Aprotinin Inhibits Plasmin-Induced Platelet Activation During Cardiopulmonary Bypass
Background In the past few years, aprotinin has been used in cardiac surgery with impressive results of reducing blood loss, but several adverse effects of aprotinin also have been reported. One of the most likely mechanisms is the inhibition of plasmin by aprotinin, although this indirect effect has not been reproduced in all experimental studies.
Methods and Results We evaluated the platelet function and fibrinolytic activity during human cardiac surgery, with or without aprotinin. During cardiopulmonary bypass (CPB) in humans without aprotinin (n=16), decrease of platelet aggregation induced by thrombin, increase of α-granule secretion of platelet and microparticle formation, and increase of plasmin/α2-antiplasmin complex (PIC) were observed. In contrast, low-dose aprotinin (1.0×106 KIU), which was administered only into the priming fluid of extracorporeal circuits (n=10), maintained platelet aggregation induced by thrombin and reduced α-granule secretion and microparticle formation of platelets during CPB. In vitro, plasmin (0.8 CU/mL) released α-granules of washed platelets, and this activation was completely inhibited by aprotinin (10 KIU/mL).
Conclusions Aprotinin has indirect effects to inhibit platelet activation, and this may partly explain the reduction of blood loss during cardiac surgery. To prevent the adverse effects, a single and minimal use of aprotinin is important. The results of in vivo and in vitro studies suggest that platelet preservation was demonstrated by the lower concentration of aprotinin (1.0×106 KIU per patient or 10 KIU/mL) compared with the concentration that inhibits plasma fibrinolysis.
Bleeding is one of the major complications after CPB, and extensive homologous transfusion may transmit viral hepatitis and acquired immunodeficiency syndrome. To reduce blood loss after cardiac surgery, platelet-protective drugs were investigated in many clinical centers.1 2 3 4 5 In the past few years, aprotinin has been used in cardiac surgery and impressive result were obtained,6 7 8 9 10 but several adverse effects of aprotinin have been reported including disseminated intravascular coagulation, acute anaphylaxis, and acute renal failure.11
Aprotinin (Trasylol; Bayer AG) is a low-molecular-weight (6512 Dalton) peptide isolated from bovine lung containing 16 different amino acids in a chain of 58 members, forming a kringle domain shape with three disulfide bonds (cysteine-cysteine bridges), and inhibits enzymatic activity of trypsin, kallikrein, and plasmin.11 Approximately 50 KIU/mL of aprotinin is required to inhibit plasma plasmin and about 200 KIU/mL to inhibit plasma kallikrein.12 The plasma half-life of aprotinin is <1 hour.13 The drug was first used in cardiac surgery in the early 1960s when it was thought that excessive postoperative bleeding was due to increased fibrinolysis.11 Additionally, aprotinin in high doses completely inhibited kallikrein-induced activation of neutrophils.14
One of the most likely mechanisms of the reduction in blood loss in cardiac surgery is the inhibition of plasmin by aprotinin. However, this indirect effect has not been reproduced in all experimental studies and remains unconfirmed in the clinical state.
In this study we focused on platelet activation by plasmin and the inhibitory effect of aprotinin on platelet activation during clinical CPB and in in vitro experiments.
From September 1995 to September 1996, platelet functions were measured in 26 adult patients who did not receive any antiplatelet drugs within 10 days before surgery and underwent CPB without blood transfusions at the University of Tsukuba Hospital. Ten patients (aprotinin group) underwent cardiac surgery with aprotinin administration only into the oxygenator priming fluid at the beginning of CPB. Sixteen patients (control group) had cardiac surgery without aprotinin. One patient in the aprotinin group needed the oxygenator exchange because of fluid leakage. Other patients underwent the operation uneventfully. The mean age and weight of the control and aprotinin groups were almost the same (Table⇓). The duration of CPB ranged between 2 and 3 hours.
