to generate a coating thickness the same as can be found on conventional cups of around 75 µm and surface roughness of approximately Ra 59.6 µm.
Selection of dual coating process parameters
Several test samples were manufactured these consisted of tensile test samples (standard dog bone shaped) which were coated with both HA and Ti HA following the grit blasting process. The result of the preliminary trails showed that it was possible to coat the samples with both HA and Ti HA and that the adhesion to the samples was good however the process distorted the samples and following the process the samples had bent by up to 2 to 3 mm. Modifications were made to the process and further trails were performed, the results of these trials were successful and produced samples with no distortion.
Following several coating experiments involving Ti and HA in different combination to determine if a Ti coating was needed or if it was possible to just have a HA coating on the back of the cup. It was determined that in order to obtain an adhesion of the coating to the cup and for the surgeon to be able to identify the cup position (post operation) it was decided to have both a Ti and HA coating on the cup. In the first instance a thin Ti coating (30 µm) was applied with a 70 µm HA coating, however, following CT scans of the samples it was determined that this could not be seen by the surgeon especially when implanted next to bone tissue
To make the cup more prominent against both soft tissue and bone the Ti coating was increased around the equator for approximately 10 mm from 30 µm to 70 µm. It was decided to only increase the coating in this rejoin so as to prevent the fins from becoming clogged with the coating media.
WP5 - Development of clinical instrumentation for new generation implant materials
Work comprised of designing and developing the alignment instruments and impaction cap needed during the insertion procedure. Following the design of the acetabulum cup a series of designs for the alignment instrumentation were generated. However, following feedback from the surgeon group, the final design of the alignment instrument was determined and consisted of a device that located in the reamed acetabulum with a location arm that hooked on the ischia spine of the pelvis. Once in position this would allow the location hole for the acetabulum cup to be drilled.
In order to impact the cup in the acetabulum, a series of designs for the impaction cap were generated. The main criterion for the impaction cap was that it fully supported the cup during the implantation procedure thus reducing the possibility of distortion to the CF-PEEK cup. A final design for the impaction cap was generated which gave full support to the cup while locating in the holes at the bottom of the horseshoe. A collet design with a central locking pin holds the cap to the cup. The collet fits into the hole which has the reverse taper to that of the collet and once the locking pin is inserted which open the collet locking the two components together.
Trials with both the alignment instrumentation and the impaction cap were performed both in the laboratory and by a surgeon during the cadaver trials. The results of these test showed that the alignment instrument was able to position the cup in the optimum position to allow for the best range of movement. The result of the testing performed with the impaction cap showed that it held the cup firmly during the impaction procedure and allowed the cup to be fitted into the acetabulum without distortion. The impaction cap also permitted the removal of the cup from the acetabulum.
Clinical instrument development
Following an in-depth study of the instrumentation used for the current implantation of the ADEPT resurfacing implant, it was decided that new instrumentation would not be required for the implantation of the new femoral component as this was similar in design. Only slight modifications to the existing instrumentation were needed in order to fit the femoral components. The main problem was with the development of the acetabulum cup instrumentation; this required the development of a location device that would allow the surgeon to fit the cup in precisely the correct position first time. This was based on the knowledge that once impacted in place it would not be possible to realign the cup without compromising the initial fixation. Therefore, it was decided to incorporate location pins on the back of the cup which would fit into template pre-drilled holes in the pelvis. The second technical challenge was that of location and support of the cup during the implantation process. To prevent damage to the relatively compliant cup during the implantation process the cup requires a large area of contact with the impaction cap. The impaction cap was therefore designed to conform to the same geometry as the internal surface of the cup which not only prevented damage but also ensured that no distortion would occur during impaction. The final obstacle to overcome was that of location of the cup and cap to allow the surgeon to manipulate the cup during the insertion process, following several design concept brainstorms and prototype modelling it was decided to use the hole in the bottom of the acetabulum cup for this purpose. Following the design of this new cap several rapid prototypes were made for analysis and with some final minor adjustments the finished design was determined and prototypes were manufactured. Following consultations with a surgeon on the design of the cup it was decided to remove the three location pins from the back of the cup as there were concerns that one or more of these could protrude through the back of the pelvis if the cup was positioned poorly during surgery. Therefore, a new location feature was designed for the back of the cup which would eliminate this possibility. The new location feature consisted of a single longer stem (pin) that protrudes down the centre of the iliac spine.
The new cup location design meant that a new design for the instrumentation was needed, therefore a design was produced where the alignment instrument locates in the reamed acetabulum but also fixes around the back of the pelvis onto the ishia spine.
