The histological effects of the implantation of different sizes of polyethylene particles in the rabbit tibia" ~
S . B. Goodmant Stanford University Medical Center, Division of Orthopedic Surgery, Stanford, California 94305-5326
V.L. Fornasier Department of Pathology, Princess Margaret Hospital, University of Toronto, Toronto, Canada J. Lee and J. Kei Stanford University Medical Center, Division of Orthopaedic Surgery, Stanford, California 94305-5326 This study examines the histological effects of different sizes of polyethylene particles implanted into the rabbit tibia. Seventeen mature New Zealand white female rabbits were allocated into three groups. Group 1 (5 rabbits) received polyethylene particles averaging approximately 16 pm in diameter, implanted into the right proximal tibia through a drill hole. Group 2 (5 animals) received particles averaging 26 pm, and Group 3 (7 rabbits) received particles averaging 67 pm. The left tibia was drilled but not implanted. Animals were sacrificed after 16 weeks. Histological analysis disclosed decreased hematopoietic activity within the left tibia1 drill hole. In all groups,
the right tibia demonstrated positively birefringent polyethylene particles surrounded by, and within (smaller particles), histiocytes and giant cells in a fibrous tissue stroma. Statistical analysis disclosed more fibrocytes and less marrow cells at the interface of Group 3 (largest particles) compared to Group 1 and 2. Larger polyethylene particles, being less readily phagocytosed, appear to produce more fibrous encapsulation, compared to particles of a smaller size. The histological reaction stimulated by the different sizes of polyethylene particles resembled the membrane surrounding loose joint arthroplasties in humans.
INTRODUCTION
The majority of prosthetic joints in current use employ ultra-high-molecularweight polyethylene (UHMWP) as one of the bearing surfaces. However, UHMWP wear debris has also been implicated as a contributing factor to The tissues harvested the development of aseptic component *Presented at the 15th Annual Meeting Of The Society For Biomaterials in Orlando, Florida May 1989. tPlease send all correspondence to: Stuart B. Goodman, M.D., Stanford University Medical Center, Division of Orthopaedic Surgery #R171, 300 Pasteur Drive, Stanford, California 94305-5326. Journal of Biomedical Materials Research, Vol. 24, 517-524 (1990) 0 1990 John Wiley & Sons, Inc. CCC 0021-9304/90/040517-08$04.00
GOODMAN ET AL.
518
at the interface of loose cemented prostheses in humans have demonstrated polyethylene particles up to 1000-2000 pm in Whereas larger particles are embedded within a fibrohistiocytic stroma, smaller particles are found within histiocytes and multinucleated giant cells. Studies in animals have confirmed that polyethylene particles can evoke a histological response that is similar to the "pseudomembrane" surrounding loose cemented joint arthroplasties in humans.'-'' Goodman et aL9 utilized polyethylene particles up to 1000 pm in diameter, whereas Howie et al." used particles of 20-200 pm, in their animal studies. The histological response of living tissue to particulate debris is governed by many factors including the specific biomaterial chosen, the size, shape and topography of the particles, the chemical structure of the surface and the surface charge.l1-I5This study examines the histological effects of implantation of three different "intermediate" sizes of UHMWP particles in the rabbit tibia. The specific sizes chosen reflect an intermediate particle size, seen in prior human and animal studies by the authors.'^^ MATERIALS AND METHODS
Seventeen mature New Zealand white female rabbits weighing 3-4 kg were used in the study. Group 1 (5 rabbits) received UHMWP particles averaging 15.68 pm (standard deviation = 14.04 pm) in diameter (80% < 19 pm). Group 2 (5 rabbits) received UHMWP particles averaging 26.14 pm (standard deviation = 37.21 pm) in diameter (58% < 19 pm). Group 3 (7 rabbits) received the largest particles of UHMWP averaging 67.29 pm (standard deviation = 148.59 pm) in diameter (0% < 19 pm). All animals received approximately the same volume of biomaterial. The particles were generated by applying a power burr to a commercially available acetabular cup made of ultra-high-molecular-weight polyethylene (Howmedica, Rutherford, NJ). The burr was cooled after 20 s of use. The debris was then sequentially sifted to yield particles of the appropriate size, using metal sieves (VWR Scientific Inc., San Francisco, CA). The size and shape of at least 10 particles in each group were analyzed using a scanning electron microscope interfaced with a computer with software for size and image analysis. The analysis disclosed that the surface roughness was similar for all three particle groups. The particles were packaged, gas sterilized with ethylene oxide, and allowed to aerate for at least 24 h. The UHMWP particles were placed through a drill hole in the proximal right tibia of experimental animals under general anesthe~ia.