- Email: [email protected]
Activation of fibrinolysis in the pericardial cavity during cardiopulmonary bypass
Activation of fibrinolysis in the pericardial cavity during cardiopulmonary bypass The clotting and fibrinolytic systems are activated by tissue factor and by tissue-type plasminogen activator in the pericardial cavity, where the thrombogenicity is greater than that of the surface of modem extracorporeal circuits. This local activation may have consequences for the systemic activation processes during cardiopulmonary bypass. To test this hypothesis, we investigated blood activation by interrupting the blood suction from the pericardial cavity during cardiopulmonary bypass in clinical coronary artery bypass operations. In blood coUected in the pericardial cavity, thrombin-antithrombin ill complex (p < 0.01), tissue-type plasminogen activator antigen (p < 0.05), fibrinogen degradation products (p < 0.01), and fibrin degradation products (p < 0.01) were significantly higher than in the systemic blood. Plasma heparin was significantly consumed in the pericardial cavity (p < 0.01). Once the pericardial blood was returned to the systemic circulation after resumed suction during cardiopulmonary bypass, thrombin-antithrombin ill complex (p < 0.05), fibrinogen degradation products (p < 0.05), and fibrin degradation product (p < 0.05) concentrations increased significantly in the systemic blood. The effects of pericardial tissue on activation of clotting and fibrinolysis were also studied in vitro. When human plasma was incubated for 5 minutes with rabbit pericardium at reduced heparin concentrations, we found significant generation of thrombin (p < 0.05) and plasmin (p < 0.05). If the thrombin inhibitor hirudin was added, plasmin generation was also inhibited (p < 0.05). The results of the clinical and experimental study are in agreement with our hypothesis that tissue factor and tissue-type plasminogen activator accelerate the activation of clotting and sequentiaUy of fibrinolysis under conditions of low heparin concentrations in the pericardial cavity and that this local activation contributes highly to the systemic activation, affecting hemostasis during cardiopulmonary bypass. Topical use of heparin in the pericardial cavity therefore seems indicated to reduce blood activation during cardiopulmonary bypass. (J THORAC CARDIOVASC SURG 1993;106:828-33)
Noriyuki Tabuchi, MD, Jacob de Haan, MSc, Piet W. Boonstra, MD, PhD, and Willem van Oeveren, PhD, Groningen, The Netherlands
Reduction of blood activation has been among the main issues in the improvement of cardiopulmonary bypass (CPB).1-4 Attempts have therefore been made to improve the biocompatibility of extracorporeal circuits (ECCs) to reduce activation of the clotting and fibrinolytic systems and to preserve platelet function.v ' The clinical results, however, are not fully satisfactory. From the Thorax Centre, University Hospital Groningen, Groningen, The Netherlands. Received for publication July 17, 1992. Accepted for publication Dec. II, 1992. Address for reprints: W. van Oeveren, PhD, Dept. of Cardiopulmonary Surgery, Research Division, University Hospital, Oostersingel 59 9713 EZ Groningen, The Netherlands. @ 1993 by Mosby-Year Book, Inc. 0022-5223/93 $1.00 +.10 12/1/45055
Copyright
828
Although the use of a heparin-coated ECC surface could reduce blood activation during CPB to some extent,6,7 blood losswas not significantly reduced. 6 Blood activation in the pericardial cavity has long been considered potentially hazardous for impairment of hemostasis.v? This blood activation is based mainly on two factors expressed in the pericardial cavity: tissue factor, a strong stimulus for the clotting system,IO, II and tissue-type plasminogen activator (t-PA), a potent activator of the fibrinolytic system. I 2-14 Blood activation in the pericardial cavity may be considerable and therefore may obscure the benefits of improved biocompatibility of ECCs. To test this hypothesis, we investigated blood activation by interrupting the blood suction from the pericardial cavity during the suturing of distal coronary anastomoses in patients undergoing elective coronary artery bypass grafting. The activation in blood from the pericardial
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5
cavity was measured separately and the systemic effect was investigated immediately after this blood was returned to the systemic circulation. Additionally, we mimickedthe effectof pericardium on blood activation by incubating human plasma with rabbit pericardial tissue. Patients and methods Clinical study. After informed consent was obtained 12 patients undergoing elective coronary artery bypass grafting entered the.study. None of the patients was older than 75 years or had evidence of severe heart failure, renal or hepatic dysfuncti~n, or bleeding diathesis. None of the patients was treated With drugs affecting the clotting system within 7 days before the operation. Operative techniques. After premedication with diazepam (10 to 15 mg), anesthesia was induced with sulfentanyl (3 to 5 ~g/kg) and pancuronium (100 to 140 ~g/kg). Ventilation was controlled by a volume-controlled respirator with air mixed with oxygen. Analgesia was provided by sulfentanyl and rnidazolan infusion. C;::~famandol (2 gm), and dexamethasone (1 mg/kg) were administered before operation. Before cannulation, bovine heparin (300 IV/kg; Leo, Emmen, The Netherlands) was injected and the addition of heparin (100 IU/kg) was repeated every 60 minutes during CPB. The activated coagulation time was determined in every patient at 5 minutes before the start of C~B and at every 60 ~inutes (International Technidyne Corp., Edison, N.J.). The activated coagulation time was confirmed to be greater than 400 seconds throughout CPB in every patient. The ECC consisted of a Cobe Excel membrane oxygenator (CO~E La~ratories, Inc., Lakewood, Colo.), and polyvinylchloride ~ubmg .. The circ~it was primed with 2000 ml oxypolygelatm (Gehfundol; Biotest AG, Dreiech, Germany) and 1500 IV bovine heparin (Leo). CPB was performed with moderate hypothermia (27 0 C nasopharyngeal temperature) with a p.ump flow of 2.4 L/m 2 per minute, maintaining a mean arterial pressure of 50 to 60 mm Hg. Myocardial preservation during aortic clamping was maintained with I L St. Thomas' Hospital cardioplegic solution (4 0 C) injected into the aortic root. No.cold s~line solution or ice slush was used in the pericardial cav~ty durmg CPB f?r topical cooling. The blood, gradually ~z~g from the surgical field into the pericardial cavity, was initially not returned to the ECC but stayed in the pericardial cavity until suturing of the distal coronary anastomoses was completed. After completion of the distal coronary anastomoses all bl~ in the pericardial cavity was returned into the ECC b; a cardiotomy sucker. Thereafter, the aortic crossclamp was released and the proximal anastomoses were performed. After CPB, heparin was neutralized by protamine chloride (3 mg/kg; Hoffman-La Roche BV, Mijdrecht, The Netherlands). Blood samples. Before the start of CPB, blood samples were taken from the radial arterial line as a control preparation. At 5 minutes before suturing of the distal anastomoses had been completed, blood samples were taken simultaneously from the pericardial cavity and from the arterial line of the oxygenator. After all blood in the pericardial cavity was returned to the ECC and to the circulation of the patient, another blood sample was taken from the arterial line of the oxygenator. After release of the aortic crossclamp, additional blood samples were collected ~y punc~ure. from the l~ft and right atria to control for the posSibleactivation of clottmg and fibrinolysis by reperfusion of the coronary and pulmonary circulations. Finally, a sample was taken at the end of CPB from the arterial line. Blood was col-
Tabuchi et al.
