 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 2004 The Journal of Bone and Joint Surgery, Inc.
Hydroxyapatite-Coated Prostheses in Total Hip and Knee Arthroplasty
1 Consultancy in Medical Devices and Biomaterials, 512 East Saddle River Road,
Ridgewood, NJ 07450. E-mail address:
boffin{at}worldnet.att.net
In support of their research or preparation of this manuscript, one or more
of the authors received grants or outside funding from Stryker Orthopaedics.
In addition, one or more of the authors received payments or other benefits or
a commitment or agreement to provide such benefits from a commercial entity
(Stryker Orthopaedics). No commercial entity paid or directed, or agreed to
pay or direct, any benefits to any research fund, foundation, educational
institution, or other charitable or nonprofit organization with which the
authors are affiliated or associated.
Hydroxyapatite-coated implants have demonstrated extensive bone apposition in animal models. The osseous interface develops even in the presence of gaps of 1 mm and relative motion of up to 500 µm. Development of implant-bone interfacial strength is due to the biological effects of released calcium and phosphate ions, although surface roughness leads to increased interface strength in the absence of interface gaps. The clinical results at fifteen years after total hip replacements have demonstrated that hydroxyapatite-coated femoral stems perform as well as, and possibly better than, other types of cementless devices, with the added benefit of providing a seal against wear debris. Hydroxyapatite-coated acetabular components must have a mechanical interlock with bone in order to take advantage of the coating effects. Clinical analyses of these types of designs at seven years have indicated good survivorship. The performance of a hydroxyapatite-coated implant depends on coating properties (thickness, porosity, hydroxyapatite content, and crystallinity), implant roughness, and overall design. The most reliable predictor of the performance of a device is success in long-term clinical studies.
Hydroxyapatite coatings for cementless fixation of total hip prostheses were introduced in 1985 by Furlong and Osborn1 and in 1986 by Geesink2. In 1996, Jaffe and Scott3 presented a Current Concepts Review of hydroxyapatite-coated hip prostheses in this journal. They concluded that, with the possible exception of smooth ("macrotextured") hemispherical acetabular shells, hydroxyapatite-coated components had proved to be equal to, and in many instances better than, non-hydroxyapatite-coated components with regard to implant fixation. Preclinical and clinical studies showed that hydroxyapatite-coated implants permitted a less intimate initial fit to bone than did other cementless devices. Despite some evidence of coating resorption in vivo, there were no indications that the coating resorption affected the long-term clinical outcome. The authors stated that a coating thickness of 50 to 75 µm was preferable, but they also reported good clinical results with coatings up to 200 µm in thickness. Their analysis was based on clinical follow-up of eight years for stems and about six years for acetabular cups. Since the 1996 review3, more information on hydroxyapatite-coated devices has appeared in the literature. The durations of follow-up in clinical series now approach fifteen years. The results associated with acetabular components of many different designs have been reported separately from the results associated with their stems, allowing a refinement of the original statements regarding hydroxyapatite fixation in the acetabulum. In addition, analyses of specimens retrieved at autopsy have provided information on the nature of the tissue-hydroxyapatite coating interface and on the fate of the coating over time. Total hip prostheses coated with mixtures of hydroxyapatite and tricalcium phosphate have been introduced. Solution-deposited coatings, briefly mentioned in the original review3, are now used clinically with porous ingrowth surfaces. Reports of hydroxyapatite and hydroxyapatite-tricalcium phosphate coatings on cementless knee components are also available4-14. Laboratory and animal studies of hydroxyapatite coatings have become more refined. For example, the relative roles of coating chemistry and surface roughness in determining the fixation strength at the implant-bone interface are better understood now than they were in 1996. These developments merit new review of hydroxyapatite coatings. The organization of this report is similar to that of the earlier review3 to aid in comparison of the two documents. We will refer specifically to the earlier article only when there is a need to do so for clarity and continuity.
Hydroxyapatite (Ca10[PO4]6[OH]2), tetracalcium phosphate, and tricalcium phosphate (Ca3[PO4]2) are examples of the various calcium phosphate compositions. Tricalcium phosphate exists in small quantities (<5%) in most so-called hydroxyapatite coatings and in larger amounts (>40%) in mixed hydroxyapatite-tricalcium phosphate coatings. Hydroxyapatite and hydroxyapatite-tricalcium phosphate coatings are commonly applied with the plasma-spray process, in which heated calcium phosphate particles are projected at high velocity in a gas stream onto the prosthesis to build up the coating15,16. Some commercially available implants have solution-deposited coatings, in which the coating is nucleated and grown on the prosthesis in solution17,18. It was shown that, for consistent performance, a hydroxyapatite coating should have low porosity, high cohesive strength, good adhesion to the substrate, moderate-to-high crystallinity, and high chemical and phase stability19. In scientific comparisons of different coating properties in vitro or in vivo, erroneous conclusions may be drawn if coating parameters are not fully defined20. For example, a comparison of the dissolution behavior of so-called hydroxyapatite coatings from six commercial sources indicated that the coatings behaved very differently when placed in the same local environment even though all were labeled as "hydroxyapatite" coatings21. Published reports of properties of commercially available calcium phosphate coatings are summarized in Table I. One manufacturer uses a hydroxyapatite-tricalcium phosphate mixture as the coating material on its ingrowth implants. The hydroxyapatite coatings have a high crystallinity and a high hydroxyapatite content to reduce the possibility of premature dissolution22,23. The porosity of the coating is also an important determinant of its dissolution rate24. Most contemporary plasma-spray coatings have a thickness of 50 to 75 µm, which seems to provide better strength than do thicker coatings, which have had a tendency toward fatigue fracture when placed under load in the laboratory25-27. However, there are commercial coatings in the 150 to 200-µm-thickness range that have no apparent clinical problems1. The substrate material of most hydroxyapatite-coated implants is titanium or titanium alloy, although some knee implants are made of cobalt-chromium alloy. Usually, the substrate surface is roughened to improve the attachment strength of the coating3,16. The coating attaches to the substrate primarily by means of mechanical interlocking generated by the impact energy between the incoming plasma-spray hydroxyapatite particles and the underlying metal substrate28. Thus, the titanium alloys, which are softer, tend to provide better coating-bond strength than does the cobalt-chromium alloy, which is harder3.
