| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2000;119:477-487
© 2000 Mosby, Inc.
CARDIOTHORACIC TRANSPLANTATION |
From the Department of General and Thoracic Surgerya and the Institute for Pathology,b University of Giessen, Giessen, Germany.
Address for reprints: Helmut Grimm, MD, PhD, Department of General and Thoracic Surgery, University of Giessen, Rudolf-Buchheim-Strasse 7, D-35392 Giessen, Germany (E-mail: helmut.grimm{at}chiru.med.uni-giessen.de ).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
secretory potential, which correlated well with a diminished immunohistologic staining of tumor necrosis factor secreted by graft-infiltrating mononuclear cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
1,2 If hyperacute rejection is prevented by complement inhibition, another type of rejection, termed delayed xenograft rejection , will proceed over the following days. This process is characterized by donor endothelial cell activation, thrombosis, and infiltration of the graft with mononuclear cells. 3
Among the factors potentially involved in delayed xenograft rejection, xenoreactive antibodies are being stressed as the cause of antibody-induced endothelial cell activation 4 or antibody-dependent cellular cytotoxicity. 5 If xenoantibodies should prove to be important in the pathogenesis of delayed xenograft rejection, the therapeutic consequences must include both the reduction of preformed (naturally occurring) xenoantibodies before xenotransplantation and the suppression of elicited xenoantibody production after transplantation.
Our study addresses the question of whether the reduction of the preformed fraction of xenoantibodies prolongs delayed xenograft rejection, provided the complement system is completely inactivated. Plasmapheresis is an established procedure to reduce preformed xenoantibody levels. However, these antibodies are reported to rebound a few days after plasmapheresis. 6 On the basis of the antibody kinetics after plasmapheresis, we xenografted decomplemented rats, whose preformed xenoantibody levels were either unmodified, reduced, or rebounding. Astonishingly, delayed xenograft rejection time was almost tripled in recipient rats with rebounding antibody levels. This finding challenges the idea that delayed xenograft rejection is mediated by naturally occurring xenoantibodies and relativizes the therapeutic strategy to reduce these xenoantibody levels before transplantation.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Treatment groups
In a preliminary set of experiments, the kinetics of preformed xenoreactive and total immunoglobulin G (IgG) or IgM antibody levels, complement activity, coagulation variables, mixed lymphocyte reaction, and mononuclear cell cytokine secretion were monitored in rats (n = 6) after plasmapheresis.
In a second set of randomized experiments (n = 6 each group), guinea pig hearts were transplanted into complement-inactivated rats with unmodified (no plasmapheresis), reduced (immediately after plasmapheresis), or rebounding xenoantibody titers (defined time point after plasmapheresis). Control rats without complement inactivation were matched to each group.
Animals
The experiments were performed after approval of the local committee of ethics and in accordance with the "Guide for the Care and Use of Laboratory Animals," as revised by the National Institutes of Health in 1985. 7
Inbred guinea pigs (BNF Giessen, 250-300 g) and rats (Sprague-Dawley Hannover, 200-250 g) served as donors and recipients in a heterotopic cardiac xenograft model. Rats used for the preliminary kinetic studies after plasmapheresis weighed 350 g for reasons of the repetitive blood withdrawals. The laboratory animals were housed in plastic cages with stainless-steel wire bottoms in a laboratory with controlled temperature (20°C), humidity (50%), and 12-hour light-dark cycle. The animals were allowed to adapt to the environment at least 1 week before transplantation. They had free access to R3-EWOS-ALAB Brood Stock Feed (ALAB, Sollentuna, Sweden).
Plasmapheresis
Plasmapheresis was performed as described elsewhere. 6 Briefly, the rats were anesthetized with phentanylcitrate 0.315 mg/kg body weight administered intramuscularly (Hypnorm, Janssen, Belgium). Polyethylene catheters (Clay Adams, Parsippany, NJ) were inserted in the right carotid artery (PE-50, inner diameter: 0.58 mm, outer diameter: 0.965 mm) and the right femoral vein (PE-60, inner diameter: 0.76 mm, outer diameter: 1.22 mm). The blood was forwarded through a peristaltic pump (Ismatec MS-Reglo 6; Ismatec, Wertheim, Germany) at a flow rate of 3 mL/min through polypropylene fiber microfilters (Fresenius SPS 1003; Fresenius, St Wendel, Germany) with an active fiber surface of 90 cm2, an active fiber length of 16 cm, and a pore diameter of 0.5 µm. The filtration rate was 0.3 mL/min. One and a half plasma volumes were exchanged and replaced by 7% human albumin in physiologic saline solution. Rats not receiving plasmapheresis were sham operated.
