Clinical data and indices of patient satisfaction have supported the use of patellar components in total knee replacement1,11,17,25,28. However, there remain serious concerns about the indications for and the efficacy of use of porous-coated metal-backed patellar components. Reported problems of loosening, failure of the metal backing, stress-shielding, and polyethylene wear3,12,19,26 have led some to avoid the use of metal-backed patellar components. In contrast, Laskin and Bucknell27 as well as Evanich et al.18 determined that these components could be clinically successful over a ten-year period. In those two series, the metal backing was placed deep (countersunk) into the patella so that it was recessed into the bone, thus avoiding possible metal-on-metal articulation with the femoral component if substantial polyethylene wear occurred. These operative techniques and component designs also allow for thicker polyethylene, which can prevent early wearing through. In the clinical studies in which metal-backed patellar components were condemned, the implants were not countersunk into the bone and the polyethylene along the outer margin of the metal backing was thin3,12,19,26.
Another improvement cited by advocates of metal-backed patellar components was the use of a femoral component in which the trochlear groove was similar in structure to the naturally occurring, highly congruent trochlear groove in the human femur. This design factor, in combination with soft-tissue balancing and use of properly sized components, prevents excessive loads from being placed on the polyethylene of the patellar component, increases the range of motion of the knee, and prevents the rapid wearing through that has been reported in other clinical series3,12,19,26.
We are aware of no quantitative postmortem study that has demonstrated whether it is possible to achieve consistent bone growth into porous-coated metal-backed patellar components. Such data are important because Firestone et al.19 attributed a rate of revision of 7 per cent (twenty-two of 334 components) to aseptic mechanical loosening of metal-backed porous-coated patellar components within thirty-five months after implantation.
An objective of the current study was to document quantitatively the bone ingrowth that had occurred in a series of porous-coated patellar components that were retrieved consecutively post mortem. Another objective was to test the hypothesis that the bone ingrowth measured in the porous coating would be equal to the amount of host cancellous bone three millimeters away from the interface. Such data are helpful in determining whether bone ingrowth can be achieved consistently and in understanding the relationship between the amount of host cancellous bone and the amount of bone ingrowth in the porous coating. The third objective was to conduct gross, radiographic, and histological analysis to determine the host-tissue response, to assess whether the implant had been damaged, and to quantitate the presence of particulate debris and radiolucent lines at the interface of this particular implant system.
Materials and Methods
Eleven patellar components (Natural Knee; Sulzer Orthopedics, Austin, Texas) were retrieved consecutively from the cadavera of eight men at autopsy. The mean age (and standard deviation) of the patients at the time of death was 74 ± 7 years (range, sixty-six to eighty-seven years). The eleven implants had been in situ for a mean of 45 ± 36 months (range, one to eighty-four months) (Table I).
All patients had had a well functioning total knee replacement according to a review of the records from the latest office visit, which included modified Hospital for Special Surgery knee scores18 (Table I), and as determined by interviews with the patient's relatives to ascertain whether the patient had had any symptoms referable to the knee at the time of death. Knee scores were recorded before the operation, at six weeks, at three months, and at one or two-year intervals thereafter. Knee scores were recorded for all but one patient three to seven months before the time of death. The exceptional patient (Case 1) died before the initial six-week follow-up examination.
The porous coating of the patellar component was a commercially pure titanium cancellous structure with a mean pore size of 530 micrometers and a porosity of 55 per cent, according to the manufacturer. The minimum thickness of the peripheral polyethylene was 1.6 millimeters and that of the central polyethylene of the five different sizes was ten millimeters. The maximum thickness of the metal backing was three millimeters in all components. Each component was torsionally stabilized with three smooth titanium-alloy pegs.
All operations were performed by one of us (A. A. H.) according to a standardized operative approach22 with use of instruments specifically designed for the implant22. The patellar components were countersunk three millimeters, and a layer of autogenous-graft bone chips was interposed between the implant and the host bone23. These bone chips were obtained from the undersurface of the resected tibial wafer with use of a patellar reamer and were spread evenly in a layer of less than one millimeter in thickness over the resected surface of the patella with the reamer in reverse mode. This layer of bone chips at the interface was unincorporated and thus visible on backscattered electron images of the specimen from the patient who had died one month postoperatively (Fig. 1).
