Histopathologic retrieval analysis of clinically failed porous tantalum osteonecrosis implants.

BACKGROUND: An Osteonecrosis Intervention Implant made of porous tantalum was recently developed to provide structural support and a compatible surface for tissue ingrowth in osteonecrotic femoral heads. From an investigational device exemption study that comprised 113 implants, we carried out a retrieval analysis of clinically failed implants. METHODS: Seventeen porous tantalum implants that had been used for the treatment of Steinberg stage-II osteonecrosis of the hip were retrieved at the time of conversion to a total hip arthroplasty. Fifteen implants that had been transected near the base of the femoral neck with the proximal portion left in situ within the femoral head underwent histopathologic analysis at an average of 13.4 months (range, three to thirty-six months) after implantation. RESULTS: Residual osteonecrosis was present in fourteen of the fifteen specimens. Fracture of the subchondral bone of the femoral head was present in all instances, and collapse of the femoral head was present in nine instances (60%). Backscattered scanning electron microscopy confirmed the presence of bone ingrowth in thirteen (87%) of the fifteen specimens. The mean extent of bone ingrowth was 1.9% (range, 0% to 4.4%). CONCLUSIONS: The retrieved implants were associated with little bone ingrowth and insufficient mechanical support of subchondral bone. The implant design, the surgical technique, its application, and the clinical characteristics of candidates for this procedure should continue to be monitored closely.

A physical vapor deposition method for controlled evaluation of biological response to biomaterial chemistry and topography.

The purpose of this study was to characterize a technique to effectively mask surface chemistry without modifying surface topography. A thin layer of titanium was deposited by physical vapor deposition (PVD) onto different biomaterial surfaces. Commercially pure titanium disks were equally divided into three groups. Disks were either polished to a mirror finish, grit blasted with alumina particles, or grit blasted and subsequently plasma sprayed with a commercial grade of hydroxyapatite (HA). A subgroup of each of these treatment types was further treated by masking the entire disk surface with a thin layer of commercially pure titanium deposited by PVD. A comparison of surface topography and chemical composition was carried out between disks within each treatment group. Canine marrow cells were seeded on all disk surfaces to determine the stability of the PVD Ti mask under culture conditions. The PVD process did not significantly alter the surface topography of any samples. The thin titanium layer completely masked the underlying chemistry of the plasma sprayed HA surface and the chemistry of the plasma vapor deposited titanium layer did not differ from that of the commercially pure titanium disks. Aliquots obtained from the media during culture did not indicate any significant differences in Ti concentration amongst the Ti and Ti-masked surfaces. The PVD application of a Ti layer on HA coatings formed a stable, durable, and homogenous layer that effectively masked the underlying surface chemistry without altering the surface topography.

Zoledronic acid causes enhancement of bone growth into porous implants.

The effect of zoledronic acid on bone ingrowth was examined in an animal model in which porous tantalum implants were placed bilaterally within the ulnae of seven dogs. Zoledronic acid in saline was administered via a single post-operative intravenous injection at a dose of 0.1 mg/kg. The ulnae were harvested six weeks after surgery. Undecalcified transverse histological sections of the implant-bone interfaces were imaged with backscattered scanning electron microscopy and the percentage of available pore space that was filled with new bone was calculated. The mean extent of bone ingrowth was 6.6% for the control implants and 12.2% for the zoledronic acid-treated implants, an absolute difference of 5.6% (95% confidence interval, 1.2 to 10.1) and a relative difference of 85% which was statistically significant. Individual islands of new bone formation within the implant pores were similar in number in both groups but were 69% larger in the zoledronic acid-treated group. The bisphosphonate zoledronic acid should be further investigated for use in accelerating or enhancing the biological fixation of implants to bone.

Acid-etched microtexture for enhancement of bone growth into porous-coated implants.

