| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2005;28:149-156
© 2005 Elsevier Science NL
a Department of Laboratory Medicine, Hallym University College of Medicine, Anyang, South Korea
b Department of Thoracic and Cardiovascular Surgery, Cancer Research Institute, Seoul National University College of Medicine, 28 Yongon-Dong, Chongno-Gu, Seoul 110-744, South Korea
c The Xenotransplantation Research Center, Seoul National University Hospital, Seoul, South Korea
Received 8 September 2004; received in revised form 15 February 2005; accepted 21 February 2005.
* Corresponding author. Address: Department of Thoracic and Cardiovascular Surgery, Cancer Research Institute, Seoul National University College of Medicine, 28 Yongon-Dong, Chongno-Gu, Seoul 110-744, South Korea. Tel.: +82 2 2072 3161; fax: +82 2 765 7117. (Email: ytkim{at}snu.ac.kr).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Materials and methods3. ResultsAppendix A. Conference...References |
|---|
Key Words: Hyperacute rejection • Pulmonary xenotransplantation • Platelet • Desmopressin • Thrombosis
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. ResultsAppendix A. Conference...References |
|---|
Intravascular thrombosis and systemic coagulopathy are other barriers for xenotransplantation [6]. Quiescent endothelium normally maintains an anticoagulant profile to the adjacent blood. However, any failure of endothelial regulators of coagulation can lead to serious procoagulant consequences. The von Willebrand factor (vWF) has been proposed as a major contributor of coagulopathy associated with xenotransplantation [7]. vWF is a protein stored in platelets and endothelial cells that is released upon endothelial cell activation [8]. vWF binds to glycoprotein Ib (GPIb) on platelets in high shear stress leading to platelet adhesion and activation. Swine vWF has a supranormal capacity to bind to human GPIb, resulting in platelet aggregation even in the absence of sheer stress [9]. This behavior contributes to the immediate sequestration of circulating platelets in the lung [2,10]. The interaction between swine vWF and human platelet involves the platelet mediated thrombus formation, which results in disseminated coagulopathy [7]. Therefore, blocking the vWF-platelet interaction could be a promising strategy to reduce initial intravascular thrombosis.
Desmopressin (1-deamino-8-D-arginine vasopressin, DDAVP), an analog of vasopressin, has been used to treat patients with a deficiency of vWF [11]. It raises the plasma levels of vWF and FVIII by releasing Weibel-Palade (WP) bodies in endothelial cells. The release of vWF from WP bodies has been demonstrated by the binding of DDAVP to a vasopressin type 2 receptor located on the surface of endothelial cells without up-regulation of endothelial vWF mRNA synthesis.
In order to deplete or reduce the amount of vWF in endothelial cells of pulmonary xenografts, we designed a novel strategy—the administration of DDAVP to donor swine before transplantation. The effects of DDAVP were evaluated to the extent of vWF depletion and inhibitory effect on pulmonary microvascular thrombosis and platelet sequestration in the swine-to-canine pulmonary xenograft model.
|
2. Materials and methods |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. ResultsAppendix A. Conference...References |
|---|
2.2. Measurement of vWF Ag content in swine lung tissue
To determine if DDAVP could deplete vWF from pulmonary endothelial cells, we obtained lung samples at the time of lung retrieval from four controls and two DDAVP-treated pigs, and measured the content of vWF Ag in the lung tissue. The tissue was immediately frozen in liquid nitrogen. Two pieces from different sites of each specimen were selected and each tissue piece was ground and treated with protein extraction buffer (50mM Tris–Cl, 150mM NaCl, 5mM EDTA, 1% triton-X 100, 1mM PMSF, pH 8.0). From each extract, the vWF Ag and total protein levels were determined. The vWF Ag level was determined by a sandwich enzyme-linked immunosorbent assay (ELISA), consisting of a capture antibody, rabbit anti-human vWF antibody (A0082, Dako Corp., Carpinteria, CA) and peroxidase-conjugated detecting antibody. The pooled normal swine plasma in serial dilutions was tested in parallel and used as a calibrator standard. The vWF Ag content expressed as the ratio of tissue vWF Ag level to total amount of protein was compared between the DDAVP treatment and control groups.
