- Email: [email protected]
Antigenicity of cortical bone allografts in dogs and effect of ethylene oxide-sterilization
Veterinary Immunology and Immunopathology 69 (1999) 47±59
Antigenicity of cortical bone allografts in dogs and effect of ethylene oxide-sterilization M. Tshamalaa, E. Coxb,*, H. De Cockc, B.M. Goddeerisb, D. Mattheeuwsa a
Department of Medicine and Clinical Biology of Small Animals, Faculty of Veterinary Medicine, RUG, Salisburylaan 133, 9820 Merelbeke, Belgium b Laboratory of Veterinary Immunology, Faculty of Veterinary Medicine, RUG, Salisburylaan 133, 9820 Merelbeke, Belgium c Laboratory of Veterinary Pathology, Faculty of Veterinary Medicine, RUG, Salisburylaan 133, 9820 Merelbeke, Belgium Received 9 July 1998; received in revised form 18 March 1999; accepted 6 April 1999
Abstract The aims of the present study were to determine the antigenicity of cortical bone allografts and the effect of ethylene oxide-sterilization (EO-sterilization). Cortical bone allografts from one donor dog were implanted in a muscle pouch in four groups of four dogs each. The grafts were either fresh, EOsterilized, demineralized or demineralized and EO-sterilized. The immune response against the grafts was determined by measuring the antibody response against surface antigens of donor cells and by the mixed lymphocyte reaction. Dogs receiving EO-sterilized grafts or bone matrix did not demonstrate an immune response. Only two of the four dogs with fresh cortical bone grafts showed a very weak immune response. This suggests a priming of the host by the fresh bone grafts. However, implanting skin grafts from the donor dog subdermally, in one dog of each of the groups, four months after implanting the bone grafts did not induce a secondary immune response. Macroscopic and histologic examination of the bone grafts five months after their implantation consistently revealed graft resorption (activity of osteoclasts) and vascularization of the fresh bone grafts, but not of EO-sterilized fresh grafts. For most EO-sterilized grafts, a strong inflammatory reaction was present in the tissues surrounding the graft and this was not apparent around the non-sterilized grafts. The absence of resorption and the presence of the inflammation seemed to be unwanted effects of the EO-sterilization. The EO-sterilisation did not affect osteoinduction since osteocytes were observed in the EO-sterilized demineralized grafts. Results indicate that cortical bone allografts used in the present study are very weak antigens and that the EOsterilization procedure used has no effect on osteoinduction, but decreases bone resorption. # 1999 Elsevier Science B.V. All rights reserved. Keywords: Cortical bone allograft; Ethylene oxide-sterilization; Antibody response; Mixed lymphocyte reaction * Corresponding author. Tel.: +32-9-2647396; fax: +32-9-2647496; e-mail: [email protected] 0165-2427/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 9 9 ) 0 0 0 4 2 - 2
48
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
1. Introduction Cortical bone transplantation is often performed to fill cortical bone defects following bone fractures or bone tumours (Friedlaender, 1991). It is not necessary that these cortical grafts contain viable cells, since they are mainly used for mechanical support and as a scaffolding for new bone formation. Therefore, dead cortical bone can be used as graft (Kerwin et al., 1991). Several preservation methods are used, such as freezing lyophilization or drying at room temperature. Contamination of the preserved grafts due to infectious diseases or as a result of manipulation is a risk that can be overcome by sterilizing the graft. The most commonly used sterilization procedures in man and animals are high-dose irradiation or ethylene oxide-sterilization (EO-sterilization) (Friedlaender, 1987). However, sterilization and preservation can change the properties of the grafts. High-dose irradiation can decrease the biomechanical properties (Spengler et al., 1978; Sugimoto et al., 1991), the antigenicity and/or the osteoinductive capacity of the grafts (Burwell, 1976), as has been described for human transplants. The current EOsterilization procedures at temperatures of 43.3±648C and pressures of 7.584 bar decrease the biomechanical properties of dog grafts (Roe et al., 1988; Wagner et al., 1994). According to Aspenberg et al. (1990), EO destroys the dose-dependent bone-induction properties, as is seen for human transplants. There are no data on the effect of EO-sterilization on the antigenicity of the bone grafts, either in man or in animals. Optimally sterilized and preserved cortical bone allografts should retain their biomechanical and osteoinductive properties and have a reduced antigenicity. Tshamala et al. (1994) described changes in the EO-sterilization procedure that resulted in preservation of the biomechanical properties of dog bone allografts. The aim of the present study was to determine the antigenicity of dog cortical bone allografts and to evaluate the effect of the EO-sterilization on dog grafts. 2. Materials and methods 2.1. Animals Nineteen adults dogs were selected without regard to sex and breed. One dog was used as a donor of cortical bone grafts and skin grafts. Furthermore, a testis cell line was prepared from this animals as a substrate for the detection of anti-DLA antibodies. One dog was sensitised with donor antigen by subdermal skin allografting (skin allografted control). A third dog was used as a blank control. The remaining 16 dogs were used as recipients They were subdivided into four groups of four dogs each. All dogs, the donor dog included, were housed individually during the experiment. The procedures and animal management were undertaken in accordance with the requirements of the animal care and ethics committee. 2.2. Bone allografts A bone segment of ca. 8 cm was surgically dissected from the distal part of the right ulna of the donor. This operation did not influence the mobility of the donor dog. The donor dog could use his operated leg without any restrictions.
