- Email: [email protected]
Tissue-type plasminogen activator and fibrin monomers synergistically cause platelet dysfunction during retransfusion of shed blood after cardiopulmonary bypass
Tissue-type plasminogen activator and fibrin monomers synergistically cause platelet dysfunction during retransfusion of shed blood after cardiopulmonary bypass Reduced hemostasis and bleeding tendency after cardiopulmonary bypass results from platelet dysfunction induced by the bypass procedure. The causes of this acquired platelet dysfunction are still subject to discussion, although, recently, greater emphasis has been placed on an overstimulated fibrinolytic system as a probable cause. In the first part of this study we assessed the effects of postoperative retransfusion of shed blood on blood loss to patients undergoing cardiopulmonary bypass. We observed that increasing concentrations of fibrinogen degradation products and tissue-type plasminogen activator stimulating activity in shed blood correlated significantly with a higher postoperative bleeding tendency (p < 0.05 for both). We further noted that retransfusion of shed blood increased the total postoperative blood loss by 43 % (925 versus 1320 ml, p < 0.05). On the basis of these clinical observations, we hypothesized that the increased bleeding tendency was caused by fibrinolysis. In the second part of this study we collected evidence in support of this hypothesis by an in vitro study, in which we introduced similar (pro)fibrinolytic activity to platelet-rich plasma and measured the influence of this treatment on platelet function indicated by ristocetin agglutination. Tissue-type plasminogen activator and fibrin monomers (tissue-type plasminogen activator stimulator) together induced severe platelet damage, resulting in a decreased ristocetin agglutination response. Therefore, we propose a fibrinolysis-related mechanism for platelet dysfunction during cardiopulmonary bypass, dependent on fibrinolytic factors such as fibrin monomers, D-dimers, and tissue-type plasminogen activator. (J THORAC CARDIOVASC SURG 1993;106:1017-23)
1. de Haan, MSc,a J. Schonberger, MD, PhD,b J. Haan, BSc,a W. van Oeveren, PhD,a and A. Eijgelaar, MD, Phl)," Groningen and Eindhoven, The Netherlands
Rtoperative blood loss after operations involving cardiopulmonary bypass CCPB) ranges from several hundred milliliters up to a few liters on occasion. Such blood losscreates a demand for homologous blood products. I, 2 From the Department of Cardiopulmonary Surgery, Research Division," University Hospital Groningen, the Department of Cardiopulmonary Surgery," Catharina Hospital, Eindhoven, The Netherlands, and the Department of Cardiopulmonary Surgery," University Hospital Groningen, Groningen, The Netherlands. Received for publication Aug. 5, 1992. Accepted for publication Feb. 23, 1993. Address for reprints: W. van Oeveren, PhD, Cardiopulmonary Surgery Research Division,University Hospital Groningen, Oostersingel59, 9713 EZ Groningen, The Netherlands. Copyright
1993 by Mosby-Year Book, Inc.
0022-5223/93 $1.00 + .10
12/1/47001
Retransfusion of shed blood after CPB is an established method of reducing the requirements for such blood products.' However, because the composition of shed blood is far from normal, the benefits from retransfusion are questionable. Shed blood contains increased concentrations of fibrinogen degradation products, high levels of plasma hemoglobin, and other particles described as debris. Moreover, shed blood is largely defibrinated and contains a considerably lower number of platelets than the circulating blood. Therefore, retransfusion of especially large amounts of shed blood is not favorable for proper hemostasis. In fact, blood loss was found to be particularly high among patients in whom retransfusion otherwise would have been the most beneficial." In accordance with an earlier study, we observed that retransfusion of large amounts of shed blood after internal mam101 7
1 01 8
de Haan et ai.
mary artery bypass grafting caused a greater total blood loss than that observed in patients not receiving shed blood, reducing the efficacy of retransfusion. In explanation, we suggested that activated systems, such as the kinin, clotting, and fibrinolytic systems, played an important role.' In the first part of this study we analyzed the changes in fibrinolytic parameters of patients undergoing uncomplicated coronary artery bypass grafting, either with or without postoperative retransfusion of shed blood. In the second part of this study we examined the effect of fibrinolytic components on platelet function demonstrated by the platelet response to activated von Willebrand factor.
Patients and methods Patients. We prospectively studied 19 patients who underwent primary coronary artery bypass grafting with either one or two internal mammary arteries. When appropriate, additional saphenous vein bypass grafting was performed. The study was approved by the medical-ethical committee of the hospital and informed consent was obtained from all patients. Patients were excluded if they required intraaortic balloon pumping or in case of severe kidney and liver function disturbances, as reflected before the operation by increased concentrations of serum creatinine (> 100 mmol/L) or an increased prothrombin time. Patients were randomly assigned to one of two groups: Group I (n = 9) patients received postoperative retransfusion of shed blood by means of a hard-shell cardiotomy reservoir (D742, Dideco S.pA, Mirandello, Italy). Group 2 (n = 10) patients received no retransfusion. In all patients, anticoagulants and platelet inhibitors were discontinued at least 5 days before the operation. Operative management. The anesthetic and surgical management of the two groups of patients was identical. Bovinelung heparin (250 IU /kg body weight) was given intravenously 5 minutes before cannulation. An additional bolus of heparin (5000 IU) was administered to the extracorporeal circuit (ECC) every half hour. The ECC consisted of a hollow-fiber oxygenator (Sarns Inc. 13M, Ann Arbor, Mich.), a cardiotomy reservoir, and an arterial line screen filter (Cobe Laboratories Inc., Lakewood, Colo.). The ECC was primed with 1500 ml urea-linked gelatin (Haemaccel, Hoechst, Behringwerke AG, Marburg, Germany) containing 5000 IU bovine lung heparin and 100 ml mannitol. Neither in the pump prime nor during the operation was any fibrinolytic inhibitor such as aprotinin or tranexamic acid used. The perfusion flow rate at moderate hypothermia of 28 0 to 32 0 C was 2.5 Lyrnin. After aortic crossclamping, cardioplegia was induced with 1000 ml of cold (4 0 C) St. Thomas' Hospital solution, and the infusion was repeated with 500 ml of solution every half hour. For external cardiac cooling about 150 ml of cold (4 0 C) saline solution was applied. The mixture of blood and cooling fluid, collected in the pericardium and pleural space, was returned to the ECC. When the hematocrit values fell below 20% during CPB, in-line hemofiltration (Fresenius BV, Den Bosch, The Netherlands) was started. At the end of the perfusion, heparin was neutralized by protamine sulfate at a dosage of one times the total dosage of heparin. After completion of the extracorporeal bypass procedure, the patients received the contents of the circuit until
The Journal of Thoracic and Cardiovascular Surgery December 1993