8 29
lected in 3.08% sodium citrate (9:1 volume/volume), stored on ice, and centrifuged at 1000gfor 10 minutes to obtain plateletpoor plasma. Laborat?ry assays. In platelet-poor plasma samples, the concentrations of t-P A antigen, thrombin-antithrombin III co~plexes (TAT), ~brin degradation products (FbDP), and fibnnoge~ deg~adatlOn products (FgDP) were determined by enzyme-hnked immunosorbent assay (Kabi Vitrum AB, Stockholm, Sweden; Behringwerke AG, Marburg, Germany; and Organon Teknika BV, Turnhout, Belgium, respectively). The he.parin ~on:~ntration ~as determined by the capacity of hepann to inhibit thrombm activity. In the presence of excess antithrombin III and thrombin, the conversion rate of a thrombin-specific substrate was determined (S2238; KabiVitrum). . In vitro st~dy. ~lood was collected from five healthy donors m 3.08% sodium ~Itrate (9:1 volume/volume), and centrifuged at I OOOg for 10mmutes to make platelet-poor plasma. Pieces of peric~r~ium were taken from anesthetized and systemically hepanmzed (100 IU /kg) rabbits. The pieces were cleaned with saline solution and cut to a standardized size shortly before the start of the experiments. All animals received care in compliance with the Dutch regulations and law regarding the care and use of laboratory animals. Generation ofthrombin. Toa I ml aliquot of human plasma, we added I 00 ~l heparin of varying concentrations (final concentrations, 1.5,2.0, and 2.5 IV/mI). These heparinized human plasma samples were mixed with I 00 ~l Ca2+ (final concentration, 12.5 mmol/L) and 100 ~l thrombin substrate (S-2238' KabiVitrum) and then incubated at 37 0 C with I em? rabbit pericardium. As a control preparation, human plasma with h~parin (.1.5 IV/mI) and Ca2+ (12.5 mmol/L) was incubated Without tissue. At the start of incubation and at 5 10 20 and 30 minutes after the start of incubation, I 00 ~l w~s t~ke~ out of the sample tube and the conversion of substrate was measured at 405 nm (Ear 400; SLT, Salzburg, Austria). The generation of thrombin activity was calculated according to the standard formula. (KabiVitrum). Generation ofplasm~n. To a I ml aliquot of human plasma, we added 100 ~l hepann of varying concentrations (final concentrations, 1.5, 2.0, and 2.5 IV/mI) or heparin (final concentration, 1.5 I~/mI) plus hirudin (final concentration, 0.5 V /ml), These hepanmzed human plasma samples were mixed with 100 ~l Ca2+ (final concentration, 12.5 mmol/L) and then incubated a~ 37 0 C with I ern? rabbit pericardium. As a control preparation, human plasma with heparin (1.5 IV/mI) and Ca2+ (12.5 mrnol/L) was incubated without tissue. At the start of incubation and at 5, 10, 20, and 30 minutes after the start of incub~tion, 80 ~l was taken out of the sample tube and 40 ~l plasmm substrate (S-2251; KabiVitrum) was added. After another 15, 30, and 60 minutes of incubation, the conversion of substrate was measured at 405 nm. Plasmin activity was calculated according to the standard formula (KabiVitrum). ~tatistic~. Statistical analysis was performed by a paired Wilcoxon ~Igned-rank test in the clinical study and by the Mann-Whitney test in the in vitro study. All data are expressed as mean ± standard deviation of the mean. A p value less than 0.05 was considered significant.
Results Clinical study. The demographic data from the patients were as follows: age, 61 ± 2 years; weight, 81 ± 3 kg; CPB time, 93 ± 10minutes;crossclamptime,
The Journal of Thoracic and Cardiovascular Surgery November 1993
8 3 a Tabuchi et al.