2.3. Assessment of expression of vWF mRNA in lung tissue
To determine if DDAVP induced the expression of vWF mRNA in vivo, the level of vWF mRNA expression was compared in both lung tissues obtained before and after the administration of DDAVP in two donor pigs. Swine endothelial cells were studied in parallel as a positive control. vWF and GAPDH mRNA were analyzed by reverse transcriptase and polymerase chain reaction (RT-PCR). The primer sequence was designed from the swine vWF cDNA sequence (sense 5'-gga ggc ctg tct act caa cg-3', antisense 5'-tga ccc tgc aga agt gac tg-3'). The change of the vWF mRNA amount was determined by a visual comparison of the band density of PCR product.
2.4. Swine-to-canine left single lung xenotransplantation
A swine, weighing about 30kg, was anesthetized with intramuscular administration of ketamine hydrochloride (20mg/kg) and an intravenous line was placed in a vein of its ear. Thiopental sodium of 200–250mg and 0.2mg of atropine were administered. The swine was then intubated and ventilated with 100% oxygen at a rate of 12 breaths per minute with a tidal volume of 15ml/kg. Arterial pressure and blood gases were monitored through a cannula placed in the femoral artery. A median sternotomy and thymectomy were performed and the pericardium was opened. The left side azygos vein was ligated and divided. The inferior and superior venae cavae were loosely encircled with heavy silk. After injection of 20,000 IU of heparin intravenously, a 7 Fr. cannula was placed in the main pulmonary artery. Five micrograms per kilogram of alprostadil (PGE1, Eglandin®, Welfide, Korea) was injected directly into the pulmonary artery 10min before harvest. The superior vena cava was ligated and the inferior vena cava and left atrial appendage were incised. Euro-Collins solution (100ml/kg) was administered into the pulmonary artery through the pulmonary arterial cannula at a height of 40cm. The inferior pulmonary ligaments were sharply divided and the trachea was divided with a stapling device. The heart and lungs were then removed en bloc and immersed in cold (4°C) saline solution. The cold ischemic time in all the animals was within 3h.
Swine-to-canine orthotopic single lung transplantations were performed in three groups: the control group, a mongrel dog transplanted with unmodified swine lung (n=3); the CVF group, CVF-pretreated dog transplanted with unmodified swine lung (n=3); the CVF+DDAVP group, CVF-pretreated dog transplanted with DDAVP treated swine lung (n=3). Nine adult male mongrel dogs weighing 26 to 42kg were used as recipients. The recipients of the CVF and CVF+DDAVP groups were given 60units/kg of CVF intravenously 24h prior to transplantation in order to deplete complement components. The recipient dogs were anesthetized with intramuscular administration of ketamine hydrochloride (20mg/kg) and were then intubated and ventilated with 100% oxygen at a rate of 12 breaths per minute with a minute ventilation of 180ml/kg. The right femoral artery was cannulated for arterial pressure monitoring. A pulse oximeter was placed on the animal tongue to trace oxygen saturation. A 5 Fr. Swan-Ganz catheter (Edward laboratory, USA) was placed through the right jugular vein and continuous pressure monitoring of the pulmonary artery and right atrium was performed. Cardiac outputs were measured using the thermodilution method. The dog was positioned at right lateral decubitus and the left thoracotomy was placed through the fourth intercostal space. A left pneumonectomy was performed and the left lung of the swine was transplanted. The left atrium was sutured with 5-0 Prolene®, and the pulmonary artery was anastomosed with 6-0 Prolene®. The bronchus was sutured with a 4-0 Prolene® suture. The left atrial auricle was cannulated and the left atrial pressure was monitored. The animal was given 2,500 IU of heparin and the pulmonary artery was unclamped at the time of reperfusion. A 12-mm ultrasonic flow meter (Transonic, USA) was placed around the left main pulmonary artery and the flow to the transplanted left lung was measured. The hemodynamic data was measured repeatedly at 10, 20, 30, 60, 90, 120, and 180min after reperfusion. Based on this data, we calculated physiological variables such as pulmonary vascular resistance and systemic vascular resistance. Serial lung biopsies, as well as blood samples, were collected during each time. After 3h of reperfusion were completed, the animal was sacrificed by giving it 20mequiv. of KCl and an autopsy was performed. The total warm ischemic time was consistently less than 60min.