Blood samples were taken from the arterial catheter before, immediately after, and 6, 9, 12, and 24 hours after plasmapheresis and then daily until day 7. After clotting at room temperature for approximately 1 hour, serum was separated from blood samples by centrifugation at 3000 rpm for 15 minutes at 0°C to 4°C and stored at –70°C until used. All sera from an individual animal were analyzed simultaneously. Heparinized blood samples (10 IU heparin per milliliter) for mixed lymphocyte reactions and the assessment of the tumor necrosis factor (TNF-) secretory potential of mononuclear cells or citrated blood samples (10% Na citrate) for the measurement of coagulation variables were withdrawn before, immediately after, and 2 days after plasmapheresis.
Circulating immunoglobulins
IgM and IgG were measured by commercially available Radial Immunodiffusion Kits according to Mancini (Camon, Wiesbaden, Germany).
Xenoreactive antibodies
Xenoreactive antibodies were measured by an enzyme-linked immunosorbent assay (ELISA) technique according to the method of Leventhal and colleagues. 8 Briefly, guinea pig platelets were isolated and lysed by freezing and thawing. Platelet membranes were extracted and centrifuged at 55,000 rpm. The supernatant was stored at –70°C until used. Guinea pig platelet membranes (100 µL/well of a 10 µg of protein per milliliter of preparation) were absorbed onto 96-well Polysor-P plates (Nunc, Napperville, Ill) for approximately 12 hours at 37°C. Serial dilutions of the primary antibody solution were added to the wells. After an incubation for 2.5 hours, the secondary antibody (rabbit anti-rat IgM [µ-chain specific] or rabbit anti-rat IgG [
-chain specific]) conjugated to alkaline phosphatase (dilution 1:500; Zymed Co, San Francisco, Calif) was added. The developing solution consisted of 1 mg/mL p -nitrophenyl phosphate and 100 mmol/L diethanolamine in 0.5 mmol/L MgCl2. The absorbance was determined at a wavelength of 405 nm.
Complement activity
Total complement activity was measured as hemolytic activity according to the method of Mayer. 9
Coagulation variables
Activated partial thromboplastin time, thrombin time, and prothrombin time were assessed from citrated whole blood (10% Na citrate), as described elsewhere. 10-12
TNF- secretory potential of mononuclear cells
Heparinized fresh whole blood (10 IU heparin/mL) was diluted 1:2 with phosphate-buffered saline solution (PBS). The peripheral blood mononuclear cell (PBMC) fraction was obtained by Ficoll-Hypaque centrifugation. The cells were washed in PBS before culturing. The PBMCs were cultured for 24 hours at 37°C at a density of 1 x 106 cells/well in RPMI supplemented with 5% (vol/vol) fetal calf serum unstimulated or after the addition of 25, 50, or 100 µg of concanavalin A. The supernatant was collected after centrifugation at 2000 rpm for 10 minutes. The concentration of TNF- in PBMC culture media was assayed by using a commercial murine TNF- ELISA kit (Endogen Inc, Boston, Mass).
Mixed lymphocyte reaction
Lymphocytes were separated from heparinized blood by density gradient centrifugation. Guinea pig stimulator cells and syngeneic rat lymphocytes were incubated with mitomycin C (final concentration, 25 µg/mL) in a humidified 5% carbon dioxide incubator at 37°C for 30 minutes. One hundred microliters of both the responder and the stimulator cell suspensions (1 x 106 cells/mL) was pipetted in each well of the microtiter plate and incubated for 5 days at 37°C (5% carbon dioxide at 90% humidity). Mitogen-stimulated controls were incubated with concanavalin A (final concentration, 25 µg/mL). After incubation with 10 µL/well bromodeoxyuridine and centrifugation at 300g for 10 minutes, the labeling medium was flicked off. After denaturation by formamide, each well was filled with 100 µL of monoclonal anti-bromodeoxyuridine. One hundred microliters of peroxidase-conjugated rabbit anti-mouse immunoglobulin was added to each well (diluted 1:1000 in PBS-Tween). After a 30-minute incubation, 100 µL of peroxidase substrate solution was added to each well. The enzymatic reaction was stopped after 10 minutes by the addition of 25 µL of 1 mol/L H2SO4. The intensity of the color reaction was read in an automatic ELISA plate reader (Titertek Multiscan MC) by using a 492-nm filter.