~'~ The left tibia functioned as a drilled, but nonimplanted control. Animals were sacrificed at 16 weeks postoperatively by barbiturate overdose (Beuthanasia-D Shering, Kenilworth, NJ), 2 cc intravenously. The upper tibiae were harvested bilaterally and fixed in buffered formaldehyde. Following fine detailed radiographs, each tibia was cut trans-
HISTOLOGICAL EFFECTS OF PE IMPLANTATION
519
versely through the drill hole site with a saw. The tibia immediately proximal to the drill hole site was decalcified in formic acid, dehydrated in progressing concentrations of alcohol, and then double embedded in celloidin-paraffin. Five- to ten-micrometer serial transverse sections were cut from the proximal tibia1 specimen, and stained with the "WHO" stain (hematoxylin, saffron, floxine, alcian green). These were used for general assessment of morphological features of the tissue. The tibia immediately distal to the drill hole site was processed by undecalcified methods by first dehydrating in progressing concentrations of alcohol and embedding in glycol methacrylate. Sections were cut on an Isomet low-speed saw, mounted on transparent plastic supports, and the free surface was ground to a hard polish with fine-grit paper. The surface was then acid etched and stained with toluidine blue. This permitted visualization of entire cross sections of the tibia with the implanted particles. Following a qualitative assessment, quantitation of specific parameters was undertaken. One hundred cells were classified along the immediate interface of the polyethylene particles and the surrounding tissue, using an eyepiece micrometer and X450 magnification (counting intervals were 50 pm). One section was counted per limb per animal. The number and the type of cells at the interface were tabulated and the resulting values were compared within each group using a paired t-test between the right and left sides. An analysis of variance was used to compare the cell counts between the three different polyethylene groups. A more stringent 0.01 level of significance was used to decrease spurious significance in counting data, as recommended by Dr.Byron Brown, biostatistician at Stanford University. RESULTS
The control left tibiae showed fatty areas with markedly decreased hematopoiesis at the operative site. Active hematopoiesis was present in the adjacent bone marrow. There was no evidence of necrosis or inflammation. The right tibiae with the positively birefringement polyethylene particles demonstrated groups of histiocytes and multinucleated foreign body giant cells surrounding the particles in a fibrous background stroma (Figs. 1 and 2). The histiocytes could be defined as small, compact polygonal or round cells with or without small birefringent particles a few micrometers in diameter. Giant cells were particularly concentrated at sharp edges of polyethylene. A peripheral zone of fibrous pseudoencapsulation surrounded the implant area, regardless of the particle size or cellular activity within. Reactive woven bone was found interspersed within this tissue reaction. Decreased hematopoietic activity surrounded the immediate implantation site, and merged with viable marrow and normal appearing endosteal bone. Statistical analysis disclosed that histiocytes, giant cells, and fibrocytes were more prevalent and marrow cells less prevalent at the immediate polyethylene interface on the right experimental side compared to the left control side in all groups (Table I).
GOODMAN ET AL.
520
Figure 1. Large polyethylene particles are clustered to the left of the photograph, surrounded primarily by fibrous tissue. Histiocytes are also seen. A smaller polyethylene fragment is present within a multinucleated giant cell in the top right of the photograph, surrounded by fat and hematopoietic tissue. Undecalcified specimen, toluidine blue stain, partial polarization, original magnification ~ 3 0 0 .