3
Heparin (IU/ml)
TAT (ng/ml)
200
150
4 FgDP (Ilg /ml)
3
**
40 FbDP (J.lg/ml)
40 t-PA (ng/ml)
30
30
*
2 2
20
100 10
50
o
o
0
Fig. 1. Plasma heparin and TAT concentrations in systemic blood (dark column) compared with blood from thepericardial cavity (light column) before admixture. Thesamples were collected at theendofsuturing ofthedistal coronary anastomosis. Double asterisk represents a significant difference between the two samples (p < 0.01). 54 ± 25 minutes; venous grafts, 0.4 ± 0.6; and arterial grafts, 1.5 ± 0.7. In total, 637 ± 32 ml shed blood was collectedin the pericardial cavityduring 57.3 ± 6.3 minutes before the end of suturing distal coronary anastomosis. In shed blood of the pericardial cavity, the plasma concentrationof heparin waslessthan half (p < 0.0I) and that of TAT was three (p < 0.01) times that in systemic circulating blood(Fig. I). Whereas t-PA antigen was two times higher (p < 0.05), the concentrationsof FgDP and FbDP were 10 times higher (p < 0.Q1) in the bloodof the pericardial cavity than in the circulation (Fig. 2). In contrast, blood collectedafter reperfusion from the left and right atria, indicating possible reperfusion-dependent blood activation, contained similar concentrationsofTATcomplexes(48 ± 41 ng/rnl versusSZ ± 49 ng/rnl) and FbDP (3061 ± 2329 ng/rnl versus 3176 ± 2622 ng/ml). Only after the blood in the pericardial cavity was returned to the systemic circulation after resumed suction were increasesof TAT (p < 0.05), FgDP (p < 0.05), and FbDP (p < 0.05) observed in systemic circulating blood (Fig. 3). In vitro study Thrombin and plasmin generation. A significant increase in thrombin and plasmin activity was observed after 5 minutes of incubation with rabbit pericardium at l.5or 2.0 IV Iml heparin (p < 0.05) (Fig. 4). In the presence of a 2.5 IV /rnl concentration of heparin, thrombin and plasmin activity was almost absent and was thus significantlyless than at 1.5 or 2.0 IV Iml concentrationsof
0+--.......
Fig. 2. Activation products in systemic blood (dark column) compared with blood from thepericardial cavity (light column) before admixture. Theplasma concentrations ofFgDP, FbDP, andt-PAantigen were measured at theendofsuturing thedistal coronary anastomosis. Asterisk and double asterisk representa significant difference between two samples (p < 0.05 and p < 0.01, respectively). heparin (p < 0.01 and p < 0.05, respectively). Without the introduction of pericardial tissue, thrombin and plasmin activity remained low. The activity of plasmin was reduced significantly, whenhirudin wasadded to samples with the lowest heparin concentration (1.5 Il.I/rnl) (p < 0.05) (Fig. 5). Discussion
Bloodactivationthrough the intrinsicclottingpathway from contact of the bloodwith the large artificialsurface of the ECC has been a main issuein the development of CPRI, 15 Evenwhen the materials of the ECC have been improved and modified by heparin coating however, the activation of the clotting system cannot be significantly suppressed during CPR6 Blood activation by tissue in pericardium should therefore be evaluated as occurring by another activationpathway during CPR Tissuefactor stimulates thrombin generation even more strongly through the extrinsicclottingpathway than doesthe ECC through the intrinsicclottingpathway.10, 11 Moreover, the activationof fibrinolysis, initiated by t-PA, is known to be accelerated lOO-fold in the presence of fibrin or fibrin monomers derived from clotting activity.!" Our study shows that the activation of fibrinolysis was more pronouncedthan that of the clottingsystemin the pericardial cavity.Even a small activationof the clotting systemcorresponded with considerable activation of fibrinolysis in the presence of abundant t-PA in the pericardial cavity. The inhibition of fibrinolysis by the addition of the thrombin inhibitor hirudin in the in vitro study supports this idea. Because of the existence of tissue factor and
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5
2500
~c
Tabuchi et at.
FgDP
TATcompTex
FbDP
6000
150
5000
2000
*
1500
*
4000
83 I
*
*
*
1000 500
0
Fig. 3. Activation products in the systemiccirculation. TAT, FgDP, and FbDP were measured before CPB (dark column), at 5 minutes before the end of suturing of the distal coronary anastomosis (medium gray column), after all bloodin the pericardialcavitywas returned to the systemiccirculation (light column), and at the end of CPB (open column). Asterisk represents a statistically significant increase after return of pericardial blood (p < 0.05).
..... E
200
..... ~
:i 'ot' W 150 0
:::)
0 II>
W
... ~
... 0
E 2
100
~
c
:a
150
c
E
100
II)
.!! e,
.c
l-
50
50
A
10
20
30
Time (min)
B
10
20
30
Time (min)
Fig. 4. Thrombin (A) and plasmin (B)generation by rabbit pericardium were measured in human plasma in the presenceof variousconcentrationsof heparin (1.5 IV Iml, closed squares; 2.0 IV Irnl, circles; 2.5, IV Iml triangles). As a control preparation, human plasma was incubated without tissue at the lowest concentration of heparin (1.5 IV /rnl, open squares). Asterisk indicates a statisticallysignificant increaseafter 5 minutes of incubation (p < 0.05). Double asterisk indicatesa significantdifferencebetweenthe highestconcentrationof heparin (2.5 IV I ml) and lower concentrations of heparin (1.5 and 2.0 IV Irnl), (p < 0.01 and p < 0.05, respectively).
t-PA, the blood activation in the pericardial cavity could therefore be far greater than the blood activation caused by the improved surfaces of modern ECCs. Clinically, this hypothesis is supported by the excellent performance of oxygenators without cardiotomy suction even under the circumstance ofdecreased systemic heparinization.?- 17, 18 It can be expected that the incubation time of blood in the pericardial cavity correlates with the intensity of blood activation. In our protocol, we retained the suctioned blood for 60 minutes during suturing of the coronary
anastomoses in the pericardial cavity, where heparin appeared to be consumed more than 50%. Our in vitro study clearly showed that this reduction of heparin resulted in suboptimal heparin concentrations, insufficient to prevent the activation of clotting and indirectly of fibrinolysis. This intensity of blood activation could be minimized if the blood pooling time in the pericardial cavity was shortened by continuous suction. Additional anticoagulation seems required, however, because our in vitro study showed that even a short incubation time
832
-E ....
