Arrificial Organs
15(?):%95, Raven Press. Ltd.. New York Q 1991 International Society for Artificial Organs
Reduced Granulocyte Activation with a Heparin-Coated Device in an In Vitro Model of Cardiopulmonary Bypass Vibeke Videm, *Leif Nilsson, ?Per Venge, and $Jan L. Svennevig Institute for Experimental Medical Research and $Surgical Department, University of Oslo, Ullevaal Hospital, Oslo, Norway, and Departments of *Thoracic Surgery and fClinical Chemistry, University Hospital, Upgsala, Sweden
Abstract: Granulocyte activation during cardiopulmonary bypass (CPB), resulting in degranulation, may have adverse effects. Fresh whole human blood and priming solution was circulated through oxygenator/tubing sets coated with functional heparin (n = 7) and through uncoated sets (n = 7) in modei CPB. Plasma concentrations of the primary granule protein myeloperoxidase (MPO) and the secondary granule protein lactofenin (LF) were measured in radioimmunoassays, and the neutrophils were counted. After 120 min, seven to nine times baseline concentrations of LF (p < 0.OOOl) were observed with both devices. Increases of MPO were also significant, but
significantly larger (p c 0.01) with the uncoated devices. There was an equivalent reduction in neutrophil numbers in both groups. MPO did not bind to heparin-coated Sephadex particles in gel chromatography. Thus, the heparin coating most likely prevented the release of potentially harmful primary granule proteins, indicating improved biocompatibility. Adhesion of neutrophils and exocytosis of LF, which may be involved in adhesion, were unaffected. Key Words: Biocompatibility-Cardiopulmonary bypass-Heparin coating-Neutrophil activationM yeloperoxidase-Lactofemn.
Contact between the blood and various nonbiologic components of an extracorporeal circuit may lead to adverse effects, as are recognized during hernodialysis (1) and after cardiopulmonary bypass (CPB) (2). T h e “whole-body inflammatory reaction” that sometimes develops after CPB comprises increased capillary permeability with fluid accumulation in the tissues, fever, bleeding diathesis, and potentially lethal dysfunction of various organs, e.g., the lungs, heart, or kidneys. A recent study has shown that elevated concentrations of interleukin-I generated by circulating monocytes shortly after CPB are linearly correlated with the postoperative increase in body temperature (3). Activation of the complement cascade also seems to be an important trigger of the extensive systemic inflammatory response (2). The activation product C5a binds to neutrophil granulocytes via a specific receptor, and binding is followed by de-
granulation and release of granule contents into the circulation. High plasma levels of granule proteins such as myeloperoxidase (MPO), lactofemn (LF), and elastase have been observed during CPB and hernodialysis (4-9). The granule components are biologically active substances with proteolytic and cytolytic properties (10). Liberation of mediators from activated granulocytes trapped in the lungs have been proposed as a cause of pulmonary dysfunction seen after hemodialysis (1). Degranulation of neutrophils and eosinophils has also been put forward as an important pathogenetic factor in the development of adult respiratory distress syndrome (11). Breda and coworkers (12) found that reperfusion of isolated ischemic pig hearts with leukocyte-depleted blood prevented ultrastructural damage and resulted in better myocardial function than when normal blood was used. Although the clinical significance has not been fully established, it seems conceivable that a sudden massive release of granule components could be detrimental to the organism and may be partly responsible for postoperative complications
-
- Received September 1990, revised November 1990. Address correspondence and reprint requests to Dr.V. Videm at Institute for Experimental Medical Research. University of Oslo. Ullevaal Hospital, N-0407 Oslo, Norway.