Conventional methods of anesthetic treatment and CPB were used in both groups. The anesthesia was induced with fentanyl citrate (10 μg/kg). Heparin (Novo Nordisk A/S) was administered at an initial dose of 3 mg/kg through venous catheters. Activated clotting time was maintained for >400 seconds throughout CPB. A membrane oxygenator (HPO-25RHF, MERA) and a roller pump were used in the extracorporeal circuit. Blood flow was maintained at 2.5 L/m2 per minute. Multidose cold crystalloid cardioplegic solution was used for myocardial preservation. Systemic hypothermia (30°C to 32°C) was maintained during aortic cross-clamping.
Five milliliters of blood was collected from patients at six separate times: at the start of the surgery; 5 minutes after heparin administration; 5 minutes, 1 hour, 2 hours, and 3 hours after the start of CPB; or at the end of CPB through an arterial blood pressure monitor catheter. Four milliliters of whole blood was mixed with 0.5 mL of ACD solution (citric acid 6.8 mmol/L, trisodium citrate 11.2 mmol/L, glucose 24 mmol/L) at room temperature. The rest of the whole blood (1 mL) was placed in a tube containing ethylenediaminetetraacetic acid (EDTA) for blood cell count. In some patients (control group, n=5; aprotinin group, n=2), 100 μL of plasma was stored at −20°C for analysis of PIC.
The blood, mixed with ACD, was centrifuged at 180g for 10 minutes at room temperature, and PRP was obtained. A 100-μL sample of PRP was fixed with 1% paraformaldehyde (E Merck) for flow cytometric analysis.
To wash the platelets or dilute the antibodies, PIPES buffer (PIPES 5 mmol/L, Dojindo; NaCl 145 mmol/L, KCl 4 mmol/L, Na2HPO4 0.5 mmol/L, MgCl2 1 mmol/L, glucose 5.5 nmol/L, and bovine albumin 3.5 mg/mL, pH 7.4) was prepared. The rest of the PRP was mixed with an equal volume of washing buffer (1:8, ACD solution:PIPES buffer) and then centrifuged at 400g for 15 minutes at room temperature. The platelet pellet was resuspended gently and washed again with the same solution. The pellet was finally suspended in PIPES buffer at the concentration of 2.0×108/mL for the assay of thrombin-induced platelet aggregation.
For in vitro studies, venous blood from healthy volunteers who had no medication for at least 10 days before donation was collected into plastic tubes containing ACD buffer. The washed platelet preparation was carried out by the same methods described in in vivo studies. The washed platelets were incubated with aprotinin or tranexamic acid (Transamin-S, Daiich Pharmacy Co) or plasminogen (Chromogenix AB) for 15 minutes at 37°C followed by incubation with human α-thrombin (Green Cross) for 5 minutes or plasmin (Chromogenix AB) for 15 minutes and then fixed with 1% paraformaldehyde.
The platelet aggregation study was performed with the use of Hema Tracer (SSR Engineering Co Ltd). The threshold concentration of thrombin (that is, the lowest concentration of the agonist capable of producing irreversible aggregation of at least 60% to 70% light transmission of buffer in 5 minutes) was determined with the sample of the start of operation. The same concentration of thrombin was used to determine the percent aggregation of washed platelets in all subsequent samples.
Expression of CD-62, CD42b, and CD41b by Flow Cytometric Analysis
Platelet surface antigen was stained with FITC-labeled or phycoerythrin-labeled mAb and analyzed by flow cytometry (FACSort; Becton Dickinson). The mAbs used in this study are as follows: CD62-PE (Becton Dickinson), a mAb that recognizes α-granule membrane protein (GMP-140) expressed on the platelet surface after platelet secretion (and we consider that this is a kind of platelet activation); CD42b (DACO A/S), a mAb against GPIb; and CD41b (TP80, Nichirei), a specific anti-GPIIb/IIIa antibody.
A 20-μL aliquot of each fixed platelet (PRP) sample was incubated with 5 μL of 1:5 diluted mAb (saturating concentration of the antibodies of DC62 or CD42b or CD41b) in the dark for 30 minutes at room temperature to allow antibody binding. The samples mixed with CD42b or CD41b were incubated again with the second antibody (FITC-conjugated goat anti-mouse IgG, DACO A/S) in the dark for 30 minutes at room temperature, and then 400 μL of ISOTON II (Coulter, Inc) solution was added to dilute the samples.