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
49
The donor bone was freed from all soft tissues and was cut into 16 pieces of ca. 0.5 1.5 cm each. Cancellous bone and bone marrow were removed using a bone curette. After rinsing the bone pieces with sterile saline solution, four fresh allografts were implanted in four dogs (group FB: dogs A to D). Four other bone pieces were sterilized with EO and, subsequently, implanted in four other dogs (group FB-EO: dogs E to H). EO-sterilization occurred by exposing the grafts to 12% ethylene oxide for 6 h at 308C and 1.4 bar (Tshamala et al., 1994). The eight remaining bone pieces were completely demineralized by incubation in 5% Na-formate buffered formic acid solution (pH 2) for eight days (bone matrices). The demineralizing solution was changed twice a day and was regularly shaken to promote the demineralization. Four of these bone matrices were implanted in four dogs (group DM: dogs I to L), the other four were EO-sterilized before implantation in four other dogs (group DM-EO: dogs M to P). All implantations were performed in a muscle pouch of the lumbar muscles. A muscle pouch was chosen to allow easy harvest of the grafts for macroscopic and histologic examination. 2.3. Skin grafts Skin was chosen as transplant, since it is easy to sample, has a high antigenicity and since skin grafting represents a frequently used experimental model of transplantation (Auchincloss and Sachs, 1989). The donor skin graft was collected from the lateral surface of the thorax of the donor dog. The skin was shaved, cleaned with chlorhexidindiglucon 4% (Hibiscrub1, I.C.I-Pharma, Destelbergen, Belgium) and subsequently disinfected with alcohol 70% for 5 min, followed by povidone-iodide (Braunol1, Braun) for 5 min. Then the skin was thoroughly rinsed with sterile saline solution to remove the disinfectant, thereafter one graft of ca. 2 cm2 was sampled. This graft was implanted subdermally in a dog (skin allografted positive control dog). Four months after bone grafting, one dog of each group received a similar skin graft from the donor dog to evaluate if they were primed by the previously implanted bone graft. 2.4. Blood sampling Blood samples for isolation of blood monomorphonuclear cells were diluted (50% v/v) in phosphate buffered saline (PBS) supplemented with 1% (v/v) penicillin/streptomycin solution (10 000 U penicillin (GibcoBRL) and 10 mg streptomycin/ml (GibcoBRL)) and 200 IU heparin/ml. Blood without anticoagulants was taken to determine the serum antibody response to cell surface antigens of the donor dog. Blood samples were taken from all dogs at 0, 6, 10 and 16 weeks after the first implantation (WPPI). In the four dogs with a second implant, samples were also taken at 4, 8, 10, and 14 weeks after this second implantation (WPSI). 2.5. Isolation of peripheral blood monomorphonuclear cells (MNC) A preexperiment was performed to assess the optimum isolation procedure for the lymphocytes. Incubation of blood with carbonyl iron and arabic gum, followed by density
50
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
gradient centrifugation and removal of plasma protein adherent cells, resulted in a 90% pure lymphocyte population. The obtained cell suspension resulted in the highest SI following concanavalin A stimulation. Carbonyl iron (average particle size 4.5±5.2 mm; Sigma Chemie, Belgium) was washed four times with PBS. Then, a 10% (w/v) suspension of carbonyl iron powder in PBS and 10% (w/v) Arabic gum (Sigma Chemie) in PBS were sterilized by autoclaving for 15 min at 1218C. Equal amounts of the carbonyl iron suspension and of Arabic gum were mixed. Blood in PBS with heparin was supplemented with 10% (v/v) of the mixture at room temperature. The carbonyl iron particles are either taken up by or adhere to phagocytes (Kwok-Choy, 1984). This results in an increased density of these cells. The blood mixture was gently shaken for 1 h at 378C, following which density gradient centrifugation was performed. The density gradient centrifugation was performed using a medium with an osmolality of 256 mOsm and a density of 1.078 g/cm3 (7.1 g Ficoll 4001 (Pharmacia Biotech, Belgium) and 9 g sodium diatrizoate (Sigma Chemie) in 100 ml distilled water). Centrifugation occurred for 30 min at 900 g and 188C, contaminating erythrocytes were lysed with Tris-buffered NH4Cl and the cells were sedimented by centrifugation at 200 g for 10 min. The cells were washed twice in PBS with 1% heat inactivated foetal bovine serum (FBS) and suspended at a concentration of 2 106 cells/ml in RPMI 1640 supplemented with 1% FBS and 1% penicillin/streptomycin. Polystyrene tissue culture flasks (Corning) of 75 cm2 were coated for 1 h at 388C with 25 ml (25% v/v) autologous plasma in RPMI 1640(GibcoBRL) containing 1% penicillin/ streptomycin. Subsequently, the flasks were washed twice with RPMI 1640. Fifteen millilitres of the cell suspension were poured into the plasma-coated flasks. After incubating for 1 h at 388C, the non-adherent cells were collected. These cells were suspended at concentrations of 5 106 or 2.5 106 cells/ml in leukocyte medium for use in the mixed lymphocyte reaction. The leukocyte medium consisted of RPMI 1640 supplemented with 10% FBS, 2 mM L-glutamine (GibcoBRL), 1 mM pyruvate (GibcoBRL), 1% non-essential amino acid solution (GibcoBRL), 100 IU/ml penicillin (GibcoBRL) and 100 mg/ml streptomycin (GibcoBRL) and 50 mM 2-mercaptoethanol. 2.6. Mixed lymphocyte reaction Aliquots of 100 ml of 5 106 responder cells/ml leukocyte medium were distributed in six wells of flat-bottomed microtitre plates. Each responder cell preparation was stimulated in duplicate with 100 ml containing 2.5 105 irradiated donor cells (allogenic proliferation) or 2.5 105 irradiated autologous cells (autologous control) or plain medium (medium control). Irradiation of the stimulator cell suspensions was performed with a Co-60 source. Cells were incubated for five days, following which the proliferation was measured. The stimulation index (SI) was calculated as the ratio of cpm of the allogenic mixed lymphocyte culture (MLC) combination (after subtraction of the cpm of the medium control) and the cpm of the autologous control.
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
51
2.7. Indirect immunofluorescence antibody test (IFAT) A plastic adherent continuous cell line was generated from testicle tissue of the donor dog. This cell line was subcultured in minimum essential medium (MEM) with Earle's salts, 10% FBS, 2 mM L-glutamine, 1 mM pyruvate, 100 IU/ml penicillin, 100 mg/ml streptomycin and 100 mg/ml kanamycin. The IFAT was performed with these cells, detached from the culture flask with trypsinEDTA (GibcoBRL), six days after subculturing. The cell suspension was washed thrice in PBS and resuspended in Earle's MEM at a concentration of 106 cells/ml. One hundred microlitres of this cell suspension were brought into the wells of a microtitre plate. One hundred microlitres of twofold dilutions of the serum samples were added to the different wells. The dilution was made in Earle's MEM with 1% gentamycin (GibcoBRL) and 5% FBS. Following incubation for 30 min at 48C, the cells were washed thrice with the medium. Subsequently, an antiserum directed against dog immunoglobulins (Goat anti-dog IgG whole molecule, refined 55325 Caplet Research Products, USA) and conjugated to fluorescein isothiocyanate was added in a correct dilution. Cells were reincubated for 30 min at 378C and, subsequently, washed again twice with the medium and a third time with PBS. The cell suspensions were observed for membrane fluorescence using a fluorescence microscope (Leitz DMRB, Leica, Germany). 2.8. Histological examination Five months after cortical bone grafting, grafts and surrounding soft tissues were harvested from the muscle pouch and fixed in 10% formalin for histological examination. Subsequently, grafts from groups I and II were demineralized and embedded in paraffin. Bone matrix harvested from groups III and IV were embedded in paraffin without further demineralisation. Sections of 5-mm thick were made and stained with hematoxylin-eosin. Skin grafts were harvested three months after implantation for histological examination. 3. Results 3.1. Antibody response against donor cell surface antigens The IFAT revealed pre-immunization antibody titres between 2 and 16. Furthermore, the titres of the negative control dog and of the donor dog varied throughout the experiment between 8 and 16 and between 2 and 4, respectively. The bone implantation only induced an increase in antibody titre in one dog (dog A) with a fresh bone allograft (group FB; Fig. 1). Its antibody titre increased from 4, at the moment of implantation, to 32, six weeks later, thereafter it decreased to two, 16 WPPI. For all other dogs, the antibody titres post the primary implantation differed maximally 1 log2 with the pre-immunization titres. The skin implantation induced a 64-fold increase in antibody titre in the skin-grafted control (Fig. 2) and 8-to 16-fold increases in the four dogs that previously were implanted
52
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Fig. 1. Mean antibody response (SEM) against donor cell surface antigens in four groups of dogs following implantation with differently treated cortical bone allografts from one donor. (WPI, weeks post-implantation).
