1999 Annual Meeting
Scientific Program
|
American Academy of Orthopaedic Surgeons
1999 Annual Meeting Scientific Program |
Saturday, February 6, 1999
08:00 AM - 10:00 AM
Location: Arena
SYMPOSIUM
J Dennis Bobyn, PhD, Montreal, QC, CANADA
Patricia A Campbell, PhD, Los Angeles, CA
Harry A McKellop, PhD, Los Angeles, CA
Thomas P Schmalzried, MD, Los Angeles, CA
Laurent Sedel, MD, Paris, FRANCE
In order to reduce the volume of particulate debris released into the periprosthetic tissues, bearing materials that wear significantly less than the standard metal-polyethylene combination should be employed. Several such material combinations are currently in clinical use: metal-on-metal, ceramic-on ceramic, zirconia-on-polyethylene and metal-on-highly crosslinked polyethylene. The clinical performance, tribological properties and biocompatibility of these alternative bearings will be discussed. New materials combinations, such as low dose radiation crosslinked polyethylene, will also be discussed.
J. Dennis Bobyn, PhD
Division of Orthopaedics, Faculty of Medicine, McGill University, Montreal General Hospital, Montreal, Quebec, Canada
Metal-metal hip bearings have traditionally been fabricated from surgical grade cobalt-chromium-molybdenum alloys (Co-Cr-Mo) because of their corrosion and wear resistance. Many of the original metal-metal implants such as the McKee-Farrar experienced problems with early loosening because of poor design and manufacture leading to equatorial seizing and high frictional torque. The center issue of wear, however, was not a cause of failure. As self-bearing materials, Co-Cr-Mo alloys are highly resistant to runaway wear, more so than self-bearing ceramics. First generation metal-metal devices that have survived out to two decades or longer are characterized by low wear rates and small changes in dimensions and surface finish.
Metal-metal articulations are currently witnessing a renaissance in several countries worldwide. More than 60,000 Sulzer metal-metal hips have been implanted since 1988. Several US companies, most notably Johnson & Johnson, Biomet, Depuy, and Wright Medical Technology, have developed implants for clinical trials. There is consensus that the femoral head should be smaller than the acetabular cup to ensure polar bearing and provide clearance for the ingress of lubricating fluids. Diametral clearances for 28 mm second generation clinical devices are generally held within the 50 to 100 micrometer range. Hip simulator studies have demonstrated the importance of tight control over dimensions, sphericity, and surface roughness. With high manufacturing quality, wear volumes of <1mm3 after several million wear test cycles can reproducibly be obtained. The typical wear response is characterized by an early period of run-in wear followed by a more steady-state period of distinctly lower wear. One of the advantages of improved bearing technology is the ability to use larger head/cup sizes without large increases in wear volumes. This can lead to improved range of motion, less impingement, and the use of very large implant sizes as required for surface replacement of the hip.
Retrieval analyses of modern total hip devices have indicated very low linear wear rates of 1-5 micrometers/year. The most likely explanation for this behaviour is elastohydrodynamic lubrication, combined with effective boundary lubrication and a wear-resistant microstructure. Theoretical predictions of lubricant film thickness using numerical models of elastohydrodynamic lubrication have indicated that complete protection of the metal surfaces during articulation can occur with the right clearance and surface roughness. There is currently no consensus on which Co-Cr-Mo alloy is preferred for self-bearing applications with the cast alloy, high carbon wrought alloy, and low carbon wrought alloys all presently in clinical use.
Characterization of the wear surfaces of simulator-tested metal-metal implants by scanning electron microscopy and atomic force microscopy has revealed several interesting features. This includes the changing morphology of alloy carbides with increasing simulator test cycles, the formation of micropits at carbide sites and within the alloy matrix, the removal of matrix material through a delamination type of process, and the finding of residual material from the grinding and polishing phases of manufacturing that could act as third-body abrasives. Further microscopic study of the subtly changes that occur at the articulating implant surfaces at long cycle intervals is required to elucidate their role in overall wear resistance and any advantages of one alloy type over another.
Preliminary indications are that metal-metal wear particles are in the 20-60 nm size range, about an order of magnitude smaller than polyethylene wear particles produced in conventional metal-polyethylene articulations. Thus, very small volumetric wear can equate to a very large number of individual particles possessing a very high surface area. This raises questions about mechanisms of macrophage-mediated osteolysis and the importance of particle size, particle area, and particle chemistry in causing bone resorption. Of important note from a clinical standpoint is that 20-30 year Ring metal-metal hip prostheses implanted without bone cement are notable by their absence of peri-implant osteolysis. Issues of cytotoxicity and the long-term local and systemic tissue response to the various Co-Cr-Mo alloy elements require further study and definition.