Hydroxyapatite and hydroxyapatite-tricalcium phosphate coatings may be applied to nonporous or porous surfaces (e.g., grit-blasted or fiber-mesh surfaces). Bone formation on an implant with a nonporous surface is termed ongrowth, whereas porous surfaces can have both ongrowth and ingrowth into the pores. Rough titanium plasma-sprayed surfaces are usually classed along with porous surfaces, and the bone attachment is frequently referred to as ingrowth. Hydroxyapatite coatings are osteoconductive. When hydroxyapatite implants were placed in bone beds with gaps between the coating and bone, bone growth across a gap of up to 1 mm was demonstrated and gap-filling in the presence of micromotion of up to 500 µm in amplitude was shown29,30. Interface micromotion above this level tended to limit the effectiveness of the coating in promoting bone apposition. When the stability of hydroxyapatite-coated implants was compared with that of uncoated implants, a stronger interface with bone was demonstrated for the coated titanium surfaces31,32. Hydroxyapatite coatings have been shown to seal the interface between the implant and bone and to prevent ingress of polyethylene particles in the presence of an initial 75-µm gap33,34. Hydroxyapatite coatings promote bone formation in compromised situations, such as ovariectomized animals35, and in canine revision models36-38. Studies of animals have shown that hydroxyapatite-tricalcium phosphate coatings accelerate bone apposition and ingrowth into porous devices in the short term but, in the longer term, push-out strength tests did not discriminate clearly between the coated porous implants and uncoated porous controls39,40. Push-out tests of implants consisting of a hydroxyapatite or hydroxyapatite-tricalcium phosphate coating on a titanium-alloy substrate in canine transcortical and intramedullary models did not discriminate between the two types of coatings at twelve and twenty-four weeks, respectively41,42. However, while the hydroxyapatite coating appeared to be relatively stable, there was considerable dissolution of the hydroxyapatite-tricalcium phosphate coating that resulted in particle release from the implants. Comparison of low and moderate-crystallinity coatings in a gap model showed little difference in ingrowth but did show differences in the coating dissolution rate, with the lower-crystallinity coating demonstrating the least stability22. Bone formation associated with a hydroxyapatite coating is believed to begin with surface dissolution of the hydroxyapatite, which releases calcium and phosphate ions into the space around the implant. Reprecipitation of carbonated apatite then occurs on the coating surface43. The hydroxyapatite binds serum proteins and cellular integrin receptors, allowing osteoblastic cells to bind to the surface44,45. Bone formation follows at both the bone and the coating surfaces16. Bone ongrowth develops more rapidly on lower-crystallinity coatings because the initial dissolution and release of calcium ions is faster than those associated with higher-crystallinity coatings23,44. Surface dissolution is therefore a driving force for bone formation, but the effect of surface roughness on bone apposition has been controversial. It has long been known that rougher surfaces demonstrate stronger interfaces with bone than do smoother surfaces in both humans and animals as long as the interface is bone ongrown46-48. In a study of dogs, Hacking et al.49 compared two types of hydroxyapatite-coated implants with regard to the strengths of their interfaces with bone. One group of implants was "as hydroxyapatite-coated," and the other was hydroxyapatite-coated and then recoated with a very thin titanium film that preserved the topography of the hydroxyapatite coating but prevented biological interaction with the hydroxyapatite coating. The interfacial strength of the recoated implants was about 80% of that achieved when the hydroxyapatite was exposed. The investigators argued that surface roughness was a larger contributor to interface strength than was the presence of the hydroxyapatite coating. It should be remembered, however, that while surface roughness is undoubtedly a factor in determining interfacial strength when an implant is in intimate contact with bone, implants often are surrounded by gaps and micromotion is present after implantation. Under these circumstances, the osteoconductive nature of the hydroxyapatite coating is a key factor in promoting the intimate bone apposition required for the attainment of interface stability. In summary, preclinical studies continue to show that a hydroxyapatite coating enhances the stability of nonporous implants by promoting early bone ongrowth even in the presence of interfacial gaps or initial interfacial instability. Hydroxyapatite-tricalcium phosphate coatings appear to have little long-term effect on ingrowth porous implants, but they do promote earlier pore-filling. For ongrowth surfaces, the chemistry and crystallinity of the hydroxyapatite coating continue to be important because a balance must be achieved between the long-term stability of the coating and the short-term ion release needed to initiate bone ongrowth. The preclinical literature does not indicate directly the optimum combination of coating characteristics that will lead to the best clinical result. Thus, we must turn to various clinical series to evaluate the results of specific hydroxyapatite coatings applied to specific implant designs.