with a bone graft. However, there was a difference in the kinetics of this response in that the highest titres were reached between 4 and 8 WPSI in the dogs previously implanted with non-sterilized or EO-sterilized fresh bone graft (dogs B and G), but only between 10 to 14 WPSI in both dogs with a previous non-sterilized or EO-sterilized demineralized bone allograft (dogs J and N). 3.2. Mixed lymphocyte reaction Before the implantation, four dogs, two in the DM group (dogs I and J) and two in the DM-EO group (dogs M and O), showed SI between 14 and 51. The other 14 dogs had an index of 5 or lower. The SI of the negative control dog did not significantly change during the experiment; it varied between 1 and 2 (Fig. 3). The bone graft implantation caused, in only two dogs of the fresh bone group (dog B and D), an increase in the SI from 2 and 1, respectively, to 6 and 6. For all other dogs, the SI did not change. The skin graft implantation induced an 18-fold increase in the SI of the skin allografted control, a 15-fold increase for the dog previously grafted with fresh bone (dog B) and a tenfold increase for the dog previously grafted with bone matrix (dog J). The increase in
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
53
Fig. 2. Antibody response following skin graft implantation of four dogs, implanted 16 weeks earlier with differently treated cortical bone allografts from one donor, in comparison with the response in the no graft and the skin graft controls. (WPI, weeks post-implantation).
these three dogs occurred seven weeks after the implantation. There was no increase in the SI of the two dogs (G and N), previously receiving an EO-sterilized graft. 4. Macroscopic appearance and histology 4.1. Bone grafts In each group, the bone graft of the dogs was not found five months after implantation (dog B, F, K and O). Furthermore, one dog of the DM group had a severe trauma 17 WPPI and was euthanatized (dog L) without harvesting its graft. All other grafts were macroscopically and histologically examined. The macroscopic examination revealed a severe-to-moderate resorption of the bone grafts only in the FB group. In the FB-EO group as well as in the DM-EO group, the resorption was moderate to weak in only one of the three dogs (dogs G and P, Table 1) and in the DM group resorption was even absent in the both dogs. In all four groups, most or all grafts were adhering to the surrounding tissues. The histological examination of the grafts consistently revealed vascularization and presence of osteocytes, except for two of the FB-EO dogs (dogs E and H, Table 1). In the
54
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Fig. 3. Mixed lymphocyte reaction following skin graft implantation of four dogs, implanted 16 weeks earlier with differently treated cortical bone allografts from one donor, in comparison with the response in the no graft and the skin graft controls. Bars represent the stimulation index (SI) at different weeks post-implantation (WPI).
tissues surrounding the EO-sterilized grafts, a strong infiltration with mainly histiocytes, and to a lesser extent lymphocytes and plasma cells, was seen (Table 1), whereas there was no or only a weak infiltration around the non-sterilized grafts. However, some polynucleated and mononucleated large macrophage-like cells were seen around two of the three non-sterilized fresh bone grafts. 4.2. Skin grafts The skin-grafted dogs of the FB (dogs B) and FB-EO group (dogs G) displayed erythema and swelling of the skin above the transplant during the second week following the skin transplantation. Such reaction was not noticed for the dog of the DM (dog J) and of the DM-EO group (dog N) (Table 2). Histological examination of these grafts revealed few differences in vascularization which varied from weak to absent. However, a clear difference was observed in the tissues surrounding the graft. In the FB dog, the graft was strongly encapsulated and the surrounding tissues were very weakly infiltrated with lymphocytes and macrophages. Whereas the capsule formation decreased in the order FB-EO, DM, DM-EO dog, the infiltration with immune cells increased and in both latter dogs, even some giant cells were observed near the grafts (Table 2).
Group
FB
FB-EO
Graft
fresh cortical bone
EO a-fresh cortical bone
Dog No.
Antibody titre WP-IMPL e
SI WP-IMPL
Macroscopic appearance
Histology d
0
0
resorption b soft tissue adherence c
vascula- osteocytes rization
leukocyte infiltration
6
6
A B C D
4 8 4 8
32 16 8 8
1 2 1 1
1 6 1 6
not found
ÿ
ÿ
E F G H
4 4 8 16
2 8 16 16
3 1 1 1
1 2 2 1
ÿ not found ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
DM
bone matrix
I J K L
8 4 8 8
8 4 8 8
51 28 2 3
9 11 2 3
ÿ ÿ not found euth f
ÿ
DM-EO
EO a-bone matrix
M N O P
8 2 4 8
8 4 8 8
14 5 23 2
17 3 15 2
ÿ ÿ not found
a
Ethylene-oxide sterilized. Strong; moderate; slight; ÿ none. c Adherent; loosely adherent; ÿ not adherent. d Strong; present; ÿ none. e Weeks post-implantation. f Euthanatized. b
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Table 1 Macroscopic and histologic appearances of bone allografts harvested 20 weeks after implantation in the lumbar muscle of dogs compared to the immune response at six weeks post-implantation
55
56
Group
FB FB-EO DM DM-EO a
Dog No.
B G J N
Antibody titre WP-IMPL a
SI WP-IMPL
0
8
0
7
8 8 4 4
64 64 8 4
0.2 0.6 0.3 0.5
15 1 9 0
Weeks post-implantation. Present; rare; ÿ absent. c Well formed; young collagen; ÿ absent. d Strong infiltration; moderate infiltration; slight infiltration. b
Local recipment skin lesions b
Histology vascularization b capsule formation c
lymphocyte infiltration d
ÿ ÿ
ÿ
ÿ ÿ
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Table 2 Macroscopic and histologic appearances of second-set skin allografts harvested three months after subdermal implantation compared to the immune response at 7±8 weeks post-implantation
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
57
5. Discussion In the present study, the immune response against the grafts was evaluated by determining the antibody response against donor cell surface antigens and by the mixed lymphocyte reaction. The induced antibodies mainly reflect differences in MHC class I antigens. The lymphocyte proliferation is mainly caused by differences in MHC class II antigens (Stevenson and Horowitz, 1992; Reinsmoen, 1993). Before implantation, animals showed antibody titres of 2±16 in the indirect immunofluorescence test. There was no correlation between the height of these preimplantation titres and the immune response following implantation. Indeed, the only dog with an antibody response had a pre-implantation titre of 4. The other dogs did not show an antibody response. Their antibody titre in subsequent serum samples differed maximally 1 log2 from the pre-implantation titre. So, the presence of these antibodies did not correlate with an immune response upon graft implantation, nor with macroscopic or histologic changes in or around the grafts. This indicates that these pre-implantation titres can be regarded as background. The nature of these background antibodies was not determined. They can be MHC-specific alloantibodies produced in response to pregnancy, blood transfusion or transplantation or it can be autoantibodies of the receiver dog (Martin and Class, 1993). The fact that several of the dogs with preimplantation antibody titres had antinuclear antibody titres of 40±160 in an antinuclear antibody immunofluorescence test using Hep-2 cells (unpublished data), seems to argue for an autoimmune nature of at least some of these antibodies. Following bone graft implantation, dogs did not show a significant increase in the mixed lymphocyte reaction even if they had SI of 14 or higher. Only two of the dogs revealed a very slight increase. Both dogs had received a fresh bone graft. These results indicate that the bone allografts in the present study are very weak antigens. Nevertheless, that some of the dogs with a fresh bone graft showed a temporary increase in antibody titre (one dog) and/or in the SI (two dogs), seems to suggest that these dogs were primed by the fresh allograft. However, the subsequent skin implantation could not confirm this since it did not result in a secondary immune response. It should be mentioned that only one of the four dogs in each group received the second implant and the dog of the fresh bone group with this implant only showed a weak increase in the SI upon contact with the bone graft and no increase in antibodies (Table 1). Priming could have been too weak to observe a secondary response. Histology can be more discriminating than immunoassays for analyzing the reaction of the host against antigenically weak implants (Burchhardt et al., 1978; Stevenson et al., 1983). Furthermore, histology of bone grafts allows to observe new bone formation. Therefore, bone grafts were sampled five months after their implantation. The examination revealed a stronger resorption of the non-sterilized fresh bone grafts than of other grafts. In addition, cells that could have been responsible for this resorption, namely large polynucleated and mononucleated macrophage-like cells, were found in two on three FB dogs. Chalmers (1959) made a similar observation. He reported that an inflammatory reaction against a bone allograft persists during four±five months, following which it disappears. This indicates that the host did react against the graft.