2.5. Complement hemolytic activity
To see the effect of CVF on the complement system, the recipients' complement hemolytic activity on rabbit erythrocytes was evaluated as previously described. Briefly, 0.1ml of 108 rabbit erythrocytes in gelatin veronal buffer containing dextrose (57mM NaCl, 5mM sodium diethyl barbiturate, 3% dextrose, 0.1% gelatin, 0.06mM CaCl2, 0.4mM MgCl2) were incubated with 0.1ml dog serum diluted in the same buffer and incubated for 60min at 37°C. After adding 1.5mL of ice-cold PBS and centrifugation for 10min at 1500g, hemolysis was determined in the supernatant by measuring the absorbance of released hemoglobin at 412nm.
2.6. Measurement of blood cell counts and coagulation parameters
We monitored the changes of the following parameters up to 3h after transplantation: blood cell counts, prothrombin time and levels of fibrinogen, antithrombin and D-dimer. The blood cell count was performed by ABCvet CBC analyzer® (ABX hematology, France) and coagulation parameters by STA Compact analyzer® (Diagnostica Stago, France). Each parameter was expressed as a percentage of baseline value prior to reperfusion.
2.7. Histologic evaluation
A series of lung biopsies of a xenograft were performed prior to and at 10, 20, 30, 60, 120, and 180min after reperfusion. The tissue samples were fixed in 10% buffered formalin, washed, dehydrated and embedded in paraffin. The embedded samples were cut into 4-µm thick sections, re-hydrated and stained with hematoxylin and eosin.
2.8. Statistical analysis
Data was expressed as the mean±SEM. Comparisons between groups were analyzed using a mixed model. Significance was accepted at P<0.05.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. ResultsAppendix A. Conference...References |
|---|
|
|
|
3.4. DDAVP administration attenuated coagulation abnormalities
Prothrombin time tended to be prolonged in the unmodified control group but there was no significant difference between groups. In contrast, fibrinogen and antithrombin levels decreased after reperfusion (P=0.0879 for fibrinogen, P=0.0034 for antithrombin, Fig. 4
). The decrease of fibrinogen and antithrombin levels tended to be attenuated in the CVF+DDAVP group. The D-dimer level was pitched occasionally during transplantation but there was no significant difference between elapsed time and group.
|
|
In this study, CVF administration depleted complement proteins and inhibited the activation of the recipient's complement system against xenografts. Serum complement hemolytic activity was maintained below a detectable range throughout the experiment and only scarce depositions of C3 fragments were seen in xenografts. However, it failed to improve the recipients’ hematologic, hemodynamic and histopathologic findings after xenotransplantation. On the contrary, the pulmonary vascular resistance increased and systemic vascular resistance decreased; the sequestration of platelets was more profound in CVF-treated recipients compared to the unmodified control group. This suggests that the hemodynamic deteriorations shown in the CVF-treated group may not be dependent on the formation of membrane attack complex on xenograft but rather relate to other factors. Theoretically, CVF administration does not completely inhibit the complement system. A small amount of activation of complement can release anaphylatoxins such as C3a and C5a, which can activate pulmonary endothelial cells. Moreover, a preformed large amount of anaphylatoxins in circulation, which have been continuously produced and released by CVF, may lead to the strong activation of pulmonary endothelial cells. It is conceivable that activation of endothelial cells is a more important cause of HPXD than complement activation itself. This would explain why the CVF group showed a more prominent increase of pulmonary vascular resistance, as well as the severer platelet sequestration than those seen in the unmodified control group. The phenomenon observed in the CVF-treated group was similar to the typical features seen in the complete discordant pulmonary xenotransplantation model. Such observations agree with the results of other researchers [4]. In terms of the xenotransplantation model, the CVF treated canine model should be considered more suitable than the unmodified model as a control.