Femoral Components The durations of implantation of hydroxyapatite-coated total hip components now approach fifteen years. Geesink50 reported the results eleven to thirteen years following the use of an Omnifit-HA femoral stem (Osteonics, Allendale, New Jersey), a metaphysis-filling design with a coating on the proximal third of the stem body, together with an Omnifit-HA threaded cup (Osteonics) in 118 hips in ninety-nine patients who were an average of fifty-three years old at the time of implantation. One patient (two hips) was lost to follow-up, and six patients (eight hips) died. At a mean of twelve years postoperatively, the survival rate was 97% (105 of 108) for the stems and 93% (100 of 108) for the cups. Distal femoral osteolysis was not seen. In 70% of the hips, there were signs of bone resorption, presumably due to a reaction to wear debris, but as this was confined to the calcar resection level it was suggested that the debris was not penetrating the fixation interface. In the United States, experience with the same hydroxyapatite-coated femoral stem began in 1988 in a multicenter study, with an investigational device exemption, approved by the Food and Drug Administration. Eighteen surgeons at fifteen sites implanted a total of 436 hydroxyapatite-coated stems between January 1988 and November 1990. They used three different types of hydroxyapatite-coated acetabular implants, including the same threaded cup implanted by Geesink50 as well as two smooth cups (Dual Radius-HA and Dual Geometry-HA; Osteonics); the study also included a porous-coated control (Dual Geometry; Osteonics). The average patient age was fifty years at the time of implantation. Clinical results from different subsets of patients in this series were reported after various periods of follow-up51-58. For example, a review of the results associated with 316 of these femoral stems (in 282 patients) at a mean of 8.1 years postoperatively revealed that only one stem (0.3%) had been revised because of aseptic loosening51. Again, there were no reported cases of intramedullary osteolysis. A study of remodeling of bone around 224 of these femoral stems at six years postoperatively showed radiographic evidence of progressive new bone formation (increased cancellous density and cortical hypertrophy) at the level of the middle and distal portions of the femoral stem52. The patterns of bone-remodeling suggested effective stress transfer from the stem to the femoral cortex. Again, intramedullary osteolysis was rare (one case), and it was concluded that the fixation provided by the hydroxyapatite coating prevented debris migration along the stem. Two (1%) of the 224 stems had been revised because of aseptic loosening. Later evaluation of the radiographic patterns of bone-remodeling around 314 stems in 274 patients at a minimum of ten years postoperatively suggested that most remodeling occurred in the first five years after the implantation and that it stabilized thereafter53 (Fig. 1). The femoral stems continued to perform well; the rate of mechanical failure (defined as aseptic loosening or radiographic findings of such loosening) was 0.5%, and there was no distal osteolysis. Analysis of only the younger patients (average age, thirty-nine years) in the entire cohort at a minimum of five years54 and ten years55 showed radiographic findings of bone-remodeling similar to those reported earlier52,53. The rates of mechanical failure in the young patients were 0% at five years and 0.9% at ten years. Proximal femoral lytic lesions were seen more frequently in younger patients (48%) than in older patients (38%), perhaps because of the greater activity level of the younger patients. Similar results for this particular hydroxyapatite-coated stem and its derivative (Omniflex-HA; Osteonics) were reported in more recent series59,60 and in the revision setting57. The clinical results in these and other cited studies are summarized in the Appendix.
There has been considerable clinical experience in Europe and Australia with the Anatomique Benoist Girard (ABG) hydroxyapatite-coated hip implant (Benoist Girard/Howmedica, Herouville, France), a titanium-alloy anatomic stem with a proximal hydroxyapatite coating applied by vacuum plasma spray onto a titanium plasma-sprayed coating. Results with these stems usually have been reported after use with the hydroxyapatite-coated ABG cup, a titanium implant with a grooved fixation surface, fixation spikes, and holes for screw placement. A clinical study of 100 hips in 100 consecutive patients (average age, sixty-three years) showed no revisions at a minimum of two years (range, twenty-four to forty-five months) postoperatively61, and a study of thirty-three hips in thirty-two patients with rheumatoid arthritis (average age, fifty-one years) showed no revisions for aseptic loosening and no radiographically unstable stems at five years postoperatively62. The International ABG Study Group described the results of 398 consecutive arthroplasties (average patient age, 63.8 years) performed at ten centers in five European countries and followed for five to seven years postoperatively63. Three stem revisions had been done because of malposition, low-grade infection, or thigh pain. No revisions took place after the second postoperative year. Radiographic evidence of bone-remodeling was most apparent during the first three years. No distal osteolysis was observed. In another study, technetium-99m bone scans made for eighty asymptomatic hips with a total hip replacement (in sixty-two patients) and twenty control hips (in ten subjects) showed that uptake in the total hip-replacement group decreased significantly (p < 0.05) from one to three months postoperatively and changed little thereafter64. The results of seventy-one ABG hip replacements performed by a single surgeon in sixty-six patients (average age, fifty-five years) showed no loose stems at an average of 4.8 years (range, two to seven years) postoperatively65. However, there was one case of endosteal cavitation in Gruen zone 266. (Figure 2 shows the seven sections of the anteroposterior radiograph of the femur delineated by Gruen et al.66.) In another study, of 100 consecutive ABG hip replacements in ninety-seven patients (average age, fifty-one years), the survival rate was 100% for the stems and 95% for the cups at an average of six years postoperatively67.
