Materials in Hip Surgery
Porous Metals for Implant Fixation
Robert M. Pilliar
• Rationale for cementless hip implants: The goal of developing cementless hip implant components has been to achieve “biological fixation” by bone ingrowth (into porous surface structures) or ongrowth (into irregular surface features) as a preferred alternative to using acrylic bone cement for implant fixation (especially for younger, physically active patients).
• For implants intended for bone ongrowth, surface features of suitable size and quantity to allow secure biological fixation, resulting in interface strength sufficient to prevent breakdown during normal activities
• For both ongrowth and ingrowth designs, limited relative movement of implant and host bone during the healing period to allow rapid formation of new bone and development of secure fixation in as short a time as possible
Porous metals serve two major uses in musculoskeletal reconstruction surgery, namely, constructs for replacing or augmenting bone, and implant surface coatings to allow implant fixation by bone ingrowth. Open-pored structures are used when needed for bone augmentation, an example being reconstruction of the acetabulum to allow acetabular cup placement. Such implants typically are highly porous with volume percent porosity in the 60% to 85% range, and have large pore openings to promote vascularity and bone ingrowth throughout. These scaffold-like structures can be formed in a variety of shapes and sizes. They may or may not serve a significant load-bearing role and typically are intended to substitute for or replace cancellous bone. Structures made of porous tantalum (Ta) or titanium (Ti) are available and are used clinically for this purpose, as are other nonmetallic implants (ceramic and polymeric). Their major role is to replace or augment bone rather than achieve fixation per se as a result of bone ingrowth, although this invariably happens. This article does not deal to any great extent with this class of porous implants, but rather with those designed to achieve secure fixation of joint replacements, hip replacements specifically, without the use of acrylic bone cement (i.e., so-called cementless implants). Other earlier reviews of this subject are recommended to the reader.1–3
The development of cementless designs for hip implants in the 1970s, formed with a porous coating with three-dimensional interconnected porosity through which new bone could form (referred to herein as bone ingrowth), or with an irregular surface with protrusions and recesses that allowed implant fixation by bone ongrowth and mechanical anchorage, was an indirect result of success in the 1960s with cemented low-friction arthroplasties.4 Bone cement as a medium for hip implant fixation was designed initially for treatment of the very elderly and of less physically active individuals suffering from debilitating joint degeneration (hip primarily) in that era. Early success of the procedure with this patient population resulted in its application to younger patients. Its use in younger, physically more active individuals tested the limits of bone cement for secure, long-term implant fixation. With consequent longer periods of use and more aggressive functional loading, the implant-supporting cement broke down more frequently as a result of microcracks, resulting in implant loosening in many cases. As a result, an unacceptably high number of failed implants requiring revision surgery were reported in the 1970s. This resulted in the search for and subsequent use of alternative methods for implant fixation and the development of strategies for forming cementless implants (initially hip replacements) that could be fixed in situ by bone ingrowth or bone ongrowth.
Early studies explored the use of fully porous metallic5 and ceramic systems6; however, the need for implant fatigue strength (for highly loaded devices such as femoral stem components) led to the development of dual-structured designs. These consisted of a strong (fully dense) metal core, which provided the required fracture and fatigue resistance, and surface zones suitable for achieving reliable fixation by mechanical interlock of newly formed bone and implant through bone ingrowth (with porous-coated implants) or ongrowth (with plasma-sprayed implants) (so-called biological fixation). Porous polymers (polyethylene,7 polysulfone8), ceramics (alumina,9 calcium aluminate6), and composites (carbon-reinforced Teflon10) were also investigated. However, the need for the porous layer to be sufficiently stiff to resist excessive distortion on loading and yet strong enough not to fracture or de-bond from the substrate upon repeated loading over many millions of loading cycles led to metals as the preferred choice for surface preparation of hip implant components. These were formed using CoCrMo alloy, Ti, or Ti alloy.11–13 These metals continue to be used for currently available porous-coated or plasma-sprayed hip implant components. Tantalum scaffold-like acetabular components (described later) have become available recently.14 Although stainless steel was considered initially for use in forming porous structures,15 its use in such implant designs was not pursued because of the greater susceptibility of that alloy to crevice corrosion compared with the other metals noted previously. Porous-coated implants with surface coatings made by sintering metal powders (CoCrMo and cp Ti primarily), diffusion bonding Ti wires (or fibers), and plasma-spraying Ti layers became available for clinical use in the 1980s; these surface designs continue to the present for cementless implant components. Currently, these cementless designs represent the preferred choice of many surgeons for use in younger patients.