An analysis of variance demonstrated that fibrocytes were more commonly found and marrow cells less commonly found at the interface in Group 3 (the largest particle size) compared to Groups 1 and 2. There were no statistically significant differences in the numbers of giant cells, histiocytes, lymphocytes, plasma cells, and polymorphonuclear leukocytes between the different UHMWP groups. DISCUSSION
Both particulate polymethylmethacrylate and polyethylene have been postulated as playing an important role in the genesis of aseptic loosening of joint arthroplasties.'-'0,'6 Each of these biomaterials has been shown to evoke a florid foreign body histiocytic and giant cell response when implanted in particulate form in the tibia of rabbits.',' Bulk forms of these materials produced a thinner, more bland fibrous delimiting membrane, with only occasional giant cells and histiocytes. Howie et al. documented the formation of a "loosening membrane" capable of bone resorption, when polyethylene particles between 20 and 200 pm in diameter were injected into the knee joints of rabbits, in which a non-weight-bearing, stable femoral
HISTOLOGICAL EFFECTS OF PE IMPLANTATION
521
Figure 2. Small polyethylene particles are seen as stringy vacant spaces surrounded by, and within numerous multinucleated giant cells and histiocytes. A small group of plasma cells can be seen in the lower center of the photograph. Occasional fibroblasts are seen surrounding larger particles. Undecalcified specimen, toluidine blue stain, original magnification x 300.
plug of methacrylate had been inserted." These studies have suggested a relationship between particulate cement and polyethylene debris, and the foreign body membrane surrounding loose joint arthroplasties. However, to our knowledge, no studies have explored the relationship between the size of the polyethylene particles and the resulting histological reaction. In vitra studies employing mouse peritoneal macrophages have demonstrated that round microspheres having a size range of 1-2 pm are phagocytosed most effectively, although microspheres up to 5 pm in diameter were seen within these cells.ll However, polyethylene particles found in the membrane of loose joint arthroplasties are not spherical and encompass a wide range of sizes, from tenths of micrometers to 1-2 mm in diameter.l These facts prompted the use of "intermediate" sized particles ranging from about 3 pm to approximately 1000 pm in this study. The fibrohistiocytic and giant cell reaction to "intermediate" size polyethylene particles was very similar over the fourfold size range tested. This may, in part, be due to the overlap of the particle sizes in the three groups. The authors attempted to obtain three separate particle sue distributions by meticulous sifting techniques. However, smaller particles often adhered to the larger ones. In reviewing the histological slides, the authors therefore attempted to concentrate their cell counts and interpretation around "aver-
1.4 (0.9) 3.8 (3.3) 87.6 (4.6)
(0)
1.2 (1.6) 12.6 (4.2)** 41.2 (11.9)***
0
0.2 2.0 4.0 1.0
21.2 (8.7)** 23.6 (9.4)** 0.6 (1.3) (0.4) (1.4) (2.7) (0.7)
Left
1.0 (1.4) 5.2 (2.6) 85.6 (4.7)
(0) (0)
0 (0) 10.8 (2.8)** 58.6 (11.7)*
0 0
0.2 2.2 5.6 0.2
(0.4) (1.1) (3.4) (0.4)
Left
8.4 (3.3)+$ 22.2 (10.7)$
Right
Group I1
0 (0) 32.7 (Y.Y)*** (2.20)*** 7
19.3 (9.4)+r 35.7 (7.2),,, 5.3 (4.4) 0 (0)
Right
Group 111
(1.7) (3.4) (0.5)
(0)
5.7 (5.3) 0.1 (0.4) 86.3 (5.3)
0 2.7 4.7 0.4
Left
Note: Mean (standard deviation); Group 1, polyethylene particles averaging 15.68 pm; Group 2, polyethylene particles averaging 37.21 pm; Group 3, polyethylene particles averaging 67.29 pm. Paired t-test. *p < 0.01, **p < 0.005, ***p < 0.0005.
Giant cells Histiocytes Lymphocytes Plasma cells Polymorphonuclear leukocytes Fibrocytes Other marrow cells
Right
Group I
TABLE I Mean Cell Counts At The Implant Interface
HISTOLOGICAL EFFECTS OF PE IMPLANTATION
523
age sized" particles within each group. However, the presence of smaller particles are a confounding variable in all the groups. In previous experiments by the authors, bulk polyethylene plugs 3.5 mm in diameter were shown to stimulate a fibrous encapsulation mernb~ane.~ However, even in the bulk specimens, histiocytes and giant cells were concentrated at areas of surface irregularities. It would appear that the size, shape, and topography of the particles are important determinants of the histological results to UHMWP particles in this study. Further experiments employing spherical UHMWP particles may separate the effects of these factors further. In summary, the fibrohistiocytic reaction stimulated by the different sizes of polyethylene particles in this study was very similar. The histological reaction closely resembled the membrane surrounding loose joint arthroplasties in humans, and underscores the importance of polyethylene debris in this process. The authors gratefully acknowledge the aid of the Department of Laboratory Animal Medicine, Stanford University; the Department of Pathology, Princess Margaret Hospital, Toronto; Dr. Byron Brown, biostatistician at Stanford University; Dr. James A. Davidson, Materials Research Director, Richards Medical Company Memphis, Tennessee; and Lela Terrazas. This work was supported in part by a grant from the Orthopaedic Research Education Foundation and Zimmer/Bristol-Myers.