The Furlong hip implant (Joint Replacement Instrumentation, London, England), a fully coated stem with a coating thickness of 200 µm, was first implanted in 19851. The coating, initially applied with plasma spray in air, was later applied with plasma spray in a vacuum. At a mean of ten years (range, nine to twelve years) after implantation of the first 100 stems together with a cemented all-polyethylene acetabular component in eighty-six patients (age, forty-five to ninety-four years), there had been no stem revisions and no radiolucent lines or granulomas around any stem68. Other studies of the same implant demonstrated few or no revisions because of aseptic loosening in younger patients at five to six years postoperatively69-71 (see Appendix). One report showed that an extensively coated hydroxyapatite stem can have good fixation to bone but may be difficult to remove if revision is necessary72. Published results are available for a variety of other hydroxyapatite-coated stem designs, including the Landos Corail implant (Landanger, Chaumont, France), a straight grit-blasted titanium-alloy stem implanted with a hemispherical grit-blasted titanium cup or a screw-fit cup. The outer surface of these components is entirely covered with a hydroxyapatite coating of 155 µm in thickness, with a 97% hydroxyapatite content, and with a crystallinity of >50% and a porosity of <10%. In a study of 323 hips in 276 patients (average age, forty-eight years), one stem was revised because of mechanical failure at the interface, twelve patients (twelve hips) died, and nineteen patients (nineteen hips) were lost to follow-up73. In a five-year follow-up study of 100 consecutive hips in eighty-six patients (average age, fifty-six years) treated with the same stem and the hydroxyapatite-coated screw-fit cup, hip function and pain scores increased to nearly normal levels during the first year and no stem subsidence or loosening was seen74. However, seventy-four of ninety-four stems had progressive double radiopaque lines in Gruen zones 1 and 8 at five years. The authors suggested that the lines indicated fibrous fixation and that compromise of the fixation could be expected; this expectation was met. In another study, of twenty hips revised in nineteen patients, osteolysis was demonstrated in the proximal region around these stems75. The Mallory-Head hip prosthesis (Biomet, Warsaw, Indiana), a titanium-alloy implant with hydroxyapatite plasma-sprayed over a plasma-sprayed titanium coating, was used with a non-hydroxyapatite-coated cementless cup in a series of sixty-three dysplastic hips in fifty patients with an average age of fifty-three years76. At the time of follow-up, at an average of seventy-five months, no osteolysis, no stem revisions, and three cup revisions were noted. In another study, the Ti-fit collarless stem (Smith and Nephew, Memphis, Tennessee), a titanium-alloy implant with a hydroxyapatite coating on its proximal third, was implanted with an Opti-fix porous-coated acetabular cup from the same manufacturer in fifty-two hips (fifty-two patients; average age, sixty-five years)77. Forty-eight stems were stable after eleven years of follow-up. There were eight cases of osteolysis but only one case of distal intramedullary osteolysis77. A ten-year clinical study of 100 Freeman femoral components (Finsbury Instruments, Leatherhead, England, and Corin, Cirencester, England), a femoral neck-retaining design with a hydroxyapatite coating of 80 to 120 µm in thickness, showed no revisions involving the stem but demonstrated one case of stem subsidence78. A five-year follow-up study of 142 hips in 136 patients (average age, 66.5 years) treated with the Austria Hip System (Logimed, Loeben, Austria), which has a hydroxyapatite coating on its proximal two-thirds, revealed five revisions of the femoral component79. Three were due to periprosthetic fracture; one, to aseptic loosening; and one, to septic loosening. The stem migrated >2 mm in twenty-nine hips, and radiographic evidence of bone loss was seen in zones 1 and 7 in all patients. It was concluded that the design of this stem (oval proximally and round distally) did not provide adequate stability and that the coating could not compensate for this shortcoming. Several studies have compared the performance of hydroxyapatite-coated devices with that of non-hydroxyapatite-coated implants. The performance of the Anatomic Hip prosthesis (Zimmer, Warsaw, Indiana) was assessed with the implant uncoated, with its fiber-mesh pads coated with hydroxyapatite-tricalcium phosphate, and with its pads and the proximal half of its stem coated with the same material80. There was little difference between the two hydroxyapatite-tricalcium phosphate-coated implants, although both showed better radiographic results and fewer revisions compared with the uncoated controls. In a prospective, randomized trial of sixty-two consecutive hips in fifty-five patients, thirty-five were treated with a Mallory-Head stem with a hydroxyapatite coating and twenty-seven were treated with the same stem without the coating81. The stems were implanted with either a RingLok or an Opti-fix acetabular component. Neither the Harris hip scores nor the stem survival rates indicated an advantage of the hydroxyapatite coating. Twenty-four patients (average age, fifty-four years) were treated bilaterally with S-ROM implants (DePuy [Johnson and Johnson]), with a 50-µm hydroxyapatite coating on the proximal part of the stem body in one hip and a 110-µm monolayer of porous beads in the contralateral hip82. At an average of fifty-two months postoperatively, there were no revisions and no differences in recovery time or bone-remodeling between the hydroxyapatite-coated and non-hydroxyapatite-coated implant groups. Similarly, a study of fifty patients in whom porous-coated IPS titanium femoral stems (DePuy, Leeds, England) had been implanted bilaterally, with a hydroxyapatite coating on one side and without the coating on the other, showed no stem subsidence or loosening and equivalent Harris hip scores after 6.6 years of follow-up83. A comparison of hydroxyapatite-coated and non-hydroxyapatite-coated grit-blasted stems of the Profile design (DePuy, Warsaw, Indiana) in fifty hips (forty-six patients) showed less subsidence of the hydroxyapatite-coated stems in the first two years84. Finally, a comparison performed at an average of 4.7 years after revisions with APR Revision Hip stems (Intermedics Orthopedics, Austin, Texas) with and without a hydroxyapatite coating in sixty-six hips (sixty-five patients) demonstrated better Harris hip scores for both pain and limp in the hips treated with the hydroxyapatite-coated stem85. However, no difference in stem survivorship was demonstrated. In summary, the favorable early experience with the Omnifit-HA femoral stem3 has continued for more than thirteen years. The pattern of bone-remodeling first reported at six years52 was maintained at the time of longer follow-up. Young patients showed excellent survival of the prosthesis but had a higher prevalence of proximal osteolysis (in Gruen zones 1 and 7) from polyethylene wear than did older patients. The stem and coating appeared to seal the interface below the resection level, as no distal osteolysis was reported. The experience with other designs has not been reported as comprehensively. Although the Furlong hip prosthesis was first implanted in 1985, only twelve-year follow-up data are available and there are few publications available for review. However, this fully coated stem with a 200-µm-thick coating does appear to perform well. In contrast, the 155-µm-thick coating of the Landanger implant has demonstrated loss of fixation to the substrate and the generation of hydroxyapatite particles. Clearly, coating thickness per se is not the reason for these failures; it is likely that the composition of the coating and its lower bond strength are responsible. The results of use of the ABG hip replacement have been followed for only seven years. The clinical results have been good for the stem, but there are indications that cup wear is a problem. Studies of other designs mostly have included only short to medium-term follow-up; they have indicated good performance except in the case of the Austria hip implant. There have been few studies comparing hydroxyapatite or hydroxyapatite-tricalcium phosphate-coated stems with uncoated stems, and those that have been published have included only short or medium-term follow-up80-85. However, it appears that hydroxyapatite and hydroxyapatite-tricalcium phosphate coatings result in greater bone formation, which provides increased stem stability, than is seen with uncoated stems. Long-term follow-up is needed to determine if this translates into better survival of the implants.