In vivo animal studies throughout the 1970s and 1980s identified certain necessary conditions for successful design and use of cementless implants. In addition to being biocompatible, these included the following requirements to ensure adequate and timely bone ingrowth or ongrowth:
1. Need for initial implant stability to avoid significant movement of the implant relative to surrounding bone following placement (a snug press-fit was recommended, or in situations in which this is not sufficient, use of ancillary anchorage devices such as screws provides initial stability).
2. Provision of suitable pore openings or surface recesses to allow uninhibited bone ingrowth or ongrowth to an extent sufficient to ensure rigid fixation during functional loading for the patient’s lifetime.
In addition, to ensure long-term reliability of these load-bearing implants, the following engineering design requirements were recommended:
It is also desirable to have the implant stiffness similar to that of adjacent host bone to avoid undesirable bone loss due to stiffness mismatch of implant and bone. Such mechanical mismatch can lead to zones of very high stress in regions of surrounding bone (e.g., at the distal tip region of a femoral stem), increasing susceptibility of bone to fracture, and zones of very low stress in other regions, causing bone loss over time due to disuse atrophy (stress shielding), also resulting in bone that is more fracture prone. A clinical follow-up study using a novel lower modulus composite stem (wrought CoCrMo core surrounded by a polymer, polyaryletherketone, and a Ti surface mesh to allow bone ingrowth—Epoch stem) placed in patients for periods out to 7 years has been reported, indicating significant reduction in bone loss compared with conventional metallic-stemmed implants (CoCrMo or Ti based), at least for the period studied.16 The need for longer-term studies was noted by investigators. A concern associated with using lower-stiffness stems, particularly in younger, active patients in whom long-term active loading is expected, is the fatigue resistance of such designs.
Implant Surface Design for Cementless Fixation
Hip implant components currently used for cementless fixation are predominantly made with either:
2. Sintered coatings to form porous structures allowing three-dimensional bone usually involving multilayered arrangements of particles or fibers, although single-layer particle designs have also been reported.17
Following successful bone ongrowth, implant-bone interfaces formed by plasma-sprayed layers (similar to grit-blasted surfaces) can resist shear forces as a result of the physical interlock of bone with surface features. However, these surfaces do not provide resistance against interfacial tensile forces. This contrasts with porous surface coatings with three-dimensional pore networks that do provide significant resistance to interface shear and tensile forces following bone ingrowth. A number of articles in the past have described as “porous” plasma spray-deposited layers nominally formed with no intended internal porosity, but such a description is misleading. For femoral stem components, which primarily are exposed to interfacial shear forces, this difference may not be clinically significant—a fact that attests to the successful and wide use of plasma spray-coated femoral stem components. However, for acetabular components, in which more complex force components may act and in which interfacial tensile stresses can develop, implants designed for bone ingrowth are preferred because they are expected to provide better long-term stability in this location compared with ongrowth designs.
As discussed later, results of in vivo animal studies have indicated a significant difference between plasma-sprayed and porous-coated surfaces (i.e., irregular surfaces vs. three-dimensional interconnected porous structures) with regard to rate of osseointegration resulting in development of secure fixation.18–20 This study is briefly summarized later and represents one of the few investigations that have focused on very early healing phenomena (i.e., within days) for bone-interfacing implants with specific focus on the effect of implant surface design on rate of osseointegration.