In addition, a measured resection technique was employed and the implant was placed medially, centered over the sagittal ridge, with use of a technique to improve tracking, described previously by Hofmann et al.24. The medialization procedure is accomplished by initially measuring the patellar thickness in order to reproduce the original thickness. The sagittal ridge then is identified and is marked on the center at the midpoint between the superior and inferior portions of the articular surface. In patients who have severe osteoarthrosis, the location of the normal ridge is approximated. A 3.2-millimeter drill-bit then is used to drill into the patella at this marked position to a depth of ten millimeters (the first mark on the calibrated drill-bit). The osteotomy guide then is placed at the medial and lateral osteocondylar junction to ensure a flat uniform surface when the patella is resected. The cut is made, and seven millimeters of bone are removed. The previously drilled hole then is identified with use of a hemostat. Next, the patellar sizer is used to identify the correct-size patellar component that will allow placement of the center of the component directly over the drill-hole, reproducing the position at the original apex of the patella while permitting a continuous rim of bone around the component. The correct-size patellar reamer then is selected and positioned so that the drill-hole is in the center of the reamer bushing. The reamer is used to remove an additional three millimeters of bone, to allow for countersinking of a ten-millimeter-thick patellar component (a 2.5-millimeter-thick metal backing and a 7.5-millimeter polyethylene thickness). The drill-guide is placed, and the appropriate drill-holes for the pegs are made to accept a metal-backed or all-polyethylene patellar component. If the final patellar component construct is as much as one millimeter thicker than the original thickness of the patella, the component is removed and the patella is reamed further to duplicate the original thickness precisely.
Post mortem, the patellae were retrieved with the implant in situ to ensure that the bone-implant interface remained undisturbed.
Gross photographs and high-resolution contact radiographs were made of the specimens, which then were fixed in 70 per cent ethanol, dehydrated in ascending grades of ethanol, infiltrated, and embedded in methylmethacrylate15,31. The polymerized blocks were cut serially into three-millimeter sections with use of a custom-made, water-cooled, high-speed, cut-off saw7. A mean of seven sections was cut along the sagittal plane of each specimen.
Analysis of High-Resolution Contact Radiographs
High-resolution contact radiographs were made of each three-millimeter section with use of a high-resolution film (Kodak SO343; Eastman Kodak, Rochester, New York) at fifty-five kilovolts and 1.0 milliampere for 1800 seconds in a Torrex-120D radiography cabinet (Scanray, Hawthorne, California)5,8,9.
The appositional bone index was measured on the high-resolution contact radiographs of all sections of each component5,8,9. With use of an eight-times-magnification photo loupe, the radiographs were viewed on a light box, and measurements were performed with use of a calibrated, handheld digital caliper (CD-6B; Mitutoyo, Tokyo, Japan). The total linear length of the porous-coated interface (ΣL1) as well as the total linear length of the bone that appeared to be in direct contact with the porous coating were measured for each section. The lengths of areas in which a radiolucent line was interposed between the bone and the implant (ΣL2) were subtracted from the length of the porous coating (ΣL1), resulting in a lower appositional bone index. The appositional bone index (ABI%) was calculated for the entire component as the percentage of bone that appeared to be in contact with the porous-coated regions of the component, with use of the equation: ABI% = ([ΣL1 - ΣL2]/ΣL1) x 100%.
Although this method is not as accurate as backscattered electron imaging because of the superimposition of multiple trabeculae (the projection effect) that naturally occurs2 in three-millimeter sections, it does provide an accurate indication of the quantity of radiolucent lines interposed between the bone and the implant and of the overall apposition of bone to the implant interface5,8. Calculation of the appositional bone index is a more accurate method for determining bone apposition and the presence of radiolucent lines than the use of clinical radiographs, which have comparatively larger projection-effect errors. The magnitude of errors caused by the projection effect is related to the thickness of the specimen2.