Acetabular Components Results associated with the acetabular components used in the investigational device exemption study referred to earlier in this review were dependent on cup type. Manley et al.58 found that, at a minimum of five years postoperatively, 1% of 131 hydroxyapatite-coated threaded cups, 2% of 109 porous cups without a hydroxyapatite coating, and 11% of 188 smooth hydroxyapatite-coated press-fit cups had been revised because of aseptic loosening. The interface failure of the smooth hydroxyapatite-coated press-fit cups that loosened most often was initiated in DeLee-Charnley86 zone 3 (Fig. 3, A). (The three zones of the acetabulum described by DeLee and Charnley are delineated in Figure 4.) Analysis of these clinical failures and comparison with published finite-element models of acetabular replacement87 led to the conclusion that the smooth fixation interface could not withstand the stresses imposed by patient activity. Furthermore, the data indicated that physical interlock between the acetabular shell and bone was a prerequisite for long-term stability of fixation in the acetabulum. The study of the Landos Corail implants73 mentioned above demonstrated poor results after use of smooth hydroxyapatite-coated acetabular components. In the initial group of 323 hips, forty-two hydroxyapatite-coated press-fit cups (13%) and nine hydroxyapatite-coated screw cups (3%) were revised73. The smooth hydroxyapatite-coated ABG cup (Benoist Girard) has seemed satisfactory after short follow-up88,89. However, in a study of 289 primary hip replacements, twenty-nine cups (10%) had been revised at an average of 54.9 months90. The causes of the revisions of the cups were osteolysis (sixteen cases), aseptic cup loosening (ten), and recurrent posterior dislocation (three). These findings reinforced the conclusion that a surface with mechanical interlock is necessary for stability in the acetabulum. In another study of the ABG cup, with a mean patient age of forty-nine years at the time of surgery, twenty-three (24%) of ninety-seven hips had failed due to osteolysis or wear at a mean of 5.75 years91.
The Mathys or RM cup (Mathys, Bettlach, Switzerland), an all-polyethylene hemispherical cup with pegs to control rotation, represented an alternative attempt at hydroxyapatite fixation. Hydroxyapatite granules were hot-pressed onto the outer surface of the cup to provide a coating of 300 µm in thickness. Clinical studies with durations of follow-up of eight and 9.4 years showed good clinical results for patients in whom the implant had survived, but severe osteolysis attributed to third-body wear from released hydroxyapatite granules was associated with some implants92,93. Retrieved components had calcium phosphate material embedded in the articulating surface of the polyethylene. Hydroxyapatite-coated and non-hydroxyapatite-coated Bi-Contact press-fit acetabular components (Aesculap, Tuttlingen, Germany) were compared in a randomized study of 153 hips (138 patients) followed for at least five years94. Three cups were revised because of aseptic loosening and one, because of recurrent dislocation. All revisions were performed with a smooth hydroxyapatite-coated cup. The Norwegian Arthroplasty Register contained an early report on the performance of Atoll press-fit and Tropic partially threaded cups (Landanger) coated with hydroxyapatite (155 µm in thickness)95. At the time of early follow-up, none of the 772 Atoll cups and one of the 1171 Tropic cups had been revised. The results in later reports were not as satisfactory. Of 191 smooth hydroxyapatite-coated press-fit Atoll cups used in primary total hip replacements in 155 patients (average age, forty-seven years), thirty-eight had been revised and three additional cups were loose at seven to ten years after the surgery96. At the revisions, the hydroxyapatite coating was found to be absent from the implants. In another report on the Atoll design, in which eighty-five hips (seventy-four patients) had been followed for ten years, twenty-six hips (twenty-two patients) had been revised and a revision was planned for six more hips97. Acetabular survivorship at ten years was 67.9% (95% confidence interval, 58.5% to 77.3%). Loss of coating was demonstrated, and it was more rapid for loose cups. The wear rates for the threaded Tropic cups accelerated with time. In a series in which 100 consecutive hips were followed for seven to nine years after total hip arthroplasty, eighteen hips with excessive polyethylene wear had a revision and a revision was planned for six more98. However, at the revisions these threaded cups were found to be firmly fixed to the bone. Other studies have shown osseous fixation of hydroxyapatite-coated threaded cups. In a recent series, 418 Arc2f hydroxyapatite-coated threaded cups (Osteonics) were fixed with supplemental bone screws in 384 patients by one surgeon, and the hips were followed for a minimum of ten years99 (Fig. 3, B). Cup survivorship was >99%, and no patient had activity-limiting pain. No radiolucent lines were found on 98% of the interpretable radiographs, and no evidence of osteolysis was seen on 95%. Geesink reported similar results with the Omnifit-HA threaded cup; they found a survivorship of 93% (100 of 108 cups) at ten years postoperatively50. A study of 173 hips (150 patients) treated with the Furlong-HA screw-ring cup (Joint Replacement Instrumentation) showed one case of osteolysis in zone 1, two cases in zone 2, one case in zone 3, and one case in all three zones at an average of 6.5 years100. There were two revisions due to polyethylene wear without loosening and three due to aseptic loosening. The authors stated that if a screw cup is to be used, hydroxyapatite coating is mandatory. Other strategies for achieving interlock in the acetabulum include use of the Secur-Fit Acetabular Cup (Osteonics), which achieves a mechanical interlock with bone through a rough, arc-deposited titanium coating beneath the hydroxyapatite surface (Fig. 3, C). In a study comparing twenty-five Secur-Fit hydroxyapatite-coated cups with twenty-five smooth Dual Radius hydroxyapatite-coated components in patient groups matched for age, sex, and preoperative diagnosis, radiographic analysis performed at a minimum of four years showed the Secure-Fit cups to be associated with fewer radiolucent lines than