References 1. R. S. Bell, J. Schatzker, V. L. Fornasier, and S. B. Goodman, "Study of implant failure in the Wagner resurfacing arthroplasty," 1. Bone J t . Surg., 67A, 1165-1174 (1985). 2. S. R. Goldring, M. Jasty, M. S. Roelke, C. M. Rourke, F. R. Bringhurst, and W. H. Harris, "Formation of a synovial-like membrane at the bone-cement interface," Arth. Rheum., 29, 836-842 (1986). 3. S. R. Goldring, A. L. Schiller, M. Roelke, C. M. Rourke, D. A. ONeil, and W. H. Harris, "The synovial-like membrane at the bone-cement interface in loose total hip replacements and its proposed role in bone lysis," 1. Bone Jt. Surg., 65A, 575-584 (1983). 4. J. M. Mirra, R. A. Marder, and H. C . Amstutz, "The pathology of failed total joint arthroplasty," Clin. Orthopaed., 170, 175-183 (1982). 5. A. Pizzofenato, "Evaluation of the tissue response to the wear products of the hip joint endo-arthroprosthesis," Biomater. Med. ]. Artificial Organs, 7(2), 257-262 (1979). 6. H. C. Willert and M. Semlitsch, "Reactions of the articular capsule to wear products of artificial joint prosthesis," 1. Biomed. Muter. Res., 11, 157-164 (1977). 7. J.S. Schatzker, S.B. Goodman, G.S. Sumner-Smith, V.L. Fornasier, B. Goften, and R. S. Bell, "Wagner resurfacing of the canine hip," Arch. Orthopued. Traum. Surg., 106, 94-101 (1987). 8. S. B. Goodman, V. L. Fornasier, and 1. Kei. "The effects of bulk versus particulate polymethylmethacrylate'on bone," Clin. Orthopued., 232, 255-262 (19881. 9. S. B. Goodman, V. L. Fornasier, and J. Kei, "The effects of bulk versus particulate ultra-high-molecular-weight polyethylene on bone," 1. Arthroplasty, Suppl., S41-S46 (1988). 10. D. W. Howie, B. Vernon-Roberts, R. Oakeshott, and B. Manthey, "A rat model of resorption of bone at the cement-bone interface in the \
,
524
GOODMAN ET AL.
11. 12. 13.
14. 15.
16.
presence of polyethylene wear particles," J. Bone J f . Surg., 70-A, 257263 (1988). Y. Tabata and Y . Ikada, "Effect of size and surface charge of polymer microspheres on their phagocytosis by macrophage," Biomaterials, 9, 356-362 (1988). J. M. Besterman and R. 8. Low, "Endocytosis a review of mechanisms and plasma membrane dynamics," J. Biochem., 210, 1-13 (1983). H. Kawaguchi, N. Koiwai, Y. Ohtsuka, M. Miyamoto, and S. Sasakawa, "Phagocytosis of latex particles by leucotytes. Dependence of phagocytosis of the size and surface potential of particles," Biomaterials, 7, 6166 (1986). C. J. Van Oss, "Phagocytosis: An overview," Meth. Enzymol., 132, 3-15 (1986). H. Nagura, J. Asai, and K. Kojima, "Studies on the mechanisms of phagocytosis. I. Effect of electric surface charge on phagocytic activity of macrophage for fixed red cells," Cell Structure Function, 2, 21-28 (1977). N. J. Jasty, W. E. Floyd, A. L. Schiller, S. R. Goldring, and W. H. Harris, "Localized osteolysis in stable, non-septic total hip replacement," 1. Bone J f . Surg., 68-A, 912-919 (1986).