Platelet α-granule secretion was monitored by detection of GMP-140 (CD62) expression on the platelet surface. The threshold level for GMP-140–negative cells was set to include 98% of nonstained platelets. Microparticles were identified by gating on GPIIb/IIIa (CD42b)-positive events and distinguished from normal-sized platelets by forward scatter size analysis (the forward scatter cutoff was set to the immediate left of the signal intact platelet population of a resting PRP sample).15 Ten thousand positive platelet events were analyzed, and microparticles were reported as a percentage of total platelet events.
To evaluate the fibrinolytic activity during CPB, PIC was measured with the frozen plasma with the use of the PIC test kit (Teijin Diagnostics).
The results are expressed as mean±SE. The effects of aprotinin on platelet aggregation, GMP-140 expression, and microparticle formation at various times during CPB were compared by Student’s unpaired t test. A value of P<.05 was considered significant.
In Vivo Studies
In the control group, platelet counts corrected by hematocrit and the expression of mean platelet surface antigen (GPIb or GPIIb/IIIa) did not show significant differences during 2 to 3 hours of CPB (Fig 1⇓, A through C). However, the thrombin-induced platelet aggregation in the control group showed significant decrease during 2 hours of CPB. In addition, the percentage of GMP-140–positive platelets increased 5 minutes after the start of CPB, and microparticles also increased throughout CPB (Fig 1⇓, D through F).
Aprotinin (1.0×106 KIU) administered at the start of CPB did not affect the platelet count corrected by hematocrit or the expression of mean platelet surface antigen GPIb or GPIIb/IIIa for 2 to 3 hours of CPB. However, the thrombin-induced platelet aggregation was maintained at a high level during CPB and showed significant differences between the aprotinin group versus the control group 5 minutes after the start of CPB (65±4% versus 43±6%, P=.010), 1 hour after the start of CPB (63±7% versus 44±6%, P=.042), and after CPB (63±5% versus 43±8%, P=.046) (Fig 1D⇑). Aprotinin also attenuated α-granule secretion of platelets and showed significant differences between the two groups at 5 minutes after the start of CPB (20.2±2.6% versus 29.5±2.8%, P=.049) (Fig 1E⇑) and reduced microparticle formation of platelets at 1 hour after the start of CPB (5.6±0.6% versus 12.8±2.2%, P=.013) (Fig 1F⇑).
Fig 2⇓ shows PIC concentration during CPB. The decrease of PIC concentration at the start of CPB may be due to hemodilution. Since the activated plasmin concentration relates to the rate of change of PIC, the increase of PIC during CPB in the control group indicates that the activation of plasminogen occurred constantly after the start of CPB. In contrast, the concentration of PIC in the aprotinin group was kept at a lower value during the first hour of CPB.
In Vitro Studies
In the study of washed platelets incubated with thrombin, the percentage of GMP-140–positive platelets increased in a concentration-dependent manner. A full activation (≈80% to 90% of platelets express GMP-140 antigen on the surfaces) was obtained by thrombin at the concentration of 0.1 unit/mL. With the same platelets, aprotinin (200 units/mL or 400 units/mL), incubated for 15 minutes in 37°C before thrombin stimulation, did not show any effects on thrombin-induced GMP-140 expression on the surfaces of platelets (Fig 3⇓).
The washed platelets, incubated with plasmin for 15 minutes at 37°C, also showed the increase of percentage of GMP-140–positive platelets (Fig 4⇓). The full secretion of α-granules of the platelets was obtained by incubation with plasmin at the concentration of >1.0 CU/mL. In the lower concentration of plasmin incubation (0.8 CU/mL), there were several minutes of nonresponse periods at the beginning of the incubation, and the maximum secretion was obtained after 15 minutes of incubation at 37°C (Fig 5⇓).