Backscattered Electron Imaging
After the radiographic analysis, the sections were ground and polished to an optical finish with use of a variable-speed grinding wheel (Buehler, Lake Bluff, Illinois)10,30 and were sputter-coated with a conductive layer of gold (Hummer VI-A; Anatech, Alexandria, Virginia) for approximately one minute. The sections then were examined with a scanning electron microscope (JSM T-330A; JEOL, Peabody, Massachusetts) with use of a backscattered electron detector (Tetra; Oxford Instruments, Cambridge, United Kingdom) at fifty times magnification. The entire porous-coated interface of the components was imaged (mean, nine fields per section and sixty-three fields per component) and analyzed with use of a semiautomated image-analysis system (Crystal; Oxford Instruments, Foster City, California)4,5,8. All images and photomicrographs were labeled according to section number and region along the entire length of the porous coating at the time that they were made (Fig. 2). This allowed them to be traced back to their point of origin, and comparison of the results for each region of a component helped us to determine if bone ingrowth was influenced by anatomical location5.
Bone ingrowth was measured as the percentage volume fraction of bone in the total volume of available pore space over all sections of all eleven components. Similarly, the periprosthetic bone was measured as the percentage volume fraction of bone directly outside of the porous coating over all sections of all components. In addition, the host bone was measured as the percentage volume fraction of bone three millimeters away from the porous-coated interface over all sections of all components (Fig. 3).
Statistical analysis with the Number Cruncher Statistical System (J. L. Hintze, Kaysville, Utah) demonstrated that not all of the data were normally distributed; therefore, a Kruskal-Wallis multiple-comparison z-value test was used to determine statistical differences between the percentage volume fractions of bone ingrowth, periprosthetic bone, and host bone. The level of significance was established as p < 0.05.
After the backscattered electron-imaging analysis was completed, the polished surfaces of thirty-two sections were attached to plastic slides (Wasatch Scientific, West Valley City, Utah), ground to a thickness of approximately fifty to seventy micrometers, stained with Sanderson rapid bone stain (Surgipath Medical Industries, Richmond, Illinois)10, and examined at 200 to 400 times magnification for the presence of a soft-tissue cellular response to particulate debris with use of polarized light microscopy. Selection and analysis of the sections were based on the presence of radiolucency, pegs, and a tissue pattern that may have suggested the presence of particulate wear debris. The regions surrounding the pegs were analyzed carefully for evidence of stress-shielding, and the outer margins of the implant were examined for osteolysis. The regions were graded according to the system of Dorr et al.14. Grades 0 through 3+ (with 3+ indicating the greatest number) were determined according to the number of metal or polyethylene particles or according to the number of giant cells or macrophages found in the observation field.
Gross analysis revealed that one of the two implants that had been in situ for eighty-four months (Case 6L) had an approximately two-millimeter-square area on the lateral side in which the polyethylene was completely worn through. No scoring of the trochlear groove of the femoral component or tissue metallosis was observed, but there was visible burnishing of the lateral aspect of the trochlear groove corresponding to the area of wear of the patellar component. There was no palpable surface roughness. The component in the right knee of the same patient, which had been in situ for seventy-two months, demonstrated some moderate cold flow of the polyethylene. The remaining nine patellar components had no observable surface or structural damage. The components in this study had 1.6 millimeters of polyethylene along the outer boundary of the metal backing.
Analysis of High-Resolution Contact Radiographs
Evaluation of the high-resolution contact radiographs of the eleven patellar components revealed a mean appositional bone index (and standard deviation) of 86 ± 12 per cent (range, 61 to 100 per cent) (Table I). A mean of 14 per cent of the porous coating was separated from the periprosthetic bone by an interposing layer of fibrous tissue, 0.5 to one millimeter thick (Fig. 4).
Small radiolucent lines occasionally were observed along the pegs, but they were not uniform along the lengths of the pegs. The radiographs revealed no signs of stress-shielding attributable to use of the titanium-alloy pegs (Fig. 5-A).