Received February 2, 1989 Accepted September 13, 1989
S . B. Goodmant Stanford University Medical Center, Division of Orthopedic Surgery, Stanford, California 94305-5326
V.L. Fornasier Department of Pathology, Princess Margaret Hospital, University of Toronto, Toronto, Canada J. Lee and J. Kei Stanford University Medical Center, Division of Orthopaedic Surgery, Stanford, California 94305-5326 This study examines the histological effects of different sizes of polyethylene particles implanted into the rabbit tibia. Seventeen mature New Zealand white female rabbits were allocated into three groups. Group 1 (5 rabbits) received polyethylene particles averaging approximately 16 pm in diameter, implanted into the right proximal tibia through a drill hole. Group 2 (5 animals) received particles averaging 26 pm, and Group 3 (7 rabbits) received particles averaging 67 pm. The left tibia was drilled but not implanted. Animals were sacrificed after 16 weeks. Histological analysis disclosed decreased hematopoietic activity within the left tibia1 drill hole. In all groups,
the right tibia demonstrated positively birefringent polyethylene particles surrounded by, and within (smaller particles), histiocytes and giant cells in a fibrous tissue stroma. Statistical analysis disclosed more fibrocytes and less marrow cells at the interface of Group 3 (largest particles) compared to Group 1 and 2. Larger polyethylene particles, being less readily phagocytosed, appear to produce more fibrous encapsulation, compared to particles of a smaller size. The histological reaction stimulated by the different sizes of polyethylene particles resembled the membrane surrounding loose joint arthroplasties in humans.
INTRODUCTION
The majority of prosthetic joints in current use employ ultra-high-molecularweight polyethylene (UHMWP) as one of the bearing surfaces. However, UHMWP wear debris has also been implicated as a contributing factor to The tissues harvested the development of aseptic component *Presented at the 15th Annual Meeting Of The Society For Biomaterials in Orlando, Florida May 1989. tPlease send all correspondence to: Stuart B. Goodman, M.D., Stanford University Medical Center, Division of Orthopaedic Surgery #R171, 300 Pasteur Drive, Stanford, California 94305-5326. Journal of Biomedical Materials Research, Vol. 24, 517-524 (1990) 0 1990 John Wiley & Sons, Inc. CCC 0021-9304/90/040517-08$04.00
GOODMAN ET AL.
518
at the interface of loose cemented prostheses in humans have demonstrated polyethylene particles up to 1000-2000 pm in Whereas larger particles are embedded within a fibrohistiocytic stroma, smaller particles are found within histiocytes and multinucleated giant cells. Studies in animals have confirmed that polyethylene particles can evoke a histological response that is similar to the "pseudomembrane" surrounding loose cemented joint arthroplasties in humans.'-'' Goodman et aL9 utilized polyethylene particles up to 1000 pm in diameter, whereas Howie et al." used particles of 20-200 pm, in their animal studies. The histological response of living tissue to particulate debris is governed by many factors including the specific biomaterial chosen, the size, shape and topography of the particles, the chemical structure of the surface and the surface charge.l1-I5This study examines the histological effects of implantation of three different "intermediate" sizes of UHMWP particles in the rabbit tibia. The specific sizes chosen reflect an intermediate particle size, seen in prior human and animal studies by the authors.'^^ MATERIALS AND METHODS
Seventeen mature New Zealand white female rabbits weighing 3-4 kg were used in the study. Group 1 (5 rabbits) received UHMWP particles averaging 15.68 pm (standard deviation = 14.04 pm) in diameter (80% < 19 pm). Group 2 (5 rabbits) received UHMWP particles averaging 26.14 pm (standard deviation = 37.21 pm) in diameter (58% < 19 pm). Group 3 (7 rabbits) received the largest particles of UHMWP averaging 67.29 pm (standard deviation = 148.59 pm) in diameter (0% < 19 pm). All animals received approximately the same volume of biomaterial. The particles were generated by applying a power burr to a commercially available acetabular cup made of ultra-high-molecular-weight polyethylene (Howmedica, Rutherford, NJ). The burr was cooled after 20 s of use. The debris was then sequentially sifted to yield particles of the appropriate size, using metal sieves (VWR Scientific Inc., San Francisco, CA). The size and shape of at least 10 particles in each group were analyzed using a scanning electron microscope interfaced with a computer with software for size and image analysis. The analysis disclosed that the surface roughness was similar for all three particle groups. The particles were packaged, gas sterilized with ethylene oxide, and allowed to aerate for at least 24 h. The UHMWP particles were placed through a drill hole in the proximal right tibia of experimental animals under general anesthe~ia.