58
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Macroscopic and microscopic examination of the skin grafts, implanted in one dog of each group to see if a secondary immune reaction would occur three months after implantation, revealed clear differences between the dogs. A capsule was formed around the skin graft in both dogs with a previous fresh bone graft implant (sterilized or nonsterilized), and was absent in both dogs receiving a bone matrix implant (sterilized or non-sterilized). This observation suggests a more severe reaction against skin grafts in the FB and the FB-EO dog compared to the DM and the DM-EO dog, and again argues for some sort of priming of the host by the fresh bone implants. Comparison between dogs receiving the non-EO-sterilized grafts and those with the EO-sterilized grafts was performed to assess the effect of EO-sterilization on the antigenicity of cortical bone allografts. However, the very weak response, or absence of immune responses after cortical bone and bone matrix implantation, did not allow observation of any differences. A difference was observed on macroscopic and histologic examination. A strong cellular infiltration, mainly of histiocytes, was seen in the tissues surrounding all six EO-sterilized implants. This severe inflammatory reaction did not occur in the tissues surrounding the non-sterilized grafts, indicating that the EO-treatment induced this aseptic inflammation. Another difference between EO-sterilized and non-sterilized grafts, seen for the fresh bone grafts, showed resorption and vascularization of the non-sterilized grafts, while two out of three EO-sterilized fresh bone grafts did not, again suggesting a negative effect of EO. The effect must have been on the osteoclast. Indeed, entry of vessels in bone grafts follows the osteoclastic activity. Furthermore, macrophage-like cells infiltrated the tissues surrounding the fresh bone grafts and not around the fresh EO-sterilized grafts. However, resorption of bone and, thus, osteoclastic activity is not needed for entry of vessels in bone matrix (Vail et al., 1994), explaining why the EO-sterilized bone matrix still showed a normal vascularization. A toxic effect of EO on macrophages could perhaps also explain the low SI in the mixed lymphocyte reaction after the skin graft implantation in both dogs previously implanted with an EO-sterilized graft. It was thought that the negative effects of the EO-sterilization were due to residual ethylene-oxide still present in the grafts, since the bone grafts were still wet just before sterilization and were implanted shortly after sterilization. It should be examined if carefully drying the grafts before sterilization could overcome this ethylene-oxide toxicity. In the present study, osteocytes were observed in only one out of three EO-sterilized fresh bone grafts, but in all the EO-sterilized bone matrices. This indicates that the used EO-sterilization procedure (12% EO at 308C and 1.4 bars for 6 h) does not affect osteoinduction. This sterilization procedure was evaluated on 40 non-sterile cortical bone segments (24 contaminated with several bacteria and 16 with Staphylococci) and rendered them all sterile (Tshamala, 1996, unpublished data). In conclusion, the results of the present study clearly show that even fresh cortical bone grafts have a low antigenicity. Furthermore, the results indicate that the EO-sterilization could be an interesting alternative for the current sterilization procedures, if the negative effect on vascularization could be overcome. Indeed, in contrast to the current EOprocedures (Aspenberg et al., 1990), the present procedure does not seem to influence the bone induction properties of the grafts.
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
59
Acknowledgements The authors thank D. Slos for the excellent technical assistance and Dr. L. Van Ham for his help during the surgery of the animals. References Aspenberg, P., Johnson, E., Thorngren, K.G., 1990. Dose-dependent reduction of bone inductive properties by ethylene oxide. J. Bone Joint Surg. 72B, 1036±1037. Auchinclosss, H., Sachs, D.H., 1989. Transplantation and graft rejection. In: Paul, W.E. (Ed.), Fundamental Immunology (second edn.). Raven Press, New York. pp. 889±922. Burchhardt, H., Jones, H., Glowczewskie, F., Rudner, C., Enneking, W.F., 1978. Freeze-dried allogeneic segmental cortical-bone grafts in dogs. J. Bone Joint Surg. 60A, 1082±1090. Burwell, R.G., 1976. The fate of freeze-dried bone allografts. Transpl. Proc., VIII, suppl. pp. 95±111. Chalmers, J., 1959. Transplantation immunity in bone homografting. J. Bone Joint Surg. 41B, 160±179. Friedlaender, G.E., 1998. Current concepts review. Bone grafts. The basic science rationale for clinical applications. J. Bone Joint Surg. 69A, 786±790. Friedlander, G.E., 1991. Bone allografts: the biological consequences of immunological events. J. Bone and Joint Surg. 73A, 1119±1122. Kerwin, S.C., Lewis, D.D., Elkins, A.D., 1991. Bone grafting and banking. Comp. Cont. Ed. Pract. Vet. 13, 1558±1566. Kwok-Choy, L., 1984. Carbonyl iron powder. In: Mishell, B.B., ShuÈgi, S.M. (Eds.), Selected Methods in Cellular Immunology. W.H. Freeman and Co, New York. pp. 179±181. Martin, S., Class, Fr., 1993. Antibodies and crossmatching for transplantation. In: Dyer, P.D., Middleton, D. (Eds.), Histocompatibility Testing: A Practical Approach. Oxford University Press, Oxford. pp. 81±106. Reinsmoen, N.L., 1993. Cellular methods, In: Dyer, P.D., Middleton, D. (Eds.), Histocompatibility Testing: A Practical Approach. Oxford University press, Oxford. pp. 143±157. Roe, S.C., Pijanowski, G.J., Johnson, A.L., 1988. Biomechanical properties of canine bone allografts: effects of preparation and storage. Am. J. Vet. Res. 49, 873±877. Spengler, D.M., Carter, D.R., Baylink, D.J., Lee, R., 1978. Mechanical properties of whole bone: effects of irradiation. Orthop. Res. Soc. 2, 207. Stevenson, S., Horowitz, M., 1992. Current concepts reviews. The response to bone allografts. J. Bone and Joint Surg. 74A, 939±950. Stevenson, S., Hohn, R.B., Templeton, J.W., 1983. Effects of tissue antigen matching on the healing of fresh cancellous bone allografts in dogs. Am. J. Vet Res. 44, 201±206. Sugimoto, M., Takahashi, S., Toguchida, J., Kotoura, Y., Shibamoto, Y., Yamamuro, T., 1991. Changes in bone after high-dose irradiation. J. Bone Joint Surg. 73B, 492±497. Tshamala, M., van Bree, H., Mattheeuws, D., 1994. Biomechanical properties of EO-sterilized and cryopreserved cortical bone allografts. VCOT 7, 25±30. Vail, T.B., Trotter, G.W., Powers, B.E., 1994. Equine demineralized bone matrix: relationship between particle size and osteoinduction. Vet. Surg. 23, 386±395. Wagner, S.D., Manley, P.A., Radasch, R.M., Hoefle, W.D., Haynes, J.S., 1994. Failure of ethylene oxidesterilized cortical allografts in two dogs. J. Am. Anim. Hosp. Assoc. 30, 181±189.