Complement depletion
CVF (Naja naja) was purchased from Sigma Chemical Company (St Louis, Mo). A 1-step procedure with fast protein liquid chromatography 13 was used to isolate phospholipase A2–free CVF. Complement-depleting activity was determined in Bellow and Cochrane units. 14 Fifty units of CVF per kilogram of body weight was administered every 8 hours in a 5-minute intravenous infusion. For that purpose, a 5-cm long, spiral-shaped piece of PE-10 catheter (Clay Adams, Parsippany, NJ) attached to a silicone tube (Silastic 0.012 in x 0.025 in, No. 602-105 HH 061999; Dow Corning Corp, Midland, Mich) was fused to a 30-cm piece of a PE-20 catheter. The silicone part of the catheter was placed in the animal’s left jugular vein, as described elsewhere, 15 and the PE-20 end was diverted to the exterior.
Transplantation technique
The abdomen was opened by a midline incision, the left kidney was removed, and the kidney vessels were cuffed, as described elsewhere. 16 The donors underwent anesthesia with pentobarbital 60 mg/kg of body weight administered intraperitoneally (Mebumal vet, Nord Vacc, Sweden). Heparin in a dose of 300 IU was injected intravenously before harvesting the heart. The grafts were flushed with cold Ringer lactate solution containing heparin in a concentration of 50 IU/mL. Immediately after being harvested, the graft was anastomized with the cuffed vessels.
Assessment of graft rejection
The transplanted hearts were palpated hourly. Rejection was considered to be complete when no more pulsations were palpable. Rejection was confirmed histologically.
Immunohistology
Cardiac xenografts were snap-frozen in isopentane. Cryostat sections of 7 µm were cut, air-dried on gelatinized slides, and fixed in acetone. The sections were incubated with anti-rat monoclonal or monospecific polyclonal antibodies directed against all leukocytes (CD45, OX-1); B cells (CD45R, OX-33); T cells (CD5, OX-19); T cells, polymorphonuclear neutrophils, and natural killer (NK) cells (W 3-13); T-helper cells (CD4, W 3-25); cytotoxic T cells (CD 8, OX-8; Camon Co, Wiesbaden, Germany); IgG, IgM, fibrin, and C3 (ICN, Irvine, CA); and TNF- (Genzyme Corp, Boston, Mass) at room temperature for 30 minutes (dilution 1:100 and 1:2500 TNF-). The second incubation with the appropriate immunoglobulin-bridging antibodies (diluted 1:20) was at room temperature for 30 minutes. After incubation with alkaline phosphatase—antialkaline phosphatase complex (diluted 1:50), the coloration was developed with alkaline-phosphatase substrate. Sections were counterstained with hemalaun. Labeled cells within 20 high-power fields per section per rat were counted by blinded observers with the aid of an ocular grid micrometer (Leitz Co, Wetzlar, Germany). Depending on the cellular count, the mononuclear cell infiltrate was considered to be scarce (+), moderate (++), or marked (+++). Cytokine labeling was expressed semiquantitatively (<1%, <5%, 10%-20%, 20%-50%, or >75% of the cells indicated) because of the presence of extracellular labeling. The specificity of labeling was assessed by using multiple controls.
Isotype-matched monoclonal antibodies or purified immunoglobulin and a control for endogenous phosphatase activity were included in each experiment, and studies were undertaken to confirm the lack of labeling of normal guinea pig hearts (n = 5) with the antibodies used. Additional samples were fixed in neutral-buffered formalin, embedded in paraffin, and sectioned for light microscopy.