We designed an in vivo study to determine if the superimposition of a microtexture on the surface of sintered titanium beads affected the extent of bone ingrowth. Cylindrical titanium intramedullary implants were coated with titanium beads to form a porous finish using commercial sintering techniques. A control group of implants was left in the as-sintered condition. The test group was etched in a boiling acidic solution to create an irregular surface over the entire porous coating. Six experimental dogs underwent simultaneous bilateral femoral intramedullary implantation of a control implant and an acid etched implant. At 12 weeks, the implants were harvested in situ and the femora processed for undecalcified, histological examination. Eight transverse serial sections for each implant were analysed by backscattered electron microscopy and the extent of bone ingrowth was quantified by computer-aided image analysis. The extent of bone ingrowth into the control implants was 15.8% while the extent of bone ingrowth into the etched implants was 25.3%, a difference of 60% that was statistically significant. These results are consistent with other research that documents the positive effect of microtextured surfaces on bone formation at an implant surface. The acid etching process developed for this study represents a simple method for enhancing the potential of commonly available porous coatings for biological fixation.

Relative contributions of chemistry and topography to the osseointegration of hydroxyapatite coatings.

The purpose of the current study was to ascertain the relative contributions of surface chemistry and topography to the osseointegration of hydroxyapatite-coated implants. A canine femoral intramedullary implant model was used to compare the osseous response to commercially pure titanium implants that were either polished, grit-blasted, plasma-sprayed with hydroxyapatite, or plasma-sprayed with hydroxyapatite and masked with a very thin layer of titanium using physical vapor deposition (titanium mask). The titanium mask isolated the chemistry of the underlying hydroxyapatite layer without functionally changing its surface topography and morphologic features. At 12 weeks, the bone-implant specimens were prepared for undecalcified thin section histologic evaluation and serial transverse sections were quantified with backscattered scanning electron microscopy for the percentage of bone apposition to the implant surface. Bone apposition averaged 3% for the polished implants and 23% for the grit-blasted implants. Bone apposition to the hydroxyapatite-coated implants averaged 74% whereas bone apposition to the titanium mask implants averaged 59%. Although there was significantly greater osseointegration with the hydroxyapatite-coated implants, 80% of the maximum bone forming response to the implant surfaces developed with the titanium mask implants. This simple, controlled experiment revealed that topography is the dominant factor governing bone apposition to hydroxyapatite-coated implants.

Effects of digestion protocols on the isolation and characterization of metal-metal wear particles. I. Analysis of particle size and shape.

Isolation of metal wear particles from hip simulator lubricants or tissues surrounding implants is a challenging problem because of small particle size, their tendency to agglomerate, and their potential for chemical degradation by digestion reagents. To provide realistic measurements of size, shape, and composition of metal wear particles, it is important to optimize particle isolation and minimize particle changes due to the effects of the reagents. In this study (Part I of II), transmission electron microscopy (TEM) was used to examine and compare the effects of different isolation protocols, using enzymes or alkaline solutions, on the size and shape of three different types of cobalt-based alloy particles produced from metal-metal bearings. The effect on particle composition was examined in a subsequent study (Part II). Large particles (<1200 nm) were generated by dry abrasion of CoCrMo alloy against itself and small particles (<300 nm) were generated by hip simulator testing of a metal-metal implant pair in the presence of either distilled-deionized water or a 95% bovine serum solution. The reagents changed particle size and to a lesser extent particle shape. For both large particles and small particles generated in water, the changes in size were more extensive after alkaline than after enzymatic protocols and increased with alkaline concentration and time in solution, up to twofold at 2 h and threefold at 48 h. However, when isolating particles from 95% serum, an initial protective effect of serum proteins and/or lipids was observed. Because of this protective effect, there was no significant difference in particle size and shape for both oval and needle-shaped particles after 2 h in 2N KOH and after enzymatic treatments. However, round particles were significantly smaller after 2 h in 2N KOH than after enzymatic treatments. Particle composition may also have been affected by the 2N KOH treatment, as suggested by a difference in particle contrast under TEM, an issue examined in detail in Part II.

Fibrous tissue ingrowth and attachment to porous tantalum.