Antigenicity of cortical bone allografts in dogs and effect of ethylene oxide-sterilization M. Tshamalaa, E. Coxb,*, H. De Cockc, B.M. Goddeerisb, D. Mattheeuwsa a
Department of Medicine and Clinical Biology of Small Animals, Faculty of Veterinary Medicine, RUG, Salisburylaan 133, 9820 Merelbeke, Belgium b Laboratory of Veterinary Immunology, Faculty of Veterinary Medicine, RUG, Salisburylaan 133, 9820 Merelbeke, Belgium c Laboratory of Veterinary Pathology, Faculty of Veterinary Medicine, RUG, Salisburylaan 133, 9820 Merelbeke, Belgium Received 9 July 1998; received in revised form 18 March 1999; accepted 6 April 1999
Abstract The aims of the present study were to determine the antigenicity of cortical bone allografts and the effect of ethylene oxide-sterilization (EO-sterilization). Cortical bone allografts from one donor dog were implanted in a muscle pouch in four groups of four dogs each. The grafts were either fresh, EOsterilized, demineralized or demineralized and EO-sterilized. The immune response against the grafts was determined by measuring the antibody response against surface antigens of donor cells and by the mixed lymphocyte reaction. Dogs receiving EO-sterilized grafts or bone matrix did not demonstrate an immune response. Only two of the four dogs with fresh cortical bone grafts showed a very weak immune response. This suggests a priming of the host by the fresh bone grafts. However, implanting skin grafts from the donor dog subdermally, in one dog of each of the groups, four months after implanting the bone grafts did not induce a secondary immune response. Macroscopic and histologic examination of the bone grafts five months after their implantation consistently revealed graft resorption (activity of osteoclasts) and vascularization of the fresh bone grafts, but not of EO-sterilized fresh grafts. For most EO-sterilized grafts, a strong inflammatory reaction was present in the tissues surrounding the graft and this was not apparent around the non-sterilized grafts. The absence of resorption and the presence of the inflammation seemed to be unwanted effects of the EO-sterilization. The EO-sterilisation did not affect osteoinduction since osteocytes were observed in the EO-sterilized demineralized grafts. Results indicate that cortical bone allografts used in the present study are very weak antigens and that the EOsterilization procedure used has no effect on osteoinduction, but decreases bone resorption. # 1999 Elsevier Science B.V. All rights reserved. Keywords: Cortical bone allograft; Ethylene oxide-sterilization; Antibody response; Mixed lymphocyte reaction * Corresponding author. Tel.: +32-9-2647396; fax: +32-9-2647496; e-mail: [email protected] 0165-2427/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 9 9 ) 0 0 0 4 2 - 2
48
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
1. Introduction Cortical bone transplantation is often performed to fill cortical bone defects following bone fractures or bone tumours (Friedlaender, 1991). It is not necessary that these cortical grafts contain viable cells, since they are mainly used for mechanical support and as a scaffolding for new bone formation. Therefore, dead cortical bone can be used as graft (Kerwin et al., 1991). Several preservation methods are used, such as freezing lyophilization or drying at room temperature. Contamination of the preserved grafts due to infectious diseases or as a result of manipulation is a risk that can be overcome by sterilizing the graft. The most commonly used sterilization procedures in man and animals are high-dose irradiation or ethylene oxide-sterilization (EO-sterilization) (Friedlaender, 1987). However, sterilization and preservation can change the properties of the grafts. High-dose irradiation can decrease the biomechanical properties (Spengler et al., 1978; Sugimoto et al., 1991), the antigenicity and/or the osteoinductive capacity of the grafts (Burwell, 1976), as has been described for human transplants. The current EOsterilization procedures at temperatures of 43.3±648C and pressures of 7.584 bar decrease the biomechanical properties of dog grafts (Roe et al., 1988; Wagner et al., 1994). According to Aspenberg et al. (1990), EO destroys the dose-dependent bone-induction properties, as is seen for human transplants. There are no data on the effect of EO-sterilization on the antigenicity of the bone grafts, either in man or in animals. Optimally sterilized and preserved cortical bone allografts should retain their biomechanical and osteoinductive properties and have a reduced antigenicity. Tshamala et al. (1994) described changes in the EO-sterilization procedure that resulted in preservation of the biomechanical properties of dog bone allografts. The aim of the present study was to determine the antigenicity of dog cortical bone allografts and to evaluate the effect of the EO-sterilization on dog grafts. 2. Materials and methods 2.1. Animals Nineteen adults dogs were selected without regard to sex and breed. One dog was used as a donor of cortical bone grafts and skin grafts. Furthermore, a testis cell line was prepared from this animals as a substrate for the detection of anti-DLA antibodies. One dog was sensitised with donor antigen by subdermal skin allografting (skin allografted control). A third dog was used as a blank control. The remaining 16 dogs were used as recipients They were subdivided into four groups of four dogs each. All dogs, the donor dog included, were housed individually during the experiment. The procedures and animal management were undertaken in accordance with the requirements of the animal care and ethics committee. 2.2. Bone allografts A bone segment of ca. 8 cm was surgically dissected from the distal part of the right ulna of the donor. This operation did not influence the mobility of the donor dog. The donor dog could use his operated leg without any restrictions.