~'~ The left tibia functioned as a drilled, but nonimplanted control. Animals were sacrificed at 16 weeks postoperatively by barbiturate overdose (Beuthanasia-D Shering, Kenilworth, NJ), 2 cc intravenously. The upper tibiae were harvested bilaterally and fixed in buffered formaldehyde. Following fine detailed radiographs, each tibia was cut trans-
HISTOLOGICAL EFFECTS OF PE IMPLANTATION
519
versely through the drill hole site with a saw. The tibia immediately proximal to the drill hole site was decalcified in formic acid, dehydrated in progressing concentrations of alcohol, and then double embedded in celloidin-paraffin. Five- to ten-micrometer serial transverse sections were cut from the proximal tibia1 specimen, and stained with the "WHO" stain (hematoxylin, saffron, floxine, alcian green). These were used for general assessment of morphological features of the tissue. The tibia immediately distal to the drill hole site was processed by undecalcified methods by first dehydrating in progressing concentrations of alcohol and embedding in glycol methacrylate. Sections were cut on an Isomet low-speed saw, mounted on transparent plastic supports, and the free surface was ground to a hard polish with fine-grit paper. The surface was then acid etched and stained with toluidine blue. This permitted visualization of entire cross sections of the tibia with the implanted particles. Following a qualitative assessment, quantitation of specific parameters was undertaken. One hundred cells were classified along the immediate interface of the polyethylene particles and the surrounding tissue, using an eyepiece micrometer and X450 magnification (counting intervals were 50 pm). One section was counted per limb per animal. The number and the type of cells at the interface were tabulated and the resulting values were compared within each group using a paired t-test between the right and left sides. An analysis of variance was used to compare the cell counts between the three different polyethylene groups. A more stringent 0.01 level of significance was used to decrease spurious significance in counting data, as recommended by Dr.Byron Brown, biostatistician at Stanford University. RESULTS
The control left tibiae showed fatty areas with markedly decreased hematopoiesis at the operative site. Active hematopoiesis was present in the adjacent bone marrow. There was no evidence of necrosis or inflammation. The right tibiae with the positively birefringement polyethylene particles demonstrated groups of histiocytes and multinucleated foreign body giant cells surrounding the particles in a fibrous background stroma (Figs. 1 and 2). The histiocytes could be defined as small, compact polygonal or round cells with or without small birefringent particles a few micrometers in diameter. Giant cells were particularly concentrated at sharp edges of polyethylene. A peripheral zone of fibrous pseudoencapsulation surrounded the implant area, regardless of the particle size or cellular activity within. Reactive woven bone was found interspersed within this tissue reaction. Decreased hematopoietic activity surrounded the immediate implantation site, and merged with viable marrow and normal appearing endosteal bone. Statistical analysis disclosed that histiocytes, giant cells, and fibrocytes were more prevalent and marrow cells less prevalent at the immediate polyethylene interface on the right experimental side compared to the left control side in all groups (Table I).
GOODMAN ET AL.
520
Figure 1. Large polyethylene particles are clustered to the left of the photograph, surrounded primarily by fibrous tissue. Histiocytes are also seen. A smaller polyethylene fragment is present within a multinucleated giant cell in the top right of the photograph, surrounded by fat and hematopoietic tissue. Undecalcified specimen, toluidine blue stain, partial polarization, original magnification ~ 3 0 0 .