the Dual Radius cups101. In an earlier series, seventy-eight patients (ninety-three hips) were treated with the Secur-Fit cup with either a cemented or a noncemented stem and, at an average of four years postoperatively, there had been no revisions due to aseptic loosening, mechanical failure, or osteolysis102. A study of Rim-Fix acetabular components (Joint Replacement Instrumentation), which have a hemispherical design with a 200-µm-thick hydroxyapatite coating and provision around the rim for screw fixation to augment stability, showed only two failures, which occurred during the eleventh and twelfth postoperative years103. The stemmed partially hydroxyapatite-coated McMinn-Link cup was found to perform satisfactorily in the revision (twenty-two cups) and complex primary replacement (seven cups) settings104. At a mean of forty-six months (range, fourteen to seventy-four months), the only failure was due to sepsis. Finally, hydroxyapatite-tricalcium phosphate coatings have been used in the acetabulum105,106. A radiostereometric analysis compared the performance of Harris-Galante-II cups with and without a hydroxyapatite-tricalcium phosphate coating at two years postoperatively105. Compared with the uncoated implants, the coated cups had migrated less around the horizontal axis. A similar study of the Trilogy cup (Zimmer) coated with hydroxyapatite-tricalcium phosphate compared thirty-four hips treated with a cup without screw-holes with thirty hips treated with a cup with a cluster of three holes for screw placement106. Radiostereometric analysis showed no difference in migration between the two groups at two years. It was concluded that screws are not necessary when this design of cup has a hydroxyapatite-tricalcium phosphate coating107. In summary, long-term success with hydroxyapatite-coated acetabular components seems to require the acquisition and maintenance of an interlock between the implant and bone. Mechanical stability is not sufficient in itself, as shown by the failures of the ABG and Tropic cups. Wear is also a consideration and is influenced by many factors, including the bearing material, component design, and patient activity. The articulating surfaces must have minimal wear, or revision is a probability. Long-term clinical survival of a hydroxyapatite-coated cup requires mechanical stability, adequacy of locking between the component and bone, and little wear, as demonstrated by the hydroxyapatite-coated Arc2f threaded cups99.
Hydroxyapatite coatings have been used on total knee replacements but on a much more limited scale compared with their use on total hip replacements. A comparison of 127 hydroxyapatite-coated Tricon-II tibial components (Smith and Nephew) with seventy grit-blasted Tricon-II tibial components showed good clinical results, with no difference between the hydroxyapatite-coated and grit-blasted components at five years postoperatively4. However, the hydroxyapatite-coated components showed much less radiolucency than did the grit-blasted components, suggesting improved fixation with the hydroxyapatite coating. In a prospective, randomized study of fifty-seven Tricon-II knee implants in which radiostereometric analysis was used to compare the performance of hydroxyapatite-coated and cemented tibial components, there were three revisions of the hydroxyapatite components: two were due to infection, and one was due to disruption of the coating5. There were no revisions of the cemented components. In a single-blind, randomized study of the PFC knee implant (Johnson and Johnson Orthopaedics, New Milton, England), radiostereometric analysis was used to compare three types of fixation after two years of follow-up6. The underside of the tibial tray had a porous coating for cementless fixation, a porous coating with a hydroxyapatite coating, or a waffle surface for use of cement. The hydroxyapatite coating was deposited by plasma spray to a 55-µm thickness. The authors concluded that there was no significant difference between the hydroxyapatite-coated and cement-fixation groups. The porous-coated tibial implants had the greatest migration. A radiostereometric analysis of migration and inducible displacement was performed to compare twenty-five knees treated with the hydroxyapatite-coated Freeman-Samuelson knee implant (Sulzer Orthopaedics, Alton, England, and Finsbury Instruments, Leatherhead, England) with twenty-six knees treated with the Miller-Galante-II knee implant (Zimmer)7. The groups were matched for age, gender, weight, degree of deformity, and smoking habits. The clinical results for the two groups were equivalent at five years postoperatively, although the Miller-Galante-II prostheses displayed more radiolucencies around the tibial stem. The Freeman-Samuelson tibial components had migrated less than the Miller-Galante-II tibial components at five years and showed less inducible displacement at one year. With revision because of loosening as the end point, the survival rate of the hydroxyapatite femoral components was 94% at a mean of ten years8. Hydroxyapatite-coated tibial components of different designs were evaluated with radiostereometric analysis at two years as part of a study of saw-blade cooling9. Comparison of the porous-coated Osteonics 7000 tibial component (Osteonics) with the hydroxyapatite-coated Duracon tibial component (Howmedica, Rutherford, New Jersey) showed no differences in clinical outcome between the two implants. The hydroxyapatite-coated components subsided less than did the porous components. It was stated that the hydroxyapatite coating had a strong positive effect on tibial component fixation. Epinette and Manley described the twelve-year experience with a variety of Osteonics knee designs with hydroxyapatite coating10. Since 1990, 400 hydroxyapatite-coated knee prostheses were implanted in 342 patients; 268 knees were followed for a minimum of five years, sixty-five were followed for a minimum of ten years, and seven were followed for more than twelve years (Fig. 5). There were only three cases of loosening and osteolysis. The cumulative survival rate at seven years was 97%, with failure for any cause as the end point. The authors commented that one of the benefits of hydroxyapatite coating in knees is the ability of the coating to fill surgically created gaps between the implant and bone.