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
49
The donor bone was freed from all soft tissues and was cut into 16 pieces of ca. 0.5 1.5 cm each. Cancellous bone and bone marrow were removed using a bone curette. After rinsing the bone pieces with sterile saline solution, four fresh allografts were implanted in four dogs (group FB: dogs A to D). Four other bone pieces were sterilized with EO and, subsequently, implanted in four other dogs (group FB-EO: dogs E to H). EO-sterilization occurred by exposing the grafts to 12% ethylene oxide for 6 h at 308C and 1.4 bar (Tshamala et al., 1994). The eight remaining bone pieces were completely demineralized by incubation in 5% Na-formate buffered formic acid solution (pH 2) for eight days (bone matrices). The demineralizing solution was changed twice a day and was regularly shaken to promote the demineralization. Four of these bone matrices were implanted in four dogs (group DM: dogs I to L), the other four were EO-sterilized before implantation in four other dogs (group DM-EO: dogs M to P). All implantations were performed in a muscle pouch of the lumbar muscles. A muscle pouch was chosen to allow easy harvest of the grafts for macroscopic and histologic examination. 2.3. Skin grafts Skin was chosen as transplant, since it is easy to sample, has a high antigenicity and since skin grafting represents a frequently used experimental model of transplantation (Auchincloss and Sachs, 1989). The donor skin graft was collected from the lateral surface of the thorax of the donor dog. The skin was shaved, cleaned with chlorhexidindiglucon 4% (Hibiscrub1, I.C.I-Pharma, Destelbergen, Belgium) and subsequently disinfected with alcohol 70% for 5 min, followed by povidone-iodide (Braunol1, Braun) for 5 min. Then the skin was thoroughly rinsed with sterile saline solution to remove the disinfectant, thereafter one graft of ca. 2 cm2 was sampled. This graft was implanted subdermally in a dog (skin allografted positive control dog). Four months after bone grafting, one dog of each group received a similar skin graft from the donor dog to evaluate if they were primed by the previously implanted bone graft. 2.4. Blood sampling Blood samples for isolation of blood monomorphonuclear cells were diluted (50% v/v) in phosphate buffered saline (PBS) supplemented with 1% (v/v) penicillin/streptomycin solution (10 000 U penicillin (GibcoBRL) and 10 mg streptomycin/ml (GibcoBRL)) and 200 IU heparin/ml. Blood without anticoagulants was taken to determine the serum antibody response to cell surface antigens of the donor dog. Blood samples were taken from all dogs at 0, 6, 10 and 16 weeks after the first implantation (WPPI). In the four dogs with a second implant, samples were also taken at 4, 8, 10, and 14 weeks after this second implantation (WPSI). 2.5. Isolation of peripheral blood monomorphonuclear cells (MNC) A preexperiment was performed to assess the optimum isolation procedure for the lymphocytes. Incubation of blood with carbonyl iron and arabic gum, followed by density
50
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
gradient centrifugation and removal of plasma protein adherent cells, resulted in a 90% pure lymphocyte population. The obtained cell suspension resulted in the highest SI following concanavalin A stimulation. Carbonyl iron (average particle size 4.5±5.2 mm; Sigma Chemie, Belgium) was washed four times with PBS. Then, a 10% (w/v) suspension of carbonyl iron powder in PBS and 10% (w/v) Arabic gum (Sigma Chemie) in PBS were sterilized by autoclaving for 15 min at 1218C. Equal amounts of the carbonyl iron suspension and of Arabic gum were mixed. Blood in PBS with heparin was supplemented with 10% (v/v) of the mixture at room temperature. The carbonyl iron particles are either taken up by or adhere to phagocytes (Kwok-Choy, 1984). This results in an increased density of these cells. The blood mixture was gently shaken for 1 h at 378C, following which density gradient centrifugation was performed. The density gradient centrifugation was performed using a medium with an osmolality of 256 mOsm and a density of 1.078 g/cm3 (7.1 g Ficoll 4001 (Pharmacia Biotech, Belgium) and 9 g sodium diatrizoate (Sigma Chemie) in 100 ml distilled water). Centrifugation occurred for 30 min at 900 g and 188C, contaminating erythrocytes were lysed with Tris-buffered NH4Cl and the cells were sedimented by centrifugation at 200 g for 10 min. The cells were washed twice in PBS with 1% heat inactivated foetal bovine serum (FBS) and suspended at a concentration of 2 106 cells/ml in RPMI 1640 supplemented with 1% FBS and 1% penicillin/streptomycin. Polystyrene tissue culture flasks (Corning) of 75 cm2 were coated for 1 h at 388C with 25 ml (25% v/v) autologous plasma in RPMI 1640(GibcoBRL) containing 1% penicillin/ streptomycin. Subsequently, the flasks were washed twice with RPMI 1640. Fifteen millilitres of the cell suspension were poured into the plasma-coated flasks. After incubating for 1 h at 388C, the non-adherent cells were collected. These cells were suspended at concentrations of 5 106 or 2.5 106 cells/ml in leukocyte medium for use in the mixed lymphocyte reaction. The leukocyte medium consisted of RPMI 1640 supplemented with 10% FBS, 2 mM L-glutamine (GibcoBRL), 1 mM pyruvate (GibcoBRL), 1% non-essential amino acid solution (GibcoBRL), 100 IU/ml penicillin (GibcoBRL) and 100 mg/ml streptomycin (GibcoBRL) and 50 mM 2-mercaptoethanol. 2.6. Mixed lymphocyte reaction Aliquots of 100 ml of 5 106 responder cells/ml leukocyte medium were distributed in six wells of flat-bottomed microtitre plates. Each responder cell preparation was stimulated in duplicate with 100 ml containing 2.5 105 irradiated donor cells (allogenic proliferation) or 2.5 105 irradiated autologous cells (autologous control) or plain medium (medium control). Irradiation of the stimulator cell suspensions was performed with a Co-60 source. Cells were incubated for five days, following which the proliferation was measured. The stimulation index (SI) was calculated as the ratio of cpm of the allogenic mixed lymphocyte culture (MLC) combination (after subtraction of the cpm of the medium control) and the cpm of the autologous control.
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
51
2.7. Indirect immunofluorescence antibody test (IFAT) A plastic adherent continuous cell line was generated from testicle tissue of the donor dog. This cell line was subcultured in minimum essential medium (MEM) with Earle's salts, 10% FBS, 2 mM L-glutamine, 1 mM pyruvate, 100 IU/ml penicillin, 100 mg/ml streptomycin and 100 mg/ml kanamycin. The IFAT was performed with these cells, detached from the culture flask with trypsinEDTA (GibcoBRL), six days after subculturing. The cell suspension was washed thrice in PBS and resuspended in Earle's MEM at a concentration of 106 cells/ml. One hundred microlitres of this cell suspension were brought into the wells of a microtitre plate. One hundred microlitres of twofold dilutions of the serum samples were added to the different wells. The dilution was made in Earle's MEM with 1% gentamycin (GibcoBRL) and 5% FBS. Following incubation for 30 min at 48C, the cells were washed thrice with the medium. Subsequently, an antiserum directed against dog immunoglobulins (Goat anti-dog IgG whole molecule, refined 55325 Caplet Research Products, USA) and conjugated to fluorescein isothiocyanate was added in a correct dilution. Cells were reincubated for 30 min at 378C and, subsequently, washed again twice with the medium and a third time with PBS. The cell suspensions were observed for membrane fluorescence using a fluorescence microscope (Leitz DMRB, Leica, Germany). 2.8. Histological examination Five months after cortical bone grafting, grafts and surrounding soft tissues were harvested from the muscle pouch and fixed in 10% formalin for histological examination. Subsequently, grafts from groups I and II were demineralized and embedded in paraffin. Bone matrix harvested from groups III and IV were embedded in paraffin without further demineralisation. Sections of 5-mm thick were made and stained with hematoxylin-eosin. Skin grafts were harvested three months after implantation for histological examination. 3. Results 3.1. Antibody response against donor cell surface antigens The IFAT revealed pre-immunization antibody titres between 2 and 16. Furthermore, the titres of the negative control dog and of the donor dog varied throughout the experiment between 8 and 16 and between 2 and 4, respectively. The bone implantation only induced an increase in antibody titre in one dog (dog A) with a fresh bone allograft (group FB; Fig. 1). Its antibody titre increased from 4, at the moment of implantation, to 32, six weeks later, thereafter it decreased to two, 16 WPPI. For all other dogs, the antibody titres post the primary implantation differed maximally 1 log2 with the pre-immunization titres. The skin implantation induced a 64-fold increase in antibody titre in the skin-grafted control (Fig. 2) and 8-to 16-fold increases in the four dogs that previously were implanted
52
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Fig. 1. Mean antibody response (SEM) against donor cell surface antigens in four groups of dogs following implantation with differently treated cortical bone allografts from one donor. (WPI, weeks post-implantation).
with a bone graft. However, there was a difference in the kinetics of this response in that the highest titres were reached between 4 and 8 WPSI in the dogs previously implanted with non-sterilized or EO-sterilized fresh bone graft (dogs B and G), but only between 10 to 14 WPSI in both dogs with a previous non-sterilized or EO-sterilized demineralized bone allograft (dogs J and N). 3.2. Mixed lymphocyte reaction Before the implantation, four dogs, two in the DM group (dogs I and J) and two in the DM-EO group (dogs M and O), showed SI between 14 and 51. The other 14 dogs had an index of 5 or lower. The SI of the negative control dog did not significantly change during the experiment; it varied between 1 and 2 (Fig. 3). The bone graft implantation caused, in only two dogs of the fresh bone group (dog B and D), an increase in the SI from 2 and 1, respectively, to 6 and 6. For all other dogs, the SI did not change. The skin graft implantation induced an 18-fold increase in the SI of the skin allografted control, a 15-fold increase for the dog previously grafted with fresh bone (dog B) and a tenfold increase for the dog previously grafted with bone matrix (dog J). The increase in
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
53
Fig. 2. Antibody response following skin graft implantation of four dogs, implanted 16 weeks earlier with differently treated cortical bone allografts from one donor, in comparison with the response in the no graft and the skin graft controls. (WPI, weeks post-implantation).