REDUCED GRANULOCYTE ACTIVATION IN CPB after surgery involving CPB. Correspondingly, the liberation of elastase during hemodialysis has been used as an indicator of the biocompatibility of dialysis membranes (8). Most probably the granulocyte activation during CPB takes place at the foreign polymer surfaces (5). Heparin coating of the device improved biocompatibility in the form of better thromboresistance (13) and represents an interesting development of extracorporeal technology. Because the foreign surfaces are entirely covered with a native substance in such a device, the present study was undertaken to compare granulocyte activation in heparin-coated and uncoated devices in vitro. MATERIALS AND METHODS
Experimental procedure Fresh whole human blood from informed volunteer donors was collected in bags (Laboratoires Travenol, La Ch$tre, France) to which 4,000 IU of heparin had been added. Fourteen Maxima hollowfiber oxygenators (Medtronic, Anaheim, California, U.S.A.) were set up with tubings and a cardiotomy reservoir in a standardized manner to form an extracorporeal circuit. Seven sets were delivered with the entire blood-contact surface coated with endpoint attached active heparin, Carrneda Bio-Active Surface (13) (Carmeda a s , Oslo, Norway). In brief, this method consists of surface amination with polyethyleneimine followed by incubation with partially degraded heparin, NaBN,CN, and NaC1, giving covalent bonding of terminal aldehyde groups of the heparin fragments to the amino groups on the surface. In a previous study, the resulting heparinized surface completely inhibited thrombin activity in defibrogenated plasma when tested with a synthetic chromogenic peptide assay (13). Fresh, normal human blood did not coagulate when exposed to the surface, and the platelet count in the exposed blood did not change significantly. The remaining seven sets were uncoated. Coated and uncoated sets were used in random order. Each set was primed with 125 ml dextran 60 mgl ml in saline, 125 ml Ringer’s acetate, 125 ml glucose 50 mg/ml, 125 ml mannitol 150 mg/ml, and 50 ml sodium hydrogen carbonate 500 mmol/L. A baseline blood sample was drawn from the bag before 1 U of blood was filled into the circuit. The temperature was maintained at 30°C by means of a heat exchanger and the bloodprimer mixture was circulated through the setup at 4 L/min for 2 h, using a nonocclusive roller pump (Gambro, Horten, Norway). Test samples were obtained at 0, 15, 30, 60,
91
and 120 min. Bloodprimer controls (n = 14) were drawn into glass tubes from the circuit at 0 min and were kept in a 30°C water bath. Control samples were drawn from these tubes at 60 and 120 min. All test and control samples were drawn in glass tubes containing EDTA and were kept on ice until analysis or centrifugation shortly thereafter. Plasma was stored at -70°C. Because MPO is a polycation and heparin is a polyanion, additional experiments were performed to investigate possible binding of MPO to the heparin coating. Sephadex G15 gel (Pharmacia Fine Chemicals, Uppsala, Sweden) was coated with h e p arin employing the same method as used in the CPB circuits (13). Pasteur pipettes were then packed with 5 cm of heparin-coated Sephadex G15. Pipettes packed with untreated Sephadex G 15 or aminated Sephadex G15 were used as controls. Three different solutions with increasing concentrations of MPO were prepared. Sample 1 contained 1.5 ml of pooled human serum and 450 p1 of a buffer solution (phosphate buffer 0.5 M 100 ml, EDTA 0.1 M 100 ml, NaN3 2% 10 ml, bovine serum albumin 10% 20 ml, NaCl 9% 50 ml, Tween 5 ml, H20millipore ad 1,OOO ml) with 0.2% N-cetyl-N,N,N-trimethyl ammonium bromide. Sample 2 contained 1.5 ml of the pooled serum and 450 p,1 of a solution with 1,600 pg/L, of purified MPO (14). Sample 3 contained 1.5 ml of the pooled serum and 450 pI of a solution with 3,200 pg/L of purified MPO. Five hundred microliters of each sample was passed through each type of gel. Test samples for measurement of MPO concentrations were obtained before and after the gel filtration. Analysis of samples The numbers of leukocytes and the hemoglobin concentration (Hgb) in the samples were determined in an electronic counter (Coulter Electronics, Harkenden, England). Neutrophil counts were calculated from May/Griinwald/Giemsa-stained blood smears. MPO and LF were quaniitated by radioimmunological technique (15,16). The results were corrected for hemodilution by multiplication with the factor: initial Hgbhample Hgb. StatiStiCS
The Friedman test (17) was used for analyses of time-dependent changes within each group. The Wilcoxon rank sum test and the Kruskal-Wallis test (17) were applied for intergroup comparisons. Results are presented as median based on Walsh numbers with the 95% nonparametric confidence interval in parentheses. Arr$ Orgons, Vol. 15, No.2. 1991
V . VIDEM ET AL.