With all metal-metal hip implants there exists the important issue of integrating the acetabular bearing surface into the overall cup design. The current approaches include molding the bearing surface directly into a polyethylene insert or avoiding polyethylene altogether by connecting the acetabular bearing directly into the metal backing with a large taper-lock. Both approaches involve modularity and therefore need to be carefully studied for mechanical stability and potential for fretting motion. Also important to consider is femoral neck impingement against the acetabular liner. This problem is aggravated by cup malposition and could cause neck notching and metal release, particularly with titanium alloy stems. Close attention to design detail and manufacturing quality must be supplemented with rigorous mechanical testing and hip simulator studies under aggressive loading conditions so that the limitations of the technology are fully understood. The potential for greatly reduced wear and problems with osteolysis clearly exists with metal-metal hip implant technology. Controlled clinical studies with each manufacturer's design should preface release for general use.
Selected References
Thomas P. Schmalzried, MD
Twenty-year performance has been reported of metal-on-metal hip articulations obtained at revision surgery after a mean implantation time of 21.3 years10. The amount of wear was too small to be measured radiographically or by the so-called shadowgraph technique. A computerized coordinate measuring machine (CMM) was used to quantify the amount of wear by assessing the sphericity of bearing surfaces and comparing the measured dimensions in multiple planes to the best-fit circle. The worst case estimate of combined femoral and acetabular linear wear was 4.2 microns per year; about twenty five times less than that typically seen with polyethylene. Table 1 summarizes the results of retrieval analyses (direct measurements) of the wear of metal-on- polyethylene and of metal-on metal bearings.
McCalden and colleagues6 have reported on the surface characteristics of 28 McKee-Farrar total hip implants retrieved after an average of 15.5 years (range; 9-25). The most common type of surface showed fine abrasive wear involving the superior or polar portion of the head and corresponding portion of the cut (Ra; 0.05+/- 0.012 u) with little change from the original surface roughness. There was no evidence of corrosive wear. Impingement wear occurred in 12 cases. Particles of bone cement were seen at the equatorial regions of the cups and could have been involved in third-body mechanisms. Using a CMM, the asphericity of the bearings averaged 13.2 +/- 8.9 u for the heads and 16.6 +/- 10.4 u for the cups compared to one unused component which measured 2.8 u for the head and 2.9 u for the cup. The type and distribution of wear had no relationship to implant survival suggesting that wear rates probably reach a stedy state in long-term implants.
McKellop and colleagues7 have analyzed the wear on 21 metal-on-metal hip replacements revised for various reasons from 1 to 25 years after implantation. There was minimal wear damage visible to the eye. One of the McKee-Farrar hips had substantially higher head and cup wear which was associated with an unusually large clearance of 1.7mm. For the other hips, the mean linear wear rates of the femoral balls were 3.3, 5.2 and 5.9 microns per year (McKee-Farrar, Ring and Muller) with mean volumetric wear rates of 1.2, 2.3 and 3.0 mm3, per year (ANOVA, p=0.02) with individual wear rates ranging from 0.1 to 5.5mm3 per year. The mean wear rate of the smaller diameter McKee-Farrar balls was about twice that of the larger diameter balls (1.4v0.7mm3 per year), although this could have been due to other factors such as differences in clearance and/or effective radius or patient activity. Considering all three designs with large diameter balls, the mean wear rate of the Mullers (3.4 mm3 per year was about 1.5 times that of the Rings (2.3 mm3 per year) and about 6 times that of the McKee-Farrars (0.7mm3 per year) (ANOVA, p=0.02). Assuming roughly equal wear on both sides of the metal-on-metal pair, the maximum volumetric wear rate for these designs was about 6 mm3 per year. There was a trend for the implants with longer service lives to have lower wear rates.
In nine cups there was insufficient non-worn hemispherical surface to allow fitting of the nominal sphere to obtain reliable profile data with the CMM. The clearances for the other hips varied from 127 to 386 microns (excluding the exceptional McKee-Farrar with a clearance of 1.7 millimeters) and within this range there was a trend for increasing wear with increased clearance (p=0.2). The McKee-Farrar with the 1.7 mm clearance had a volumetric wear rate of 11.2mm3 per year which was about 16 times the rate of the other large diameter McKee-Farrars.
Kothari and colleagues5 reported on 22 McKee-Farrar total hips retrieved after 7 to 23 years. The radial clearance varied from 3 to 112 u. The cumulative (head and cup) volumetric wear ranged from 5 to more than 100 mm3 with volumetric wear rates ranging from 0.5 to more than 8 mm3 per year. Two distinct wear patterns were seen; polar and more peripheral (relatively equatorial). Of the implants that failed with acetabular loosening, those with polar bearing survived an average of 13.2 years and those with more peripheral wear were revised at an average of 8.7 years (p=0.07). Further, the wear of the pairs with polar bearing was generally less than other others. This work supports earlier studies which suggested that the location of the contact zone is an important variable in the success or failure of the implant15. Several bearings had clearances which would theoretically allow hydyrodynamic lubrication but no correlation was found between clearance and measured wear. Further studies are needed to define the macrogeometric variables effecting lubrication and wear of metal-on-metal hips.