Statistics
Differences between the groups in graft survival, PBMC cytokine release, and mixed lymphocyte reaction were analyzed by using the Student t test after a normal distribution had been confirmed by using the Kolmogorov-Smirnov test. The Mann-Whitney U test was used to compare immunohistologic differences between the different groups. The Bonferroni procedure was applied to adjust for repeated comparisons. For all other determinations, 1-way analysis of variance with the Tukey honestly-significant-difference post hoc test was used to test for differences between the various groups after a normal distribution had been confirmed with the Kolmogorov-Smirnov test. Values are presented as means ± SD.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
, A ) followed by a slight rebound.
|
|
PBMC TNF- secretory potential
Secretion of TNF- by PBMCs immediately after plasmapheresis was maximal on stimulation with 12.5 µg of concanavalin A (285 ± 60 pg/106 PBMCs). Two days after plasmapheresis, maximum TNF- secretion was 75 ± 11 pg/106 PBMCs (P = .0098) after stimulation with 25 µg of concanavalin A (Fig 3).
|
|
|
|
Rats treated with CVF rejected the grafts after 31.8 ± 2 hours (P < .001 vs unmodified recipients). CVF combined with pretransplant plasmapheresis caused a transplant survival of 28 ± 6.1 hours (P < .001 vs plasmapheresis, P > .2 vs CVF).
Plasmapheresis 2 days before transplantation combined with CVF immediately after transplantation prolonged graft survival to 72 ± 9.7 hours (P < .001 vs plasmapheresis 2 days before transplantation, P < .001 vs plasmapheresis plus CVF).
Histologic and immunohistologic evaluation of rejected xenografts
In unmodified rats and rats treated with plasmapheresis, extensive hemorrhage and edema were seen throughout the myocardium of rejected xenografts besides capillary and venular congestion and platelet thrombi. The immunohistologic examination revealed focal capillary deposits of IgM and diffuse deposits of C3 and fibrin layers along endothelial surfaces of small and large blood vessels. IgG was scattered in the interstitium of the tissue. No infiltrating cells were detected (Table II).
|
|
Plasmapheresis in combination with CVF did not essentially change the immunohistologic rejection pattern, apart from diminished IgM deposits along endothelial walls, when both treatments were applied simultaneously.
TNF- labeling of intragraft mononuclear cells, however, was markedly increased (>75% of infiltrating mononuclear cells) in the group treated with plasmapheresis and CVF simultaneously (Fig 6
, A ) compared with the group that underwent plasmapheresis 2 days before CVF application and xenotransplantation (10%-20% of infiltrating mononuclear cells; Fig 6, B ).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
1 and through the antibody-independent alternative pathway, with the influence of the latter varying with the species combination. 2 Inactivation of both the classical and alternative pathway by CVF abrogates hyperacute rejection, giving way to a milder rejection process called delayed xenograft rejection, which is terminated after several days. 3 We found, as others did before, 3,17 the immunohistologic pattern of delayed xenograft rejection to be characterized by lacking C3 and abundant immunoglobulin deposits along the endothelial surfaces and a marked mononuclear cell infiltrate, which is predominated by activated macrophages and NK cells and associated with large amounts of cytokines (eg, TNF-). Thus there are mechanisms apart from complement activation that lead to a somewhat prolonged xenograft rejection and presumably precede T cell–mediated defense processes known from acute rejection. One of the initial steps of this delayed form of xenograft rejection seems to be donor organ endothelial cell activation, which is characterized by transcriptional induction of genes and subsequent protein synthesis, resulting in the expression of adhesion molecules, cytokines, and procoagulant molecules. 18 Apart from cytokines or mitogens, 19 xenoantibodies have been shown in vitro to activate endothelial cells. 20 On the other hand, xenoreactive antibodies are reported to facilitate the adhesion of NK cells to endothelial cells and the eventual lysis of endothelial cells in in vitro studies. 21,22 Under the hypothesis that xenoreactive antibodies are involved in delayed xenograft rejection, the pretransplant reduction of naturally occurring preformed xenoantibodies, the posttransplant inhibition of induced xenoantibody production, or both should further prolong delayed xenograft rejection. Our study concentrates on the preformed fraction of xenoantibodies and does not examine the role of specific xenoantibodies induced after xenografting.