these three dogs occurred seven weeks after the implantation. There was no increase in the SI of the two dogs (G and N), previously receiving an EO-sterilized graft. 4. Macroscopic appearance and histology 4.1. Bone grafts In each group, the bone graft of the dogs was not found five months after implantation (dog B, F, K and O). Furthermore, one dog of the DM group had a severe trauma 17 WPPI and was euthanatized (dog L) without harvesting its graft. All other grafts were macroscopically and histologically examined. The macroscopic examination revealed a severe-to-moderate resorption of the bone grafts only in the FB group. In the FB-EO group as well as in the DM-EO group, the resorption was moderate to weak in only one of the three dogs (dogs G and P, Table 1) and in the DM group resorption was even absent in the both dogs. In all four groups, most or all grafts were adhering to the surrounding tissues. The histological examination of the grafts consistently revealed vascularization and presence of osteocytes, except for two of the FB-EO dogs (dogs E and H, Table 1). In the
54
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Fig. 3. Mixed lymphocyte reaction following skin graft implantation of four dogs, implanted 16 weeks earlier with differently treated cortical bone allografts from one donor, in comparison with the response in the no graft and the skin graft controls. Bars represent the stimulation index (SI) at different weeks post-implantation (WPI).
tissues surrounding the EO-sterilized grafts, a strong infiltration with mainly histiocytes, and to a lesser extent lymphocytes and plasma cells, was seen (Table 1), whereas there was no or only a weak infiltration around the non-sterilized grafts. However, some polynucleated and mononucleated large macrophage-like cells were seen around two of the three non-sterilized fresh bone grafts. 4.2. Skin grafts The skin-grafted dogs of the FB (dogs B) and FB-EO group (dogs G) displayed erythema and swelling of the skin above the transplant during the second week following the skin transplantation. Such reaction was not noticed for the dog of the DM (dog J) and of the DM-EO group (dog N) (Table 2). Histological examination of these grafts revealed few differences in vascularization which varied from weak to absent. However, a clear difference was observed in the tissues surrounding the graft. In the FB dog, the graft was strongly encapsulated and the surrounding tissues were very weakly infiltrated with lymphocytes and macrophages. Whereas the capsule formation decreased in the order FB-EO, DM, DM-EO dog, the infiltration with immune cells increased and in both latter dogs, even some giant cells were observed near the grafts (Table 2).
Group
FB
FB-EO
Graft
fresh cortical bone
EO a-fresh cortical bone
Dog No.
Antibody titre WP-IMPL e
SI WP-IMPL
Macroscopic appearance
Histology d
0
0
resorption b soft tissue adherence c
vascula- osteocytes rization
leukocyte infiltration
6
6
A B C D
4 8 4 8
32 16 8 8
1 2 1 1
1 6 1 6
not found
ÿ
ÿ
E F G H
4 4 8 16
2 8 16 16
3 1 1 1
1 2 2 1
ÿ not found ÿ
ÿ
ÿ
ÿ
ÿ
ÿ
DM
bone matrix
I J K L
8 4 8 8
8 4 8 8
51 28 2 3
9 11 2 3
ÿ ÿ not found euth f
ÿ
DM-EO
EO a-bone matrix
M N O P
8 2 4 8
8 4 8 8
14 5 23 2
17 3 15 2
ÿ ÿ not found
a
Ethylene-oxide sterilized. Strong; moderate; slight; ÿ none. c Adherent; loosely adherent; ÿ not adherent. d Strong; present; ÿ none. e Weeks post-implantation. f Euthanatized. b
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Table 1 Macroscopic and histologic appearances of bone allografts harvested 20 weeks after implantation in the lumbar muscle of dogs compared to the immune response at six weeks post-implantation
55
56
Group
FB FB-EO DM DM-EO a
Dog No.
B G J N
Antibody titre WP-IMPL a
SI WP-IMPL
0
8
0
7
8 8 4 4
64 64 8 4
0.2 0.6 0.3 0.5
15 1 9 0
Weeks post-implantation. Present; rare; ÿ absent. c Well formed; young collagen; ÿ absent. d Strong infiltration; moderate infiltration; slight infiltration. b
Local recipment skin lesions b
Histology vascularization b capsule formation c
lymphocyte infiltration d
ÿ ÿ
ÿ
ÿ ÿ
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Table 2 Macroscopic and histologic appearances of second-set skin allografts harvested three months after subdermal implantation compared to the immune response at 7±8 weeks post-implantation
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
57
5. Discussion In the present study, the immune response against the grafts was evaluated by determining the antibody response against donor cell surface antigens and by the mixed lymphocyte reaction. The induced antibodies mainly reflect differences in MHC class I antigens. The lymphocyte proliferation is mainly caused by differences in MHC class II antigens (Stevenson and Horowitz, 1992; Reinsmoen, 1993). Before implantation, animals showed antibody titres of 2±16 in the indirect immunofluorescence test. There was no correlation between the height of these preimplantation titres and the immune response following implantation. Indeed, the only dog with an antibody response had a pre-implantation titre of 4. The other dogs did not show an antibody response. Their antibody titre in subsequent serum samples differed maximally 1 log2 from the pre-implantation titre. So, the presence of these antibodies did not correlate with an immune response upon graft implantation, nor with macroscopic or histologic changes in or around the grafts. This indicates that these pre-implantation titres can be regarded as background. The nature of these background antibodies was not determined. They can be MHC-specific alloantibodies produced in response to pregnancy, blood transfusion or transplantation or it can be autoantibodies of the receiver dog (Martin and Class, 1993). The fact that several of the dogs with preimplantation antibody titres had antinuclear antibody titres of 40±160 in an antinuclear antibody immunofluorescence test using Hep-2 cells (unpublished data), seems to argue for an autoimmune nature of at least some of these antibodies. Following bone graft implantation, dogs did not show a significant increase in the mixed lymphocyte reaction even if they had SI of 14 or higher. Only two of the dogs revealed a very slight increase. Both dogs had received a fresh bone graft. These results indicate that the bone allografts in the present study are very weak antigens. Nevertheless, that some of the dogs with a fresh bone graft showed a temporary increase in antibody titre (one dog) and/or in the SI (two dogs), seems to suggest that these dogs were primed by the fresh allograft. However, the subsequent skin implantation could not confirm this since it did not result in a secondary immune response. It should be mentioned that only one of the four dogs in each group received the second implant and the dog of the fresh bone group with this implant only showed a weak increase in the SI upon contact with the bone graft and no increase in antibodies (Table 1). Priming could have been too weak to observe a secondary response. Histology can be more discriminating than immunoassays for analyzing the reaction of the host against antigenically weak implants (Burchhardt et al., 1978; Stevenson et al., 1983). Furthermore, histology of bone grafts allows to observe new bone formation. Therefore, bone grafts were sampled five months after their implantation. The examination revealed a stronger resorption of the non-sterilized fresh bone grafts than of other grafts. In addition, cells that could have been responsible for this resorption, namely large polynucleated and mononucleated macrophage-like cells, were found in two on three FB dogs. Chalmers (1959) made a similar observation. He reported that an inflammatory reaction against a bone allograft persists during four±five months, following which it disappears. This indicates that the host did react against the graft.