the left atrial filling pressure was optimal, according to the prebypass diastolic pulmonary artery pressure. The residual volume in the ECC after aortic decannulation was collected in blood transfusion bags and retransfused gradually in the intensive care unit. Blood loss and management of retransfusion. In the patients of group 1 (retransfused), the mediastinal and pleural chest tubes were connected to a polyvinylchloride tube and attached to the inlet port of the cardiotomy reservoir as soon as protamine infusion was completed. The cardiotomy reservoirof the ECC was then used as an autotransfusion system. Suction with a negative pressure of 15 to 20 ern H 20 was applied at the vacuum port of the cardiotomy reservoir. A pump with a built-in bubble and clot detector (Imed Inc., Oxford, United Kingdom) was attached to the lower outlet port of the reservoir. During a maximum period of 8 hours, collected blood was retransfused to the patient via an Imed pump. The infusion rate of the Imed pump was adjusted each hour, corresponding to the blood levelin the cardiotomy reservoir, permitting a minimal flow of 50 ml/hr, In group 2 (control, receivingno retransfused shed blood), the shed blood was collected in a Pleur-evac system (Deknatel GmbH, Neustadt, Germany). The chest tubes were removed 18 to 24 hours after the operation. Biochemical and hematologic methods. During and after the operation, blood samples were collected from each patient at nine specific intervals: 25 minutes before CPB (after anesthesia); 5 minutes before CPB (after heparinization); after 30 minutes of CPB; just after release of the aortic crossclamp; 15 minutes after protamine sulfate administration; and at 0, 2, 4, and 8 hours on arrival in the intensive care unit. Additional samples were collected from the retransfusion reservoir 15 minutes after protamine sulfate administration and at 0, 2, 4, and 8 hours after arrival in the intensive care unit. Blood sampling. Three milliliters of blood was collected in ethylenediaminetetraacetic acid (final concentration 0.01 moll L) to assess red blood cell, platelet, and leukocyte counts. Three milliliters of blood was collected in aprotinin-hirudincitrate medium (final concentrations of 100 KIU, 0.05 U, and 0.308%, respectively). These blood samples were then centrifuged at 1000 g for 10 minutes to obtain platelet-poor plasma. Platelet-poor plasma was stored at -80 0 C until further processing for the determination of fibrinogen degradation products by enzyme-linked immunosorbent assay (Organon Teknika, Turnhout, Belgium). Two milliliters of blood were collected in citrate medium (0.308% sodium citrate, final concentration) and centrifuged at 1000 g for 10 minutes to obtain platelet-poor plasma for determination of tissue-type plasminogen activator (t-PA) antigen (enzyme-linked immunosorbent assay, Kabi Diagnostica, Stockholm, Sweden). In addition, the potential for t-PA stimulation was determined. In this assay, purified t-PA, plasminogen, plasmin-substrate S2403 (Kabi Diagnostics, Stockholm, Sweden), and a twenty-fold diluted plasma sample were incubated for 60 minutes at 37 0 C. The plasma concentration of t-PA stimulators, such as fibrin monomers, determines the activation rate of topA in converting plasminogen into the substratecleaving enzyme plasmin. The yellowcolor generated was measured at 405 nm in a spectrophotometer (EAR 400, SLTLabinstruments, Salzburg, Austria). In vitro study. In accordance with the declaration of Helsinki, whole blood was obtained from five healthy volunteers who
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6
had not received aspirinwithin 14daysbeforethe donation. The blood was collected in neutralized acid-citrate-dextrose medium. The platelet-rich plasma was separated from the remaining blood by centrifugation (100 g, 15 minutes)by the removal of the platelet-containing upper layer.Thereafter, the platelets wereisolated bygelfiltration ofthe platelet-rich plasmathrough a Sepharose 2B column (Pharmacia-LKB, Woerden, The Netherlands) by elution. They werecollected with a phosphate buffer, 50 mmol/L, pH 7.4,containing citricacid, 36 mrnol/L, potassium chloride, 5 mmol/L, sodiumchloride, 103 mmol/L, glucose 5 mmol/L, bovine serum albumin Y, 3.5 mg/rnl (Sigma Chemical Co., St. Louis, Mo.). The remaining blood was again centrifuged (1000 g, 10 minutes) to obtain platelet-poor plasma and red blood cell concentrate. The platelet-poor plasma was passed through a 0.22 ILm filter to obtain platelet free plasma. Plasminogen activatorinhibitorwasinhibited to ensure standardized and functional concentrations oft-PA in the tests. Theplatelet-free plasma,therefore, wasacidified byadding 10% volume of sodiumphosphate, 1.25 mol/L (pH 3.9), incubated for I minute, and neutralized afterward by adding sodium phosphate, 2.5 mol/L (pH 8.3),to a final pH of7.4. For further testing, gel-fixed platelets were mixed with this pretreated platelet-free plasmaand exposed to the fibrinolytic components. Introduction of t-P A and fibrin monomers to gel-filtered platelet/platelet-free plasma mixtures. Platelets in the gelfiltered platelet/platelet-free plasma (80:20%) mixtures were exposed for a period of 10 minutesat 37° C to various concentrations of fibrin monomers (t-PA stimulator, Kabi) ranging from 0 to 20 ILg/ml and/or t-PA (Kabi) ranging from 0 to 40 IV/m\' Ristocetin agglutination response. Platelet response to ristocetin was determined in 500 ILl reconstituted platelet-rich plasma (10% normal plasma was added to the gel-filtered platelet/platelet-free plasma mixtures) by measuring the changein optical densityof the platelet-rich plasmacausedby the von Willebrand factor-mediated platelet agglutination triggered with ristocetin (0.6 mg/rnl final concentration). Alternately, heparinwasadded (2 V /rnl) to assess its influence on the ristocetin response (Chrono-Log, Chrono-Log Corp., Havertown, Pa.). The response was quantified by both the steepest angle (a) measured during the agglutination process and the final height obtained after the agglutination process (H). Finally, the response wasexpressed as the percentage ofthe baseline ristocetin response as observed for an untreatedcontrol sample: Ristocetin response = (a X H)treated/(O' X H)untreated X 100% Statistics. All values are expressed as mean and standard error of the mean.Statisticalanalysis was accomplished either by the unpaired Student's t test if the two study groups were compared or bythe pairedStudent's t test if changeswithin one studygroup were analyzed. A p value less than 0.05 was considered to be significant. Results Operation and patients. The patient data concerning crossclamp time, CPB time, and age did not differ significantly between the two groups (Table I). Total and net blood loss. The total postoperative blood loss in the retransfusion group (1320 ± 162 ml, range
de Haan et al.
10 19
Table I. Demographic data and clinical data during CPB of both study groups Control (n = 10) Age (yr) Perfusion time (min) Crossclamp time (min) Total blood loss (rnl) Net blood loss (m!) Hemoglobin, before retransfusion (mmoljL) Hemoglobin, after 4 hr retransfusion (mmoljL)
58 81 49 925 925 5.9
± ± ± ± ± ±
Retransfused (n= 9) Significance
3.5 53 ± 2.7 11.5 84 ± 12.8 7.5 47 ± 6.0 78 1320 ± 162 78 568 ± 94 0.3 5.8 ± 0.4
6.3 ± 0.2
6.8 ± 0.5
NS NS NS
p<0.05 p
NS
values are expressed as means and standard error ofthe mean. Significance is calculated according to the unpaired Student's t test. NS, Not significant.