Factors Influencing Bone Ingrowth/ongrowth
The development of as rapid as possible bone ingrowth or ongrowth to achieve secure implant fixation represents a primary goal in the design and use of cementless implants, because this increases the likelihood that successful biological fixation will occur. A number of factors have been shown to influence the rate of bone formation and development of implant fixation. These include relative micromovement of implant and bone during early healing, vascularity at the implant site, implant surface geometry, pore size and possibly shape for porous-coated implants, closeness of fit of implant relative to bone, and the effects of mechanical stimulation on bone formation.
Surface Design—ingrowth Versus Ongrowth: the Effects of Local Tissue Strain on Osteogenesis
A significant difference exists between plasma-sprayed or grit-blasted (ongrowth) and sintered porous-coated (ingrowth) implants in terms of rate of development of rigid implant fixation. This conclusion is based on a study using a rabbit implant model to determine the nature of tissues forming within the implant-bone interface zone and the interface strength and stiffness at very early periods following implant placement (e.g., 4 to 16 days following implantation).18 Press-fitted implants (porous-coated or plasma-sprayed) were placed with a snug initial fit transversely in rabbit femoral condyle sites. The healing response for sintered porous-coated and plasma-sprayed implants was compared. Tapered truncated conical-shaped implants were used with a 300-micron (approximate)-thick porous coating formed by sintering Ti6Al4V alloy powders (45- to 150-micron size range) or Ti plasma-sprayed deposit of approximately 30 microns in thickness. The sintered porous coating had approximately 35 volume percent porosity and consisted of two to three layers of particles sintered to form 50- to 200-micron interconnected pores. The tapered implants (5-degree taper) were self-seating and allowed a good press-fit on placement. The taper shape minimized friction effects at the implant-host bone interface during pull-out testing, so that a sensitive assessment of the mechanical characteristics of interface zone fixation by newly formed tissues was possible. Animals were sacrificed at 4-, 8-, and 16-day periods. Implant fixation after these periods (as well as an initial zero time period that allowed confirmation of the effectiveness and similarity of the initial press-fit anchorage for the two designs) was compared by mechanical pull-out testing (to determine interface shear strength and interface stiffness, as indicated by the slope of the load-displacement curve), histology, and SEM examination (secondary and back-scattered electron imaging).
Mechanical pull-out tests indicated that despite similar initial (time-zero) pull-out resistance (due to snug press-fitting), sintered porous-coated implants exhibited significantly higher pull-out forces and higher interface stiffness at day 4 and day 8. No significant difference was noted at the 16-day period. Examination of interface zones by light microscopy, back-scattered SEM imaging of ground and polished sections, and secondary electron emission imaging of the surface of pulled-out implants indicated localized bone formation within some of the pores of the sintered coating by day 8. This contrasted with the absence of any new bone interlocking with the surface features of plasma-sprayed implants at that time period. The day 4 porous-coated samples (prior to formation of any mineralized tissue) showed a collagen matrix network interwoven throughout the porous structure, which is consistent with the higher pull-out strength and interface stiffness observed at day 4 and may have contributed to earlier bone formation by day 8.