This study determined the soft tissue attachment strength and extent of ingrowth to a porous tantalum biomaterial. Eight dorsal subcutaneous implants (in two dogs) were evaluated at 4, 8, and 16 weeks. Upon retrieval, all implants were surrounded completely by adherent soft tissue. Implants were harvested with a tissue flap on the cutaneous aspect and peel tested in a servo-hydraulic tensile test machine at a rate of 5 mm/min. Following testing, implants were dehydrated in a solution of basic fuschin, defatted, embedded in methylmethacrylate, and processed for thin-section histology. At 4, 8, and 16 weeks, the attachment strength to porous tantalum was 61, 71, and 89 g/mm respectively. Histologic analysis showed complete tissue ingrowth throughout the porous tantalum implant. Blood vessels were visible at the interface of and within the porous tantalum material. Tissue maturity and vascularity increased with time. The tissue attachment strength to porous tantalum was three- to six-fold greater than was reported in a similar study with porous beads. This study demonstrated that porous tantalum permits rapid ingrowth of vascularized soft tissue, and attains soft tissue attachment strengths greater than with porous beads.

The Otto Aufranc Award. Wear and lubrication of metal-on-metal hip implants.

The implication of polyethylene wear particles as the dominant cause of periprosthetic osteolysis has created a resurgence of interest in metal-on-metal implants for total hip arthroplasty because of their potential for improved wear performance. Twenty-two cobalt chromium molybdenum metal-on-metal implants were custom-manufactured and tested in a hip simulator. Accelerated wear occurred within the first million cycles followed by a marked decrease in wear rate to low steady-state values. The volumetric wear at 3 million cycles was very small, ranging from 0.15 to 2.56 mm3 for all implants tested. Larger head-cup clearance and increased surface roughness were associated with increased wear. Independent effects on wear of material processing (wrought, cast) and carbon content were not identified. Implant wear decreased with increasing lambda ratio, a parameter used to relate lubricant film thickness to surface roughness, suggesting some degree of fluid film lubrication during testing. This study provided important insight into the design and engineering parameters that affect the wear behavior of metal-on-metal hip implants and indicated that high quality manufacturing can reproducibly lead to very low wear.

Characteristics of bone ingrowth and interface mechanics of a new porous tantalum biomaterial.

We have studied the characteristics of bone ingrowth of a new porous tantalum biomaterial in a simple transcortical canine model using cylindrical implants 5 x 10 mm in size. The material was 75% to 80% porous by volume and had a repeating arrangement of slender interconnecting struts which formed a regular array of dodecahedron-shaped pores. We performed histological studies on two types of material, one with a smaller pore size averaging 430 microm at 4, 16 and 52 weeks and the other with a larger pore size averaging 650 microm at 2, 3, 4, 16 and 52 weeks. Mechanical push-out tests at 4 and 16 weeks were used to assess the shear strength of the bone-implant interface on implants of the smaller pore size. The extent of filling of the pores of the tantalum material with new bone increased from 13% at two weeks to between 42% and 53% at four weeks. By 16 and 52 weeks the average extent of bone ingrowth ranged from 63% to 80%. The tissue response to the small and large pore sizes was similar, with regions of contact between bone and implant increasing with time and with evidence of Haversian remodelling within the pores at later periods. Mechanical tests at four weeks indicated a minimum shear fixation strength of 18.5 MPa, substantially higher than has been obtained with other porous materials with less volumetric porosity. This porous tantalum biomaterial has desirable characteristics for bone ingrowth; further studies are warranted to ascertain its potential for clinical reconstructive orthopaedics.

The osseous response to corundum blasted implant surfaces in a canine hip model.

The purpose of this study was to examine the radiographic and histologic response to corundum blasted implant surfaces of varying roughness in a canine total hip arthroplasty model. Three types of tapered femoral implants were made from titanium alloy and were identical in every respect except surface finish. The entire surface of the femoral implant had a 2.9-, 4.2-, or 6.7-micron average surface roughness (Ra) from blasting with 60-, 24-, or 16-grit corundum particles, respectively. Twenty-two stems in 11 dogs were evaluated at 6 months. Twenty-one of the stems showed osseointegration, whereas in one stem a fibrous interface developed. Abundant new periimplant bone formation occurred, particularly within the intramedullary canal where trabeculae spanned implant to endosteal cortex gaps as large as 5 mm. Bone apposition with the 60-, 24-, and 16-grit stems averaged 31.7%, 32%, and 27.9%, respectively; the differences were not statistically significant. However, the pattern of new bone formation was different in that the average length of each region of bone apposition for the 60- and 24-grit surfaces was 50% greater than that for the coarser 16-grit surface. The observations of this study indicate that because of their highly osteoconductive nature, corundum blasted surfaces represent an important and valuable technology for the design of noncemented implants.