All
730 to 2110 ml) proved to be significantly higher (p < 0.05) than the blood loss in the control group (925 ± 78 ml, range 650 to 1370 ml) (Fig. 1). The net blood loss, comprising the total loss minus the retransfused volume, was significantly lower in the retransfusion group (589 ± 89 ml, p < 0.05) than in the control group. No significant reduction of blood product transfusion was obtained by retransfusion (0.8 ± 0.4 red blood cells, 0.1 ± 0.1 fresh frozen plasma) if compared with that of the control group (0.8 ± 0.3 red blood cells, 0.2 ± 0.2 fresh frozen plasma). Hematology. Blood cell counts (white blood cells, platelets, red blood cells) and hematocrit values during the CPB procedure and intensive care period did not differ significantly between the two groups (data not shown). Also, hemoglobin concentrations during and after the CPB period were similar in the two groups, showing only a minor and not statistically significant higher concentration after retransfusion (Table I). Fibrinogen degradation products. In both groups, the concentrations of fibrinogen degradation products in circulating blood rose immediately at the start of CPB, increasing to the greatest degree after unclamping of the aorta, to four times the baseline values (Fig. 2). After reaching the highest values between protamine infusion and admission of the patient to the intensive care unit, concentrations of fibrinogen degradation products in the control group decreased significantly (p < 0.05) to baseline values within 10 hours after the start of CPB. In contrast to measurements in the control group, however, concentrations of fibrinogen degradation products in the retransfusion group increased significantly once more (p < 0.05) soon after the retransfusion of the shed blood was started; concentrations reached a maximum within 4 to 6 hours after the start of CPB.
1020
The Journal of Thoracic and Cardiovascular Surgery
de Haan et al.
December 1993
1500 30
20
1000
...·r
.. .' L....1'· '
10
500
......··r........·········1
relransfusion
l2!J----=..:...~-----;
o
Fig. 1. Postoperative blood loss of retransfused group of patients (dark) compared with control group (light). Including the retransfused amount of blood in the total bloodloss shows a significantly higher bloodlossin the retransfusion group than in the controlgroup. The net bloodlosswassignificantly higher in the controlgroup than in the retransfusion group. (*p < 0.05 by Student's unpaired t test.)
5000
•
4000
~
4
6
8
10
12
houl'8
o
I
2
3000
2000
1000
0
l2!J -2
0
2
relransfusion 4
6
8
10
12
houl'8
Fig. 2. Fibrinogen degradation products (FgDP) during and after CPB with (closed squares) or without retransfusion (open circles). The dashed line represents the FgDP levels of the retransfusedbloodin the reservoir. After CPB, statisticaIIy significant differences were observed between the control and retransfusion groups. (*p < 0.05 by Student's unpaired t test.) t-PA Antigen and t-PA stimulating potential. Circulating concentrations of t-PA antigen were similar in the two groups of patients, approximating 10 ng/rnl. The concentrations of t-PA antigen increased at the onset of CPB, reaching maximum values of 25 ng/rnl after unclamping of the aorta, and remained elevated during the first 8 postoperative hours (Fig. 3).
Fig. 3. t-PA Antigen during and after CPS with (closed squares) or withoutretransfusion (open circles). The dashed line representsthe t-PA level of the retransfusedbloodin the reserVOIr. The t-PA stimulating potential in both groups reached its highest value even before the onset of CPB, remained high during CPB, and returned to preoperative values between unclamping of the aorta and 4 hours after the start of CPB (Fig. 4). During the latter period, however, t-PA stimulating potential in the patients who underwent retransfusion was higher than in the control group. Furthermore, t-PA stimulating potential in the shed blood increased significantly and reached a maximum 6 hours after the start of CPB. Ristocetin agglutination response. Compared with untreated reconstituted platelet-rich plasma, the ristocetin agglutination response of platelet-rich plasma treated with heparin, fibrin monomers, or t-PA decreased slightly, but to a statistically significant degree (p < 0.05), to values of70% (Fig. 5). However, with the introduction of both fibrin monomers and t-PA, the ristocetin agglutination response decreased to 25%, a highly significant change (p < 0.01). Discussion Retransfusion of a large amount of shed blood after CPB increases the total blood loss significantly. We have hypothesized that this reduced hemostasis is related to the high level of fibrinolytic stimulation in the retransfused shed blood, a position supported by the correlation of postoperative concentrations of fibrinogen degradation products and the total postoperative blood loss in all patients exceeding average blood loss (Fig. 6). In patients with moderate or low blood loss this correlation did not exist, indicating that in these patients bleeding is mainly caused by other factors. This increased concentration of circulating fibrinogen degradation products, observed in
The Journal of Thoracic and
de Haan et al.
Cardiovascular Surgery
10 2 1
Volume 106, Number 6
400
./1"
300
.·····1·····'.
.. •
200
100
../ *
-,'.
*
..••
I *
100
o
[£!]f-----===:::.:....------l o
6
2
8
10
75
.5
I!
•e• o •Go f
r8lransfusion
-2
..z3
50
25
12
hou,.
Fig. 4. Level of t-PA stimulation during and after CPB with (closed squares) or without retransfusion (open circles). The dashed line represents the level of t-PA stimulation in the retransfused blood, increasing significantly to maximum levels between 4 and 6 hours after the start of CPB. (*P < 0.05 comparedwithend-CPBvalue,calculatedbypairedStudent's t test)
the patients during retransfusion, is tenfold the retransfused amount of fibrinogen degradation products in the shed blood. Hence another mechanism responsible for increased concentrations of fibrinogen degradation products in the blood of retransfused patients must exist, mediated by renewed fibrinolytic activity and resulting in a reduced hemostasis. In general, fibrinolysis during clinical interventions is dependent mainly on plasminogen activation by t-PA6; indeed, the concentration of t-PA observed in the retransfusion group was high. However, t-PA concentrations in the control group were similar; therefore, another component in the shed blood, increasing the fibrinolytic activity during retransfusion, was suspected. This other factor appeared to be the high potential for t-PA stimulation, as observed in the shed blood during retransfusion. Because the shed blood is defibrinated because of prior clotting activity in the pleural space and thoracic wound area," its t-PA stimulating potential is derived primarily from fibrin monomers. After binding, fibrin as well as fibrin monomers stimulate t-PA activity l500-fold as compared with free t-pA,8 However, these fibrin monomers could not be measured in the patients' plasma after infusion of the shed blood, despite the increased fibrinolytic activity indicated by increased levels of fibrinogen degradation products. These findings suggest that clotting products such as fibrin monomers can bind to platelets, similar to plasminogen? and fibrinogen degradation products.!" and are scavenged from plasma. If so, these fibrin monomers cause a high level of plasmin activity localized on the platelet surface. I I, 12 Moreover, this surface-bound
o
conlrol
heparin
FM
FM+l-PA
l-PA
Fig. 5. Ristocetin agglutination response of normal platelets after a challenge to various stimulators such as heparin (2 Ulml), fibrin monomers (FM, 20 f.Lg/ml), t-PA (40 IUlml) or a combination of fibrin monomers (20 f.Lg/ml) and t-PA (40 IVIml). Compared with control (100%), the introduction of heparin, fibrin monomers, or t-PA caused a significantly decreasedresponse. Introductionoft-PA as wellas fibrinmonomers caused a highly significant decreased response (*p < 0.05; **p
< 0.01).
2500--r---------------'7'"'"'"1
<:-
.§.