Plasmapheresis proved to be a nonspecific but effective method to reduce natural xenoantibodies. In our preliminary kinetic studies the initial reduction of preformed xenoantibodies to below 30% of their original level was followed by an antibody rebound 2 days after plasmapheresis, which confirmed analogous results in the literature. 6 Like others, 23 we found a correlation between the titers of naturally occurring antiendothelial cell antibodies and the concentrations of IgM because the kinetics of both exactly paralleled each other after plasmapheresis. Supporting this finding, IgM, and not IgG, turned out to be the isotype of preformed xenoreactive antibodies in our model, as shown in the thrombocyte-targeted ELISA. There seem to be species-related differences in the isotype of naturally occurring xenoreactive antibodies. Although preformed rat anti-guinea pig antibodies are commonly reported to belong to the IgM subclass, 24 naturally occurring human anti-pig antibodies are believed to be mostly IgG. 25 On the other hand, specific xenoantibodies induced after xenotransplantation by immunization seem to belong to both the IgG and IgM classes. 26 Because of its nonspecificity, plasmapheresis reduced not only antibody levels but also coagulation factors and complement components, resulting in a temporary impairment of both the coagulation system and the complement activity, which were basically restored within 2 days.
Pretransplant plasmapheresis prolonged hyperacute rejection in our model from minutes to several hours by reducing preformed xenoantibodies and the activities of both the complement and the coagulation systems. Two days after plasmapheresis, when the above-mentioned parameters are renormalized, hyperacute rejection occurs in the same fulminant way as in unmodified recipients, giving clinical proof that the parameters accounting for hyperacute rejection were sufficiently restored.
The situation changes completely when plasmapheresis is administered in both above-mentioned time settings, provided hyperacute rejection has been abrogated by CVF. If combined with CVF-induced complement inactivation, plasmapheresis exerted immediately after transplantation did not result in a prolongation of xenograft survival nor in an alteration of the immunohistologic rejection pattern apart from reduced immunoglobulin deposits along endothelia. However, delayed xenograft rejection was prolonged from an average of 28 to 72 hours when rats received plasmapheresis 2 days before CVF application and subsequent transplantation. From the perspective of a pathogenetic role of preformed antibodies in delayed xenograft rejection, this finding was absolutely surprising because in animals treated with plasmapheresis and CVF simultaneously, xenotransplantation was performed in a decomplemented organism with preformed xenoantibody levels minimized, whereas rats undergoing plasmapheresis 2 days before the application of CVF and grafting represented a decomplemented host with renormalized xenoantibody levels, even being on the threshold of a rebound production. Although these results do not exclude endothelial cell activation nor antibody-dependent cellular cytotoxicity mechanisms by preformed xenoantibodies, which were reported previously, 20,21 they question the relevance of these processes in delayed xenograft rejection, at least within the observed time range of 72 hours. Meanwhile, there are rising doubts that preformed xenoreactive antibodies are pathogenetically involved in delayed xenograft rejection. NK cells have been found to activate endothelial cells with and without the involvement of natural xenoantibodies, 18 probably because of a partial overlap of the xenogeneic structures recognized by NK cells and xenoreactive natural antibodies. 27 NK cells were also shown to adhere to and efficiently lyse xenogeneic endothelium both in the presence and absence of xenoreactive natural antibodies. 21 Interactions between accessory molecule receptor-ligand pairs on NK cells, macrophages and xenogeneic endothelium seem to be of critical importance in delayed xenograft rejection. 28 These processes might be readily enhanced because NK cells proliferate in the presence of xenogeneic endothelial cells. 29 Although our study data question the necessity to reduce natural xenoantibodies before transplantation to prolong delayed xenograft rejection, they do not deal with a possible involvement of induced xenoantibodies in this process. In comparison with preformed xenoantibodies, which are known to be polyreactive, 30 endothelial cell activation and antibody-dependent cellular cytotoxicity mechanisms caused by the much more specific induced xenoantibodies 31 might be relevant pathogenetic factors in delayed xenograft rejection.