Design of the impaction cap
In order to impact the new cup in the acetabulum a new impaction cap was needed as methods used for metallic cups were unsuitable for the PEEK. Therefore, a series of designs for fixing the impaction cap to the cup were generated. These looked at attaching the impaction cap to the cup through the hole in the bottom of the cup.
Following the development of the impaction cap, prototype caps were manufactured for test and analysis. Following the manufacture of the prototype components, injection moulded cup were attached to the impaction caps and implanted into cadavers to test the system.
WP6: Mechanical testing and surface analysis
During this WP, a number of physical tests and simulations were developed for the verification of the design and material integrity of the femoral and acetabular components. In order to conduct the work, a number of key factors were examined:
- the mechanical strength of the ceramic components under in-vivo loading
- the geometric conformity of the acetabular components after implantation,
- the dynamic friction in the bearing between femoral and acetabular components,
- the predicted tribological conditions in the bearing couple,
- the loads experienced by femoral and acetabular components under surgical loading,
- the loads required to remove the acetabular cup when implanted, and
- the resistance to lever-out loads of the acetabular cup when implanted.
Ceramic component testing
Physical testing of representative prototype ceramic heads was conducted using worst case support and implant positioning for three main stress concentrations identified under in-vivo loading, as reported and simulated in WP2. The heads were loaded in servo-hydraulic test machines, and the fracture load recorded.
Results
The heads were required to sustain in excess of 16 kN axial burst loading. Two heads sustained 50 kN without fracture, and the third fractured above 47kN. The stems were required to sustain in excess of 1.27 kN, and sustained a minimum of 3.38 kN. The heads were required to sustain in excess of 5.6 kN axial load, and sustained a minimum of 13.73 kN.
Geometric evaluation of the femoral components
Uncoated and coated cups were implanted into a range of under-reamed hemispherical bores in representative foam biomechanical test material.
The cup deformation was measured, and the resulting frictional torque generated in internal-external rotation articulation was recorded.
Further tests were conducted for the bearings using full gait and stair climbing / descent loading cycles, incorporating flexion-extension and internal-external rotation articulations. This was conducted with a novel test platform incorporating a multi-axis robot to apply the required load magnitude and direction cycles, taken from the literature.
Side loading tests to evaluate the effect of incorrect fitting
Tests and computational analyses were conducted to consider the influence of cup malpositioning upon the integrity of the implant-bone interface and the bearing performance.
First, a test was conducted where cups were pressed-in to representative foam blocks. The cups were then levered out by applying force on the cup ri, representing an impingement load in-vivo, as a result of cup-femoral neck impingement contact, which would be a possibility if the cup were malpositioned.
Second, a finite element analysis was conducted to predict the tribological conditions in the cup and the proximity of the contact patch in the bearing to the edge of the cup's bearing surface. Edge loading on the hemispherical part of the bearing surface would be no different than for currently used sub-hemispherical resurfacing cups, but the novel aspect of this cup, without clinical precedence, was the bearing surface cut-out, so this was the focus of the investigation. The cup was oriented with relatively steep abduction according to current surgical recommendations, and extreme anterior axial rotation. Then the contact patch location was predicted simply with computer-aided design (CAD) geometric analysis, and the contact pressure was predicted in more detail with a FE analysis model. This was conducted for load cases representing normal gait, stair ascent and descent, and deep-flexion rising from seated. The peak joint contact force instant of each cycle was applied in quasi-static models. Analyses were run for the same range of bearing clearances tested previously (250 µm, 500 µm and 750 µm). The peak stress in the cup was also predicted in these scenarios, to support WP2's prior cup strength analysis.
Results
In physical testing, average torques of 16 Nm to 34 Nm were required to lever the cup out of its supporting foam block, with the lever-out torque increasing with a greater interference fit between the bone and the cup. The lever out resistance was found to be considerably lower when uncoated cups were tested.
In the CAD analysis, the contact patch was predicted to lie within the bearing surface for gait and stair ascent / descent activities. However, in the deep flexion 'rising from seated' load case, the contact patch was predicted to reach the edge of the acetabular notch in the bearing surface.
The peak contact pressure predicted was 21.6 MPa in stair ascent, when the bearing clearance was largest (750 µm). The contact patch was predicted to reach the edge of the bearing surface during rising from seated when the patch was largest (250 µm clearance), but owing to the lower magnitude of the joint contact force in this case, the pressure was not expected to exceed 9.3 MPa, which was lower than the peak contact pressure for the stair descent load case.