The Journal of Thoracic and Cardiovascular Surgery November 1993
Tabuchi et al.
150
~
0
It) I
W
,..
0
--
100
>C
*
C
E Ul
*
.!! D.
*
50
O~----~----,.....------r
o
10
20 30 Time (min)
Fig. 5. Plasmin generation inhibited was by the addition of hirudin. Plasmin generation was measured in heparinized human plasma (1.5 IV Im1, squares) accordingto the incubation time with rabbit pericardium,withor without hirudin (0.5 V Im1, diamonds). Asterisk represents significant difference between sampleswith and without hirudin (p < 0.05). results in considerable thrombin and plasmin generation under conditions of low heparin concentration. These results raise the question of whether the systemic anticoagulation was sufficient in our present study. In our protocol, we used an initial heparin dose of 300 IV Ikg, followed by an additional dose of 100 IV Ikg every 60 minutes. Also, the activated clotting time was kept at greater then 400 seconds throughout CPRI9 The acceptable minimum value of anticoagulation is still under debate, however, because subclinical coagulation activity always exists during CPB and does not correlate with systemic heparin concentrations.l? We think that this discrepancy may be a consequence of the peculiar blood activation in the pericardial cavity. Even if anticoagulation is sufficient in the ECC, it is not necessarily sufficient in the pericardial cavity. If the biocompatibility of ECC surfaces become better, the discrepancy in thrombogenicity between ECC and tissue will become larger. To manage this discrepancy, it may be preferable to use topical heparin in the pericardial cavity to maintain a sufficient local anticoagulation. The FDPs generated in the pericardial cavity are returned to the systemic circulation during CPB and thereby affect hemostasis through interference with platelet receptors, fibrinogen binding to platelets, and clot formation.P: 21 Also, potentially active products from the
fibrinolytic system may be reinfused as a result of a shortage of plasmin inhibitors in the pericardial blood.22 In our technique of CPB, the amount of blood coming back from the pericardial cavity was about I L throughout the whole CPB period in an uncomplicated coronary artery bypass grafting operation. The local activation in the pericardial cavity may thus considerably contribute to total blood activation during CPR In the in vitro study, we used rabbit tissue to activate human plasma, a documented laboratory practice. 10 It is also known that there is a cross-reaction oft-PA between human and the other mammalian species." Our in vitro set-up therefore essentially mimics the clinical situation. The cells in the mesothelium are known to secrete physiologically abundant t-PA to avoid adhesion between organs, 12,13 The fact that the blood remaining in the thoracic cavity or in the pericardial cavity is defibrinated in the patients with chest trauma exemplifies the potent fibrinolytic activity there. 23 Our study demonstrates that the same process may be going on in the pericardial cavity during CPB. In conclusion, we found fulminant fibrinolysis initiated by clotting activation in blood of the pericardial cavity in the presence oftissue factor and t-PA. This high level of blood activation was expressed because heparin was locally consumed in the pericardial cavity. This local blood activation significantly contributed to the total blood activation during CPB. The pericardial blood therefore should be aspirated continuously to minimize blood activation and the topical use of heparin should be considered to maintain sufficient anticoagulation in the pericardial cavity, which appears to be the most thrombogenic site during CPR We thank J. Haan and F. Wei for technicalassistance. REFERENCES 1. van OeverenW, WildevuurCRH, KazatchkineMD. Biocompatibility of extracorporea1 circuits in heart surgery. Transfus Sci 1990;11:5-33. 2. Harker LA. Bleeding after cardiopulmonary bypass. N Engl J Med 1986;314:1446-7. 3. KirklinJK, Westab1y S, Blackstone EH, Kirklin JW, ChenowethDE, Pacifico AD. Complementand the damaging effects ofcardiopulmonary bypass. J THORAC CARDIOVASC SURG 1983;86:845-57. 4. van Oeveren W, Kazatchine MD, Descamps-Latscha B, Maillet F, Fischer E, Carpentier A. Deleterious effectof cardiopulmonary bypass: a prospective study of bubble versus membrane oxygenation. J THORAC CARDIOVASC SURG 1985;89:888-9. 5. Videm V, MollnesTE, Garred P, Svennevig JL. Biocompatibilityof extracorporeal circulation. J THORAC CARDIOVASC SURG 1991;101:654-60.
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5
6. Gu YJ, vanOeveren W, van der Kamp KWHJ, Akkerman C, Boonstra PW, Wildevuur CRH. Heparin-coating of extracorporeal circuits reduces thrombin formation in patients undergoing cardiopulmonary bypass. Perfusion 1991;6:221-5. 7. von Segesser LK, Weiss BM, Garcia E, von Felten A, Turina M. Reduction and elimination of systemic heparinization during cardiopulmonary bypass. J THORAC CARDIOVASC SURG 1992;103:790-8. 8. Okies JE, Goodnight SH, Litchford B, Connell RS, Starr A. Effects of infusion of cardiotomy suction blood during extracorporeal circulation for coronary artery surgery. J THORAC CARDIOVASC SURG 1977;74:440-4. 9. Boonstra PW, van ImhoffGW, Eysman L. Reduced platelet activation and improved hemostasis after controlled cardiotomy suction during clinical membrane oxygenator perfusions. J THORAC CARDIOVASC SURG 1985;89:900-6. 10. Marlar RA, Kleiss AJ, Griffin JH. An alternative extrinsic pathway of human blood coagulation. Blood 1982;60: 1353-9. 11. Nemerson Y. Tissue factor and hemostasis. Blood 1988; 71:1-8. 12. van Hinsbergh VWM, Kooistra T, Scheffer MA, van BoekelJH, van Muijen GNP. Characterization and fibrinolytic properties of human omental tissue mesothelial cells: comparison with endothelial cells. Blood 1990;75:1490-7. 13. Whitaker D, Papadimitrou M, Walters MNI. The mesothelium: its fibrinolytic properties. J Pathol 1982; 136:291-9.