Backscattered Electron Imaging
Backscattered electron imaging revealed approximation of the bone to the porous coating and the pegs in most regions (Fig. 5-B). The volume fraction of bone growth into the porous coating averaged 13 ± 9 per cent (range, 0 to 30 per cent) for the eleven components (Table I).
The volume fraction of periprosthetic bone directly adjacent to the porous coating averaged 29 ± 14 per cent (range, 11 to 48 per cent). This volume fraction was significantly (average, 15 per cent) higher than that of bone growth into the porous coating (p < 0.05, Kruskal-Wallis multiple-comparison z-value test). Ten of the eleven components had less bone in the porous coating than in the periprosthetic region (Table I).
The mean volume fraction of host bone three millimeters away from the interface was 19 ± 8 per cent (range, 8 to 31 per cent). We detected no significant difference (p > 0.05, Kruskal-Wallis test), with the numbers available, between the volume fractions of periprosthetic and host bone.
Both components from one cadaver (Case 1) had been in situ for a very short period of time (one month). If these two components are eliminated from the analysis, the mean volume fraction of bone ingrowth was 16 ± 7 per cent (range, 5 to 30 per cent); the mean volume fraction of periprosthetic bone, 31 ± 14 per cent (range, 16 to 48 per cent); and the mean volume fraction of host bone, 18 ± 8 per cent (range, 8 to 31 per cent). The data continue to support the hypothesis that the amount of bone growth into the porous coating would not be measurably different than the amount of host bone. The data also indicate that, when a porous-coated device is implanted into human cancellous bone, the amount of bone available at that anatomical location will determine the amount of bone that can grow into the porous coating. The volume fraction of periprosthetic bone remained significantly higher than those of bone ingrowth and host bone (p < 0.05).
The sectional and regional analysis of bone ingrowth did not demonstrate any preference for a particular region along the surface of the implant. The bone growth into the porous coating averaged 13 ± 9 per cent (range, 1 to 28 per cent) in the lateral region, 14 ± 10 per cent (range, 0 to 32 per cent) in the central region, and 11 ± 8 per cent (range, 0 to 28 per cent) in the medial region (regions 1, 2, and 3, respectively, in Fig. 2). With the numbers available, no significant difference was detected among the volume fractions in the three regions (p > 0.05). Similarly, no significant difference was detected among the volume fractions of either periprosthetic bone or host bone in the three regions.
Backscattered electron imaging demonstrated that autogenous-graft bone chips placed at the interface at the time of implantation had integrated with the host bone by three months (Fig. 6-A). In some regions, the host bone had not integrated with the porous coating and a layer of fibrous connective tissue was interposed between the bone and the surface of the implant (Fig. 6-B). This was confirmed with light microscopy.
Particles of debris consisting of polyethylene (grade 1+) and metal (grade 2+) were observed along the outer boundary of the eighty-four-month specimen obtained from one cadaver (Case 6L) (Fig. 7). These particles were contained within a 0.2 to 0.5-millimeter outer margin along the perimeter of the interface between the metal backing and the bone. Few macrophages or giant cells were observed (grade 1+ for both).
In the remaining ten specimens, predominantly bone marrow and bone were abutted against the porous coating. Occasionally, fibrous connective tissue was observed in the porous coating near the outer margins; polyethylene particles (grade 1+) were found intermittently in this tissue (Fig. 8). Fibrous tissue containing polyethylene debris was found in less than 5 per cent of the total available surface area of the porous coating.
No particles of debris were evident in the areas of bone ingrowth. Limited regions of fibrous-tissue attachment demonstrated a predominant fibroblast-and-collagen network within the porous coating.