92
NC diff
RESULTS Neutrophil counts are shown in Fig. 1. The initial values for the coated and uncoated groups were not significantly different. During the experiment there was a significant reduction of neutrophil counts in both groups. The reduction was equivalent for uncoated and coated devices (Fig. 2). MPO concentrations increased significantly in both groups (Fig. 3). In the controls there was a slight but significant rise after 120 min. The increases from 0 to 60 min and from 0 to 120 min were significantly lower in the coated than in the uncoated group (Fig. 4). LF concentrations rose significantly during the experiment in both oxygenator groups and in the controls (Fig. 5 ) . There were no signifkant differences between the uncoated and coated groups (Fig. 6), but the levels in both oxygenator groups were significantly higher than in the controls after 60 and 120 min. When serum containing various known amounts of MPO was passed through columns of heparincoated or uncoated Sephadex particles, the recovery of MPO was close to 100% (Table 1). In contrast, the aminated Sephadex particles bound approximately 90% of the MPO present in each
2
0
uncoated heparin coated
1
0 0-60
0-120 min
FIG.2. Reduction in neutrophil counts (NCdiw (median with 95% confidence intervals) after circulation for 60 and 120 min during in vitro cardiopulmonary bypass. Differences between coated and uncoated groups are not statistically significant.
DISCUSSION
6-
0 uncoated heparincoated
7 -
6 -
4 1
I
I
T
I
I
I
1
1
0
15
30
60
120
mln
FIG. 1. Neutrophil counts (NC)(median with 95% confidence intervals) versus circulation time through heparin-coated and uncoated cardiopulmonary bypass circuits in vitro. Asterisks denote statistically significant changes compared with initial value: 'p < 0.05, "p < 0.01, ""p < 0.0001.
Arfi/Organs. bol I S . N o 2 . 1991
In the present study there was an equivalent reduction in granulocyte counts in CPB circuits coated with functionally active heparin and in uncoated circuits. The plasma MPO concentrations were approximately five times higher in the uncoated group, whereas there were no significant differences in LF concentrations. One possible explanation for the reduced circulating levels of MPO in the heparin-coated group was binding of MPO to the heparin. However, no binding affinity to'heparin was found in the control experiments, in which MPO-enriched serum was passed through heparincoated Sephadex gel. The validity of the control experiments is supported by the fact that MPO showed a high affinity to the aminated gel. Thus, the authors infer that MPO liberation from the granulocytes was significantly lower after heparin coating. Most investigators find a transient drop in the number of circulating neutrophils during CPB (18,19). The initial fall during perfusion may be explained by adhesion to polymer surfaces and filters or by trapping of granulocytes in the lung circulation. Because there was no filter in the CPB circuit
REDUCED GRANULOCYTE ACTIVATION IN CPB
93
3058
700
600
FIG. 3. Plasma concentrations of myeloperoxidase (MPO) (median with 95% confidence intervals) versus circulation time through heparincoated and uncoated circuits during in vitro cardiopulmonary bypass and in control samples. Asterisks denote statistically significant changes compared with initial value: ‘p < 0.05, ”p < 0.01, **-p < o.Ooo1.
400
300
0
15
30
60
120
min
in the present study, the decline in neutrophil numbers was most probably caused by adhesion to the polymer surfaces. The similar magnitude of the reduction with coated and uncoated devices indicates that the differences in surface properties did not influence neutrophil adherence. Boxer et al. (20) observed neutropenia and margination of neutrophils in the microvasculature of the hamster cheek pouch after infusion of LF, providing evidence that L F is involved in neutrophil adherence to the endothelial cell. Based on the equivalent L F increases in the two oxygenator groups, the authors postulate that LF may be involved in the neutrophil adherence to polymer surfaces, whether they are coated with heparin or not. The mechanisms of neutrophil activation during extracorporeal circulation are not fully understood. Craddock and co-workers (1) allege that activated complement is responsible. Other studies show that activation of complement and granulocytes are not always parallel phenomena. Horl et al. (21) found considerable granulocyte activation accompanied by only little complement activation during hemodialysis, which points at a complement-independent
mechanism. Hdlgren et al. ( 5 ) found no granulocyte-activating properties of serum from dialyzed patients and assumed that degranulation in this setting took place locally from cells attached to the membranes. The authors found that the liberation of L F from secondary granules was not affected by heparin coating. On the other hand, the release of MPO from primary granules was substantially reduced. According to Gallin (lo), exocytosis of secondary granules occurs with milder stimuli such as nonspecific surface adhesion, whereas ‘exocytosis of primary granules generally requires more powerful stimuli. The heparin-coated surface seems to permit neutrophil adhesion and liberation of L F but not further activation with release of the more potent primary granule components such as MPO. The polymer surface is thus rendered more “endothelium-like.” The mechanism for this effect of the heparin coating is not clear. CONCLUSION It is concluded that modification of the polymer surfaces in a model CPB circuit by heparin coating
Arrf Organs. Vol.
IS, No. 2 , 1991
V . VIDEM ET AL.
94
0 uncoated
700
600
1
0 uncoated I3 heparin coated
c) heparin coated
w control
control 1426
2938
1618
1000
-
800
-
500
600
400 -
400
1
200
T
C
0-60
0-120, min
0-60
2006 1037
flG. 4. Increase in myeloperoxidase concentratiqns (MPO diff) (median with 95% confidence intervals) after circulation for 60 and 120 rnin during in vitro cardiopulmonary bypass and in control samples. Statistically significant differences after So rnin: uncoated vs. coated p < 0.001, uncoated vs. control p < 0.001, coated vs. control p < 0.05; after 720 min: uncoated vs. coated p < 0.01. uncoated vs. control p < O.OOO1, coated vs. control p < 0.01.
0-120 min
FIG. 6. Increase in plasma lactoferrin concentrations (LF diff) (median with 95% confidence intervals) after circulation for 60 and 120 rnin during in vitro cardiopulmonary bypass and in control samples. Statistically significant differences after 60 min: uncoated vs. coated not significant (NS),uncoated vs. control p < 0.001, coated vs. control p < 0.01; after 720 min: uncoated vs. coated NS, uncoated vs. control p < 0.0001, coated vs. control p < O.OOO1.
LF
(PS-J 1)
2209
lo00
heperincoated
900Boo700
FIG. 5. Plasma concentrations of lactoferrin (LF)(median with 95% confidehce intervals) versus circulation time through heparin-coated and uncoated circuits during in vitro cardiopulmonary bypass and in control samples. Asterisks denote statistically significant changes compared with initial value: 'p < 0.05, "p < 0.01, 49-p < 0.0001.
-
boo-
I
0
15
30
60
120
mln
ArrfOrgons. V o l . I S . N o . 2. 1991
REDUCED GRANULOCYTE ACTIVATION IN CPB TABLE 1. Concentration of myeloperoxidase (micrograms per liter) in three human serum samples before and after filtration through three Sephadex gel preparations Before filtration After Sephadex GI5 After aminated Sephadex G15 After heparin-coated Sephadex GI5
Sample 1
Sample 2
Sample 3
280 252
643 63 1
1,250 1,300
34
85
165
385
686
1,370
leads to alteration of the granulocyte response. Adhesion of neutrophils was unaffected, whereas further activation was significantly reduced. The authors interpret these findings as an indication of improved biocompatibility, which may prove clinically advantageous in cardiac surgery. Acknowledgment: We are grateful for excellent technical assistance from Mr. Hans Nielsen and the staff of the Surgical Department Laboratory and the Blood Bank,U1levaal Hospital, and Ms. Kerstin Lindblad, Department of CIinical Chemistry, University Hospital, Uppsala. The study was supported by Anders Jahres Fund for the Promotion of Science, the Norwegian Council on Cardiovascular Research, and the Swedish Medical Research Council.