.. •
2000
~
'8 ~
•
•
lie
1500
.--
~_/
-I.··
»"• • • •
1000
• •
SOO +---.,...----r--~--____,.____-____i
o
2000
4000
6000
6000
10000
FgDP (ng/mll
Fig. 6. Total postoperative blood loss and concentrations of circulating fibrinogen degradation products (FgDP) after start of retransfusion and at 2, 4, and 8 hoursafter CPB in all patients exceeding mean blood loss of 925 ml were significantly correlated according simple regression statistics (r 2 = 0.52, p
< 0.01).
plasmin activity is uncontrolled because of its low accessibility to a2-antiplasmin.13-15 The plasmin activity can thus cause platelet damage and dysfunction, expressed primarily by a downregulation of glycoprotein IbjIX
1022
de Haan et al.
complexes'< 17 and affected glycoprotein Ib-von Willebrand factor interactions.!" Our hypothesis also offers an explanation for the observed platelet dysfunction during CPB, 19,20 because increased concentrations of t-PA and t-PA stimulator are observed at that time as well. We tested this hypothesis in the second (in vitro) part of this study with the introduction of t-PA and fibrin monomers to normal platelets and measured the effect of this treatment on ristocetin agglutination. The platelet response to ristocetin reflects the capacity of platelets to adhere to the subendothelium of a damaged blood vessel and is based on activation of von Willebrand factor. Synergistically, t-PA and fibrin monomers cause within 10 minutes a reduction in glycoprotein Ib-dependent platelet function of75%. The concentrations oft-PA and fibrin monomers used in vitro are similar to those measured in vivo. Others have shown t-PA-induced platelet dysfunction in plasma as well, but at concentrations of at least 400 ng/rnl, 12 whereas we used moderate concentrations up to 60 ng/rnl. Therefore, we consider these in vitro results illustrative of the process of hemostatic derangement in vivo, caused by moderate concentrations of t-PA and fibrin monomers, resulting in generation of plasmin on the surface of platelets. This model explains several observations: I. Within the first postoperative hours, hemostasis recovers. Massive replenishment of damaged platelets by new platelets is not anticipated within this short period, and therefore recovery of the affected platelets is most likely. Michelson and Barnard'" show the feasibility of this recovery after plasmin treatment and the subsequent introduction of aprotinin. 2. A longer CPB time causes a higher postoperative blood loss. Lu and associates" showed that plasminrelated platelet inhibition is enhanced by hypothermia, and others demonstrated the (moderate) digestion of glycoprotein Ib proteins by plasmin. Thus a long hypothermic CPB period will cause more (irreversible) platelet damage and postoperative blood loss. 3. Because all effects described are related to t-PAdependent plasmin generation and action, the blood-saving effects of fibrinolytic inhibitors such as aprotinin, tranexamic acid, and e-aminocaproic acid can be explained. They all are able, in contrast to the physiologic inhibitor £x2-antiplasmin,13 to inhibit plasmin, generated by t-PA and plasminogen on fibrin. The moderate concentrations of t-PA and fibrin monomers indicate as well that it is difficult to prevent platelet damage during CPB. Therefore, blood from the wound area, comprising the main source for fibrin monomers, should be retransfused in limited quantities (Schonberger et aI., personal communication, 1992), not retransfused at all, or washed
The Journal of Thoracic and Cardiovascular Surgery December 1993
by use of cell-saving devices before retransfusion. During CPB, however, because the combined interaction of fibrin monomers and t-PA with platelets cannot be prevented in any direct way, platelet preservation can be achieved with the use of antifibrinolytic drugs such as aprotinin" or tranexamic acid. 22 Because both these drugs are presumed capable to inhibit surface-localized plasmin.P they are most effective for preservation of hemostasis,19,22,24 eliminating the need for postoperative retransfusion of shed blood. We express our gratitude to Jeffrey W. Kolff for reading the manuscript and improving the linguistics. REFERENCES I. Mammen EF, Koets MH, Washington BC, et al. Hemostasis changes during cardiopulmonary bypass surgery. Semin Thromb Hemost 1985;11:281-92. 2. Boonstra PW, van ImhoffGW, Eysman L, et al. Reduced platelet activation and improved hemostasis after controlled cardiotomy suction during clinical membrane oxygenator perfusions. J THORAC CARDIOVASC SURG 1985;89:900-6. 3. Cosgrove DM, Amiot DM, Meserko JJ. An improved technique for autotransfusion of shed mediastinal blood. Ann Thorac Surg 1985;40:519-20. 4. Adan A, Brutel de la Riviere A, Haas F, van Zalk A, de Nooy E. Autotransfusion of drainage mediastinal blood after cardiac surgery: a reappraisal. J THORAC CARDlOVASC SURG 1988;36:10-4. 5. Schonberger J, Everts P, Bredee, et al. The effect of postoperative normovolemic anemia and autotransfusion on blood saving after internal mammary artery bypass surgery. Perfusion [In press]. 6. Stibbe J, Kluft C, Brommer EJP, Gomes M, de Jong DS, Nauta J. Enhanced fibrinolytic activity during cardiopulmonary bypass in open-heart surgery in man is caused by extrinsic (tissue-type) plasminogen activator. Eur J Clin Invest 1984;14:375-82. 7. Tabuchi N, de Haan J, Boonstra PW, van Oeveren W. Activation of fibrinolysis in the pericardiaI activity during cardiopulmonary bypass. J THORAC CARDIOVASC SURG [In press]. 8. Hoylaerts M, Rijken HR, Collen D. Kinetics of the activation of plasminogen by human tissue plasminogen activator: role of fibrin. J Bioi Chern 1982;257:2912-9. 9. Adelman B, Rizk A, Hanners E. Plasminogen interactions with platelets in plasma. Blood 1988;75:1530-5. 10. Thorsen LI, Brosstad F, Gogstad G, Sletten K, Solum NO. Competitions between fibrinogen with its degradation products for interactions with the platelet-fibrinogen receptor. Thromb Res 1986;44:611-23. II. Verheijen JH, Nieuwenhuizen W, Wijngaards G. Activation of plasminogen by tissue activator is increased specifically in the presence of certain soluble fibrin(ogen) fragments. Thromb Res 1982;27:377-85.
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6
12. Stricker RB, Wong D, Shiu DT, Reyes PT, Shuman MA. Activation of plasminogen by tissue plasminogen activator on normal and thrombasthenic platelets: effects on surface proteins and platelet aggregation. Blood 1986;68:275-80. 13. Miles LA, Ginsberg MH, White JG, Plow EF. Plasminogen interacts with human platelets through two distinct mechanisms. J Clin Invest 1986;77:2001-9. 14. Miles LA, Plow EF. Binding and activation of plasminogen on the platelet surface. J Bioi Chern 1985;260:4303-11. 15. Miles LA, Dahlberg CM, Plow EF. The cell-binding domains of plasminogen and their function in plasma. J Bioi Chern 1988;263:11928-34. 16. Michelson AD, Barnard MR. Plasmin-induced redistribution of platelet glycoprotein lb. Blood 1990;76:2005-10. 17. Cramer EM, Lu H, Caen JP, Soria C, Berndt MC, Tenza D. Differential redistribution of platelet glycoproteins Ib and lIb-IlIa after plasmin stimulation. Blood 1991;77:6949. 18. Adelman B, Michelson AD, Loscalzo J, Greenberg J, Handin RI. Plasmin effect on platelet glycoprotein Ib-von Willebrand factor interactions. Blood 1985;65:32-40.
de Haan et al.