Tissue response to porous tantalum acetabular cups: a canine model.

This study evaluated the osseous tissue response to a noncemented metal-backed acetabular component made of a new porous tantalum biomaterial. Eleven dogs with bilateral total hip arthroplasties (22 acetabular implants) were studied for a period of 6 months. Thin section histology, high-resolution radiography, and backscattered scanning electron microscopy revealed that all 22 implants had stable bone-implant interfaces. Regions of bone ingrowth were present in all histologic sections. The depth of bone ingrowth varied from 0.2 mm to the maximal limit of 2 mm. Analyzing contiguous regions of interest across the full bone-implant interface, the mean bone ingrowth for all sections was 16.8% +/- 5.7%. In the peripheral regions of the cup where bone-implant contact was most consistent, bone ingrowth averaged 25.1% +/- 10.1%. The data indicate that the porous tantalum material is effective for biologic fixation in the dog and may provide a suitable alternative to other porous materials used in acetabular cup design.

Effect of flexibility of the femoral stem on bone-remodeling and fixation of the stem in a canine total hip arthroplasty model without cement.

The purpose of this study was to compare, with regard to fixation of the implant and femoral bone resorption, two fully porous-coated stems of different stiffnesses in a canine total hip arthroplasty model. A bilateral arthroplasty was carried out with insertion of a titanium-alloy stem (which had stiffness properties comparable with those of the canine femur) on one side and with insertion of a composite stem (which was three to fivefold more flexible than the canine femur) on the contralateral side. Eight femora were evaluated at six months and eight, at eighteen months after the operation, to determine the extent of bone ingrowth, periprosthetic cortical area, intracortical porosity, and bone-remodeling. Despite the markedly greater flexibility of the composite stems, no significant difference could be detected (with the numbers available), with regard to the overall degree of femoral stress-shielding, cortical area, or cortical porosity, between these stems and the stiffer, titanium-alloy stems at either time-period. However, the composite stems had less bone ingrowth and more formation of radiopaque lines than did the titanium-alloy stems. At eighteen months, the values for bone ingrowth were 9.7 +/- 5.38 percent (mean and standard deviation) for the composite stems compared with 28.1 +/- 5.31 percent for the titanium-alloy stems (p = 0.003). Furthermore, the histological sections from the femora containing a composite stem showed radiopaque lines indicative of fibrous ingrowth approximately threefold more often than did those from the femora containing a titanium-alloy stem (p = 0.02).

Kinematics of the MATCO hip simulator and issues related to wear testing of metal-metal implants.

Metal-metal hip implants have been used clinically in Europe to reduce the risk of wear particle induced osteolysis. Joint simulator devices could provide useful information for design improvement of the modern generation of metal-metal hip implants. Early wear results for metal-metal hip specimens were obtained using a MATCO hip simulator. A detailed kinematic analysis was developed for the MATCO simulator and applied to two of the wear experiments to predict the starting surface motion, contact zone and lubricant film thickness. It was shown that points on cup surfaces were not subjected to a reciprocating interaction with the head during wear at the beginning of testing but as wear proceeded, it was suggested that, in some cases, reciprocating interaction did occur on the cup surface. Comparison between simulator and in vivo kinematics suggested a more realistic representation for cup than for head wear. In the simulator, the Hertzian contact zone moved in a circular path over the cup surface and changed in size in correspondence with the applied load. Elastohydrodynamic lubrication was considered to be possible in the simulator, with estimated fluid film thickness as great as 0.1 micron. However, such thick films were not likely to have occurred at the start of the two wear tests which were examined in detail, although some mixed film lubrication might have accounted for the relatively low wear of one of the specimens. The inclusion of kinematic details, contact mechanics and elastohydrodynamic lubrication analysis in simulator testing protocols and in design of metal metal hip implants was recommended.