The effect of a hydroxyapatite coating on the micromotion of the Interax knee (Howmedica International, Limerick, Ireland) was evaluated in a prospective, randomized, double-blind study of eleven components fixed with cement as well as ten hydroxyapatite-coated components and ten non-hydroxyapatite-coated porous components fixed without cement11. The hydroxyapatite-coated and cemented components were reported to have less micromotion than the porous components. A randomized study of Miller-Galante-II knee replacements demonstrated similar results with regard to tilt and subsidence when hydroxyapatite-tricalcium phosphate-coated tibial components were compared with porous components in forty knees (thirty-six patients)12. Although the clinical results for the two groups were the same at two years, there were more radiolucent lines under the tibial tray and around the stem of the uncoated implants. In a study of NexGen knee replacements (Zimmer), forty-six uncoated implants fixed with screws were compared with forty-six hydroxyapatite-tricalcium phosphate-coated implants fixed without screws13. At twelve months postoperatively, there was radiolucency around 56% of the uncoated femoral and 33% of the uncoated tibial components and around only one tibial component with a hydroxyapatite-tricalcium phosphate coating. Hydroxyapatite coatings have also been used for fixation of unicondylar knee replacements (Fig. 6). In a study by Epinette et al.14, 523 Unix knee replacements (Osteonics) in 457 patients were followed for seven years. The Unix tibial component is a modular design with a flat titanium tray that has been grit-blasted on its fixation interface and coated with hydroxyapatite (95% hydroxyapatite, 50-µm thickness). The tray has multiple screw-holes and a blade-keel that is inserted into the tibial eminence. The femoral component is a cobalt-chromium self-centering design coated with the same hydroxyapatite finish. In the study by Epinette et al., the survival rate was 93.9% with revision for any reason as the end point and 99.6% with aseptic loosening as the end point. Rehabilitation after treatment with the Unix replacement was found to be more rapid than that after treatment with either a cemented unicondylar knee replacement of the same design or a total knee replacement.
Hydroxyapatite and hydroxyapatite-tricalcium phosphate coatings appear to offer advantages in cementless fixation of femoral and tibial components. Radiostereometric analysis has demonstrated less migration and less inducible displacement of coated components compared with cementless components without a hydroxyapatite or hydroxyapatite-tricalcium phosphate coating105-107. Hydroxyapatite and hydroxyapatite-tricalcium phosphate-coated components have demonstrated less radiolucency, with bone filling gaps around the implant10. The stability of hydroxyapatite or hydroxyapatite-tricalcium phosphate-coated knee components appears comparable with that of cemented components. This finding suggests that long-term fixation of hydroxyapatite-coated implants will be maintained in the knee.
Five fully functional Omnifit-HA total hip stems (Osteonics) were analyzed following the deaths of the patients108. These implants had been in place for between five and twenty-five months. The hydroxyapatite coating was identified on each stem, and the mean thickness was between 26 and 44 µm compared with an initial thickness of 50 µm. There was a variable amount of apposition of bone (32% to 78% of the available surface per section). Bone was most prevalent on the surface that was close to the endosteal surface of the femur. Histologic examination showed occasional foci of bone-remodeling, including osteoclast-mediated removal of the hydroxyapatite coating. There were also areas of bone directly in contact with the metal substrate. A few particles of hydroxyapatite were present within macrophages in the adjacent bone marrow. Gross examination of the retrieved stems showed that attached bone seams followed the surface features of the hydroxyapatite-coated fixation regions of these implants. Similar findings have been recorded at revisions of well-fixed nonporous hydroxyapatite-coated stems, once the stem was removed from the femur. Figure 7 shows the bone bed associated with an Omnifit-HA stem that was removed during an acetabular revision after having been in place for thirteen years109. The surgeon noted that the stem was well-fixed at the time of the revision.
A histologic analysis was performed of two hydroxyapatite-coated Dual Geometry and three hydroxyapatite-coated threaded cups (Osteonics) that had been in place for between 4.5 and twenty-five months and had performed well up to the time of retrieval after the patient's death110. Bone apposition was mainly in areas of load transmission—i.e., at the periphery of the Dual Geometry components and at the peaks of the threads in the screw cups. Hydroxyapatite coating was detected over the majority of the implant surface. Loss of the hydroxyapatite coating was found to be more probable in the valleys of the threads on the screw-cups where stress-shielding occurred. There have been several reports on the tissue around ABG hydroxyapatite-coated total hip prostheses. In a study of five stems that had been retrieved at autopsy after functioning normally for between 3.3 and 6.2 years, four of the stems showed an extensive direct bond between the hydroxyapatite coating and the bone or between the prosthesis and the bone, without a fibrous or inflammatory interface111. The fifth stem showed progressive loosening due to osteolysis in Gruen zones 1B to 7A following polyethylene wear. The few hydroxyapatite particles that were observed were adjacent to the metaphyseal part of the stem, and none were noted distal to the coating. There were signs of active phagocytosis in areas where the bone marrow was adjacent to the coating. The amount of hydroxyapatite was greatest in the distal metaphyseal regions, indicating that the loss of coating may have been due to bone-remodeling. An analysis of the ABG hydroxyapatite-coated acetabular components, together with an additional cup retrieved at 6.6 years, was published separately112. All components had been clinically successful. There was no difference in bone contact among the three DeLee-Charnley zones. The thickness of the coating decreased with increasing duration of implantation. Osteoclastic resorption of the hydroxyapatite was seen histologically. Few hydroxyapatite particles were noted, and those that were seen were adjacent to the cup. However, the empty screw-holes contained polyethylene particles. No inflammatory reaction to the hydroxyapatite was noted. The amount of bone contact remained stable even in areas where the hydroxyapatite had completely disappeared. A case report on an infection at the site of an arthroplasty described the fate of the hydroxyapatite coating113. The infection developed shortly after the surgery and was successfully treated with the implant kept in place. The patient died from unrelated causes at six years postoperatively. The ABG hip prosthesis and the surrounding tissue were retrieved. Analysis of the stem showed that it was well fixed in the femur, and