Tabuchi et al.
833
14. Collen D. Human tissue-type plasminogen activator. Circulation 1985;72:18-20. 15. Heirnark RL, Kurachi K, Fujikawa K, Davis EW. Surface activation of blood coagulation, fibrinolysis and kinin formation. Nature 1980;286:456-60. 16. Lucas FV, Miller ML. The fibrinolytic system. Cleve Clin J Med 1988;55:531-41. 17. Moront MG, Katz NM, Keszer M, et al. Extracorporeal membrane oxygenation for neonatal respiratory failure. J THORAC CARDIOVASC SURG 1989;97:706-14. 18. Kundu SK, Salley SO, Whittlesey GC, Klein MD. Extracorporeal membrane oxygenation without anticoagulation. J Lab Clin Med 1989;114:58-62. 19. Gravlee GP, Haddon WS, Rothberger HK, et al. Heparin dosing a!l
Noriyuki Tabuchi, MD, Jacob de Haan, MSc, Piet W. Boonstra, MD, PhD, and Willem van Oeveren, PhD, Groningen, The Netherlands
Reduction of blood activation has been among the main issues in the improvement of cardiopulmonary bypass (CPB).1-4 Attempts have therefore been made to improve the biocompatibility of extracorporeal circuits (ECCs) to reduce activation of the clotting and fibrinolytic systems and to preserve platelet function.v ' The clinical results, however, are not fully satisfactory. From the Thorax Centre, University Hospital Groningen, Groningen, The Netherlands. Received for publication July 17, 1992. Accepted for publication Dec. II, 1992. Address for reprints: W. van Oeveren, PhD, Dept. of Cardiopulmonary Surgery, Research Division, University Hospital, Oostersingel 59 9713 EZ Groningen, The Netherlands. @ 1993 by Mosby-Year Book, Inc. 0022-5223/93 $1.00 +.10 12/1/45055
Copyright
828
Although the use of a heparin-coated ECC surface could reduce blood activation during CPB to some extent,6,7 blood losswas not significantly reduced. 6 Blood activation in the pericardial cavity has long been considered potentially hazardous for impairment of hemostasis.v? This blood activation is based mainly on two factors expressed in the pericardial cavity: tissue factor, a strong stimulus for the clotting system,IO, II and tissue-type plasminogen activator (t-PA), a potent activator of the fibrinolytic system. I 2-14 Blood activation in the pericardial cavity may be considerable and therefore may obscure the benefits of improved biocompatibility of ECCs. To test this hypothesis, we investigated blood activation by interrupting the blood suction from the pericardial cavity during the suturing of distal coronary anastomoses in patients undergoing elective coronary artery bypass grafting. The activation in blood from the pericardial
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5
cavity was measured separately and the systemic effect was investigated immediately after this blood was returned to the systemic circulation. Additionally, we mimickedthe effectof pericardium on blood activation by incubating human plasma with rabbit pericardial tissue. Patients and methods Clinical study. After informed consent was obtained 12 patients undergoing elective coronary artery bypass grafting entered the.study. None of the patients was older than 75 years or had evidence of severe heart failure, renal or hepatic dysfuncti~n, or bleeding diathesis. None of the patients was treated With drugs affecting the clotting system within 7 days before the operation. Operative techniques. After premedication with diazepam (10 to 15 mg), anesthesia was induced with sulfentanyl (3 to 5 ~g/kg) and pancuronium (100 to 140 ~g/kg). Ventilation was controlled by a volume-controlled respirator with air mixed with oxygen. Analgesia was provided by sulfentanyl and rnidazolan infusion. C;::~famandol (2 gm), and dexamethasone (1 mg/kg) were administered before operation. Before cannulation, bovine heparin (300 IV/kg; Leo, Emmen, The Netherlands) was injected and the addition of heparin (100 IU/kg) was repeated every 60 minutes during CPB. The activated coagulation time was determined in every patient at 5 minutes before the start of C~B and at every 60 ~inutes (International Technidyne Corp., Edison, N.J.). The activated coagulation time was confirmed to be greater than 400 seconds throughout CPB in every patient. The ECC consisted of a Cobe Excel membrane oxygenator (CO~E La~ratories, Inc., Lakewood, Colo.), and polyvinylchloride ~ubmg .. The circ~it was primed with 2000 ml oxypolygelatm (Gehfundol; Biotest AG, Dreiech, Germany) and 1500 IV bovine heparin (Leo). CPB was performed with moderate hypothermia (27 0 C nasopharyngeal temperature) with a p.ump flow of 2.4 L/m 2 per minute, maintaining a mean arterial pressure of 50 to 60 mm Hg. Myocardial preservation during aortic clamping was maintained with I L St. Thomas' Hospital cardioplegic solution (4 0 C) injected into the aortic root. No.cold s~line solution or ice slush was used in the pericardial cav~ty durmg CPB f?r topical cooling. The blood, gradually ~z~g from the surgical field into the pericardial cavity, was initially not returned to the ECC but stayed in the pericardial cavity until suturing of the distal coronary anastomoses was completed. After completion of the distal coronary anastomoses all bl~ in the pericardial cavity was returned into the ECC b; a cardiotomy sucker. Thereafter, the aortic crossclamp was released and the proximal anastomoses were performed. After CPB, heparin was neutralized by protamine chloride (3 mg/kg; Hoffman-La Roche BV, Mijdrecht, The Netherlands). Blood samples. Before the start of CPB, blood samples were taken from the radial arterial line as a control preparation. At 5 minutes before suturing of the distal anastomoses had been completed, blood samples were taken simultaneously from the pericardial cavity and from the arterial line of the oxygenator. After all blood in the pericardial cavity was returned to the ECC and to the circulation of the patient, another blood sample was taken from the arterial line of the oxygenator. After release of the aortic crossclamp, additional blood samples were collected ~y punc~ure. from the l~ft and right atria to control for the posSibleactivation of clottmg and fibrinolysis by reperfusion of the coronary and pulmonary circulations. Finally, a sample was taken at the end of CPB from the arterial line. Blood was col-
Tabuchi et al.