Engineering issues and wear performance of metal on metal hip implants.

A major concern in total hip arthroplasty is the generation of polyethylene wear particles at the articulating surfaces and resulting macrophage mediated periimplant osteolysis. There is renewed interest in metal on metal bearings as a solution to this problem in view of their potential for greatly improved wear performance. Using a commercially available hip simulator, the wear performance of metal on metal femoral head and acetabular cup combinations was evaluated and various parameters affecting metal on metal implant wear were identified. Nine implants custom manufactured from 2 medical grades of CoCrMo alloy (ASTM F1537-95 and F75-92) were tested within bovine serum as the lubricant to 3 million cycles (equivalent to approximately 3 years of service in vivo). The progressive wear of the components was determined by gravimetric methods at approximately every 300,000 cycles. The wear rates were characterized by an initial period of accelerated wear after which a lower steady state wear rate was observed for subsequent cycles. The presence of calcium phosphate films on the component surfaces, the microstructure of the lower carbon, wrought alloy, and increased effective radii (decreased diametral clearances) were identified as factors that may be favorable to improved wear performance. The extent of the effect on wear of each parameter, however, cannot be discerned at this point and necessitates a study in which parametric changes are more tightly controlled. The present study suggests that the use of metal on metal articulating surfaces may mitigate the problem of osteolysis by offering improved wear performance.

Comparison of alloys and designs in a hip simulator study of metal on metal implants.

Previous studies of metal on metal hip implants have shown that it is possible to obtain substantially lower volumetric wear than with metal on polyethylene pairings. To work toward design optimization, the gravimetric wear of serum lubricated, metal on metal implants was examined in a hip simulator apparatus. Seventeen implant specimens were fabricated from 3 alloys in various combinations of diameter and clearance and wear tested as many as 3 million cycles. Theory was developed to predict the thickness of elastohydrodynamic lubricant films, the linear wear, and wear zone geometry. To help interpret the data, implant diameter and clearance were combined into a single parameter called effective radius. Implants with larger effective radii (6-11 m) tended to experience lower wear, and theory suggested that this was a result of the correspondingly thicker elastohydrodynamic lubricant films providing an enhanced mixed film lubrication. As in most wear testing, the results were scattered but, overall, a low carbon, wrought alloy had lower wear than either a high carbon, cast alloy or a low carbon, commercial alloy. Because of the small number of specimens tested and possible variations in tribologic parameters such as calcium phosphate film formation and surface microgeometry, there was low statistical reliability of the results.

Fatigue behavior of titanium femoral hip prosthesis with proximal sleeve-stem modularity.

Modular hip prostheses are increasing in variety and utilization. Component stability, high endurance limit, and minimal particulate debris generation are critical for long-term clinical success. The purpose of this study was to characterize the fatigue response and evaluate the in vitro potential for component motion and wear of the S-ROM¿, a Ti-6Al-4V hip prothesis with a modular design based on a Morse taper connection. A fatigue jib was designed to simulate fixation of the device at the sleeve-bone interface only with distal support mainly against the lateral endosteal cortex. Two series of tests were performed in air at room temperature: one with direct vertical loading (to produce high bending moments in the coronal plane) and one with a compounding loading angle directed at 15 degrees out-of-plane (to include torsional physiological loads). Applied loads using a servohydraulic test machine ranged from 5 x BW (body weight) to 9 x BW (1 x BW = 73 kg, approximately 160 lb) at 10 Hz on an Instron apparatus. No mechanical failures were observed on the 11-mm size stems below 6 x BW for in-plane vertical loading, and at or below 7 x BW for out-of-plane loading. Using displacement monitoring with a sensitivity of 35 mum, no measurable slippage or relative motion was detected between the stem and sleeve when they were properly assembled. Examination of the contact areas with scanning electron microscopy releaved random surface modification (an indication of fretting or burnishing) with occasional evidence of transfer of material between stem and sleeve.(ABSTRACT TRUNCATED AT 250 WORDS)

Concerns with modularity in total hip arthroplasty.