Schmidt and colleagues12 reported an average linear wear of 17 McKee-Farrar femoral components of 6.6 u per year (range; 0.1 to 28.0) retrieved after an average of 14.5 years (range; 0.2 to 22.0) and the average linear wear of 13 McKee-Farrar cups of 4.9 u (range; 0.2 to 30.0) retrieved at an average of 16.3 years (range; 6.7 to 22). The average linear wear rate of 10 Muller metal-on-metal femoral components was 2.0 u per year (range; 0.7 to 6.2) after an average of 16.6 years (range; 9.0 to 25.0) and the average linear wear of 14 Muller cups was 2.0 u (range; 0.0 to 3.9) at an average of 15.7 years (range; 9 to 25). The average clearance of the bearing couples was 0.12 +/- 0.04 mm for the McKee-Farrars and 0.21 +/- 0.05 mm for the Mullers. The wear rate of the metal-on-metal bearings had a tendency to decrease with time in situ. This is consistent with an initial "conditioning phase" of more rapid wear over the first 1-2 years (running-in) which is then followed by a lower steady-state wear rate. Thus, the overall wear rate decreases with time in situ. One should exercise caution in assigning a cause and effect relationship to this observation; it may be that hip bearings with lower wear survive in the long-term while those with poorer wear characteristics fail.
Willert and colleagues17 reported wear analyses of six pairs of metal-on-metal bearing total hips. These include 3 McKee-Farrar hips retreived after an average of 17.9 years (range; 9.0-23.5) 2 Mullers and one Huggler retrieved after 6.1 years (range; 2.3 to 10.0) (Table 1). The data on the Mullers and the Huggler were combined for this summary. The average linear wear of the McKee-Farrar cups was 5.2 u (range; 3.5-11.1). The average linear wear rate of the Mullers and the Huggler femoral component was 3.8 u per year (range; 1;3 to 6.2) and the average linear wear of cuos was 4.4 u (range; 1.5 to 7.9). The calculated volumetric wear rates were 8.6 mm3 per year (range; 0.66 to 22.36) for the mcKee-Farrars and 2.5 mm3 per year (range; 0.22 to 5.98) for the others.
Second-Generation Metal-on-Metal Bearings
The Metasul (Sulzer Medical Technology, Winterthur, Switzerland) total hip bearing was developed using a carbide-containing, forged, cobalt-chromium-molybdenum allow (Prostasul-21WF) which reportedly has a clearance of 0.15 mm for the 28 mm articulation8. Over 100,000 Metasul bearings have been implanted with a variety of Sulzer hip systems. This metal-on-metal bearing technology has also been extended to large diameter surface replacement components14. Clinical results of total hip systems with these bearings have generally been very good1,14,16. There are no reports of any reoperations for a problem directly attributable to the metal-on-metal articulation. There has been no gross evidence of metal wear, such as metal staining of tissues and no or few meal particles seen in histologic sections14. There have, however, been reoperations for infection, heterotopic ossification, instability and aseptic loosening. Schmidt and colleagues12 reported 17 pairs and 10 heads retrieved and sent for analysis by the Sulzer laboratory. The average time in situ was only 1.7 years for the heads and 1.3 years for the cups. The mean linear wear rates for the heads was 11 microns per year and 8 microns per year for the cups. Hips revised for dislocation demonstrate higher wear, likely due to damage of the head from the dislocation(s). In analyzing the wear rates of all 44 components, there appears to be an initial higher wear rate of 15-20 microns per year per component (the initial conditioning phase or running-in). This is then followed by a decreasing wear rate towards a lower long-term rate of about 2-5 microns per year per component.
The reported United States experience with the Metasul bearing has been limited to a single-surgeon investigation1. Between 1991 and 1994, 74 Metasul bearings in the Weber cemented cup have been implanted in 5 combination with 38 cemented APR stems, 27 cementless APR-II stems of which 3 have modular diaphyseal sleeves and 19 APR-IIT stems (Anatomic Porous Replacement, Intermedics Orthopaedics, Austin, TX). With up to four years follow-up (average 2.2), the clinical results have been good to excellent and no hips have loosened. Twenty-seven of these patients have a contralateral metal on plastic bearing hip of similar design and none of these patients could detect a difference between the two hips.
As we explore alternatives to metal-on-polyethylene bearings in total hip replacement we must keep in mind the lessons of the past. Clinical success and failure are multifactorial. The chance of success with new hard-on-hard bearings will be increased if the bearing is combined with well-established femoral stems and acetabular shells so as to minimize other variables in the clinical outcome equation, for example, a Charnley stem with a metal-on-metal bearing.