bone was in contact with 37% of the implant. The authors suggested that, in view of the excellent bone contact, loss of the coating was probably caused by bone turnover rather than coating resorption due to reduced pH at the site of the infection. A comparative analysis of porous, grit-blasted, and hydroxyapatite-coated surface features on the Biometric hip (Biomet, Bridgend, United Kingdom) was carried out on specimens retrieved post mortem114. There was significantly (p < 0.05) more ingrowth and attachment of bone to the hydroxyapatite coating than to the porous coating. There was no difference between the porous and grit-blasted surfaces in terms of bone contact. Also, no significant difference in the volume of the hydroxyapatite was noted with the passage of time, and there were no adverse consequences related to the hydroxyapatite coating. A case report described the histologic characteristics of the tissue around a well-functioning Titan Corail hip implant (Landanger) with a 155-µm-thick hydroxyapatite coating that had been retrieved at autopsy fifty-two months after implantation115. No fibrous tissue was evident, and bone was in contact with 39.9% of the perimeter of the proximal third of the stem, 62.8% of the middle third, and 65.2% of the distal third. No debris was found in adjacent tissues or in the articulating surface of the polyethylene. In comparison, a study of the Landanger type of hydroxyapatite coating revealed that it could be a source of hydroxyapatite particles116. Biopsy specimens taken from around twenty Landos Corail hip replacements showed large flakes of hydroxyapatite117. It has been pointed out that such flakes were specific to the thick Landos Corail coating and that particle release was due to the poorer bond strength of that coating (20 to 30 MPa) compared with the bond strength provided by other manufacturers (62 to 65 Mpa)117. The histologic findings are in agreement with the results of animal studies. Bone is often found against the hydroxyapatite coating. The coating generally decreases in thickness over time. The loss of coating is greatest in areas where bone-remodeling is greatest, indicating a cellular mechanism for coating loss. Fixation, or contact between the bone and implant, can be maintained in the face of coating loss. The importance of suitable coating characteristics was seen with the Landos Corail coatings, the composition and strength of which were inadequate. The lower bond strength resulted in loss of coating and particle release with the possible sequelae of third-body wear and loss of fixation.
The literature continues to provide strong clinical evidence supporting the use of hydroxyapatite coatings for fixation of the femoral stems of total hip replacements and suggests that the results with those implants are at least equal to those reported in association with other cementless designs used for any indication for total hip replacement. Experience with many different designs of hydroxyapatite-coated femoral components followed for more than ten years now shows that hydroxyapatite coatings with controlled properties do not need to be applied over a roughened or ingrowth metal substrate surface in order to achieve long-term implant stability. Indeed, the absence of ingrowth or interlock between the bone and stem may be an asset if it is necessary to remove the stem later. With regard to challenging total hip replacements such as those performed in active, young individuals (less than fifty years old), non-ingrowth hydroxyapatite-coated hip stems perhaps are the new standard against which others should be judged. However, some clinical studies of implants with thick ongrowth coatings have shown that such success can be achieved only with proper control of coating properties during manufacture. Hydroxyapatite-tricalcium phosphate mixtures have been used to promote early ingrowth into porous structures in some hips, but the effects on stem fixation seem to be limited to the enhancement of early fixation; there is little effect on long-term implant survival.
Unlike the case with the femoral stem, lack of interlock between the bone
and implant in the acetabulum is now known to be detrimental to long-term
implant survival. Early studies of hydroxyapatite-coated acetabular components
showed disappointing results. Because of the lack of inherent stability
provided by the implant geometry, press-fit sockets that relied solely on the
hydroxyapatite coating for fixation did not remain stable once much of the
coating had been removed by biological activity. After the need for adequate
stress transfer between the implant and bone through osseous interlock was
understood, several different fixation surfaces were used as the coating
substrate to achieve the necessary stability. With these designs, the bone
adaptation properties of hydroxyapatite coatings are used to provide filling
of the pores or threads on the underlying metal surface, and the implant
geometry then provides long-term interface stability. The survival rate of
The results of total knee replacements with hydroxyapatite-coated components have been promising after durations of follow-up of ten years or more. Demonstrated measures of less subsidence of the tibial component and greater resistance to inducible motion suggest stable fixation. Anecdotal reports of gap-filling between the tibial base-plate and the site of the tibial resection are encouraging. Excellent femoral and tibial survival rates at more than ten years following unicondylar knee replacements suggest that the coating is effective in this setting. However, experience with hydroxyapatite coatings in the knee is more limited than that in the hip. Multicenter studies with longer durations of follow-up and larger numbers of patients are required to demonstrate superiority over cemented total and unicompartmental devices. In the years following the review of this subject by Jaffe and Scott3, those authors' statement that the experimental and intermediate-term clinical results associated with hydroxyapatite-coated femoral stems in younger patients are encouraging appears to be supported by positive clinical results fifteen years after implantation. Evidence published since that review suggests that use of hydroxyapatite coatings has achieved successful fixation of acetabular components and total and unicondylar knee replacements at ten years or more after implantation. Histologic findings from autopsies and retrieval specimens continue to confirm preclinical findings of bone apposition to these coatings. A decrease in coating thickness with time and remodeling of the hydroxyapatite by osteoclasts reaffirm the need for inherently stable implant geometries matched to the biomechanical needs of the implant site.
Tables presenting the results in clinical series of total hip arthroplasties involving use of hydroxyapatite-coated components are available with the electronic versions of this article, on our web site at jbjs.org (go to the article citation and click on "Supplementary Material") and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM).
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
99% reported
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |