Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Liver disease is characterized by complex alterations in the hemostatic system, including changes in blood platelets, blood coagulation, and the fibrinolytic system.1, 2 During liver transplantation, substantial additional changes in the hemostatic system occur.3, 4 Traditionally, it has been thought that the hemostatic alterations in patients with liver disease result in a bleeding tendency, but this concept is gradually changing. Recent studies have shown that many hemostatic changes that may promote bleeding (eg, reduced levels of procoagulant factors) are compensated for by changes that promote clot formation (eg, reduced levels of anticoagulant factors).5–7 Even though new information about coagulation in patients with cirrhosis has changed the way in which some investigators view hemostatic balance,1, 8, 9 critics of the rebalanced theory suggest that further evidence is necessary.10 Preliminary laboratory and clinical evidence suggests that the hemostatic system of patients with cirrhosis has sufficient capacity to produce a stable platelet plug and sufficient thrombin to prevent excessive bleeding. However, this rebalanced hemostatic system presumably has narrower hemostatic reserve in comparison with the balance in healthy individuals, and the balance can turn more easily toward bleeding or thrombosis.11 In stable patients with liver disease, we and others have recently shown that the primary hemostatic system, in which blood platelets play a central role, may also not be as abnormal as previously assumed.12, 13 Although patients with cirrhosis frequently have reduced platelet counts, increased plasma levels of von Willebrand factor (VWF) compensate for this in vivo.5 Therefore, low platelet numbers in patients with cirrhosis may not necessarily reflect defective primary hemostasis and may not require correction by the infusion of platelet concentrates prior to an invasive procedure.
Nevertheless, routine diagnostic tests used to monitor coagulation during liver transplantation frequently point to defective hemostasis and may prompt the surgical team to transfuse blood products. In addition, 2 observations have pointed toward a deterioration of platelet function during liver transplantation, providing an argument to support the use of blood products such as platelet transfusions. First, after graft reperfusion, a hyperfibrinolytic state may develop because of the release of tissue-type plasminogen activator (tPA) from the graft.14, 15 The resulting formation of plasmin is potentially capable of proteolysis of key platelet receptors, which could compromise platelet function and number.16 Second, after reperfusion, platelets are trapped in the liver graft, presumably because of adhesion to activated endothelium.17 It is possible that the platelets that are activated by the graft stay in circulation but can no longer be activated and therefore do not participate in hemostasis. These platelets may be recognized by the expression of activation markers on the platelets such as P-selectin and activated αIIbβ3. Measurement of the platelet count alone is not likely to reflect the platelet-dependent hemostatic capacity.
The hypothesis of deteriorating platelet function during liver transplantation by the aforementioned mechanisms has never been formally tested. This may in part be explained by difficulties in assessing platelet function, which requires specialized diagnostic tests in fresh blood. Furthermore, reports of platelet function parameters during liver transplantation in older literature may no longer reflect the situation in patients transplanted today. In the last 15 years, substantial changes in anesthesiological care, surgical techniques, and organ preservation have contributed to a steady reduction in transfusion requirements.18 Consequently, the perioperative hemostatic status of patients has changed substantially.
In this study, we aimed to describe key aspects of platelet functionality in patients undergoing liver transplantation to come to a more rational approach to platelet transfusion. We used a combination of approaches to study properties of platelet function during and after liver transplantation. We measured activation markers on the platelet itself, platelet-released activation markers in plasma, and expression of constitutively expressed platelet receptors. In contrast to the hypotheses described previously, the combined results of our studies indicate that platelet function is well preserved throughout the transplantation. These findings provide additional evidence for a restrictive use of platelet transfusions during liver transplantation.
CTAD, citrate, theophylline, adenosine, and dipyridamole; DBD, donation after brain death; DCD, donor after cardiac death; ELISA, enzyme-linked immunosorbent assay; FFP, fresh frozen plasma; MELD, Model for End-Stage Liver Disease; PF4, platelet factor 4; RBC, red blood cell; sGPV, soluble glycoprotein V; tPA, tissue-type plasminogen activator; VWF, von Willebrand factor.
PATIENTS AND METHODS
Twenty adult patients undergoing liver transplantation between April 2007 and March 2008, who gave informed consent, were included in this study. The median follow-up was 7 months (range, 2–13 months). At the start of this study, we intended to study platelet activation markers separately in patients who did or did not receive aprotinin (Trasylol) during transplantation, and we initially randomly assigned patients to an aprotinin or placebo group. However, during the course of our study, aprotinin was withdrawn from the market, and we therefore decided to continue the study with only patients that did not receive aprotinin during their transplant procedure. Patient characteristics, including indications for transplantation, and surgical variables are provided in Table 1. Two patients required urgent retransplantation on postoperative day 3 for primary nonfunction and hepatic artery thrombosis, respectively. From these patients, only blood samples taken until postoperative day 3 were included in this study. Neither patient was re-entered into this study after retransplantation. In July 2008, 19 patients were still alive with well-functioning grafts; 1 patient died at postoperative day 60 because of severe graft-versus-host disease.
Table 1. Patient Characteristics and Surgical Variables (n = 20)
NOTE: The data are presented as number (percentage) for categorical variables and as median (range) for continuous variables.
Abbreviations: DBD, donation after brain death; DCD, donor after cardiac death (non–heart-beating donor); FFP, fresh frozen plasma; MELD, Model for End-Stage Liver Disease; RBC, red blood cell.
Acute liver failure
Preoperative MELD score
Type of donor
Type of transplantation
Cold ischemia time (minutes)
Warm ischemia time (minutes)
RBC transfusion (number of patients)
FFP transfusion (number of patients)
Platelet transfusion (number of patients)
Preoperative laboratory values
Platelet count (×109/L)
Prothrombin time (seconds)
Blood samples were taken at the following time points during and after surgery: 30 minutes after induction of anesthesia, 30 minutes after the start of the anhepatic phase, 30 minutes after graft reperfusion, at the end of surgery, and on postoperative days 1, 5, and 10. Blood samples were obtained from a dedicated nonheparinized arterial line and were drawn into 3.2% sodium citrate (9:1 vol/vol) or into a mixture of citrate, theophylline, adenosine, and dipyridamole (CTAD). Some of the citrated whole blood was processed for flow cytometry (as discussed in the next section); the remainder of the citrated blood and the CTAD blood was processed to obtain platelet-poor plasma. To obtain platelet-poor plasma, samples were centrifuged once at 1000g for 10 minutes, after which the samples were stored at −80°C until use.
Aliquots of whole blood were processed immediately after the blood draw for flow cytometry analysis. Samples were incubated with fluorescently labeled (fluorescein isothiocyanate or phycoerythrin) antibodies against P-selectin (BD Pharmingen, Breda, The Netherlands), glycoprotein Ibα (Zebra Bioscience, Enschede, The Netherlands), integrin αIIbβ3 (Zebra Bioscience), and activated integrin αIIbβ3 (BD Biosciences, Breda, The Netherlands) for 30 minutes at room temperature. Subsequently, the samples were fixed with a commercially available fixative (Thrombofix platelet stabilizer, Beckman Coulter, Mijdrecht, The Netherlands) and stored at room temperature until the analysis, which was performed within 24 hours after the start of surgery. The samples were analyzed with flow cytometry (FACSCalibur, BD Biosciences). Platelets were gated from whole blood scatter plots, and the median fluorescent signal was recorded. For each analysis, we used platelets from healthy volunteers, which were or were not activated ex vivo with thrombin receptor activating peptide, as positive and negative controls.
Soluble Activation Markers
The soluble proteolytic fragment of glycoprotein Ib (glycocalicin) was measured in citrated plasma by an in-house enzyme-linked immunosorbent assay (ELISA) using a monoclonal capturing antibody, a polyclonal detection antibody, and a horseradish peroxidase–labeled secondary goat anti-rabbit antibody. A standard curve was constructed with glycocalicin purified from human plasma as described.19
The soluble proteolytic fragment of glycoprotein V [soluble glycoprotein V (sGPV)] was measured in citrated plasma with a commercially available ELISA (Asserachrom, Roche Diagnostics, Woerden, The Netherlands). Levels of β-thromboglobulin and platelet factor 4, which are proteins secreted from platelet α-granules upon activation, were measured in CTAD plasma with commercially available ELISAs (Asserachrom, Roche Diagnostics, Woerden, The Netherlands, and R&D Systems, Abingdon, United Kingdom).
Normal values for all ELISAs were established by the measurement of levels in individual CTAD plasma samples from healthy volunteers.
To evaluate differences between different time points and the baseline measurement (the start of surgery), a linear mixed effect model was used. This model takes the correlation between the repeated measurements within a single patient into account. P values < 0.05 were considered statistically significant. All analyses were performed with the statistical software package SPSS 15.0 (SPSS, Inc., Chicago, Ill.)