8 29
lected in 3.08% sodium citrate (9:1 volume/volume), stored on ice, and centrifuged at 1000gfor 10 minutes to obtain plateletpoor plasma. Laborat?ry assays. In platelet-poor plasma samples, the concentrations of t-P A antigen, thrombin-antithrombin III co~plexes (TAT), ~brin degradation products (FbDP), and fibnnoge~ deg~adatlOn products (FgDP) were determined by enzyme-hnked immunosorbent assay (Kabi Vitrum AB, Stockholm, Sweden; Behringwerke AG, Marburg, Germany; and Organon Teknika BV, Turnhout, Belgium, respectively). The he.parin ~on:~ntration ~as determined by the capacity of hepann to inhibit thrombm activity. In the presence of excess antithrombin III and thrombin, the conversion rate of a thrombin-specific substrate was determined (S2238; KabiVitrum). . In vitro st~dy. ~lood was collected from five healthy donors m 3.08% sodium ~Itrate (9:1 volume/volume), and centrifuged at I OOOg for 10mmutes to make platelet-poor plasma. Pieces of peric~r~ium were taken from anesthetized and systemically hepanmzed (100 IU /kg) rabbits. The pieces were cleaned with saline solution and cut to a standardized size shortly before the start of the experiments. All animals received care in compliance with the Dutch regulations and law regarding the care and use of laboratory animals. Generation ofthrombin. Toa I ml aliquot of human plasma, we added I 00 ~l heparin of varying concentrations (final concentrations, 1.5,2.0, and 2.5 IV/mI). These heparinized human plasma samples were mixed with I 00 ~l Ca2+ (final concentration, 12.5 mmol/L) and 100 ~l thrombin substrate (S-2238' KabiVitrum) and then incubated at 37 0 C with I em? rabbit pericardium. As a control preparation, human plasma with h~parin (.1.5 IV/mI) and Ca2+ (12.5 mmol/L) was incubated Without tissue. At the start of incubation and at 5 10 20 and 30 minutes after the start of incubation, I 00 ~l w~s t~ke~ out of the sample tube and the conversion of substrate was measured at 405 nm (Ear 400; SLT, Salzburg, Austria). The generation of thrombin activity was calculated according to the standard formula. (KabiVitrum). Generation ofplasm~n. To a I ml aliquot of human plasma, we added 100 ~l hepann of varying concentrations (final concentrations, 1.5, 2.0, and 2.5 IV/mI) or heparin (final concentration, 1.5 I~/mI) plus hirudin (final concentration, 0.5 V /ml), These hepanmzed human plasma samples were mixed with 100 ~l Ca2+ (final concentration, 12.5 mmol/L) and then incubated a~ 37 0 C with I ern? rabbit pericardium. As a control preparation, human plasma with heparin (1.5 IV/mI) and Ca2+ (12.5 mrnol/L) was incubated without tissue. At the start of incubation and at 5, 10, 20, and 30 minutes after the start of incub~tion, 80 ~l was taken out of the sample tube and 40 ~l plasmm substrate (S-2251; KabiVitrum) was added. After another 15, 30, and 60 minutes of incubation, the conversion of substrate was measured at 405 nm. Plasmin activity was calculated according to the standard formula (KabiVitrum). ~tatistic~. Statistical analysis was performed by a paired Wilcoxon ~Igned-rank test in the clinical study and by the Mann-Whitney test in the in vitro study. All data are expressed as mean ± standard deviation of the mean. A p value less than 0.05 was considered significant.
Results Clinical study. The demographic data from the patients were as follows: age, 61 ± 2 years; weight, 81 ± 3 kg; CPB time, 93 ± 10minutes;crossclamptime,
The Journal of Thoracic and Cardiovascular Surgery November 1993
8 3 a Tabuchi et al.
3
Heparin (IU/ml)
TAT (ng/ml)
200
150
4 FgDP (Ilg /ml)
3
**
40 FbDP (J.lg/ml)
40 t-PA (ng/ml)
30
30
*
2 2
20
100 10
50
o
o
0
Fig. 1. Plasma heparin and TAT concentrations in systemic blood (dark column) compared with blood from thepericardial cavity (light column) before admixture. Thesamples were collected at theendofsuturing ofthedistal coronary anastomosis. Double asterisk represents a significant difference between the two samples (p < 0.01). 54 ± 25 minutes; venous grafts, 0.4 ± 0.6; and arterial grafts, 1.5 ± 0.7. In total, 637 ± 32 ml shed blood was collectedin the pericardial cavityduring 57.3 ± 6.3 minutes before the end of suturing distal coronary anastomosis. In shed blood of the pericardial cavity, the plasma concentrationof heparin waslessthan half (p < 0.0I) and that of TAT was three (p < 0.01) times that in systemic circulating blood(Fig. I). Whereas t-PA antigen was two times higher (p < 0.05), the concentrationsof FgDP and FbDP were 10 times higher (p < 0.Q1) in the bloodof the pericardial cavity than in the circulation (Fig. 2). In contrast, blood collectedafter reperfusion from the left and right atria, indicating possible reperfusion-dependent blood activation, contained similar concentrationsofTATcomplexes(48 ± 41 ng/rnl versusSZ ± 49 ng/rnl) and FbDP (3061 ± 2329 ng/rnl versus 3176 ± 2622 ng/ml). Only after the blood in the pericardial cavity was returned to the systemic circulation after resumed suction were increasesof TAT (p < 0.05), FgDP (p < 0.05), and FbDP (p < 0.05) observed in systemic circulating blood (Fig. 3). In vitro study Thrombin and plasmin generation. A significant increase in thrombin and plasmin activity was observed after 5 minutes of incubation with rabbit pericardium at l.5or 2.0 IV Iml heparin (p < 0.05) (Fig. 4). In the presence of a 2.5 IV /rnl concentration of heparin, thrombin and plasmin activity was almost absent and was thus significantlyless than at 1.5 or 2.0 IV Iml concentrationsof
0+--.......