A combined computational modelling study and physical test programme was conducted to assess the strength of the femoral head implant under surgical impaction loading. A prototype head prosthesis was cemented onto a representative shape and stiffness foam biomechanical test material cylinder, and attached to a load cell. A head impactor instrument used in current resurfacing surgery was obtained, and modified to feature a slide hammer.
The force-time characteristic of the impact, and the calculated cumulative impulse over time, were analysed in comparison to the FE model results, for model validation. Two models had been produced, including and excluding the mass and stiffness influence of the surgeon's hand, holding the impactor. Ten impact loads were required to verify the head's strength against impaction loads.
Results
During the tests, the head was impacted ten times without fracture. The force-time characteristic recorded in the physical tests and the calculated cumulative impulse both showed good correlation with both basic and hand-damped FE models in terms of peak force and duration. The basic (undamped) model over-predicted the peak force by 49 %, but agreed closely with the cumulative impulse. The hand-damped model only over-estimated the peak force by 23 % but underestimated the cumulative impulse by 16 %, so the basic model was selected as a worst-case.
WP7: Tribological and geometry study
WP7 aimed to 'optimise the wear and friction performance' of the prosthesis components. Deliverables were:
- to provide details for the wear and friction systems and methods ready for tribological testing, and
- to derive the optimal geometry for the ceramic on polymer bearing surfaces with respect to their tribological performance.
Two main approaches were used to assess the design's tribological characteristics:
- FE analysis of the contacting bodies, coupled with theoretical contact calculations, and
- wear testing using a 10-station ProSim hip wear simulator.
The theoretical calculations were used to predict the performance of three bearing clearance options for the proposed Biolox delta ceramic on carbon fibre reinforced PEEK (CFRPEEK) prosthesis. Then, the wear simulator testing was conducted to assess the performance of the preferred design relative to two traditional bearing couples: ceramic on ultra-high molecular weight polyethylene (UHMWPE) and ceramic on highly cross linked polyethylene (HXLPE).
The theoretical predictions indicated that the ENDURE bearing is likely to operate in a boundary lubrication regime, and literature review showed that this is consistent with typically employed bearing clearances with this material combination, which are insufficient to achieve fluid film lubrication. The literature also indicates that friction factor in similar ceramic-CF-PEEK bearings is independent of clearance, provided the clearance is above 250 µm. Therefore the design clearance must be sufficient that implantation and service deformations do not reduce the in-vivo clearance to below this level.
FE indicated that the deformation under normal service loading is likely to be small in comparison to the bearing clearance. Deformation under press-fitted implantation is likely to be higher, and should be the focus of further research.
Finally, wear testing results for simple hemispherical total hip replacement cups of the correct material combination and clearance were obtained which showed no significant difference in wear between CFRPEEK cups and XLPE cups representative of those in clinical use. For this WP, we have combined the tasks:
7.1 Established testing systems
7.2 Literature search and review of similar designs and testing methods and task
7.3 Wear simulation testing in to one report that outlines all of the work performed as these task were so closely related.
Hard-on-soft hip replacement bearings (principally metal-on-ultra-high-molecular-weight-polyethylene (UHMWPE)) have been seen to function well up to 25 years in-vivo. However, the use of this bearing combination is unsuitable for younger patients and resurfacing. For patients below the age of 50 years the survivorship rate is only 79.3 % at 11 years. This is due to the patient's increased activity levels, resulting higher loads and greater number of cycles experienced by the implants adding to the chances of revision surgery. Osteolysis, the body's immune response to polyethylene wear formed through local mechanical damage on the contact surface has been suggested as the major contributing factor to the failure of total hip replacements. Wear particles are generated due to articulating mixed / boundary lubricated contact between the bearing surfaces. In order to reduce the incidence of osteolysis, the wear properties of the artificial joint need to be improved to reduce the volume and bioactivity of the wear debris generated.
Hard-on-hard (MoM or ceramic-on-ceramic (CoC)) bearings were introduced to encourage fluid film lubrication, in which the load is fully supported by a lubricating film, eliminating contact between the bearing surfaces to extend the lifetime of the prosthesis. Nevertheless, achieving fluid film lubrication in a MoM bearing is not instantaneous and involves an initial 'running-in' period in which wear rate is high. The film formation mechanism begins when the MoM components are in relative motion and are loaded. At first implantation, a full lubricant film cannot develop due to small polar bearing point contact between the new head and cup surfaces. Postoperative activities result in wear occurring at a high rate and the area of contact between the surfaces increases. This 'running in' phase ends when the clearance in the contact patch reduces towards zero and the geometry is most favourable for fluid-film lubrication. As migration into the 'steady state' phase occurs and the wear rate slows the contact patch continues to develop at a reduced rate. The effective clearance within the contact patch and within the surrounding entraining geometry is then slowly consumed by wear. Although the wear rate during the steady state is low there is concern over the initial high running in wear, releasing metal ions which are dispersed leading to systemic exposure and potentially have a negative effect on the patient's health.