References
Laurent Sedel, MD
Chairman, Orthopedics Department and Orthopedics Laboratory
University Hospital Lariboisiere and
University Paris 7 (Denis Diderot) France
Alumina/Alumina couple was first implanted in 1970 by Pierre Boutin. Hard on hard material had the theoretical advantage of very low wear, debris generation and low friction if some technical details are fulfilled. Alumina ceramic being highly oxidized is also highly compatible in bulk or particulate forms. Fracture toughness and wear are directly related to the material quality that resumed in low grain size (2 to 3 µ ±1), high purity, high density, low porosity: Conical sleeving has been used to anchor the head on the stem. Metal cone technology is very demanding regarding cone angle, roughness, length. Some alumina head fracture have been related to poor cone technology. Many different alumina design and quality have been implanted in the past. These design such as Mittlemeier one included cementless screw in ring, cementless stem and large head. It resulted in many early failures not directly related to alumina material but to wrong design and difficult surgery. Since the early eighties, high security level has been obtained. If fracture risk still remain of some concern, its calculated risk at worst could be in the order of one per two thousand for a ten year period. We started our trial in 1977 with a cemented plain alumina socket and a cemented titanium alloy stem collared, smooth and anodized.
Clinical results are available over a twenty year period. Cemented acetabular fixation and its use whatever the age of the patient resulted in 83% survivorship at 10 years and 70% at 15 years. Nevertheless, results were excellent in young age population with 86% survivorship at 15 years in patient of less than 50 years of age and the stem survivorship depicted 97% at 14 years. Osteolysis was encountered in less than 1% and only related to early socket loosening in patients who refused to be reoperated. These patients having a metal against ceramic rubbing problem after the ceramic had tilted. This pioneering period was utilized to document data's concerning clinical results and after revision for failures to analyze retrieved alumina components, retrieved tissue analysis. During the last ten years, survivorship analysis and clinical results on pain, range of motion experienced quite identical results to other implanted materials except the fact that osteolysis was never encountered and providing that the patient is allowed to perform any activity without limitation. The results already published confirmed the very in vivo low wear: with an average of less than 1µ per year for linear wear except if the prostheses had tilted before revision. Histological studies confirmed the excellent biological tolerance of alumina ceramic debris which were few and gave rise to pure fibrocytic reaction without any macrophages or giant cells.
We modified socket fixation ten years ago to a metal backed press fit alumina. The alumina liner being held by conical sleeving as well as the head. We did select patients young, active, heavy with strenuous activities and limit the use of this material to these type of patients. With this material we could expect an exceptional long term survival without any activity limitation. Fracture risk is now evaluated at 1/2000 for a 10 years period. As osteolysis was not encountered by us or by other authors, if revision is necessary for any reason, it becomes a very easy procedure without the need for bone reconstruction. After more than 2000 alumina on alumina couples implanted, we arrive at the conclusion that this material is specially dedicated to young and active people, retaining alumina on polyethylene or metal on polyethylene for elderly or less active patients.
References
Harry McKellop, PhD and Fu-Wen Shen, PhD
The J. Vernon Luck Center for Orthopaedic Research
Orthopaedic Hospital, Los Angeles, CA
The majority of total hip prostheses implanted in the past three decades have included an acetabular cup of ultra-high molecular weight polyethylene (UHMWPE) articulating against a femoral ball of cobalt-chromium alloy. The resultant wear of the polyethylene bearing produces billions of sub-micron sized wear particles annually4,8. In some cases, the resultant foreign body response leads to bone resorption (osteolysis) and, eventually, gross loosening of the components14. Wear is a particular concern in a young or active patient, who may face one or more revisions with accumulative bone loss. Substantially reducing the number of particles released by improving the wear resistance of the polyethylene could substantially extend the clinical lifespan of total hip prostheses.
Crosslinking of polyethylene, which can be done using chemical or radiation techniques, improves its abrasion resistance in some industrial applications and, potentially, for hip prostheses. Using a ball-on-flat wear tester, Rose et al13. showed that the wear resistance of high-density polyethylene was improved by crosslinking with peroxide during molding, and the wear resistance of UHMW polyethylene was improved by crosslinking with gamma irradiation at 2.5 to 5.0 Mrad. Similarly, Grobbelaar et al.3 reported that high-density polyethylene radiation crosslinked at 15 Mrad showed almost a 30% improvement wear resistance in a sand-slurry abrasion test. In contrast, Streicher16 reported that the wear resistance of UHMWPE in sand slurry abrasion decreased after electron beam irradiation (up to 30 Mrad).
Clinically, Oonishi et al.12 implanted hip prostheses with UHMW polyethylene acetabular cups that had been subjected to 100 Mrads of gamma radiation in air, bearing against either cobalt-chrome or alumina ceramic femoral balls. Recently11, Oonishi reported that the steady-state radiographic wear rate of the cups with the cobalt-chrome balls was as low as 0.06 mm/year. This was somewhat lower than the range of 0.1 to 0.2 mm per year typical of conventional UHMW polyethylene that has been crosslinked as a byproduct of sterilization using gamma irradiation (i.e., 2.5 to 4 Mrad in air)7, and was substantially lower than the average of 0.29 mm/yr for Oonishi’s patients with non-irradiated cups. Similarly, Wroblewski et al.20 implanted "crosslinked" cups (probably chemically) with alumina ceramic balls and, after an initial period of unusually rapid wear, observed radiographic wear rates about 0.02 mm/yr, compared to 0.07 mm/yr for conventional UHMWPE implants in the same series against metal balls.