Fig 6⇓ shows that aprotinin inhibited the α-granule secretion induced by plasmin. In this experiment, washed platelets were incubated with aprotinin for 15 minutes at 37°C followed by incubation with plasmin (0.8 CU/mL) for 15 minutes at 37°C. The complete inhibition of the α-granule secretion was obtained by 10 KIU/mL of aprotinin. Tranexamic acid and plasminogen also reduced the GMP-140 expression on the surfaces of washed platelets induced by plasmin in a concentration-dependent manner (Fig 7⇓).
Platelet analysis by flow cytometry and thrombin-induced platelet aggregation (platelet count adjusted) have an advantage that the data have no need to be corrected by hemodilution. The value of the percentage of GMP-140–positive platelets and microparticles in the samples at the start of surgery was relatively higher than those in healthy donors (GMP-140, 4.5±2.5%; microparticles, 3.6±1.4%; n=10). This may be due to the time loss between the collection of samples and the platelet fixation by 1% paraformaldehyde.
Van Oeveren and associates16 demonstrated the decrease of GPIb (receptor to von Willebrand factor) by 50% in the untreated patients during CPB, whereas GPIb did not decrease in the aprotinin-treated patients (6.0×106 KIU or 2.0×106 KIU per patient).17 To the contrary, Winters and associates18 reported that during pharmacological fibrinolysis, the inhibition of platelet function in plasma was not due to degradation of platelet adhesive receptors. In our study, GPIb and GPIIb/IIIa receptors on the surface of platelets were not changed significantly before and during bypass. The increase of microparticle formation may explain the decrease of GPIb or GPIIb/IIIa receptors per platelet during CPB, even though the concentration of the receptors on the platelet surface was not changed.
Wachtfogel et al14 suggested that the protective effect of aprotinin on platelets is probably indirect and is due to inhibition of various platelet agonists that are produced during CPB, but they did not show the effect of plasmin to platelets because endothelial cells are not present and plasmin is not generated in their simulated extracorporeal circulation model.
During CPB, many factors including thrombin, ADP, collagen, epinephrine, temperature, share stress, and contact with a synthetic surface may activate the platelets. To evaluate the α-granule secretion by thrombin in vitro, platelets were washed to prevent coagulation and incubated with thrombin. The percentage of GMP-140–positive platelets was increased by thrombin, although aprotinin had no effect on the activation of platelets by thrombin (Fig 3⇑). ADP also increased GMP-140 expression on the surfaces of platelets in PRP, and aprotinin did not show any effect (data not shown).
At this point, we tested the effect of plasmin on platelets because aprotinin is one of the plasmin inhibitors. As shown in Figs 4⇑ and 5⇑, plasmin enhanced the α-granule secretion of washed platelets but required longer times to get the maximum secretion of platelets than did thrombin or ADP.19 Mean fluorescence intensity of GPIb-FITC and GPIIb/IIIa-FITC on the surface of washed platelets incubated with plasmin did not show significant differences in in vitro study. The enhancement of α-granule secretion induced by plasmin was not detected in PRP, which contains a large amount of α2-plasmin inhibitors. Washed platelet aggregation induced by plasmin (0.8 CU/mL) was observed at 15 to 20 minutes of incubation at 37°C, and Ca2+ (3 mmol/L) attenuated this aggregation (data not shown). Winters et al18 and others also reported that platelet adhesive receptors (GPIIb/IIIa, GPIb, GPIa) were degraded by plasmin at 37°C in the absence of exogenous Ca2+. The mechanism of these results was not clear, but platelet activation by plasmin has a time lag in response and depends on Ca2+ and temperature.20 21
Aprotinin inhibits the catalytic activity of plasmin at the dose of 50 KIU/mL, and this was one of the reasons that a high dose (6×106 KIU per patient) of aprotinin was administered in patients undergoing cardiac surgery.11 We used 1.0×106 KIU of aprotinin in patients, and the platelets were preserved during 2 to 3 hours of CPB. Platelet functions were not measured after 3 hours of CPB because blood transfusion was done in most cases. The additional aprotinin might be required to protect platelets after 2 hours of CPB. In the in vitro study, 10 KIU/mL of aprotinin completely inhibited the enhancement of the washed platelet α-granule secretion induced by plasmin (Fig 6⇑). Tranexamic acid binds to lysin binding sites and inhibits the binding of plasmin to fibrin.22 23 The activation of washed platelets by plasmin was completely inhibited by tranexamic acid at a concentration of 2.5 mg/mL (Fig 7⇑). Plasminogen also inhibited platelet activation by plasmin, but the complete inhibition was not achieved even with plasminogen, with a concentration 4 times higher than plasmin (Fig 7⇑). These indicate that both the catalytic center and the lysin binding sites of plasmin are required to release α-granules of washed platelets, and plasminogen showed competitive binding to the surface of platelets with plasmin. If plasmin binds to the surface of platelets, and the triplets of plasmin, plasminogen activator, and platelets (or platelet fibrin) are formed,24 plasmin on the surface of platelets may be resistant to the inactivation by α2-antiplasmin in plasma, and lower concentrations of plasmin may activate the platelet.