58
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
Macroscopic and microscopic examination of the skin grafts, implanted in one dog of each group to see if a secondary immune reaction would occur three months after implantation, revealed clear differences between the dogs. A capsule was formed around the skin graft in both dogs with a previous fresh bone graft implant (sterilized or nonsterilized), and was absent in both dogs receiving a bone matrix implant (sterilized or non-sterilized). This observation suggests a more severe reaction against skin grafts in the FB and the FB-EO dog compared to the DM and the DM-EO dog, and again argues for some sort of priming of the host by the fresh bone implants. Comparison between dogs receiving the non-EO-sterilized grafts and those with the EO-sterilized grafts was performed to assess the effect of EO-sterilization on the antigenicity of cortical bone allografts. However, the very weak response, or absence of immune responses after cortical bone and bone matrix implantation, did not allow observation of any differences. A difference was observed on macroscopic and histologic examination. A strong cellular infiltration, mainly of histiocytes, was seen in the tissues surrounding all six EO-sterilized implants. This severe inflammatory reaction did not occur in the tissues surrounding the non-sterilized grafts, indicating that the EO-treatment induced this aseptic inflammation. Another difference between EO-sterilized and non-sterilized grafts, seen for the fresh bone grafts, showed resorption and vascularization of the non-sterilized grafts, while two out of three EO-sterilized fresh bone grafts did not, again suggesting a negative effect of EO. The effect must have been on the osteoclast. Indeed, entry of vessels in bone grafts follows the osteoclastic activity. Furthermore, macrophage-like cells infiltrated the tissues surrounding the fresh bone grafts and not around the fresh EO-sterilized grafts. However, resorption of bone and, thus, osteoclastic activity is not needed for entry of vessels in bone matrix (Vail et al., 1994), explaining why the EO-sterilized bone matrix still showed a normal vascularization. A toxic effect of EO on macrophages could perhaps also explain the low SI in the mixed lymphocyte reaction after the skin graft implantation in both dogs previously implanted with an EO-sterilized graft. It was thought that the negative effects of the EO-sterilization were due to residual ethylene-oxide still present in the grafts, since the bone grafts were still wet just before sterilization and were implanted shortly after sterilization. It should be examined if carefully drying the grafts before sterilization could overcome this ethylene-oxide toxicity. In the present study, osteocytes were observed in only one out of three EO-sterilized fresh bone grafts, but in all the EO-sterilized bone matrices. This indicates that the used EO-sterilization procedure (12% EO at 308C and 1.4 bars for 6 h) does not affect osteoinduction. This sterilization procedure was evaluated on 40 non-sterile cortical bone segments (24 contaminated with several bacteria and 16 with Staphylococci) and rendered them all sterile (Tshamala, 1996, unpublished data). In conclusion, the results of the present study clearly show that even fresh cortical bone grafts have a low antigenicity. Furthermore, the results indicate that the EO-sterilization could be an interesting alternative for the current sterilization procedures, if the negative effect on vascularization could be overcome. Indeed, in contrast to the current EOprocedures (Aspenberg et al., 1990), the present procedure does not seem to influence the bone induction properties of the grafts.
M. Tshamala et al. / Veterinary Immunology and Immunopathology 69 (1999) 47±59
59
Acknowledgements The authors thank D. Slos for the excellent technical assistance and Dr. L. Van Ham for his help during the surgery of the animals. References Aspenberg, P., Johnson, E., Thorngren, K.G., 1990. Dose-dependent reduction of bone inductive properties by ethylene oxide. J. Bone Joint Surg. 72B, 1036±1037. Auchinclosss, H., Sachs, D.H., 1989. Transplantation and graft rejection. In: Paul, W.E. (Ed.), Fundamental Immunology (second edn.). Raven Press, New York. pp. 889±922. Burchhardt, H., Jones, H., Glowczewskie, F., Rudner, C., Enneking, W.F., 1978. Freeze-dried allogeneic segmental cortical-bone grafts in dogs. J. Bone Joint Surg. 60A, 1082±1090. Burwell, R.G., 1976. The fate of freeze-dried bone allografts. Transpl. Proc., VIII, suppl. pp. 95±111. Chalmers, J., 1959. Transplantation immunity in bone homografting. J. Bone Joint Surg. 41B, 160±179. Friedlaender, G.E., 1998. Current concepts review. Bone grafts. The basic science rationale for clinical applications. J. Bone Joint Surg. 69A, 786±790. Friedlander, G.E., 1991. Bone allografts: the biological consequences of immunological events. J. Bone and Joint Surg. 73A, 1119±1122. Kerwin, S.C., Lewis, D.D., Elkins, A.D., 1991. Bone grafting and banking. Comp. Cont. Ed. Pract. Vet. 13, 1558±1566. Kwok-Choy, L., 1984. Carbonyl iron powder. In: Mishell, B.B., ShuÈgi, S.M. (Eds.), Selected Methods in Cellular Immunology. W.H. Freeman and Co, New York. pp. 179±181. Martin, S., Class, Fr., 1993. Antibodies and crossmatching for transplantation. In: Dyer, P.D., Middleton, D. (Eds.), Histocompatibility Testing: A Practical Approach. Oxford University Press, Oxford. pp. 81±106. Reinsmoen, N.L., 1993. Cellular methods, In: Dyer, P.D., Middleton, D. (Eds.), Histocompatibility Testing: A Practical Approach. Oxford University press, Oxford. pp. 143±157. Roe, S.C., Pijanowski, G.J., Johnson, A.L., 1988. Biomechanical properties of canine bone allografts: effects of preparation and storage. Am. J. Vet. Res. 49, 873±877. Spengler, D.M., Carter, D.R., Baylink, D.J., Lee, R., 1978. Mechanical properties of whole bone: effects of irradiation. Orthop. Res. Soc. 2, 207. Stevenson, S., Horowitz, M., 1992. Current concepts reviews. The response to bone allografts. J. Bone and Joint Surg. 74A, 939±950. Stevenson, S., Hohn, R.B., Templeton, J.W., 1983. Effects of tissue antigen matching on the healing of fresh cancellous bone allografts in dogs. Am. J. Vet Res. 44, 201±206. Sugimoto, M., Takahashi, S., Toguchida, J., Kotoura, Y., Shibamoto, Y., Yamamuro, T., 1991. Changes in bone after high-dose irradiation. J. Bone Joint Surg. 73B, 492±497. Tshamala, M., van Bree, H., Mattheeuws, D., 1994. Biomechanical properties of EO-sterilized and cryopreserved cortical bone allografts. VCOT 7, 25±30. Vail, T.B., Trotter, G.W., Powers, B.E., 1994. Equine demineralized bone matrix: relationship between particle size and osteoinduction. Vet. Surg. 23, 386±395. Wagner, S.D., Manley, P.A., Radasch, R.M., Hoefle, W.D., Haynes, J.S., 1994. Failure of ethylene oxidesterilized cortical allografts in two dogs. J. Am. Anim. Hosp. Assoc. 30, 181±189.