In addition to potentially improving the wear resistance, crosslinking can adversely affect other physical properties (such as tensile strength)16, and the free radicals generated by the irradiation can render the polyethylene susceptible to long-term oxidation. Consequently, some manufacturers now advocate non-irradiation sterilization (Et-O or gas plasma) to avoid oxidation altogether. Others continue to sterilize with gamma radiation in order to crosslink the polymer for improved wear resistance, but with the components sealed in a low-oxygen packet (e.g., inert gas, vacuum or with an oxygen scavenger). The immediate result of this approach is less oxidation, and a lower wear rate than for non-irradiated polyethylene (Fig.1). However, free radicals remaining in the irradiated polyethylene can oxidize over time, substantially increasing the wear rate (Fig. 2).
In order to obtain the "best of both worlds," the polyethylene can be crosslinked prior to its being fabricated into an acetabular cup. For example, an extruded bar can be crosslinked using conventional gamma irradiation and then heated above the melt-temperature long enough (e.g., 150 °C for 16 hours) to extinguish the residual free radicals15. The resulting oxidized surface layer of the bar is then discarded during machining of the acetabular cup, providing a polyethylene with superior resistance to wear and to long-term oxidation6. If the crosslinking gamma dose is increased just above the 2.5 to 4 Mrad range normally used for radiation sterilization, the wear in a hip simulator test becomes immeasurably small (Fig. 3). This is a substantially lower dose than used in Oonishi's study12. For clinical use, cups pre-crosslinked in this manner can be sterilized with ethylene oxide or gas plasma to avoid the re-introduction of free radicals that would occur during radiation sterilization.
Improvements in the wear resistance of UHMWPE after radiation crosslinking have been reported by others1,9,18. For example, Muratoglu et al. crosslinked the polyethylene using electron beam irradiation, either while heated10 or at room temperature followed by remelting9, and Wang and colleagues18 used gamma radiation to sterilize and crosslink finished polyethylene cups while they were packaged in nitrogen, and then reduced the residual free radicals by heating the cups while still in the nitrogen packets to 50 °C (i.e., below the melting point) for six days. With both techniques, the wear decreased with dose comparably to Fig. 3. As indicated by the close correspondence between the plots of wear rate and swell ratio (Fig. 3), the minimal wear rate corresponds to saturation of the crosslinking in the polyethylene16. Clarke et al.1 also reported negligible wear of polyethylene cups tested in a hip simulator, using doses of 50 to 200 Mrad, i.e., far above the saturation point for crosslinking.
Apparently, crosslinking improves the wear resistance of UHMWPE to the type of crossing-path motion that occurs in an acetabular cup5,19. Without crosslinking, the polyethylene molecules may become oriented parallel to the sliding direction during one part of the walking cycle, such that they are more susceptible to fracture in a crossing direction during a later part of the cycle. Consistent with this model, Wang et al.,18 reported that radiation crosslinking of polyethylene conferred relatively little additional wear resistance to tibial plateaus tested in a knee simulator that subjected the components to nearly unidirectional motion. Thus, the usefulness of crosslinking, and the optimum dose, depends on the type of motion to which the polyethylene component is subjected.
Radiation crosslinking reduces the tensile strength and elongation below that for non-crosslinked polyethylene. However, for doses greater than or equal to 10 Mrad, the tensile properties are above the minimums specified in ASTM F648-96 (Fig 4). In addition, the ASTM values serve as a definition of UHMW polyethylene, but do not necessarily indicate the minimum properties that are needed for a polyethylene component to function in-Vivo without experiencing mechanical failure. Historically, the 2.5 to 4.0 Mrad gamma radiation dose that was used to sterilize the majority of UHMW polyethylene components in clinical use also reduced the tensile properties below those of the non-crosslinked material. Until recently, this sterilization was done in air, and without any thermal treatment to remove the resultant free radicals, such that the components frequently underwent substantial additional reductions in their physical properties (Fig. 4) during years of shelf storage and/or in-vivo17. In contrast, since remelting removes the residual free radicals, acetabular cups fabricated from the crosslinked-remelted polyethylene have a wear resistance superior to cups of non-crosslinked polyethylene, and should not experience the long-term degradation of their wear resistance and other mechanical properties2 that has been typical of the radiation-sterilized implants in past clinical use (Fig. 4).