Finite element models representing the two interface zone geometries were then developed to enable prediction of local tissue strains.19,20 According to Carter’s tissue differentiation hypothesis, predicted strain states corresponded to significant differences in the osteogenic potential of the two designs.21 This analysis suggested that three-dimensional open-pored coatings offer an advantage in terms of rate of fixation by ingrowth when compared with ongrowth onto plasma-sprayed surfaces. In addition to the strain state, other factors such as vascularity and local biochemical and biological environment may differ significantly. Nevertheless, the effects of biomechanics as determined by implant surface design appeared to significantly influence cellular events during the healing process. Lower distortional strains were predicted within some pore regions compared with tissues next to the plasma-sprayed layer. It was proposed that pore architecture protected tissues that initially formed at the interface region (i.e., clot, collagen fibers, and cell infiltrate) from imposed forces, thereby resulting in lower distortional strains that, according to the tissue differentiation hypothesis, would favor osteogenesis. This suggests a preferred peri-implant stress/strain environment for rapid bone formation, which is consistent with the concept that mechanical stimulation under controlled levels of imposed cyclical forces promotes osteogenesis. Studies have also indicated the potential benefits of a three-dimensional porous network for enhanced osteoinduction.22,23
Relative Micromovement and Bone Ingrowth/Ongrowth
Bone ingrowth into the porous surfaces of cementless implants has been compared with bone formation during primary fracture healing. A necessary condition for successful bone ingrowth (as with primary fracture healing) is mechanical stability at the implant-bone junction. Reported animal studies have showed that with excessive relative movement at the implant-bone interface, bone ingrowth does not occur, but, rather, fibrous tissue develops.24 This may result in fixation through a pseudoligamentous attachment if a collagen fiber structure forms throughout the porous network.25 With very large relative movement, however, fibrous tissue encapsulation of the implant results.26 Studies to determine the quantitative limits of relative movement causing bone or fibrous tissue attachment or fibrous tissue encapsulation have been reported. In a canine model study using porous-surfaced Ti6Al4V implants (average pore size ≈100 microns with 35 volume percent porosity) placed in healed mandibular premolar sites, it was shown that bone ingrowth occurred if imposed shear displacement at the implant interface was less than 50 microns. With relative displacement of approximately 150 microns, fibrous tissue encapsulation resulted, while fibrous tissue ingrowth and development of a pseudoligamentous attachment were observed for relative displacements between 50 and 150 microns.26,27 The findings of other studies28,29 using Ti fiber mesh-coated implants appear consistent with these results. The different structures (fixation by bone ingrowth, pseudoligamentous attachment, and fibrous encapsulation) are readily distinguished radiographically,30 and images observed in animal studies have been related to light microscopy (histology) and scanning electron microscopic assessments.25
Although excessive movement under load can inhibit and even prevent bone ingrowth, some level of mechanical stimulation during the postimplantation healing period may be beneficial for faster bone formation. This is consistent with observations of enhanced osteosynthesis during application of controlled levels of repeated mechanical force during fracture healing.31
For successful biological fixation of cementless implants, it is essential to achieve secure initial implant stabilization to minimize risk of disruption of the implant-bone interface, preventing or slowing osteogenesis. Several different strategies may be used to achieve this condition, including achievement of a snug press-fit followed by limited loading for an appropriate period (i.e., 3 to 4 months), or protection of the interface through use of an adjuvant method of implant fixation such as screws—the most common method used for initial stabilization of acetabular cup components. Recent porous-coated implant designs have attempted to improve initial press-fit fixation by using more irregularly shaped (asymmetrical) powder particles to form porous coatings that more firmly grip initially when press-fitted into a prepared site (see Fig. 9-3C).
Pore Geometry Effect
Pore size is known to affect bone ingrowth rate. As the discussion on relative movement suggests, prevention of excessive relative movement of porous-coated and plasma-sprayed implants is a necessary condition for bone formation. Thus, the influence of implant design on rate of bony ingrowth is noted because this determines the potential length of exposure of the cementless interface to disruptive forces that could result in excessive relative movement. Early studies by Bobyn and associates32 showed that pore size affects the rate of bone ingrowth. Porous-coated implants formed by sintering CoCrMo alloy particles of four different pore sizes were implanted transversely across the cortex of canine femurs, and fixation strengths and interface structures were assessed by mechanical push-out testing and histology at 4-, 8-, and 12-week periods. In these studies, resistance to push-out developed most rapidly for samples having pores in the 50- to 400-micron size range. For finer pore-sized samples (20 to 50 microns), bone ingrowth was inhibited and maximum interface shear strength (as measured by push-out testing) was lower at all time points. Samples with coatings with pore size of 400 to 800 microns, although eventually approaching fixation strength similar to the 50- to 200- and the 200- to 400-micron samples, required a significantly longer time to do so (longer than 12 weeks versus 8 weeks for the 50- to 200- and 200- to 400-micron pore-sized coatings). This pore size dependence of the rate of bone ingrowth may be related to the different microenvironments present within pores of different sizes and the effect that this has on osteogenesis.