The results of this study demonstrate that bone ingrowth and skeletal attachment can be achieved in commercially pure titanium porous-coated metal-backed patellar components. The fixation by bone ingrowth was maintained for a mean of forty-five months and was not found to be significantly different, with the numbers available, from the amount of host cancellous bone in the patellae. This finding supported the stated hypothesis. These results differ from those of studies of metal-backed patellar components by Firestone et al.19 and by Kobs and Lachiewicz26, who attributed clinical complications within the first thirty-five months after implantation to the absence of bone ingrowth. In the current series, the appositional bone index indicated that a mean of 86 per cent of the porous-coated surface was not associated with radiolucent lines on high-resolution contact radiographs and appeared to be in contact with the periprosthetic bone. This represents a large region of surface contact between the porous coating of the implant and the skeleton, and, in addition to the presence of bone ingrowth in the porous coating, it could be interpreted as confirmation that the implants were clinically stable.
In a recent study, from our laboratory, of seven metal-backed acetabular components that had the same porous coating as those in the current study and that were retrieved post mortem8, all components had less bone in the porous coating than in the periprosthetic region (mean, 12 ± 6 per cent compared with 36 ± 11 per cent). This finding was attributed to the influence of the autogenous-graft bone-chip layer used at the time of implantation or to the bone chips that formed during the reaming process8. There was also less bone in the porous coating than in the periprosthetic region (mean, 6 ± 2 per cent compared with 15 ± 5 per cent) in a study, from our laboratory as well, of eight porous-coated metal-backed tibial components that were retrieved post mortem5. The present investigation also demonstrated that the amount of bone ingrowth in the porous coating of the components that were retrieved after a long time in situ was less than that in the periprosthetic region (mean, 16 ± 7 per cent compared with 31 ± 14 per cent). Because we detected no significant difference, with the numbers available, between the amount of bone ingrowth and the amount of host cancellous bone, our hypothesis that the bone volume would be equal in both regions was supported. Although these data are interesting, consideration must be given to how much bone is needed in the porous coating to provide long-term fixation and patient satisfaction.
It appears, on the basis of quantitative studies of components retrieved post mortem from our laboratory5,8 and the quantitative studies of porous-coated implants by Pidhorz et al.29 (eleven components) and Engh et al.16 (nine components), that the amount of bone ingrowth needed to attach a porous-coated implant to the skeleton ranges from 6 to 24 per cent, possibly depending on the anatomical location of the implant. Future quantitative investigations of postmortem specimens may demonstrate a relationship between the amount of human cancellous bone ingrowth in the porous coating and the amount of such bone at a specific anatomical location. This information may help orthopaedic surgeons to decide whether to perform total joint replacement with cement or to use a porous-coated device or a so-called hybrid prosthesis (one or two components inserted with cement and the others porous-coated), depending on the amount of cancellous bone available at each location. The data from the current investigation and those from previous postmortem studies conducted by our research team5,8 suggest that human cancellous bone ingrowth is biologically limited to an amount that is equal to or slightly less than the amount present at each anatomical location. This assumes that no bone is impacted into the porous coating during implantation. Hofmann et al.21 emphasized that impaction of bone into a porous-coated device at the time of the operation can mislead investigators with regard to the biological potential for bone ingrowth and can lead them to overestimate the capacity of human cancellous bone to grow into such a device.
The poor clinical results associated with use of cobalt-chromium-alloy porous-coated metal-backed patellar components that were reported by Firestone et al.19 may be attributable to two main factors. The pegs of the porous-coated anatomical system are porous-coated and are therefore prone to stress-shielding. Hedley et al.20 observed that porous-coated pegs produced dramatic stress-shielding, preventing bone growth into the remainder of the porous coating on the undersurface of the implant. The high rate of radiolucency in the study by Firestone et al., indicating lack of bone ingrowth, could be due to stress-shielding by the porous-coated pegs. The gaps between the undersurface of the implant and the host bone may in turn have contributed to the instability of the components. Although the smooth pegs of the patellar components in the current investigation were in close approximation with the host bone, stress-shielding had not occurred. The other factor contributing to the difference in the results between the current study and that of Firestone et al.19 may have been our use of autogenous-graft bone chips at the time of the operation to encourage bone ingrowth, as advocated by Hofmann et al.22,23.