A recent approach to improve the success of a joint prosthesis is to improve the wear resistance and mechanical properties of the acetabular cup material. CFRPEEK has been suggested as a potential bearing material due to its high wear resistance and mechanical properties which will allow the material to function alone in an acetabular component without the requirement for a metal backing shell for support. In addition, PEEK's proven biocompatibility indicates long term use without an adverse reaction in-vivo due to the material's excellent chemical stability which is a particular problem associated with the wear debris produced by UHMWPE and metal cups. Wang et al. (1999) carried out an investigation into the influence of carbon fibre content and wear resistance. The study concluded that PEEK resin blended with 30 w/w% of PITCH carbon fibre gives the optimal wear resistance, five times lower than 10 % reinforcement and three times lower if 50 % reinforcement was used.
Pin on plate studies have shown a combination of CFRPEEK and ceramic exhibits a five times reduction in wear factor compared to UHMWPE and CoCrMo combinations. Unlike, hip simulators pin on plate experiments do not replicate the conditions in-vivo. Instead the test is useful as a material screening technique to compare new potential materials.
To gain a greater understanding of wear rates of PEEK in a set up that more closely simulates the hip motion and loads through the components, hip simulator studies have been performed. The first hip simulator results were reported by researchers from Howmedica who performed an investigation on injection moulded 30 % discontinuous pitch carbon fibre PEEK acetabular cups articulating against zirconia heads. The CFRPEEK combination exhibited two orders of magnitude less wear than the ceramic-UHMWPE combination after 10 million cycles.
Materials and methods
Test 1: Theoretical analysis and modelling
Wear test evidence from the literature indicates that CF-PEEK cup prostheses will operate in mixed-boundary lubrication. The following is an analysis of the contact conditions in the proposed ENDURE cup, aiming to identify a target clearance, contributing to the detailed design. This study employed theoretical predictions and an FE analysis model. Three 48 mm diameter cups were analysed: the nominal cup as drawn (clearance 750 µm), plus variants with 500 µm and 250 µm. The contact pressure distribution and patch location was predicted under loads representing gait, stair ascent, stair descent and rising from seated.
The FE model from WP6 report 1 was re-used for this analysis, based on the ENDURE cup set in a cylindrical block of PMMA and loaded with a rigid spherical head.
Considering micro-elastohydrodynamic lubrication (uEHL), formulae from Hamrock and Dowson were employed to predict the lubricant film thickness 'hmin' in spherical-spherical bearings and to calculate the lambda ratio of the loaded bearing which indicates its likely lubrication regime. The lambda ratio compares the lubricant film thickness to the maximum surface roughness. A lambda ratio above three implies the achievement of fluid film lubrication, below 1 implies boundary lubrication and between 1 and 3 indicates a mixed lubrication regime. These formulae were input into an Excel spreadsheet and verified using example results given by Dowson.
The approach used in this study was to predict the minimum film thickness for several bearing clearances for the full size range of ENDURE cups, using the same service loading from the Dowson paper. The surface roughness of the CF-PEEK injection mouldings was not known at the time of writing. Therefore, with a surface roughness for the ceramic estimated as Ra=20 nm (range 7 nm - 50 nm for ceramic femoral heads in general), the equations were used to predict the threshold CF-PEEK surface roughness values to produce boundary or fluid film lubrication regimes.
With the uEHL calculation predictions, the results parameter was the maximum CF-PEEK roughness to permit mixed or fluid film lubrication under gait loading, across the full ENDURE size range. With the FE Model, the main output parameters were the peak contact pressure and the radial deformation of the cup rim, indicating the additional clearance required in the unimplanted prosthesis. This was calculated as the designed rim diameter minus the diameter of the maximum inscribed circle (MIC) fitted to the rim of the loaded, deformed cup. The MIC diameter was calculated using an algorithm written into a post--processing subroutine, which performed similar measurements to roundness testing machines.
Test 2: Physical wear testing
To assess the suitability of CFRPEEK as an acetabular component material and assess a few of the conditions described previously, long term wear tests were performed on pitch-based CFRPEEK cups articulating against Biolox Delta alumina femoral heads. Identical HXLPE and UHMWPE acetabular cups articulating against Biolox Delta alumina femoral heads were also tested for comparison.