Modularity is being diversified in total hip prostheses to increase surgical latitude in optimizing implant fixation and adjusting hip biomechanics. However, several problems have been clearly identified with implant modularity. First generation metal-backed acetabular components have shown deficiencies in the locking mechanism, the congruency and extent of polyethylene liner support, and polyethylene thickness, all of which have been implicated in accelerated polyethylene wear and failure. Evidence of screw motion against the metal backing, release of particulate material, and focal osteolysis have also been observed. At the head/neck junction evidence of corrosion and fretting has been documented with both similar-metal and mixed-metal taper combinations. Femoral prostheses with other sites of modularity present additional concerns with regard to mechanical integrity and generation of particulate debris by fretting. The modular junctions of three hip prostheses, the S-ROM, Infinity, and RMHS, were subjected to wet environment high cycle mechanical testing in a worst-case loading scenario. Preliminary results at relatively low loads up to three times body weight indicated gross stability of the modular junctions with evidence of minor fretting damage. Analysis of water solutions surrounding the modular junctions after ten to 20 million loading cycles yielded counts of one to three micron sized particles totalling several hundred thousand to several million. It is unknown what quantity of particulate material is sufficient to cause macrophage-mediated osteolysis or whether the debris from modular junctions can cause third-body wear of the articulating surfaces. Modular hip prostheses should be examined under stringent test conditions in order to characterize their fretting behavior and establish their mechanical limitations.

Mechanical compatibility of noncemented hip prostheses with the human femur.

It is generally accepted that more flexible implants are needed to reduce stress shielding and postoperative thigh pain. However, there is no detailed information on the stiffness of currently used implants relative to the human femur. The purpose of this study was to determine the stiffness characteristics (bending, torsional, and axial) of human femora relative to commercially available prostheses as a first step in assessing the mechanical compatibility of the implants. This was achieved by computerized tomography scanning of a collection of human femora from proximal to distal at 10 mm intervals, digitizing the cross-sectional contours, and calculating the stiffness characteristics of each section using standard beam theory. The results show that significant stiffness mismatches exist, especially for larger stem sizes and for stems fabricated from cobalt-chrome alloy. Interestingly, certain implant stiffness values are lower than those of the femur for stems up to 15 mm in diameter, substantially so if the implant is made from titanium alloy and incorporates design features that reduce area and moments of inertia. The data suggest that only larger implant sizes need to be adjusted for increased flexibility compared with current stands.

The effect of stem stiffness on femoral bone resorption after canine porous-coated total hip arthroplasty.

Bilateral noncemented total hip arthroplasty (THA) was produced in dogs to determine the effect of stem stiffness on stress-related bone resorption. Two porous-coated femoral implants of substantially different stiffnesses were designed for direct comparison. One was manufactured from cobalt-chromium (CoCr) alloy, the other from titanium alloy. The titanium stem was hollowed out to a wall thickness of 1 mm to further reduce its stiffness. The cumulative stiffness differences were about 5.4-fold axially and 3.6-fold in bending and torsion. Staged bilateral THA was performed on eight dogs. Each dog received a stiff CoCr stem on one side and a flexible titanium stem on the other. After death, the femora were removed and processed for undecalcified thin-section histology. Bone ingrowth and remodeling were quantified by computer-aided image analysis and compared between stem designs. All femoral specimens showed bone ingrowth fixation of both stiff and flexible stems along the implant length. Tetracycline labeling indicated active bone turnover in the femoral cortex and in regions of ingrowth. However, gross differences in femoral bone remodeling were observed both roentgenographically and histologically. Femora with the flexible stems consistently showed much less bone resorption than those with the stiff stems. Quantitative analysis of paired cross-sections indicated an average of 25%-35% more cortical bone area in the femora with flexible stems. Severe resorption of the cortex in the midstem region occurred in three of the femora with the stiff stems but in none with the flexible stems. Stem stiffness strongly influences bone remodeling. The flexible stem results in more uniform load transfer and less stress shielding.