Histologically, there were small regions of fibrous connective tissue containing particles of metal and polyethylene along the outer margins of one of the longer-term specimens. Regions of fibrous tissue were also seen between the porous coating and the periprosthetic bone in other specimens, but these were sporadic and could not be attributed to micromotion or to a tissue response associated with particles of wear debris. It is more likely that the fibrous tissue in these small regions was due to gaps that were not filled with autogenous-graft bone chips at the time of the operation, thus allowing the layer of fibrous tissue to form6.
A potential concern associated with use of metal-backed patellar components is that the peripheral polyethylene may be insufficient. The minimum thickness of the peripheral polyethylene in the design that was examined in this study was 1.6 millimeters and that of the central polyethylene of the five different sizes was ten millimeters. The thickness of the peripheral polyethylene was later increased to a minimum of three millimeters. It is important to stress that cold flow of the polyethylene and wear of the patellar component are not due solely to the thickness of the peripheral polyethylene but also to the distribution of contact pressure. Collier et al.13 performed contact-pressure analysis on six patellar components and found that a metal-backed design had the lowest contact pressure because of its congruent geometry. The contact pressures of the other five, less congruent designs exceeded the yield strength of polyethylene by as much as twofold. Theiss et al.32 reported that the design of the trochlear groove has a substantial role in the rate of patellofemoral complications, with a high rate of failure being associated with a shallow groove. Additionally, Collier et al.13, in a gross analysis of retrieved components, reported severe polyethylene wear in five of twelve all-polyethylene and forty-five (39 per cent) of 115 metal-backed patellar components.
The results of the current postmortem study, combined with the previously reported clinical data18, suggest that metal-backed patellar components, when properly designed and implanted, should perform well clinically. Evanich et al.18 reported that 290 (96 per cent) of 302 metal-backed patellar components inserted without cement survived, with no clinical or radiographic evidence of loosening after follow-up of six to ten years. Those authors also noted that metal-backed patellar components with smooth pegs and a non-protruding porous coating, similar to those in the current study, could be easily revised by placing a 6.4-millimeter osteotome on the edge of the component, at its junction with the bone, and then striking the osteotome with a metal mallet. This allowed easy extraction and preservation of bone when revision was performed in active patients who had a life expectancy of more than twenty-five years after the implantation18. Loss of bone due to revision of a cemented polyethylene patellar component or of a metal-backed patellar component with porous-coated pegs and a protruding porous coating can be severe enough to force the surgeon to perform a patellectomy.
It is necessary to countersink the metal backing of the patellar component so that the edge of the backing is not exposed, with the attendant risk of metal-on-metal articulation. Should maltracking or subluxation occur, the clinical outcome in a patient who has been managed with this operative technique and implant design is similar to that obtained with use of polyethylene components inserted with cement18. As with the latter components, subluxation or maltracking could cause excessive wear or cold flow and the bone would be the next sequential material to articulate with the femoral component, alerting the orthopaedic surgeon to future potential clinical complications.
The results of the current study of cadavera agree with the successful six to ten-year clinical results reported by Evanich et al.18. Problems with metal-backed patellar components can be expected if the operative technique and design principles described earlier are not followed; however, if they are adhered to, good clinical results and consistent bone ingrowth appear to be achievable18. The data in the current study and the corresponding clinical results18 are encouraging. However, the broad clinical use of metal-backed patellar components awaits a prospective, randomized, double-blind clinical study comparing the results of replacements with cementing of the patellar component with those of replacements in which the patella is not resurfaced. Such an investigation is necessary in order to understand the long-term importance of resurfacing of the patella with a metal-backed component in total knee replacement.
*One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund or foundation, educational institution, or other non-profit organization with which one or more of the authors is associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding sources were the Department of Veterans Affairs Medical Research Funds, Veterans Affairs Medical Center, and the Department of Orthopaedics, University of Utah School of Medicine, Salt Lake City, Utah, and Sulzer Orthopedics, Incorporated, Austin, Texas.