1023
19. van Oeveren W, Harder HP, Roozendaal KJ, Eijsman L, Wildevuur CRH. Aprotinin protects platelets against the initial effect of cardiopulmonary bypass. J THORAC CAROlOVASC SURG 1990;99:788-97. 20. van Oeveren W, Eijsman L, Roozendaal KJ, Wildevuur CRH. Platelet preservation by aprotinin during CPB. Lancet 1988;2:644. 21. Lu H, Soria C, Cramer EM, et al. Temperature dependence of plasmin-induced activation or inhibition of human platelets. Blood 1991;77:996-1005. 22. Horrow JC, Hlavacek J, Strong MD, et al. Prophylactic tranexamicacid decreases bleeding after cardiac operations. J THORAC CARDIOVASC SURG 1990;99:70-4. 23. Stephens RW, Pollanen J, Tapiovaara H, et al. Activation of pro-urokinase and plasminogen on human sarcoma cells: a proteolytic system with surface-bound reactants. J Cell Bioi 1989;108:1987-95. 24. Edmunds LH, Niewiarowski S, Colman RW. Invited letter concerning: aprotinin. J THORAC CARDIOVASC SURG 1991;101:1103-10.
1. de Haan, MSc,a J. Schonberger, MD, PhD,b J. Haan, BSc,a W. van Oeveren, PhD,a and A. Eijgelaar, MD, Phl)," Groningen and Eindhoven, The Netherlands
Rtoperative blood loss after operations involving cardiopulmonary bypass CCPB) ranges from several hundred milliliters up to a few liters on occasion. Such blood losscreates a demand for homologous blood products. I, 2 From the Department of Cardiopulmonary Surgery, Research Division," University Hospital Groningen, the Department of Cardiopulmonary Surgery," Catharina Hospital, Eindhoven, The Netherlands, and the Department of Cardiopulmonary Surgery," University Hospital Groningen, Groningen, The Netherlands. Received for publication Aug. 5, 1992. Accepted for publication Feb. 23, 1993. Address for reprints: W. van Oeveren, PhD, Cardiopulmonary Surgery Research Division,University Hospital Groningen, Oostersingel59, 9713 EZ Groningen, The Netherlands. Copyright
1993 by Mosby-Year Book, Inc.
0022-5223/93 $1.00 + .10
12/1/47001
Retransfusion of shed blood after CPB is an established method of reducing the requirements for such blood products.' However, because the composition of shed blood is far from normal, the benefits from retransfusion are questionable. Shed blood contains increased concentrations of fibrinogen degradation products, high levels of plasma hemoglobin, and other particles described as debris. Moreover, shed blood is largely defibrinated and contains a considerably lower number of platelets than the circulating blood. Therefore, retransfusion of especially large amounts of shed blood is not favorable for proper hemostasis. In fact, blood loss was found to be particularly high among patients in whom retransfusion otherwise would have been the most beneficial." In accordance with an earlier study, we observed that retransfusion of large amounts of shed blood after internal mam101 7
1 01 8
de Haan et ai.
mary artery bypass grafting caused a greater total blood loss than that observed in patients not receiving shed blood, reducing the efficacy of retransfusion. In explanation, we suggested that activated systems, such as the kinin, clotting, and fibrinolytic systems, played an important role.' In the first part of this study we analyzed the changes in fibrinolytic parameters of patients undergoing uncomplicated coronary artery bypass grafting, either with or without postoperative retransfusion of shed blood. In the second part of this study we examined the effect of fibrinolytic components on platelet function demonstrated by the platelet response to activated von Willebrand factor.
Patients and methods Patients. We prospectively studied 19 patients who underwent primary coronary artery bypass grafting with either one or two internal mammary arteries. When appropriate, additional saphenous vein bypass grafting was performed. The study was approved by the medical-ethical committee of the hospital and informed consent was obtained from all patients. Patients were excluded if they required intraaortic balloon pumping or in case of severe kidney and liver function disturbances, as reflected before the operation by increased concentrations of serum creatinine (> 100 mmol/L) or an increased prothrombin time. Patients were randomly assigned to one of two groups: Group I (n = 9) patients received postoperative retransfusion of shed blood by means of a hard-shell cardiotomy reservoir (D742, Dideco S.pA, Mirandello, Italy). Group 2 (n = 10) patients received no retransfusion. In all patients, anticoagulants and platelet inhibitors were discontinued at least 5 days before the operation. Operative management. The anesthetic and surgical management of the two groups of patients was identical. Bovinelung heparin (250 IU /kg body weight) was given intravenously 5 minutes before cannulation. An additional bolus of heparin (5000 IU) was administered to the extracorporeal circuit (ECC) every half hour. The ECC consisted of a hollow-fiber oxygenator (Sarns Inc. 13M, Ann Arbor, Mich.), a cardiotomy reservoir, and an arterial line screen filter (Cobe Laboratories Inc., Lakewood, Colo.). The ECC was primed with 1500 ml urea-linked gelatin (Haemaccel, Hoechst, Behringwerke AG, Marburg, Germany) containing 5000 IU bovine lung heparin and 100 ml mannitol. Neither in the pump prime nor during the operation was any fibrinolytic inhibitor such as aprotinin or tranexamic acid used. The perfusion flow rate at moderate hypothermia of 28 0 to 32 0 C was 2.5 Lyrnin. After aortic crossclamping, cardioplegia was induced with 1000 ml of cold (4 0 C) St. Thomas' Hospital solution, and the infusion was repeated with 500 ml of solution every half hour. For external cardiac cooling about 150 ml of cold (4 0 C) saline solution was applied. The mixture of blood and cooling fluid, collected in the pericardium and pleural space, was returned to the ECC. When the hematocrit values fell below 20% during CPB, in-line hemofiltration (Fresenius BV, Den Bosch, The Netherlands) was started. At the end of the perfusion, heparin was neutralized by protamine sulfate at a dosage of one times the total dosage of heparin. After completion of the extracorporeal bypass procedure, the patients received the contents of the circuit until
The Journal of Thoracic and Cardiovascular Surgery December 1993
the left atrial filling pressure was optimal, according to the prebypass diastolic pulmonary artery pressure. The residual volume in the ECC after aortic decannulation was collected in blood transfusion bags and retransfused gradually in the intensive care unit. Blood loss and management of retransfusion. In the patients of group 1 (retransfused), the mediastinal and pleural chest tubes were connected to a polyvinylchloride tube and attached to the inlet port of the cardiotomy reservoir as soon as protamine infusion was completed. The cardiotomy reservoirof the ECC was then used as an autotransfusion system. Suction with a negative pressure of 15 to 20 ern H 20 was applied at the vacuum port of the cardiotomy reservoir. A pump with a built-in bubble and clot detector (Imed Inc., Oxford, United Kingdom) was attached to the lower outlet port of the reservoir. During a maximum period of 8 hours, collected blood was retransfused to the patient via an Imed pump. The infusion rate of the Imed pump was adjusted each hour, corresponding to the blood levelin the cardiotomy reservoir, permitting a minimal flow of 50 ml/hr, In group 2 (control, receivingno retransfused shed blood), the shed blood was collected in a Pleur-evac system (Deknatel GmbH, Neustadt, Germany). The chest tubes were removed 18 to 24 hours after the operation. Biochemical and hematologic methods. During and after the operation, blood samples were collected from each patient at nine specific intervals: 25 minutes before CPB (after anesthesia); 5 minutes before CPB (after heparinization); after 