Is There Evidence for Systemic Platelet Activation During Liver Transplantation?
To assess platelet activation during liver transplantation, 3 distinct approaches were used. First, we measured levels of P-selectin and activated αIIbβ3 on the platelet surface. P-selectin resides in the α-granules in resting platelets but is translocated to the surface upon activation, whereas the aggregation receptor αIIbβ3 is in a fully inactive form on a resting platelet. No increase in P-selectin or activated αIIbβ3 expression was observed during surgery (Fig. 1A,B). Even after graft reperfusion, which has been hypothesized to result in massive platelet activation due to exposure of platelets to ischemically damaged endothelium, no increase in these activation markers was observed. Moreover, the well-described drop in the platelet count after reperfusion, which has been ascribed to trapping of platelets inside the graft, was not observed in this series, even after the exclusion of those patients receiving platelet concentrates (Fig. 1C). Postoperatively, the platelet count increased slowly, reaching values that were significantly higher than baseline values on postoperative day 10 (P < 0.001). After surgery, levels of P-selectin and activated αIIbβ3 decreased slightly, reaching a statistically significant difference on postoperative day 10 for activated αIIbβ3 (P = 0.004).
At the start of surgery, the mean P-selectin value was only slightly higher than the mean value measured in 5 healthy volunteers (3.2 versus 2.5 FU), and the same was true for the mean value for activated αIIbβ3 (2.5 FU at the start of surgery versus 1.9 in healthy volunteers). Mean values at postoperative day 10 were comparable to levels in healthy volunteers.
In a second approach, we measured plasma levels of β-thromboglobulin and platelet factor 4. These proteins are excreted from platelet α-granules upon stimulation. In accordance with the data on P-selectin and activated αIIbβ3, we did not observe appreciable increases in plasma levels of β-thromboglobulin and platelet factor 4 during surgery (Fig. 2A,B). In fact, levels of platelet factor 4 decreased slightly after reperfusion (P < 0.01 versus baseline). Postoperatively, levels of both markers substantially dropped in comparison with baseline values (P < 0.001 for β-thromboglobulin on days 1, 5, and 10 versus the start of surgery and P < 0.001 for platelet factor 4 at the end of surgery and on days 1, 5, and 10 versus the start of surgery).
Finally, we measured plasma levels of glycocalicin and sGPV, both part of the abundant GPIb/IX/V complex, which are cleaved from the platelet upon platelet activation. After graft reperfusion, plasma levels of glycocalicin decreased slightly but significantly, and levels remained slightly lower than baseline values until day 10 (P < 0.01 for reperfusion, end of surgery, day 1, and day 5 versus the start of surgery). Levels of sGPV remained stable during surgery but decreased significantly during the first postoperative week (P < 0.001 for days 1, 5, and 10 versus the start of surgery; Fig. 3A,B).
At the start of surgery, plasma levels of all 4 soluble markers of platelet activation were substantially higher than values found in healthy individuals (normal levels are given in the figure legends).
Is There Evidence for Proteolysis of Key Platelet Receptors During Liver Transplantation?
We measured the expression of glycoprotein Ibα and the total amount (activated and resting) of αIIbβ3 on the platelet surface in samples taken during and after liver transplantation (Fig. 4). Expression of these proteins is known to decrease following proteolysis by proteases such as plasmin, and the expression of GPIbα is also diminished as a result of platelet activation. However, even after reperfusion, in which a hyperfibrinolytic state was well recognized in at least some of the patients, we did not observe a decrease in expression of the 2 receptors. Also, in the postoperative period, the levels of both receptors were not significantly different from baseline values.
In this comprehensive analysis, we have searched for evidence of platelet activation and a subsequent decrease in platelet function during liver transplantation by measuring various markers of platelet activation. We found no evidence of activation of circulating blood platelets in patients undergoing liver transplantation. Furthermore, we found no evidence for proteolytic cleavage of key platelet receptors in these patients. These combined results suggest that platelet function does not deteriorate as a result of the physiological challenges encountered during the transplant procedure. In combination with our previous finding that low platelet counts in patients with cirrhosis are compensated for by increased plasma levels of VWF,5 we conclude that platelet function and therefore primary hemostasis remain adequate during the course of liver transplantation.