An analysis of variance demonstrated that fibrocytes were more commonly found and marrow cells less commonly found at the interface in Group 3 (the largest particle size) compared to Groups 1 and 2. There were no statistically significant differences in the numbers of giant cells, histiocytes, lymphocytes, plasma cells, and polymorphonuclear leukocytes between the different UHMWP groups. DISCUSSION
Both particulate polymethylmethacrylate and polyethylene have been postulated as playing an important role in the genesis of aseptic loosening of joint arthroplasties.'-'0,'6 Each of these biomaterials has been shown to evoke a florid foreign body histiocytic and giant cell response when implanted in particulate form in the tibia of rabbits.',' Bulk forms of these materials produced a thinner, more bland fibrous delimiting membrane, with only occasional giant cells and histiocytes. Howie et al. documented the formation of a "loosening membrane" capable of bone resorption, when polyethylene particles between 20 and 200 pm in diameter were injected into the knee joints of rabbits, in which a non-weight-bearing, stable femoral
HISTOLOGICAL EFFECTS OF PE IMPLANTATION
521
Figure 2. Small polyethylene particles are seen as stringy vacant spaces surrounded by, and within numerous multinucleated giant cells and histiocytes. A small group of plasma cells can be seen in the lower center of the photograph. Occasional fibroblasts are seen surrounding larger particles. Undecalcified specimen, toluidine blue stain, original magnification x 300.
plug of methacrylate had been inserted." These studies have suggested a relationship between particulate cement and polyethylene debris, and the foreign body membrane surrounding loose joint arthroplasties. However, to our knowledge, no studies have explored the relationship between the size of the polyethylene particles and the resulting histological reaction. In vitra studies employing mouse peritoneal macrophages have demonstrated that round microspheres having a size range of 1-2 pm are phagocytosed most effectively, although microspheres up to 5 pm in diameter were seen within these cells.ll However, polyethylene particles found in the membrane of loose joint arthroplasties are not spherical and encompass a wide range of sizes, from tenths of micrometers to 1-2 mm in diameter.l These facts prompted the use of "intermediate" sized particles ranging from about 3 pm to approximately 1000 pm in this study. The fibrohistiocytic and giant cell reaction to "intermediate" size polyethylene particles was very similar over the fourfold size range tested. This may, in part, be due to the overlap of the particle sizes in the three groups. The authors attempted to obtain three separate particle sue distributions by meticulous sifting techniques. However, smaller particles often adhered to the larger ones. In reviewing the histological slides, the authors therefore attempted to concentrate their cell counts and interpretation around "aver-
1.4 (0.9) 3.8 (3.3) 87.6 (4.6)
(0)
1.2 (1.6) 12.6 (4.2)** 41.2 (11.9)***
0
0.2 2.0 4.0 1.0
21.2 (8.7)** 23.6 (9.4)** 0.6 (1.3) (0.4) (1.4) (2.7) (0.7)
Left
1.0 (1.4) 5.2 (2.6) 85.6 (4.7)
(0) (0)
0 (0) 10.8 (2.8)** 58.6 (11.7)*
0 0
0.2 2.2 5.6 0.2
(0.4) (1.1) (3.4) (0.4)
Left
8.4 (3.3)+$ 22.2 (10.7)$
Right
Group I1
0 (0) 32.7 (Y.Y)*** (2.20)*** 7
19.3 (9.4)+r 35.7 (7.2),,, 5.3 (4.4) 0 (0)
Right
Group 111
(1.7) (3.4) (0.5)
(0)
5.7 (5.3) 0.1 (0.4) 86.3 (5.3)
0 2.7 4.7 0.4
Left
Note: Mean (standard deviation); Group 1, polyethylene particles averaging 15.68 pm; Group 2, polyethylene particles averaging 37.21 pm; Group 3, polyethylene particles averaging 67.29 pm. Paired t-test. *p < 0.01, **p < 0.005, ***p < 0.0005.
Giant cells Histiocytes Lymphocytes Plasma cells Polymorphonuclear leukocytes Fibrocytes Other marrow cells
Right
Group I
TABLE I Mean Cell Counts At The Implant Interface
HISTOLOGICAL EFFECTS OF PE IMPLANTATION
523
age sized" particles within each group. However, the presence of smaller particles are a confounding variable in all the groups. In previous experiments by the authors, bulk polyethylene plugs 3.5 mm in diameter were shown to stimulate a fibrous encapsulation mernb~ane.~ However, even in the bulk specimens, histiocytes and giant cells were concentrated at areas of surface irregularities. It would appear that the size, shape, and topography of the particles are important determinants of the histological results to UHMWP particles in this study. Further experiments employing spherical UHMWP particles may separate the effects of these factors further. In summary, the fibrohistiocytic reaction stimulated by the different sizes of polyethylene particles in this study was very similar. The histological reaction closely resembled the membrane surrounding loose joint arthroplasties in humans, and underscores the importance of polyethylene debris in this process. The authors gratefully acknowledge the aid of the Department of Laboratory Animal Medicine, Stanford University; the Department of Pathology, Princess Margaret Hospital, Toronto; Dr. Byron Brown, biostatistician at Stanford University; Dr. James A. Davidson, Materials Research Director, Richards Medical Company Memphis, Tennessee; and Lela Terrazas. This work was supported in part by a grant from the Orthopaedic Research Education Foundation and Zimmer/Bristol-Myers.