30 minutes of CPB; just after release of the aortic crossclamp; 15 minutes after protamine sulfate administration; and at 0, 2, 4, and 8 hours on arrival in the intensive care unit. Additional samples were collected from the retransfusion reservoir 15 minutes after protamine sulfate administration and at 0, 2, 4, and 8 hours after arrival in the intensive care unit. Blood sampling. Three milliliters of blood was collected in ethylenediaminetetraacetic acid (final concentration 0.01 moll L) to assess red blood cell, platelet, and leukocyte counts. Three milliliters of blood was collected in aprotinin-hirudincitrate medium (final concentrations of 100 KIU, 0.05 U, and 0.308%, respectively). These blood samples were then centrifuged at 1000 g for 10 minutes to obtain platelet-poor plasma. Platelet-poor plasma was stored at -80 0 C until further processing for the determination of fibrinogen degradation products by enzyme-linked immunosorbent assay (Organon Teknika, Turnhout, Belgium). Two milliliters of blood were collected in citrate medium (0.308% sodium citrate, final concentration) and centrifuged at 1000 g for 10 minutes to obtain platelet-poor plasma for determination of tissue-type plasminogen activator (t-PA) antigen (enzyme-linked immunosorbent assay, Kabi Diagnostica, Stockholm, Sweden). In addition, the potential for t-PA stimulation was determined. In this assay, purified t-PA, plasminogen, plasmin-substrate S2403 (Kabi Diagnostics, Stockholm, Sweden), and a twenty-fold diluted plasma sample were incubated for 60 minutes at 37 0 C. The plasma concentration of t-PA stimulators, such as fibrin monomers, determines the activation rate of topA in converting plasminogen into the substratecleaving enzyme plasmin. The yellowcolor generated was measured at 405 nm in a spectrophotometer (EAR 400, SLTLabinstruments, Salzburg, Austria). In vitro study. In accordance with the declaration of Helsinki, whole blood was obtained from five healthy volunteers who
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6
had not received aspirinwithin 14daysbeforethe donation. The blood was collected in neutralized acid-citrate-dextrose medium. The platelet-rich plasma was separated from the remaining blood by centrifugation (100 g, 15 minutes)by the removal of the platelet-containing upper layer.Thereafter, the platelets wereisolated bygelfiltration ofthe platelet-rich plasmathrough a Sepharose 2B column (Pharmacia-LKB, Woerden, The Netherlands) by elution. They werecollected with a phosphate buffer, 50 mmol/L, pH 7.4,containing citricacid, 36 mrnol/L, potassium chloride, 5 mmol/L, sodiumchloride, 103 mmol/L, glucose 5 mmol/L, bovine serum albumin Y, 3.5 mg/rnl (Sigma Chemical Co., St. Louis, Mo.). The remaining blood was again centrifuged (1000 g, 10 minutes) to obtain platelet-poor plasma and red blood cell concentrate. The platelet-poor plasma was passed through a 0.22 ILm filter to obtain platelet free plasma. Plasminogen activatorinhibitorwasinhibited to ensure standardized and functional concentrations oft-PA in the tests. Theplatelet-free plasma,therefore, wasacidified byadding 10% volume of sodiumphosphate, 1.25 mol/L (pH 3.9), incubated for I minute, and neutralized afterward by adding sodium phosphate, 2.5 mol/L (pH 8.3),to a final pH of7.4. For further testing, gel-fixed platelets were mixed with this pretreated platelet-free plasmaand exposed to the fibrinolytic components. Introduction of t-P A and fibrin monomers to gel-filtered platelet/platelet-free plasma mixtures. Platelets in the gelfiltered platelet/platelet-free plasma (80:20%) mixtures were exposed for a period of 10 minutesat 37° C to various concentrations of fibrin monomers (t-PA stimulator, Kabi) ranging from 0 to 20 ILg/ml and/or t-PA (Kabi) ranging from 0 to 40 IV/m\' Ristocetin agglutination response. Platelet response to ristocetin was determined in 500 ILl reconstituted platelet-rich plasma (10% normal plasma was added to the gel-filtered platelet/platelet-free plasma mixtures) by measuring the changein optical densityof the platelet-rich plasmacausedby the von Willebrand factor-mediated platelet agglutination triggered with ristocetin (0.6 mg/rnl final concentration). Alternately, heparinwasadded (2 V /rnl) to assess its influence on the ristocetin response (Chrono-Log, Chrono-Log Corp., Havertown, Pa.). The response was quantified by both the steepest angle (a) measured during the agglutination process and the final height obtained after the agglutination process (H). Finally, the response wasexpressed as the percentage ofthe baseline ristocetin response as observed for an untreatedcontrol sample: Ristocetin response = (a X H)treated/(O' X H)untreated X 100% Statistics. All values are expressed as mean and standard error of the mean.Statisticalanalysis was accomplished either by the unpaired Student's t test if the two study groups were compared or bythe pairedStudent's t test if changeswithin one studygroup were analyzed. A p value less than 0.05 was considered to be significant. Results Operation and patients. The patient data concerning crossclamp time, CPB time, and age did not differ significantly between the two groups (Table I). Total and net blood loss. The total postoperative blood loss in the retransfusion group (1320 ± 162 ml, range
de Haan et al.
10 19
Table I. Demographic data and clinical data during CPB of both study groups Control (n = 10) Age (yr) Perfusion time (min) Crossclamp time (min) Total blood loss (rnl) Net blood loss (m!) Hemoglobin, before retransfusion (mmoljL) Hemoglobin, after 4 hr retransfusion (mmoljL)
58 81 49 925 925 5.9
± ± ± ± ± ±
Retransfused (n= 9) Significance
3.5 53 ± 2.7 11.5 84 ± 12.8 7.5 47 ± 6.0 78 1320 ± 162 78 568 ± 94 0.3 5.8 ± 0.4
6.3 ± 0.2
6.8 ± 0.5
NS NS NS
p<0.05 p
NS
values are expressed as means and standard error ofthe mean. Significance is calculated according to the unpaired Student's t test. NS, Not significant.
All
730 to 2110 ml) proved to be significantly higher (p < 0.05) than the blood loss in the control group (925 ± 78 ml, range 650 to 1370 ml) (Fig. 1). The net blood loss, comprising the total loss minus the retransfused volume, was significantly lower in the retransfusion group (589 ± 89 ml, p < 0.05) than in the control group. No significant reduction of blood product transfusion was obtained by retransfusion (0.8 ± 0.4 red blood cells, 0.1 ± 0.1 fresh frozen plasma) if compared with that of the control group (0.8 ± 0.3 red blood cells, 0.2 ± 0.2 fresh frozen plasma). Hematology. Blood cell counts (white blood cells, platelets, red blood cells) and hematocrit values during the CPB procedure and intensive care period did not differ significantly between the two groups (data not shown). Also, hemoglobin concentrations during and after the CPB period were similar in the two groups, showing only a minor and not statistically significant higher concentration after retransfusion (Table I). Fibrinogen degradation products. In both groups, the concentrations of fibrinogen degradation products in circulating blood rose immediately at the start of CPB, increasing to the greatest degree after unclamping of the aorta, to four times the baseline values (Fig. 2). After reaching the highest values between protamine infusion and admission of the patient to the intensive care unit, concentrations of fibrinogen degradation products in the control group decreased significantly (p < 0.05) to baseline values within 10 hours after the start of CPB. In contrast to measurements in the control group, however, concentrations of fibrinogen degradation products in the retransfusion group increased significantly once more (p < 0.05) soon after the retransfusion of the shed blood was started; concentrations reached a maximum within 4 to 6 hours after the start of CPB.