Acknowledgement: Supported by NIH grant #AR40996 and the Los Angeles Orthopaedic Hospital Foundation
References
References
Patricia A. Campbell, PhD
There are numerous published definitions and descriptions of biocompatibility, but all of them have in common the requirement that, to be considered biocompatible, a material must be tolerated in the body without causing an adverse host response. Adverse responses have been classified as metabolic, bacteriologic, immunologic and neoplastic2. Although the biomaterials used clinically in total joint arthroplasty have been tested and approved for biocompatibility, a review of the literature shows examples of each of the above classifications of adverse effects, although, fortunately, the numbers of such reports are relatively small compared to the high numbers of components implanted each year.
The earliest uses of materials for implants and fixation devices often provided dramatic examples of adverse host responses and high failure rates from materials that were either biologically incompatible, functionally incompatible or both. For example, the implantation of pure metals resulted in marked tissue redness, swelling and induration with histological effects including inflammatory cell infiltration, tissue degeneration and necrosis. The early failure of a polyethylene cup arthroplasty highlighted the importance of distinguishing between the biocompatibility of the bulk materials of an implant and the wear products of that implant, and Newman and Scales (1951)39 observed that, while polyethylene was chemically inert, the size and shape of the particles play an important role in the tissue response. Although he did not use the term biocompatibility, Charnley was also well aware that materials "become degraded by the chemical environment of the body, to liberate secondary products or cause their physical structure to be altered and become unsuitable for their original purpose", as he learned first hand from the failure of the Teflon acetabular cups8. However, he considered that if there were no signs of intolerance to a material after a period of four years in vivo, it should be regarded as safe.
Early histological studies of periprosthetic tissues described the accumulation of wear particles of various kinds within macrophages and giant cells with the resultant formation of granulomas, often with associated necrosis and thickening of the tissue5,9,51. Autopsy studies showed that wear particles had traveled to, and lodged within, the draining lymph nodes which had undergone histiocytic or necrotic changes8,51. Using the most sophisticated equipment available at the time, the particles were identified with spectroscopes, measured with television image analyzers and the tissue reaction in large numbers of pathological samples described at length4. Despite conflicting opinions as to which particulate material was the most harmful, there was consensus in the general concern over the accumulation of prosthetic wear particles in periprosthetic tissues. The conclusion of many such reports was that the long term consequences of these particles were unknown. In 1982, in a treatise on the biocompatibility of orthopaedic implants, Williams noted that the "key factor in the biocompatibility of joint prostheses is the response of the tissue provoked by the materials" and that, while the implant materials were well tolerated, there was the possibility of an adverse reaction to the release of ions from the metals or to wear particles of any of the materials resulting in an inflammatory response52.
There has been extensive research in this area; animal and cell culture models of the effects of metal ions and metallic and polymeric wear particles have verified these possibilities, and numerous reports have verified their clinical reality. For example, in a series of experiments spanning more than a decade, Goodman et al. distinguished between the fibrous encapsulating tissue around bulk PMMA and UHMWPE and the florid foreign body granulomas induced by the particulate materials, and went on to show that the tissues and cells exposed to particulate UHMWPE produced various inflammatory mediators associated with osteoclast activation and bone resorption17,18,19,20,21. The same osteoclast-activating properties of cells derived from clinically retrieved tissues were demonstrated in a number of centers16,19,25,31,44 and the association between UHMWPE particles and bone loss became stronger. It is now generally accepted that the majority of cases of aseptic loosening and osteolysis are caused by the macrophagic response to UHMWPE wear particles. The purpose of this symposium is to discuss alternative bearing materials that will avoid this problem without introducing new, unforeseen problems.
Metal on metal (M-M) implants have been re-introduced using improved manufacturing procedures and designs but employing essentially the same cobalt chrome alloy that was used in the first generation of M-M implants such as the McKee-Farrar and Ring designs. At the time that those early designs were put into clinical use, there was concern about the effects of CoCr corrosion and wear products in patients. Studies of the levels of cobalt and chrome in the hair, blood and urine showed that metal levels in patients with M-M total hips were higher than in patients with metal-UHMWPE11. This finding has been repeated in more recent studies of serum and urine metal levels that employed analytical instruments with increased sensitivity and which used stringent methods to avoid contamination of the samples29,30. These studies showed that cobalt ions are rapidly transported from the implant site and mostly eliminated in the urine, while chromium tends to be stored in the tissues and eliminated more slowly. However, while the release of Co and Cr ions from M-M THRs has been verified, the clinical outcome of this finding is still unclear. The issue of host hypersensitivity to these elements remains a concern, but a review of the literature has indicated that, while such reactions can occur56, the incidence seems to be very low37.
In vitro studies of the effects of metallic particles and ions were also conducted around the time of the clinical implantation of first generation M-M THRs. Many of those studies emphasized the carcinogenicity of the metals35 and there were numerous reports of high rates of tumor formation in rats exposed to powders or salts of C and Cr24,46,50. Fortunately, the extent of the risk to humans that these early rat studies suggested has not been borne out by the clinical incidence of sarcomas in M-M THR patients which is extremely low34 and epidemiological studies have not verified an increased risk of cancers in total joint patient48.