Intravascular thrombosis and disseminated coagulopathy are two of the main pathologies of pulmonary hyperacute rejection in xenotransplantation. Depletion of xenoreactive antibodies or inhibition of complement activation only partially improves xenograft functions [3] and both cannot abolish coagulopathy. Although the mechanism of coagulopathy in pulmonary xenotransplantation is still not fully understood, it has been proposed that the activation of endothelial cells secondary only to the binding of antibodies, including anti 
Gal or anti-non Gal antibodies, or a sublytic amount of complement activation play an important role [13,14]. Endothelial activation enhances intravascular thrombosis and platelet sequestration by expressing adhesion molecules, producing vasoconstrictors and releasing secretory granules containing platelet activators. Swine vWF which is released from endothelial cells binds to human platelets, resulting in platelet activation and aggregation. It was reported that approximately 82% of the platelets disappeared from the perfusate during the first minute of perfusion of swine lung with human blood [2]. Therapy with anti-GPIb monoclonal antibody inhibited platelet deposition on the xenograft [15] and inhibitors for the interaction between GPIb or GPIIIa and vWF attenuated the increase of pulmonary vascular resistance and release of thrombin and histamine [15]. However, an anti-platelet monoclonal antibody can lead to sequestration of platelets to the reticuloendothelial system [16] and thus, non-selective inhibition of the recipient's platelet function can cause possible hemorrhagic complications or systemic shock. Therefore, it is necessary to develop strategies to modulate vWF on the part of the donor.
DDAVP has been used to treat vWF-deficient patients because it induces the release of vWF from endothelial cells without up-regulation of vWF synthesis in endothelial cells [17]. The data showed that pre-treatment of DDAVP prevented platelet sequestration, reduced reactive leukocytosis, and attenuated subsequent coagulopathy demonstrated by the fall of both fibrinogen and antithrombin levels. The effect of DDAVP administration on preventing platelet activation was surprisingly successful considering that canine platelets have not been reported to have a strong interaction with swine vWF, unlike a primate's platelets. We believe that the effect of DDAVP would be more prominent if we were to use a discordant primate as a recipient.
We speculated that the beneficial effect of DDAVP might be caused by reducing the release of vWF from the xenograft endothelial cells because DDAVP administration reduced the amount of vWF in swine lung tissue to 50% in our experiment and also the association between vWF release and platelet activation had been already proven [16]. However, there is a possibility that vWF may not be the only player for the favorable effect of DDAVP. The alteration of P-selectin expression after DDAVP treatment might have played a certain role because P-selectin and vWF are contained in WP bodies of endothelial cells together. To clarify this, the relationships between the amount of vWF depletion and extent of platelet activation should be investigated further and the effect of DDAVP on P-selectin expression of endothelial cells also should be studied.
There is evidence that a xenoreactive natural antibody, predominately IgM, binds to the xenograft and activates the complement. Subsequent deposition of IgM, IgG, C1q, C3, C4, C9 will lead to the rapid failure of xenograft [3,18]. However, whether xenoreactive antibodies actually remain bound to swine endothelium is not certain [19]. It was demonstrated that, although xenoreactive antibodies can bind to pulmonary xenografts, many of these antibodies are shed as a form of antibody-antigen complex [20]. On the other hand, an acute pulmonary xenograft injury was characterized by a rise of pulmonary vascular resistance, which is associated with the profuse release of thromboxane A2 [21]. Thromboxane A2 is produced from arachidonate in platelets by the aspirin sensitive cyclooxygenase pathway [22]. Collins et al. demonstrated that elimination of a pulmonary intravascular macrophage with liposome chlodronate also reduces the synthesis of thromboxane [21]. An activated macrophage in the lung may trigger endothelial cells to release vWF, which sequentially induces binding, aggregation of platelets, and the synthesis and release of thromboxane into circulation from the platelets. Thus, strategies to reduce interaction between endothelial cells and platelets would be beneficial in preventing the rapid increase of pulmonary vascular resistance, maintaining barrier function of pulmonary endothelial cells, as well as ameliorating platelet sequestration.