1020
The Journal of Thoracic and Cardiovascular Surgery
de Haan et al.
December 1993
1500 30
20
1000
...·r
.. .' L....1'· '
10
500
......··r........·········1
relransfusion
l2!J----=..:...~-----;
o
Fig. 1. Postoperative blood loss of retransfused group of patients (dark) compared with control group (light). Including the retransfused amount of blood in the total bloodloss shows a significantly higher bloodlossin the retransfusion group than in the controlgroup. The net bloodlosswassignificantly higher in the controlgroup than in the retransfusion group. (*p < 0.05 by Student's unpaired t test.)
5000
•
4000
~
4
6
8
10
12
houl'8
o
I
2
3000
2000
1000
0
l2!J -2
0
2
relransfusion 4
6
8
10
12
houl'8
Fig. 2. Fibrinogen degradation products (FgDP) during and after CPB with (closed squares) or without retransfusion (open circles). The dashed line represents the FgDP levels of the retransfusedbloodin the reservoir. After CPB, statisticaIIy significant differences were observed between the control and retransfusion groups. (*p < 0.05 by Student's unpaired t test.) t-PA Antigen and t-PA stimulating potential. Circulating concentrations of t-PA antigen were similar in the two groups of patients, approximating 10 ng/rnl. The concentrations of t-PA antigen increased at the onset of CPB, reaching maximum values of 25 ng/rnl after unclamping of the aorta, and remained elevated during the first 8 postoperative hours (Fig. 3).
Fig. 3. t-PA Antigen during and after CPS with (closed squares) or withoutretransfusion (open circles). The dashed line representsthe t-PA level of the retransfusedbloodin the reserVOIr. The t-PA stimulating potential in both groups reached its highest value even before the onset of CPB, remained high during CPB, and returned to preoperative values between unclamping of the aorta and 4 hours after the start of CPB (Fig. 4). During the latter period, however, t-PA stimulating potential in the patients who underwent retransfusion was higher than in the control group. Furthermore, t-PA stimulating potential in the shed blood increased significantly and reached a maximum 6 hours after the start of CPB. Ristocetin agglutination response. Compared with untreated reconstituted platelet-rich plasma, the ristocetin agglutination response of platelet-rich plasma treated with heparin, fibrin monomers, or t-PA decreased slightly, but to a statistically significant degree (p < 0.05), to values of70% (Fig. 5). However, with the introduction of both fibrin monomers and t-PA, the ristocetin agglutination response decreased to 25%, a highly significant change (p < 0.01). Discussion Retransfusion of a large amount of shed blood after CPB increases the total blood loss significantly. We have hypothesized that this reduced hemostasis is related to the high level of fibrinolytic stimulation in the retransfused shed blood, a position supported by the correlation of postoperative concentrations of fibrinogen degradation products and the total postoperative blood loss in all patients exceeding average blood loss (Fig. 6). In patients with moderate or low blood loss this correlation did not exist, indicating that in these patients bleeding is mainly caused by other factors. This increased concentration of circulating fibrinogen degradation products, observed in
The Journal of Thoracic and
de Haan et al.
Cardiovascular Surgery
10 2 1
Volume 106, Number 6
400
./1"
300
.·····1·····'.
.. •
200
100
../ *
-,'.
*
..••
I *
100
o
[£!]f-----===:::.:....------l o
6
2
8
10
75
.5
I!
•e• o •Go f
r8lransfusion
-2
..z3
50
25
12
hou,.
Fig. 4. Level of t-PA stimulation during and after CPB with (closed squares) or without retransfusion (open circles). The dashed line represents the level of t-PA stimulation in the retransfused blood, increasing significantly to maximum levels between 4 and 6 hours after the start of CPB. (*P < 0.05 comparedwithend-CPBvalue,calculatedbypairedStudent's t test)
the patients during retransfusion, is tenfold the retransfused amount of fibrinogen degradation products in the shed blood. Hence another mechanism responsible for increased concentrations of fibrinogen degradation products in the blood of retransfused patients must exist, mediated by renewed fibrinolytic activity and resulting in a reduced hemostasis. In general, fibrinolysis during clinical interventions is dependent mainly on plasminogen activation by t-PA6; indeed, the concentration of t-PA observed in the retransfusion group was high. However, t-PA concentrations in the control group were similar; therefore, another component in the shed blood, increasing the fibrinolytic activity during retransfusion, was suspected. This other factor appeared to be the high potential for t-PA stimulation, as observed in the shed blood during retransfusion. Because the shed blood is defibrinated because of prior clotting activity in the pleural space and thoracic wound area," its t-PA stimulating potential is derived primarily from fibrin monomers. After binding, fibrin as well as fibrin monomers stimulate t-PA activity l500-fold as compared with free t-pA,8 However, these fibrin monomers could not be measured in the patients' plasma after infusion of the shed blood, despite the increased fibrinolytic activity indicated by increased levels of fibrinogen degradation products. These findings suggest that clotting products such as fibrin monomers can bind to platelets, similar to plasminogen? and fibrinogen degradation products.!" and are scavenged from plasma. If so, these fibrin monomers cause a high level of plasmin activity localized on the platelet surface. I I, 12 Moreover, this surface-bound
o
conlrol
heparin
FM
FM+l-PA
l-PA
Fig. 5. Ristocetin agglutination response of normal platelets after a challenge to various stimulators such as heparin (2 Ulml), fibrin monomers (FM, 20 f.Lg/ml), t-PA (40 IUlml) or a combination of fibrin monomers (20 f.Lg/ml) and t-PA (40 IVIml). Compared with control (100%), the introduction of heparin, fibrin monomers, or t-PA caused a significantly decreasedresponse. Introductionoft-PA as wellas fibrinmonomers caused a highly significant decreased response (*p < 0.05; **p
< 0.01).
2500--r---------------'7'"'"'"1
<:-
.§.
.. •
2000
~
'8 ~
•
•
lie
1500
.--
~_/
-I.··
»"• • • •
1000
• •
SOO +---.,...----r--~--____,.____-____i
o
2000
4000
6000
6000
10000
FgDP (ng/mll
Fig. 6. Total postoperative blood loss and concentrations of circulating fibrinogen degradation products (FgDP) after start of retransfusion and at 2, 4, and 8 hoursafter CPB in all patients exceeding mean blood loss of 925 ml were significantly correlated according simple regression statistics (r 2 = 0.52, p
< 0.01).
plasmin activity is uncontrolled because of its low accessibility to a2-antiplasmin.13-15 The plasmin activity can thus cause platelet damage and dysfunction, expressed primarily by a downregulation of glycoprotein IbjIX
1022
de Haan et al.