Investigation performed at the Veterans Affairs Medical Center and the Department of Orthopaedics, University of Utah, Salt Lake City
1. Abraham, W.; Buchanan, J. R.; Daubert, H.; Greer, R. B., III; and Keefer, J.: Should the patella be resurfaced in total knee arthroplasty? Efficacy of patellar resurfacing. Clin. Orthop.
, 236: 128-134, 1988.
2. Bachus, K. N., and Bloebaum, R. D.: Projection effect errors in biomaterials and bone research. Cells and Mater.
, 2: 347-355, 1992.
3. Berry, D. J., and Rand, J. A.: Isolated patellar component revision of total knee arthroplasty. Clin. Orthop.
, 286: 110-115, 1993.
4. Bloebaum, R. D.; Rubman, M. H.; and Hofmann, A. A.: Bone ingrowth into porous-coated tibial components implanted with autograft bone chips. Analysis of ten consecutively retrieved implants. J. Arthroplasty
, 7: 483-493, 1992.
5. Bloebaum, R. D.; Bachus, K. N.; Jensen, J. W.; and Hofmann, A. A.: Postmortem analysis of consecutively retrieved asymmetric porous-coated tibial components. J. Arthroplasty
, 12: 920-929, 1997.
6. Bloebaum, R. D.; Bachus, K. N.; Momberger, N. G.; and Hofmann, A. A.: Mineral apposition rates of human cancellous bone at the interface of porous coated implants. J. Biomed. Mater. Res.
, 28: 537-544, 1994.
7. Bloebaum, R. D.; Merrell, M.; Gustke, K.; and Simmons, M.: Retrieval analysis of a hydroxyapatite-coated hip prosthesis. Clin. Orthop.
, 267: 97-102, 1991.
8. Bloebaum, R. D.; Mihalopoulus, N. L.; Jensen, J. W.; and Dorr, L. D.: Postmortem analysis of bone growth into porous-coated acetabular components. J. Bone and Joint Surg.
, 79-A: 1013-1022, July 1997.
9. Bloebaum, R. D.; Rhodes, D. M.; Rubman, M. H.; and Hofmann, A. A.: Bilateral tibial components of different cementless designs and materials. Microradiographic, backscattered imaging, and histologic analysis. Clin. Orthop.
, 268: 179-187, 1991.
10. Bloebaum, R. D.; Sanderson, C.; McCarvill, S.; and Campbell, P.: Plastic slides in the preparation of implant and tissue for interface analysis. J. Histotechnol.
, 12: 307-310, 1989.
11. Boyd, A. D., Jr.; Ewald, F. C.; Thomas, W. H.; Poss, R.; and Sledge, C. B.: Long-term complications after total knee arthroplasty with or without resurfacing of the patella. J. Bone and Joint Surg.
, 75-A: 674-681, May 1993.
12. Buechel, F. F.; Pappas, M. J.; and Makris, G.: Evaluation of contact stress in metal-backed patellar replacements. A predictor of survivorship. Clin. Orthop.
, 273: 190-197, 1991.
13. Collier, J. P.; McNamara, J. L.; Surprenant, V. A.; Jensen, R. E.; and Surprenant, H. P.: All-polyethylene patellar components are not the answer. Clin. Orthop.
, 273: 198-203, 1991.
14. Dorr, L. D.; Bloebaum, R.; Emmanual, J.; and Meldrum, R.: Histologic, biochemical, and ion analysis of tissue and fluids retrieved during total hip arthroplasty. Clin. Orthop.
, 261: 82-95, 1990.
15. Emmanual, J.; Hornbeck, C.; and Bloebaum, R. D.: A polymethyl methacrylate method for large specimens of mineralized bone with implants. Stain Technol.
, 62: 401-410, 1987.
16. Engh, C. A.; Zettl-Schaffer, K. F.; Kukita, Y.; Sweet, D.; Jasty, M.; and Bragdon, C.: Histological and radiographic assessment of well functioning porous-coated acetabular components. A human postmortem retrieval study. J. Bone and Joint Surg.
, 75-A: 814-824, June 1993.