In the current study, all measured markers of platelet activation were substantially elevated at the start of surgery in comparison with normal values. Although this could indicate ongoing platelet activation in these patients, an alternative explanation is an accumulation of these markers due to the inability of the diseased liver to clear these markers.19–21 After liver transplantation, plasma levels of all soluble markers decreased substantially, and this suggested that platelets were in a slightly activated state before liver transplantation. An alternative explanation could be that clearance of these markers by the liver is improved and normalizes after transplantation. Presumably both phenomena play a role, as a pre-activated state of platelets is evidenced by slightly enhanced expression of P-selectin and activated αIIbβ3 in samples taken at the start of surgery in comparison with samples taken after transplantation. This pre-activated state of platelets in patients with liver disease has been reported previously.22–24 Although we have not measured platelet functional capacity in this study, we have previously shown intrinsic platelet activity in patients with cirrhosis to be indistinguishable from that of healthy controls.12 However, a recent study in which platelets were activated ex vivo showed a substantial defect in the activatory capacity of platelet samples taken before and during liver transplantation, and further study is required to resolve the discrepancies between this study and our studies.25
Although the evolution of platelet function during liver transplantation was poorly documented in the past, 2 well-recognized previous observations had suggested that platelet function deteriorates significantly during liver transplantation, especially after graft reperfusion.15, 17 First, liver transplantation has been associated with activation of the fibrinolytic system, which is mainly caused by an increase in plasma levels of tPA during the anhepatic phase and after graft reperfusion. This release of tPA may potentially result in the generation of large amounts of plasmin, contributing to the proteolytic cleavage of platelet receptors.17 Second, it has been demonstrated that platelets adhere to and become activated by the ischemically damaged graft.17 Previous studies in a porcine liver transplant model have shown that liver transplantation is associated with substantial activation and consumption of blood platelets.26 In addition, a progressive increase in plasma levels of the platelet activation markers β-thromboglobulin and platelet factor 4 has been reported previously in human liver transplantation. These historical findings are in sharp contrast with the current study, which used more sophisticated laboratory methods to demonstrate platelet activation and dysfunction. In the current study, we observed a rather stable and well-preserved function of platelets during and after liver transplant surgery. Especially after reperfusion of the graft, we found no evidence of substantial platelet activation or proteolysis of platelet receptors.
An explanation for the current findings could be improvements in organ preservation in the last decade as well as the policy in recent years of keeping the cold ischemia times as short as possible. This may have resulted in a reduction in activation of the liver endothelium in comparison with earlier time periods, and this could explain why in current clinical practice there is no evidence of substantial platelet activation or dysfunction in patients undergoing liver transplantation.
Recent reports from several centers indicate that a substantial proportion of patients (up to 50%, depending on the center) can undergo liver transplantation without the need for any platelet transfusion, despite low preoperative platelet counts,18, 27 and this supports our finding that platelet function during liver transplantation appears preserved. A very restrictive use of blood products, especially of platelet concentrates, would probably never be possible if platelet function or other components of hemostasis were severely affected. Furthermore, it was recently shown that transfusion of platelet concentrates during liver transplantation is associated with increased mortality due to an increased incidence of acute respiratory distress syndrome.28, 29 In the basis of the results of the present study and the available clinical evidence, we advocate restrictive use of blood products in these patients.
One may argue that although there is no evidence for a substantial activation of platelets and a subsequent decrease of platelet function during liver transplantation, most patients requiring liver transplantation have low platelet counts to start with, and this could still affect primary hemostatic function, justifying the transfusion of platelets. In this context, it is, however, important to note that our recent study has indicated that thrombocytopenia in patients with cirrhosis is (at least partially) compensated for by high plasma levels of the platelet adhesive protein VWF.5 In addition, we have recently shown that the functional capacity of VWF increases significantly during liver transplantation, and this provides an additional compensatory mechanism for the reduced number of circulating platelets.30
In conclusion, we found no evidence for platelet activation or proteolysis of key platelet receptors in patients undergoing liver transplantation. These findings are in sharp contrast to earlier studies using less specific laboratory methods, which suggested that platelet function deteriorates during liver transplantation. Altogether, these findings suggest that platelet function may be well preserved during liver transplantation; therefore, there might be a reduction in the indication for prophylactic transfusion of platelet concentrates before and during the procedure to improve primary hemostasis. Although different studies have provided evidence of serious detrimental effects of platelet transfusions on outcome after liver transplantation,18, 28, 29 we believe that prophylactic platelet transfusions should be avoided as much as possible and be given therapeutically only in case of excessive bleeding.
The authors acknowledge Ans A. M. Hagenaars for her assistance with the data collection.