The wear samples were nominal diameter 36 mm hip prostheses of Biolox Delta ceramic (CeramTec GmbH, Germany) heads against acetabular cups of three different materials - CFRPEEK (Invibio Ltd, UK), UHMWPE and HXLPE. Testing samples included four CFRPEEK, three UHMWPE and three HXLPE cups and were soaked in deionised water for a minimum six weeks before the wear test.
The test was carried out according to ISO14242 with the following conditions:
- Load: 50 - 3000 N single axis twin peak Paul type load curve
- Movement: ±10° internal / external and -15° to +30° extension / flexion
- Frequency: 1 Hz
- The cups were installed with an inclination angle of 35° in the medial-lateral plane and a single dynamic force was applied along the vertical axis
- Fluid: 25 % newborn calf serum with 0.1 % sodium azide
- The test was carried out for five million cycles
- The solution was changed at least every 0.5 million cycles
- Interruptions were made at 0.5 1, 2, 3, 4 and 5 million cycles for gravimetric determination of the wear and worn surface analysis
- Prior to the gravimetric measurement of wear, components were washed in detergent water, thoroughly rinsed in water and deionised water, cleaned in alcohol using an ultrasonic bath for at least 10 minutes to ensure removal of all traces of debris and lubricant, and then left in an atmosphere controlled room for at least 24 h to dry and thermally stabilise
- Wear was determined gravimetrically using an electronic balance (Sartorius ME235S, Germany) to a precision of 0.01 mg
- The gravimetrical wear was converted into geometrical wear using specific density of 1.41 for CFRPEEK, 0.934 for UHMWPE and HXLPE, and 4.36 for Biolox Delta ceramic
- The worn surfaces were analysed using the Alicona InfiniteFocus 3D optical microscope.
Results and discussion
Micro-elasto-hydrodynamic (µEHL) equation results
µEHL calculations led to the prediction of minimum CF-PEEK Ra roughness values to permit mixed and fluid film lubrication regimes to be achieved. These show that for the full size range, mixed lubrication should be achievable with Ra values of 32.9 nm, 48.3 nm and 86.6 nm for the 250 µm, 500 µm and 750 µm clearance cups respectively. Due to its smallest equivalent radius, the lowest film thickness was generated in the smallest size. Fluid film lubrication was predicted to be achievable with an Ra value of 21.8 nm for the 250 µm clearance cup. FFL would not be achievable for the smaller cups in the size range with clearances of 500 µm and 750 µm, because the 20 nm Ra roughness of the ceramic would take up a sufficient proportion of the lubricant film thickness that separation of the surfaces would not be possible.
Scholes et al measured the roughness of the CFR-PEEK MITCH cup during 25 million wear cycles. The initial Ra was approximately 3300 nm, and reduced to approximately 393 nm after the wear test. If these roughnesses are typical of what can be achieved in manufacturing of the ENDURE cup, boundary lubrication is to be expected.
Recent literature studies have reported that clinically used ceramic-on-CF-PEEK bearing couples (MITCH) typically operate in boundary or mixed lubrication regimes, so the failure to achieve FFL may not be of concern. The MITCH cup, which this literature data refers to, had a larger nominal diametric clearance (0.80 - 0.91 µm) but greater ability to deform with its full cut-out horseshoe shape. The wear simulator data reported would probably not include the deformation resulting from press-fitting, so represents a worse case (high clearance). Older literature studies from the MITCH development process tested lower clearances and saw high friction with zero clearance, but no influence of clearance on friction for clearances between 250 µm and 750 µm.
The implication is that in the fully hemispherical, less flexible ENDURE design, the selection of bearing design clearance should be informed by prevention of cup clamping, reducing in-vivo clearance to zero.
WP8 Sterilisation and biocompatibility
This WP addresses the serialisation, packaging and biocompatibility of the new hip components and the instrumentation needed for its insertion. During this WP, a review of the materials to be used in the construction of both the femoral and acetabulum components was performed. Following a literature review, it is clear that the materials selected for the implants have already history in the medical field in similar applications and therefore problems with the sterilisation and biocompatibility were considered to be low risk. As part of the work performed, the materials to be used and the possible configuration of the coatings to be applied was identified. Having identified the materials to be used, a list of tests that would be needed in order to ascertain the biocompatibility was generated. Another key part of this work package was the development of handling and packaging protocols to allow the final products to be assembled where necessary and packaged in a form the was both cost effective and easy to serialise.
For further details please refer to the related project report.