Fig. 2. Activation products in systemic blood (dark column) compared with blood from thepericardial cavity (light column) before admixture. Theplasma concentrations ofFgDP, FbDP, andt-PAantigen were measured at theendofsuturing thedistal coronary anastomosis. Asterisk and double asterisk representa significant difference between two samples (p < 0.05 and p < 0.01, respectively). heparin (p < 0.01 and p < 0.05, respectively). Without the introduction of pericardial tissue, thrombin and plasmin activity remained low. The activity of plasmin was reduced significantly, whenhirudin wasadded to samples with the lowest heparin concentration (1.5 Il.I/rnl) (p < 0.05) (Fig. 5). Discussion
Bloodactivationthrough the intrinsicclottingpathway from contact of the bloodwith the large artificialsurface of the ECC has been a main issuein the development of CPRI, 15 Evenwhen the materials of the ECC have been improved and modified by heparin coating however, the activation of the clotting system cannot be significantly suppressed during CPR6 Blood activation by tissue in pericardium should therefore be evaluated as occurring by another activationpathway during CPR Tissuefactor stimulates thrombin generation even more strongly through the extrinsicclottingpathway than doesthe ECC through the intrinsicclottingpathway.10, 11 Moreover, the activationof fibrinolysis, initiated by t-PA, is known to be accelerated lOO-fold in the presence of fibrin or fibrin monomers derived from clotting activity.!" Our study shows that the activation of fibrinolysis was more pronouncedthan that of the clottingsystemin the pericardial cavity.Even a small activationof the clotting systemcorresponded with considerable activation of fibrinolysis in the presence of abundant t-PA in the pericardial cavity. The inhibition of fibrinolysis by the addition of the thrombin inhibitor hirudin in the in vitro study supports this idea. Because of the existence of tissue factor and
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5
2500
~c
Tabuchi et at.
FgDP
TATcompTex
FbDP
6000
150
5000
2000
*
1500
*
4000
83 I
*
*
*
1000 500
0
Fig. 3. Activation products in the systemiccirculation. TAT, FgDP, and FbDP were measured before CPB (dark column), at 5 minutes before the end of suturing of the distal coronary anastomosis (medium gray column), after all bloodin the pericardialcavitywas returned to the systemiccirculation (light column), and at the end of CPB (open column). Asterisk represents a statistically significant increase after return of pericardial blood (p < 0.05).
..... E
200
..... ~
:i 'ot' W 150 0
:::)
0 II>
W
... ~
... 0
E 2
100
~
c
:a
150
c
E
100
II)
.!! e,
.c
l-
50
50
A
10
20
30
Time (min)
B
10
20
30
Time (min)
Fig. 4. Thrombin (A) and plasmin (B)generation by rabbit pericardium were measured in human plasma in the presenceof variousconcentrationsof heparin (1.5 IV Iml, closed squares; 2.0 IV Irnl, circles; 2.5, IV Iml triangles). As a control preparation, human plasma was incubated without tissue at the lowest concentration of heparin (1.5 IV /rnl, open squares). Asterisk indicates a statisticallysignificant increaseafter 5 minutes of incubation (p < 0.05). Double asterisk indicatesa significantdifferencebetweenthe highestconcentrationof heparin (2.5 IV I ml) and lower concentrations of heparin (1.5 and 2.0 IV Irnl), (p < 0.01 and p < 0.05, respectively).
t-PA, the blood activation in the pericardial cavity could therefore be far greater than the blood activation caused by the improved surfaces of modern ECCs. Clinically, this hypothesis is supported by the excellent performance of oxygenators without cardiotomy suction even under the circumstance ofdecreased systemic heparinization.?- 17, 18 It can be expected that the incubation time of blood in the pericardial cavity correlates with the intensity of blood activation. In our protocol, we retained the suctioned blood for 60 minutes during suturing of the coronary
anastomoses in the pericardial cavity, where heparin appeared to be consumed more than 50%. Our in vitro study clearly showed that this reduction of heparin resulted in suboptimal heparin concentrations, insufficient to prevent the activation of clotting and indirectly of fibrinolysis. This intensity of blood activation could be minimized if the blood pooling time in the pericardial cavity was shortened by continuous suction. Additional anticoagulation seems required, however, because our in vitro study showed that even a short incubation time
832
-E ....
The Journal of Thoracic and Cardiovascular Surgery November 1993
Tabuchi et al.
150
~
0
It) I
W
,..
0
--
100
>C
*
C
E Ul
*
.!! D.