complexes'< 17 and affected glycoprotein Ib-von Willebrand factor interactions.!" Our hypothesis also offers an explanation for the observed platelet dysfunction during CPB, 19,20 because increased concentrations of t-PA and t-PA stimulator are observed at that time as well. We tested this hypothesis in the second (in vitro) part of this study with the introduction of t-PA and fibrin monomers to normal platelets and measured the effect of this treatment on ristocetin agglutination. The platelet response to ristocetin reflects the capacity of platelets to adhere to the subendothelium of a damaged blood vessel and is based on activation of von Willebrand factor. Synergistically, t-PA and fibrin monomers cause within 10 minutes a reduction in glycoprotein Ib-dependent platelet function of75%. The concentrations oft-PA and fibrin monomers used in vitro are similar to those measured in vivo. Others have shown t-PA-induced platelet dysfunction in plasma as well, but at concentrations of at least 400 ng/rnl, 12 whereas we used moderate concentrations up to 60 ng/rnl. Therefore, we consider these in vitro results illustrative of the process of hemostatic derangement in vivo, caused by moderate concentrations of t-PA and fibrin monomers, resulting in generation of plasmin on the surface of platelets. This model explains several observations: I. Within the first postoperative hours, hemostasis recovers. Massive replenishment of damaged platelets by new platelets is not anticipated within this short period, and therefore recovery of the affected platelets is most likely. Michelson and Barnard'" show the feasibility of this recovery after plasmin treatment and the subsequent introduction of aprotinin. 2. A longer CPB time causes a higher postoperative blood loss. Lu and associates" showed that plasminrelated platelet inhibition is enhanced by hypothermia, and others demonstrated the (moderate) digestion of glycoprotein Ib proteins by plasmin. Thus a long hypothermic CPB period will cause more (irreversible) platelet damage and postoperative blood loss. 3. Because all effects described are related to t-PAdependent plasmin generation and action, the blood-saving effects of fibrinolytic inhibitors such as aprotinin, tranexamic acid, and e-aminocaproic acid can be explained. They all are able, in contrast to the physiologic inhibitor £x2-antiplasmin,13 to inhibit plasmin, generated by t-PA and plasminogen on fibrin. The moderate concentrations of t-PA and fibrin monomers indicate as well that it is difficult to prevent platelet damage during CPB. Therefore, blood from the wound area, comprising the main source for fibrin monomers, should be retransfused in limited quantities (Schonberger et aI., personal communication, 1992), not retransfused at all, or washed
The Journal of Thoracic and Cardiovascular Surgery December 1993
by use of cell-saving devices before retransfusion. During CPB, however, because the combined interaction of fibrin monomers and t-PA with platelets cannot be prevented in any direct way, platelet preservation can be achieved with the use of antifibrinolytic drugs such as aprotinin" or tranexamic acid. 22 Because both these drugs are presumed capable to inhibit surface-localized plasmin.P they are most effective for preservation of hemostasis,19,22,24 eliminating the need for postoperative retransfusion of shed blood. We express our gratitude to Jeffrey W. Kolff for reading the manuscript and improving the linguistics. REFERENCES I. Mammen EF, Koets MH, Washington BC, et al. Hemostasis changes during cardiopulmonary bypass surgery. Semin Thromb Hemost 1985;11:281-92. 2. Boonstra PW, van ImhoffGW, Eysman L, et al. Reduced platelet activation and improved hemostasis after controlled cardiotomy suction during clinical membrane oxygenator perfusions. J THORAC CARDIOVASC SURG 1985;89:900-6. 3. Cosgrove DM, Amiot DM, Meserko JJ. An improved technique for autotransfusion of shed mediastinal blood. Ann Thorac Surg 1985;40:519-20. 4. Adan A, Brutel de la Riviere A, Haas F, van Zalk A, de Nooy E. Autotransfusion of drainage mediastinal blood after cardiac surgery: a reappraisal. J THORAC CARDlOVASC SURG 1988;36:10-4. 5. Schonberger J, Everts P, Bredee, et al. The effect of postoperative normovolemic anemia and autotransfusion on blood saving after internal mammary artery bypass surgery. Perfusion [In press]. 6. Stibbe J, Kluft C, Brommer EJP, Gomes M, de Jong DS, Nauta J. Enhanced fibrinolytic activity during cardiopulmonary bypass in open-heart surgery in man is caused by extrinsic (tissue-type) plasminogen activator. Eur J Clin Invest 1984;14:375-82. 7. Tabuchi N, de Haan J, Boonstra PW, van Oeveren W. Activation of fibrinolysis in the pericardiaI activity during cardiopulmonary bypass. J THORAC CARDIOVASC SURG [In press]. 8. Hoylaerts M, Rijken HR, Collen D. Kinetics of the activation of plasminogen by human tissue plasminogen activator: role of fibrin. J Bioi Chern 1982;257:2912-9. 9. Adelman B, Rizk A, Hanners E. Plasminogen interactions with platelets in plasma. Blood 1988;75:1530-5. 10. Thorsen LI, Brosstad F, Gogstad G, Sletten K, Solum NO. Competitions between fibrinogen with its degradation products for interactions with the platelet-fibrinogen receptor. Thromb Res 1986;44:611-23. II. Verheijen JH, Nieuwenhuizen W, Wijngaards G. Activation of plasminogen by tissue activator is increased specifically in the presence of certain soluble fibrin(ogen) fragments. Thromb Res 1982;27:377-85.
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 6
12. Stricker RB, Wong D, Shiu DT, Reyes PT, Shuman MA. Activation of plasminogen by tissue plasminogen activator on normal and thrombasthenic platelets: effects on surface proteins and platelet aggregation. Blood 1986;68:275-80. 13. Miles LA, Ginsberg MH, White JG, Plow EF. Plasminogen interacts with human platelets through two distinct mechanisms. J Clin Invest 1986;77:2001-9. 14. Miles LA, Plow EF. Binding and activation of plasminogen on the platelet surface. J Bioi Chern 1985;260:4303-11. 15. Miles LA, Dahlberg CM, Plow EF. The cell-binding domains of plasminogen and their function in plasma. J Bioi Chern 1988;263:11928-34. 16. Michelson AD, Barnard MR. Plasmin-induced redistribution of platelet glycoprotein lb. Blood 1990;76:2005-10. 17. Cramer EM, Lu H, Caen JP, Soria C, Berndt MC, Tenza D. Differential redistribution of platelet glycoproteins Ib and lIb-IlIa after plasmin stimulation. Blood 1991;77:6949. 18. Adelman B, Michelson AD, Loscalzo J, Greenberg J, Handin RI. Plasmin effect on platelet glycoprotein Ib-von Willebrand factor interactions. Blood 1985;65:32-40.
de Haan et al.
1023
19. van Oeveren W, Harder HP, Roozendaal KJ, Eijsman L, Wildevuur CRH. Aprotinin protects platelets against the initial effect of cardiopulmonary bypass. J THORAC CAROlOVASC SURG 1990;99:788-97. 20. van Oeveren W, Eijsman L, Roozendaal KJ, Wildevuur CRH. Platelet preservation by aprotinin during CPB. Lancet 1988;2:644. 21. Lu H, Soria C, Cramer EM, et al. Temperature dependence of plasmin-induced activation or inhibition of human platelets. Blood 1991;77:996-1005. 22. Horrow JC, Hlavacek J, Strong MD, et al. Prophylactic tranexamicacid decreases bleeding after cardiac operations. J THORAC CARDIOVASC SURG 1990;99:70-4. 23. Stephens RW, Pollanen J, Tapiovaara H, et al. Activation of pro-urokinase and plasminogen on human sarcoma cells: a proteolytic system with surface-bound reactants. J Cell Bioi 1989;108:1987-95. 24. Edmunds LH, Niewiarowski S, Colman RW. Invited letter concerning: aprotinin. J THORAC CARDIOVASC SURG 1991;101:1103-10.