Other early studies assessed the cellular toxicity of Co and Cr and reported various effects including morphological changes, release of enzymes such as lactate dehydrogenase and glucose-6 dehydrogenase, and cell death following exposure to particles or salts in cell culture models13,42. Cobalt was often noted to be more toxic than Cr or the CoCr alloy. Those studies often employed doses of metals far in excess of those seen clinically and used forms of particles that are unlikely to occur in patients with implants. Several authors have pointed out that metallic wear or corrosion products released in vivo form organometallic complexes with proteins, and it is to these that the body responds2,41,47,55. Furthermore, Co and Cr ions exist in different valences, as well as different chemical forms, including soluble organometallic complexes, solid wear particles, matrix-bound organometallic complexes, intracellular and extracellular soluble or precipitated metal compounds. Their biological effects at local and remote sites may differ according to these various physical and chemical states2.
Although the actual nature of the metallic wear particles and organometallic complexes formed in vivo is still under investigation, recent studies have attempted to utilize more clinically relevant models, for example, by exposing particles to serum prior to use in vitro, using particles extracted from periprosthetic tissues, or in doses similar to those occurring in vivo23,33,45. While experiments conducted with variable cell types, particles and analytical methods often lead to conflicting results27, it appears that, at clinically relevant concentrations, the wear products from CoCr implants are able to modulate cytokine expression in macrophages21,22 ,27,49 and may have inhibitory effects upon osteoblasts1, neutrophils43 and T cells15 but do not induce cytotoxic effects1,26,49. In contrast to the latter observation from cell culture models is the frequent observation of necrosis in the periprosthstic tissues around M-M THRs6,12,14,32. However, a direct correlation between necrosis and visible metallic particles has not been established and extensive necrosis is also seen in tissues around failed metal-UHMWPE total hips12.
Alumina ceramic bearing surfaces have also been proposed as low wearing alternatives to metal-UHMWPE. Bulk alumina is reported to be chemically inert and therefore to have excellent biocompatibility10. The response to particulate alumina in periprosthetic tissues has been described as less intense than that typically seen around metal-UHMWPE total hips10,28,40 but the particles still induced a foreign body response. Isolated cases of osteolysis have been reported clinically, often when excessive debris was present3,40,54,57. Cell culture studies have indicated that particulate alumina is not toxic36, but can stimulate the release of inflammatory mediators such as IL-1, IL-6 and TNF in a dose dependent manner38, although one study has shown that the level of TNF release was less than that provoked by high density polyethylene particles7.
Just as the materials and designs of implants used in total hip arthroplasty have evolved over the last three decades, our ability to monitor the bone and tissue response to them has evolved. The list of cytokines released by macrophages that have phagocytosed biomaterial particulates or been exposed to media containing biomaterials, is constantly growing, and the cellular and chemical mechanisms by which these interactions occur are slowly being unraveled. How have these advances into the potentially adverse effects of particles changed the concept of biomaterial compatibility? Williams has suggested that biocompatibility should now be defined as "the ability of a material to perform with an appropriate response in a specific application. This means that a material that has all the characteristics of biocompatibility under one set of conditions may show different and possibly inappropriate responses under different conditions"53. Since there are clearly adverse responses to wear particles of all of the biomaterials in clinical use, the onus rests not only on the implant manufacturers to provide implants with optimal designs and material quality for wear resistance, but also on the surgeons to utilize those implants appropriately for optimal long term use in patients.
References
Harlan C. Amstutz, MD
Surface replacement represents a significant development in the evolution of hip arthroplasty. It is a direct descendant of the cup arthroplasty originally conceived by Smith-Petersen (1948). Surface replacement is a bone conserving alternative to total hip arthroplasty that restores normal joint biomechanics and load transfer and ensures joint stability. In the mid 1960s, Maurice Müller used a metal-on-metal (Co-Cr-Mo) resurfacing system which was a press-fit (Müller and Boitzy 1968). Despite satisfactory early results, this system was abandoned because of loosening of the components. In 1970, Gerard in France also implanted metal-on-metal resurfacing prostheses with motion occurring not only between the components but between the components and bone (Gerard 1978; Gerard et al. 1974). Since total hip systems using fixation with cement were very successful in the short term, it was only after the exploration of revision of the stems that resurfacing became refined with acrylic fixation in the mid 1970's in five different countries (Amstutz 1991).