*
50
O~----~----,.....------r
o
10
20 30 Time (min)
Fig. 5. Plasmin generation inhibited was by the addition of hirudin. Plasmin generation was measured in heparinized human plasma (1.5 IV Im1, squares) accordingto the incubation time with rabbit pericardium,withor without hirudin (0.5 V Im1, diamonds). Asterisk represents significant difference between sampleswith and without hirudin (p < 0.05). results in considerable thrombin and plasmin generation under conditions of low heparin concentration. These results raise the question of whether the systemic anticoagulation was sufficient in our present study. In our protocol, we used an initial heparin dose of 300 IV Ikg, followed by an additional dose of 100 IV Ikg every 60 minutes. Also, the activated clotting time was kept at greater then 400 seconds throughout CPRI9 The acceptable minimum value of anticoagulation is still under debate, however, because subclinical coagulation activity always exists during CPB and does not correlate with systemic heparin concentrations.l? We think that this discrepancy may be a consequence of the peculiar blood activation in the pericardial cavity. Even if anticoagulation is sufficient in the ECC, it is not necessarily sufficient in the pericardial cavity. If the biocompatibility of ECC surfaces become better, the discrepancy in thrombogenicity between ECC and tissue will become larger. To manage this discrepancy, it may be preferable to use topical heparin in the pericardial cavity to maintain a sufficient local anticoagulation. The FDPs generated in the pericardial cavity are returned to the systemic circulation during CPB and thereby affect hemostasis through interference with platelet receptors, fibrinogen binding to platelets, and clot formation.P: 21 Also, potentially active products from the
fibrinolytic system may be reinfused as a result of a shortage of plasmin inhibitors in the pericardial blood.22 In our technique of CPB, the amount of blood coming back from the pericardial cavity was about I L throughout the whole CPB period in an uncomplicated coronary artery bypass grafting operation. The local activation in the pericardial cavity may thus considerably contribute to total blood activation during CPR In the in vitro study, we used rabbit tissue to activate human plasma, a documented laboratory practice. 10 It is also known that there is a cross-reaction oft-PA between human and the other mammalian species." Our in vitro set-up therefore essentially mimics the clinical situation. The cells in the mesothelium are known to secrete physiologically abundant t-PA to avoid adhesion between organs, 12,13 The fact that the blood remaining in the thoracic cavity or in the pericardial cavity is defibrinated in the patients with chest trauma exemplifies the potent fibrinolytic activity there. 23 Our study demonstrates that the same process may be going on in the pericardial cavity during CPB. In conclusion, we found fulminant fibrinolysis initiated by clotting activation in blood of the pericardial cavity in the presence oftissue factor and t-PA. This high level of blood activation was expressed because heparin was locally consumed in the pericardial cavity. This local blood activation significantly contributed to the total blood activation during CPB. The pericardial blood therefore should be aspirated continuously to minimize blood activation and the topical use of heparin should be considered to maintain sufficient anticoagulation in the pericardial cavity, which appears to be the most thrombogenic site during CPR We thank J. Haan and F. Wei for technicalassistance. REFERENCES 1. van OeverenW, WildevuurCRH, KazatchkineMD. Biocompatibility of extracorporea1 circuits in heart surgery. Transfus Sci 1990;11:5-33. 2. Harker LA. Bleeding after cardiopulmonary bypass. N Engl J Med 1986;314:1446-7. 3. KirklinJK, Westab1y S, Blackstone EH, Kirklin JW, ChenowethDE, Pacifico AD. Complementand the damaging effects ofcardiopulmonary bypass. J THORAC CARDIOVASC SURG 1983;86:845-57. 4. van Oeveren W, Kazatchine MD, Descamps-Latscha B, Maillet F, Fischer E, Carpentier A. Deleterious effectof cardiopulmonary bypass: a prospective study of bubble versus membrane oxygenation. J THORAC CARDIOVASC SURG 1985;89:888-9. 5. Videm V, MollnesTE, Garred P, Svennevig JL. Biocompatibilityof extracorporeal circulation. J THORAC CARDIOVASC SURG 1991;101:654-60.
The Journal of Thoracic and Cardiovascular Surgery Volume 106, Number 5
6. Gu YJ, vanOeveren W, van der Kamp KWHJ, Akkerman C, Boonstra PW, Wildevuur CRH. Heparin-coating of extracorporeal circuits reduces thrombin formation in patients undergoing cardiopulmonary bypass. Perfusion 1991;6:221-5. 7. von Segesser LK, Weiss BM, Garcia E, von Felten A, Turina M. Reduction and elimination of systemic heparinization during cardiopulmonary bypass. J THORAC CARDIOVASC SURG 1992;103:790-8. 8. Okies JE, Goodnight SH, Litchford B, Connell RS, Starr A. Effects of infusion of cardiotomy suction blood during extracorporeal circulation for coronary artery surgery. J THORAC CARDIOVASC SURG 1977;74:440-4. 9. Boonstra PW, van ImhoffGW, Eysman L. Reduced platelet activation and improved hemostasis after controlled cardiotomy suction during clinical membrane oxygenator perfusions. J THORAC CARDIOVASC SURG 1985;89:900-6. 10. Marlar RA, Kleiss AJ, Griffin JH. An alternative extrinsic pathway of human blood coagulation. Blood 1982;60: 1353-9. 11. Nemerson Y. Tissue factor and hemostasis. Blood 1988; 71:1-8. 12. van Hinsbergh VWM, Kooistra T, Scheffer MA, van BoekelJH, van Muijen GNP. Characterization and fibrinolytic properties of human omental tissue mesothelial cells: comparison with endothelial cells. Blood 1990;75:1490-7. 13. Whitaker D, Papadimitrou M, Walters MNI. The mesothelium: its fibrinolytic properties. J Pathol 1982; 136:291-9.
Tabuchi et al.
833
14. Collen D. Human tissue-type plasminogen activator. Circulation 1985;72:18-20. 15. Heirnark RL, Kurachi K, Fujikawa K, Davis EW. Surface activation of blood coagulation, fibrinolysis and kinin formation. Nature 1980;286:456-60. 16. Lucas FV, Miller ML. The fibrinolytic system. Cleve Clin J Med 1988;55:531-41. 17. Moront MG, Katz NM, Keszer M, et al. Extracorporeal membrane oxygenation for neonatal respiratory failure. J THORAC CARDIOVASC SURG 1989;97:706-14. 18. Kundu SK, Salley SO, Whittlesey GC, Klein MD. Extracorporeal membrane oxygenation without anticoagulation. J Lab Clin Med 1989;114:58-62. 19. Gravlee GP, Haddon WS, Rothberger HK, et al. Heparin dosing a!l