As we measured pulmonary vascular resistance while the right lung was perfused, the increase of pulmonary vascular resistance observed in the CVF group was primarily caused by the canine native right lung. The starting point of the PVR measurement was when the left pulmonary artery was being clamped just before the reperfusion, and thus represented right pulmonary vascular resistance. After reperfusion, the PVR in both the unmodified and CVF+DDAVP groups decreased to a small degree and maintained its level thereafter. This result suggested that DDAVP administration, which achieved the depletion of vWF, played an important role in preventing a rise of PVR in the native recipient's right lung which was presented in the CVF group. It has been well studied in the swine-to-baboon model that the immune complex between the swine vWF and the xenoreactive natural antibody from the baboon is shed into the blood from the transplanted xenograft and causes a disseminated intravascular coagulopathy (DIC) [7]. We hypothesized that the vWF immune complex secreted from the transplanted xenograft could cause DIC. These immune complexes could further activate the complement system, generate anaphylatoxins and cause systemic shock. As another explanation, increased amounts of anaphylatoxins in CVF treated recipients could have led to the release of vWF from endothelial cells and subsequent activation of platelets, which induced profuse release of thromboxane A2. An elevated level of thromboxane A2 in circulation could be responsible for the increase of PVR in the native recipient's right lung. In the CVF+DDAVP group, the depletion of vWF secretion might have ameliorated such release of thromboxane A2 and its consequent systemic effect. A decrease of platelet count, increase of leukocyte count and decrease of fibrinogen and antithrombin levels were observed in both the unmodified and CVF groups while those findings were not observed in the CVF+DDAVP group. Those results can also be explained in the same context. Even though the beneficial effect of DDAVP was transitory, it may prove to be worthwhile as the reduction of the initial inflammation would be beneficial to prolong overall graft survival.
The swine vWF molecule can be a target for gene alteration in a genetically modified swine production project for xenotransplantation. However, considering its role in maintaining physiological hemostasis, complete aberration of the vWF gene would not be desirable. Transient inhibition of gene expression might be an alternative way to control vWF. How to achieve timely and well-tuned control on the extent of vWF expression is the next huddle to overcome.
Our data suggested that swine vWF plays an important role in HPXD and the depletion of donor vWF would be a safe and effective strategy to prevent microvascular thrombosis in xenografts. However, the number of each experiment group was rather small. Moreover, our experimental model was a swine-to-canine transplantation which is not completely discordant and might not be appropriate for evaluating the agents or strategies to prevent HPXD in regard to natural antibodies. Before reaching a final conclusion, we need to confirm our postulation through ex-vivo perfusion using human blood or the swine-to-baboon xenotransplantation model. An experiment to define a quantitative relationship between the beneficial effect of DDAVP and extent of vWF depletion is also necessary in order to be clarified. In this study, we confirmed that DDAVP pre-treatment is a very effective and safe preconditioning regimen for pulmonary xenotransplantation experiments.
|
Appendix A. Conference discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. ResultsAppendix A. Conference...References |
|---|
Dr Kim : Each group has 3 animals.
Dr Van Raemdonck : Is that not too small to show any differences?
Dr Kim : Even though we did not use the baboon, this study took a lot of our time, effort, and expense. I think for a large animal model, 3 in each group are reasonable.
If we had used baboons, for example, 1 or 2 for each group would have been enough. As our model is not a discordant xenotransplantation model, but a concordant model, we want to know what will happen if we apply the same model using a baboon or human. For this finding, we are now working on the ex-vivo perfusion model. Once we have baboon facilities in my country in the future, we are planning to do that type of research.
Dr Van Raemdonck : Can you repeat what this drug DVAP is doing. What is the working mechanism?
Dr Kim : That is the drug which has been used for Type I von Willebrand's disease. For Type I von Willebrand's disease, the patients have a smaller amount of von Willebrand factor in their blood, and if we use the injection of vasopressin, it causes the endothelial cells, especially from the lung, to secrete the von Willebrand's factor from the WP body. And what became apparent is that although several treatments work fine at the beginning, it did not increase serum von Willebrand factor after repeated injections, which means the von Willebrand factor had been depleted from the WP body. We tested this with several lung biopsies, treating a pig with DDAVP for 10 days. What we found is 2 days are enough to deplete pulmonary endothelial von Willebrand factor.
Dr Van Raemdonck : Did you look at the effect of this drug in an allotransplant model?
Dr Kim : Actually, we did not.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Presented at the joint 18th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 12th Annual Meeting of the European Society of Thoracic Surgeons, Leipzig, Germany, September 12–15, 2004. |
References |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. ResultsAppendix A. Conference...References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