17. Enis, J. E.; Gardner, R.; Robledo, M. A.; Latta, L.; and Smith, R.: Comparison of patellar resurfacing versus nonresurfacing in bilateral total knee arthroplasty. Clin. Orthop.
, 260: 38-42, 1990.
18. Evanich, C. J.; Tkach, T. K.; von Glinski, S.; Camargo, M. P.; and Hofmann, A. A.: 6- to 10-year experience using countersunk metal-backed patellas. J. Arthroplasty
, 12: 149-154, 1997.
19. Firestone, T. P.; Teeny, S. M.; Krackow, K. A.; and Hungerford, D. S.: The clinical and roentgenographic results of cementless porous-coated patellar fixation. Clin. Orthop.
, 273: 184-189, 1991.
20. Hedley, A. K.; Clarke, I. C.; Kozinn, S. C.; Coster, I.; Gruen, T.; and Amstutz, H. C.: Porous ingrowth fixation of the femoral component in a canine surface replacement of the hip. Clin. Orthop.
, 163: 300-311, 1982.
21. Hofmann, A. A.; Bloebaum, R. D.; and Bachus, K. N.: Progression of human bone ingrowth into porous-coated implants. Rate of bone ingrowth in humans. Acta Orthop. Scandinavica
, 68: 161-166, 1997.
22. Hofmann, A. A.; Murdock, L. E.; Wyatt, R. W. B.; and Alpert, J. P.: Total knee arthroplasty. Two- to four-year experience using an asymmetric tibia tray and a deep trochlear-grooved femoral component. Clin. Orthop.
, 269: 78-88, 1991.
23. Hofmann, A. A.; Bloebaum, R. D.; Rubman, M. H.; Bachus, K. N.; and Plaster, R. L.: Microscopic analysis of autograft bone applied at the interface of porous-coated devices in human cancellous bone. Internat. Orthop.
, 16: 349-358, 1992.
24. Hofmann, A. A.; Tkach, T. K.; Evanich, C. J.; Camargo, M. P.; and Zhang, Y.: Patellar component medialization in total knee arthroplasty. J. Arthroplasty
, 12: 155-160, 1997.
25. Keblish, P. A.; Varma, A. K.; and Greenwald, A. S.: Patellar resurfacing or retention in total knee arthroplasty. A prospective study of patients with bilateral replacements. J. Bone and Joint Surg.
, 76-B(6): 930-937, 1994.
26. Kobs, J. K., and Lachiewicz, P. F.: Hybrid total knee arthroplasty. Two- to five-year results using the Miller-Galante prosthesis. Clin. Orthop.
, 286: 78-87, 1993.
27. Laskin, R. S., and Bucknell, A.: The use of metal-backed patellar prostheses in total knee arthroplasty. Clin. Orthop.
, 260: 52-55, 1990.
28. Levitsky, K. A.; Harris, W. J.; McManus, J.; and Scott, R. D.: Total knee arthroplasty without patellar resurfacing. Clinical outcomes and long-term follow-up evaluation. Clin. Orthop.
, 286: 116-121, 1993.
29. Pidhorz, L. E.; Urban, R. M.; Jacobs, J. J.; Sumner, D. R.; and Galante, J. O.: A quantitative study of bone and soft tissues in cementless porous-coated acetabular components retrieved at autopsy. J. Arthroplasty
, 8: 213-225, 1993.
30. Rhodes, D. M.; Sanderson, C.; and Bloebaum, R. D.: Glue-glove technique for use in grinding and polishing thin specimens. J. Histotechnol.
, 13: 63-64, 1990.
31. Sanderson, C., and Kitabayashi, L. R.: Parallel experience of two different laboratories with the initiator Perkadox 16 for polymerization of methylmethacrylates. J. Histotechnol.
, 17: 343-348, 1994.
Copyright 1998 by The Journal of Bone and Joint Surgery, Incorporated
32. Theiss, S. M.; Kitziger, K. J.; Lotke, P. S.; and Lotke, P. A.: Component design affecting patellofemoral complications after total knee arthroplasty. Clin. Orthop.
, 326: 183-187, 1996.