References 1. R. S. Bell, J. Schatzker, V. L. Fornasier, and S. B. Goodman, "Study of implant failure in the Wagner resurfacing arthroplasty," 1. Bone J t . Surg., 67A, 1165-1174 (1985). 2. S. R. Goldring, M. Jasty, M. S. Roelke, C. M. Rourke, F. R. Bringhurst, and W. H. Harris, "Formation of a synovial-like membrane at the bone-cement interface," Arth. Rheum., 29, 836-842 (1986). 3. S. R. Goldring, A. L. Schiller, M. Roelke, C. M. Rourke, D. A. ONeil, and W. H. Harris, "The synovial-like membrane at the bone-cement interface in loose total hip replacements and its proposed role in bone lysis," 1. Bone Jt. Surg., 65A, 575-584 (1983). 4. J. M. Mirra, R. A. Marder, and H. C . Amstutz, "The pathology of failed total joint arthroplasty," Clin. Orthopaed., 170, 175-183 (1982). 5. A. Pizzofenato, "Evaluation of the tissue response to the wear products of the hip joint endo-arthroprosthesis," Biomater. Med. ]. Artificial Organs, 7(2), 257-262 (1979). 6. H. C. Willert and M. Semlitsch, "Reactions of the articular capsule to wear products of artificial joint prosthesis," 1. Biomed. Muter. Res., 11, 157-164 (1977). 7. J.S. Schatzker, S.B. Goodman, G.S. Sumner-Smith, V.L. Fornasier, B. Goften, and R. S. Bell, "Wagner resurfacing of the canine hip," Arch. Orthopued. Traum. Surg., 106, 94-101 (1987). 8. S. B. Goodman, V. L. Fornasier, and 1. Kei. "The effects of bulk versus particulate polymethylmethacrylate'on bone," Clin. Orthopued., 232, 255-262 (19881. 9. S. B. Goodman, V. L. Fornasier, and J. Kei, "The effects of bulk versus particulate ultra-high-molecular-weight polyethylene on bone," 1. Arthroplasty, Suppl., S41-S46 (1988). 10. D. W. Howie, B. Vernon-Roberts, R. Oakeshott, and B. Manthey, "A rat model of resorption of bone at the cement-bone interface in the \
,
524
GOODMAN ET AL.
11. 12. 13.
14. 15.
16.
presence of polyethylene wear particles," J. Bone J f . Surg., 70-A, 257263 (1988). Y. Tabata and Y . Ikada, "Effect of size and surface charge of polymer microspheres on their phagocytosis by macrophage," Biomaterials, 9, 356-362 (1988). J. M. Besterman and R. 8. Low, "Endocytosis a review of mechanisms and plasma membrane dynamics," J. Biochem., 210, 1-13 (1983). H. Kawaguchi, N. Koiwai, Y. Ohtsuka, M. Miyamoto, and S. Sasakawa, "Phagocytosis of latex particles by leucotytes. Dependence of phagocytosis of the size and surface potential of particles," Biomaterials, 7, 6166 (1986). C. J. Van Oss, "Phagocytosis: An overview," Meth. Enzymol., 132, 3-15 (1986). H. Nagura, J. Asai, and K. Kojima, "Studies on the mechanisms of phagocytosis. I. Effect of electric surface charge on phagocytic activity of macrophage for fixed red cells," Cell Structure Function, 2, 21-28 (1977). N. J. Jasty, W. E. Floyd, A. L. Schiller, S. R. Goldring, and W. H. Harris, "Localized osteolysis in stable, non-septic total hip replacement," 1. Bone J f . Surg., 68-A, 912-919 (1986).
Received February 2, 1989 Accepted September 13, 1989