Clemow and colleagues33 investigated the effects of pore size on implant fixation in cortical and cancellous bone using porous-coated Ti6Al4V rods implanted in canine femurs. Three different coatings of equivalent porosity (36% to 40% by volume) with average pore size of 175, 225, or 375 microns were investigated. Implants were placed for a 6-month period, after which pull-out force was measured. Results for implants interfacing both cortical and cancellous bone showed that strength of fixation increased with decreasing pore size. This dependence correlated with measured volume of bone ingrowth. Investigators concluded that decreasing pore size beyond the minimum pore size necessary for bone ingrowth resulted in higher interfacial shear strength.
Pore Shape/Surface Morphology
Micron- and nano-sized surface features have effects on both osteoconduction and osteoinduction. A study by Fujibayashi and co-workers22 showed that more complex pore shapes (i.e., porous Ti structures formed by plasma spraying compared with pressure-bonded Ti fibers) resulted in enhanced osteoinduction if the implants were appropriately chemically and thermally treated to make the Ti “bioactive.” Others have reported no significant effects of pore shape on bone ingrowth34 (for those coatings included in the study).
Materials for Forming Porous Structures
Although sintered porous coatings made from polymers, ceramics, and metals have been investigated in animal studies, only metals are commonly used currently for making implants because of the superior fracture and fatigue resistance of metals, their acceptable corrosion resistance and biocompatibility, and their ability to readily form porous-coated structures over substrates with a number of fairly straightforward techniques. Of the metallic biomaterials available for use in orthopedics, 316-L stainless steel, although it is considered suitable for some other implants, is not recommended for forming cementless implants (either sintered porous-coated or plasma spray-coated) because of its greater susceptibility to crevice corrosion with the more complex surface geometry of the coatings. Currently, porous-coated hip implant components are made from CoCrMo, cpTi, or Ti6Al4V alloy powders and Ti short wires/fibers. For plasma spray-coated implants, Ti coatings are most common. Because of their greater osseointegration potential,35,36 Ti and Ti alloys are presently favored. Tantalum is also used for making some implants for fixation by bone ingrowth. Surface modification resulting in the deposit of calcium phosphate films and layers onto Ti substrates has been shown to promote osteoconduction.37–40 A calcium phosphate surface layer combined with a three-dimensional pore structure has been suggested as enhancing osteoinduction.22
Stress Shielding and Implant Fixation
Stress shielding with rigidly fixed implants can occur if (1) bone and implant of sufficient length are appropriately aligned parallel to the direction of an applied force, (2) they are rigidly fixed to each other over a sufficient length for significant force transfer from bone to implant, and (3) the implant is much stiffer than adjacent bone. Resulting bone loss due to reduced stresses acting in bone over periods of months or years makes the bone more susceptible to fracture. To minimize stress shielding with porous-coated implants, some femoral stem components are designed with porous-coated regions limited to the proximal portions of the stems. Judicious limitations on the extent of porous coating do not compromise implant fixation and long-term stability following bone ingrowth.28 Stress shielding can also be avoided by using lower-stiffness stems. Selection of Ti alloys with their lower modulus compared with CoCrMo alloys (110 GPa c.f. 220 GPa) has been rationalized in this way, but it is unlikely that this results in a significant difference. This is supported by results of a canine study comparing bone loss due to stress shielding by stainless steel onlay plates (E = 200 GPa, similar to CoCrMo) versus Ti alloy plates (E = 110 GPa). After 6-month implantation periods, the structure of bone next to the two implants was virtually the same, displaying significant bone loss under the plates.41 Composite-structured and hollow tubular stems have been suggested as possible ways of avoiding stress shielding.42 However, the fatigue characteristic of such designs is a concern. This continues to be an area of active investigation. As previously noted, clinical investigation of a novel CoCrMo + polyaryletherketone + Ti mesh composite femoral stem having lower stiffness (≈50% of that of an equivalently dimensioned Ti stem) revealed that it has been shown to significantly reduce bone loss, yielding encouraging results, at least over a 7-year patient follow-up period.16