Our results demonstrate that plasmapheresis administered immediately before transplantation with the intention to minimize preformed xenoantibodies or to impair coagulation activity is not an effective therapeutic procedure to prolong delayed xenograft rejection in decomplemented hosts. However, the question remains why plasmapheresis proved effective when exerted 2 days before xenotransplantation. Apparently, plasmapheresis leads to a retarded and delayed reduction of parameters that account for xenograft failure beyond hyperacute rejection.
Specific cellular rejection mechanisms can be ruled out because the immunohistologic examination of the xenografts rejected after an average of 72 hours presented only mild T-cell infiltrates far apart from the cellular rejection pattern known from allograft rejection. Besides, major T-cell involvement is not to be expected in such an early stage of rejection.
On the other hand, xenogeneic helper T-cell responses were demonstrated to be weaker than allogeneic responses, at least by some authors. 1 In accordance, no relevant proliferation of rat lymphocytes on co-culture with guinea pig lymphocytes was to be observed in our mixed lymphocyte reaction. The failure to proliferate may be due to a defective CD4+ T helper cell-xenoantigen interaction or the failure of certain lymphokines to function across species differences, as suggested by the observation that primary xenogeneic cytotoxic responses were reconstituted by exogenous syngeneic, but not xenogeneic, growth factors. 32 Thus it may well be that T cell–mediated responses to xenografts are weaker than those to allografts, although there are conflicting results in models other than guinea pig to rat. 33
Our grafts prone to delayed xenograft rejection contained an infiltrate of mononuclear cells other than T cells. This corresponds well with other studies, which found polymorphonuclear neutrophils, macrophages, and NK cells to be the predominant types of graft-infiltrating cells in delayed xenograft rejection. 1 Adoptive transfer of macrophages from rat recipients of a first xenograft to a CVF-treated naive rat has been shown to accelerate the rejection of a discordant xenograft of the same donor species. 34 Even in the absence of xenoantibodies, T cell– and NK cell–activated macrophages were found to be able to reject xenografts, probably because of TNF-–associated endothelial cell activation. 35 Exposure of cultured endothelial cells to TNF- has been demonstrated to cause a series of changes in the cells, including the expression of new cell surface proteins and modifications in the permeability of the cell monolayer to plasma proteins and blood cells, which together are called endothelial cell activation . 36
In fact, the TNF- secretory potential of mononuclear cells proved to be markedly reduced 2 days after plasmapheresis compared with immediately after plasmapheresis. Seemingly, the procedure of plasmapheresis stimulates mononuclear cells to increase the spontaneous TNF- secretory output, which may come true for other cytokines as well, eventuating in a retarded exhaustion 2 days thereafter. The kinetics of cytokine secretion after plasmapheresis has hitherto not been studied, but previous studies indicated a rise in plasma anaphylatoxins on hemodialysis. 37 Among other factors that have not been examined in this study, the delayed secretory exhaustion of mononuclear cells might explain the prolonged xenograft rejection 2 days after plasmapheresis because of the impaired TNF- secretion on the stimulus of xenotransplantation. Correspondingly, TNF- was, to a much lesser degree, stained immunohistologically compared with the experimental group, in which plasmapheresis was carried out immediately before grafting.
In summary, delayed xenograft rejection appears to proceed in the absence of detectable complement activation. The role of xenoantibodies in this process is currently being examined. Our study results strongly argue against a pivotal role of the preformed fraction of xenoreactive antibodies, provided the complement system is completely inactivated, whereas adoptive transfer experiments substantiate the involvement of induced xenoantibodies. 38 Thus the therapeutic strategy to reduce preformed xenoantibodies before xenografting is challenged. Emphasis has to be put on inhibiting the production of the more specific elicited xenoantibodies after transplantation.
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
and diffuse endothelial activation. Transplant Proc 1993;25:2932. [Medline]Gal antibodies in human serum and their relation to serum cytotoxicity against pig cells. Xenontransplantation 1994;1:58-65. . Blood Coagul Fibrinolysis 1997;8:39-47. [Medline]This article has been cited by other articles:
|
|
 |
|
  T. Hasegawa, K. Okada, Y. Takano, Y. Hiraishi, and Y. Okita Autologous fibrin-coated small-caliber vascular prostheses improve antithrombogenicity by reducing immunologic response J. Thorac. Cardiovasc. Surg., May 1, 2007; 133(5): 1268 - 1276. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