Resurfacing procedures addressed the problem of preserving femoral bone stock at the initial operation and also had the potential of easy revision since the femoral canal was not violated. The early results were encouraging but the longer follow-up was often disappointing (Ahnfelt et al. 1990; Bell et al. 1985; Head 1981; Howie et al. 1990a, 1990b). Although failure is multifactorial, it is now understood that bearing wear debris induced osteolysis is the main problem (Amstutz et al. 1994; Bell et al. 1985; Howie et al. 1990a, 1990b, 1993; Mai et al. 1996; Nasser et al 1990; Schmalzried et al. 1992, 1994). Surface replacement failure was primarily due to the high volumetric polyethylene wear, secondary to a large prosthetic head and not due to high frictional torque, neck fractures or osteonecrosis of the femoral head. The role of polyethylene debris in the failure of surface replacements is further amplified by our results of surface hemiarthroplasty for Ficat Stage III or early stage IV osteonecrosis of the femoral head (Amstutz et al. 1994; Grecula et al. 1995) where in the absence of polyethylene, no loosening or osteolysis was observed after more than 16 years postoperatively (Amstutz 1997). Those hips which have required reoperation were revised for groin pain associated with deterioration of the acetabular cartilage. The retrieved femoral head specimens showed no evidence of osteolysis, were viable and the cement-bone interface remained intact. The concern that resurfacing of the arthritic femoral head would cause osteonecrosis has not been substantiated by our human retrieval studies. More recently it has been verified that the femoral head remains viable in the vast majority of cases (Howie et al. 1993).
The New Era Of Full Surface Replacement
Recent research suggests that the volumetric wear of cast cobalt chrome alloy bearings is 40 to 100 times less than that of the metal-on-polyethylene combination (McKellop et al. 1996; Semlitsch et al. 1989).
Three metal-on-metal surface replacements have been developed, two in Europe by Heinz Wagner in Germany (Wagner and Wagner 1996) and Derek McMinn in England (McMinn et al. 1996). Both systems were initially all-cementless. The Wagner design has a bearing surface of forged cobalt-chrome alloy (F799 with high carbon content) and has a grit blasted, titanium alloy carrier with macro features for fixation to bone. The initial McMinn surface replacement was cast cobalt-chrome alloy, uncoated press-fit both on the femoral and acetabular side.
In 1996, we began custom implantation with new devices designed to minimize wear and optimize fixation. The acetabular component is one piece and has sintered beads (102 µ avg., 50-200 µ - 38% porosity) on the outer dimension designed for interference fitting to obtain initial stability even in the smaller diameter dysplastic acetabulum while bone ingrowth occurs. The femoral component is patterned after the THARIES™ chamfered cylindrical design but with a short stem to ensure precision reaming and a fixation with a uniform cement mantle. The component is identical to that used for hemisurface replacement but with improved sphericity and surface finish bearing tolerances to minimize friction and wear.
Our short term experience with the hybrid system is promising. Two hybrid designs were used. The first twenty-one were modified McMinns. The McMinn acetabular bearing component was cemented into an interference fit Porous Surface Replacement (Depuy) uncemented cup. The Conserve Plus(Wright Medical Technology) acetabular component is one piece and designed for interference fitting to obtain initial stability. The first 94 patients (21 McMinns; 72 Conserve Plus) which underwent this procedure had an average age of 48 years old (range 15-69) at the time of surgery. Follow-up ranges from 4 months to 5 years. There were 62 males, 33 females. The primary diagnosis was osteoarthritis in 52, osteonecrosis in 19, development dysplasia of the hip in 9, inflammatory in 6, trauma in 5, Athrokatadysis in 2, and SCFE in 2. There were 3 complications: subluxation and wire removal from the greater trochanter in one patient, one infection which required direct exchange of components, one heterotopic ossification requiring excision and one dislocation. With a failure rate of 9.4% at 2 years with the THARIES and none due to technique or components in on our metal on metal surface replacement, we have substantially reduced the short-term failures and minimized the complications. These results combined with hemiresurfacing prosthesis surviving seventeen years and PSR cups 15 years without loosening, we are encouraged that the long-term results will be better than the THARIES and enable our patients to maintain a very active lifestyle without compromising any further hip surgery whenever needed.
We believe that the lessons learned about design and technique of implantation of resurfacing components, combined with modern precision manufacturing of metal-on-metal bearing surfaces, have ushered in a new era of surface replacement of the hip. Since the volumetric wear reduction is substantial, we anticipate greatly improved durability. Further, it is our hypothesis that the metallic wear debris is much better tolerated by the tissues around the implant than is the polyethylene. We have histological data to support this premise although we need more information with respect to the overall body response (Doorn et al. 1996).
The advantages of a surface replacement include the initial preservation of head and neck bone (Capello et al. 1978) and the more physiological sound stress transfer to the femur through the neck which avoids proximal femoral stress shielding. There has never been a reported case of thigh pain following surface replacement because the femoral canal is not entered. A major advantage, particularly for the young and active, is that there has been no penetration of the femoral canal should a revision be required in the future. There are some disadvantages because the devices require precision manufacturing to optimize and further minimize the wear of the surfaces. Precision inevitably results in increased production costs but these can certainly be reduced with improved manufacturing capabilities. Further, the operation remains technically more difficult than total hip replacement because we have to ream around the head and neck and the procedure takes longer. However, the advantages of minimal morbidity, more physiologic procedure offset this. Anatomically bone is maintained, biologically and biomechanically bone is preserved and revision is relatively simple should it be necessary.
References
| Home | 1999 Meeting | Scientific Program |