Platelet activation by plasmin is one of the serious problems in the treatment of patients who have received thrombolysis therapy because tissue plasminogen activator might cause another thromboembolism after the recanalization of blood vessels.25 26 27 Identification of plasmin receptors on the surface of platelets and control of platelet activation by plasmin are important subjects.
The conversion of plasminogen to plasmin was caused during CPB, and aprotinin protected platelets during CPB in vivo and inhibited the washed platelet α-granule secretion induced by plasmin in in vitro studies. This suggests that aprotinin inhibits plasmin-induced platelet activation, and this may partly explain the reduction of blood loss during cardiac surgery.
Selected Abbreviations and Acronyms
|KIU||=||kallikrein inactivator units|
|mAb||=||monoclonal antibody (antibodies)|
- Received October 31, 1996.
- Revision received December 17, 1996.
- Accepted January 2, 1997.
- Copyright © 1997 by American Heart Association
Addonizio VP Jr, Strauss JF III, Colman RW, Edmunds LH Jr. Effect of prostaglandin E1 on platelet loss during in vitro cardiopulmonary bypass. Circ Res. 1979;44:350-357.
Shigeta O, Gluszko P, Downing SW, Lu W, Niewiarowski S, Edmunds LH Jr. Protection of platelets during long-term extracorporeal membrane oxygenation in sheep with a single dose of a disintegrin. Circulation. 1992;86(suppl II):II-398-II-404.
Kaller H, Patzschke K, Wegner LA, Horster FA. Pharmacokinetic observations following intravenous administration of radioactive labelled aprotinin in volunteers. Eur J Drug Metab Pharmacokinet. 1978;2:79-85.
Winters KJ, Eisenberg PR, Jaffe AS, Santoro SA. Dependence of plasmin-mediated degradation of platelet adhesive receptors on temperature and Ca+. Blood. 1990;76:1546-1557.
Schafer AI, Maas AK, Ware JA, Johnson PC, Rittenhouse SE, Salzman EW. Platelet protein phosphorylation, elevation of cytosolic calcium, and inositol phospholipid breakdown in platelet activation induced by plasmin. J Clin Invest. 1986;78:73-79.
Gouin I, Lecompte T, Morel MC, Lebrazi J, Modderman PW, Kaplan C, Samama MM. In vitro effect of plasmin on human platelet function in plasma: inhibition of aggregation caused by fibrinogenolysis. Circulation. 1992;85:935-941.
Lu H, Soria C, Cramer EM, Soria J, Maclouf J, Perrot JY, Li H, Commin PL, Schumann F, Regnier O, Lijnen RH, Caen JP. Temperature dependence of plasmin-induced activation or inhibition of human platelets. Blood. 1991;77:996-1005.
Plow EF, Herren T, Redlitz A, Miles LA, Jane L, Hoover-Plow JL. The cell biology of the plasminogen system. FASEB J. 1995;9:939-945.
Montrucchio G, Bergerone S, Bussolino F, Alto G, Sylvester L, Lucia E, Cravat A, Cravat M, Leo MD, Emanuelli G, Camussi G. Streptokinase induces intravascular release of platelet-activating factor on patients with acute myocardial infarction and stimulates its synthesis by cultured human endothelial cells. Circulation. 1993;88:1476-1483.