REFERENCES I . Craddock PR, Fehr J, Brigham KL, Kronenberg RS, Jacob HS. Complement and leukocyte-mediated pulmonary dysfunction in hemodialysis. New Engl J Med 1977;2%:769-74. 2. Kirklin JK, Westaby S, Blackstone EH. Kirklin JW, Chenoweth DE, Pacific0 AD. Complement and the damaging effects of cardiopulmonary bypass. J Thorac Cardiovasc SUrg 1983;86:845--57. 3. Haeffner-Cavaillon N, Roussellier N, Ponzio 0, Carreno MP, Laude M, Carpentier A, Kazatchkine MD. Induction of interleukin-1 production in patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1989;98:1100-6. 4. Nilsson L, Brunnkvist S, Nilsson U, NysMm SO, Tydtn H, Venge P, Aberg T. Activation of inflammatory systems during cardiopulmonary bypass. Scand J Thorac Cardiovasc Surg 1988;22:51-3. 5 . Hallgren R, Venge P, Wikstr8m B. Hemodialysis-induced increase in serum lactofemn and serum eosinophil cationic
95
protein as signs of local neutrophil and eosinophil degranulation. Nephron 1981;29:233-8. 6. Wachtfogel YT, Kurich U, Greenplate J, Gluszko P, Abrams W. Weinbaum G, Wenger RK, Rucinski B. Niewiarowski S, Edmunds LH, Colman RW. Human neutrophil degranulation during extracorporeal circulation. Blood 1987;69:32430. 7. Antonsen S, Brandslund J. Clemensen S , Soefeldt S , Madsen T, Alstrup P. Neutrophil lysosomal enzyme release and complement activation during cardiopulmonary bypass. Scand J Thorac Cardiovasc Surg 1987;21:41-52. 8. Knudsen F, Nielsen AH, Pedersen JO, Jersild C. On the kinetics of the complement activation, leukopenia and gmnulocyte-elastase release induced by haemodialysis. Scand J CIin Lob Invest 1985;45:75M. 9. Riegel W, Spillner B, Schlosser V, Hdrl WH. Plasma levels of main granulocyte components during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;95:1014-9. 10. Gallin JI. Neutrophil specific granules: a fuse that ignites the inflammatory response. Clin Res 19&;32:320-8. 1 1 . HUgren R, Borg T, Venge P, Modig J. Signs of neutrophil and eosinophil activation in adult respiratory distress syndrome. Crir Care Med 1984;12:14-8. 12. Breda MA, Drinkwater DC,Laks H, Bhuta S, Como AF, Davtyan HG, Chang P. Prevention of reperfusion injury in the neonatal heart with leukocyte-depleted blood. J Thorac Cardiovasc Surg 1989;654-65. 13. Larm 0.Larsson R, Olsson P. A new non-thrombogenic surface prepared by selective covalent binding of heparin via a modified reducing terminal residue. Biomater Med Devices Arrif Organs 1983;1 1:161-73. 14. Olsson I, Venge P. Cationic proteins of human granulocytes. 11. Separation of the cationic proteins of the granules of leucemic myeloicf cells. Blood 1974;44:235-46. IS. Olofsson T, Olsson J, Venge P. Myeloperoxidase and lactofemn of blood neutrophils and plasma in chronic granulocytic leukemia. Scand J Haemafol 1977;18:113-20. 16. Olofsson T, Olsson I, Venge P, Elgefors B. Serum myeloperoxidase and lactofemn in neutropenia. Scand J Haemato1 1977;1 8 : 7 M . 17. Conover WJ. Practical nonparmerric statistics, 2nd ed. New York: Wiley, 1980:229-37, 299-300. 18. Fosse E, Mollnes TE, Ingvaldsen B. Complement activation during major operations with or without cardiopulmonary bypass. J Thorac Cardiovasc Surg 1987;93:860-6. 19. van Oeveren W, Kazatchkine MD,Dcscamps-Latscha B, Maillet F, Fischer E, Carpentier A, Wildevuur CRH. Deleterious effects of cardiopulmonary bypass. J Thorac Cardiovasc Surg 1985;89888-99. 20. Boxer LA, Bjarksttn B, BjBrk J, Yang HH, Allen JM, Baehner RL. Neutropenia induced by systemic infusion of lactofenin. J Lub CIin Med 1982;99:866-72. 21. Hbrl WH, Riegel W, Schollmeyer P, Rautenberg W, Neumann S. Different complement and granulocyte activation in patients dialyzed with PMMA dialyzers. Clin Nephrol 1986;25:30&7.
Artf Organs. Vol. 15. No. 2, 1991
15(?):%95, Raven Press. Ltd.. New York Q 1991 International Society for Artificial Organs
Reduced Granulocyte Activation with a Heparin-Coated Device in an In Vitro Model of Cardiopulmonary Bypass Vibeke Videm, *Leif Nilsson, ?Per Venge, and $Jan L. Svennevig Institute for Experimental Medical Research and $Surgical Department, University of Oslo, Ullevaal Hospital, Oslo, Norway, and Departments of *Thoracic Surgery and fClinical Chemistry, University Hospital, Upgsala, Sweden
Abstract: Granulocyte activation during cardiopulmonary bypass (CPB), resulting in degranulation, may have adverse effects. Fresh whole human blood and priming solution was circulated through oxygenator/tubing sets coated with functional heparin (n = 7) and through uncoated sets (n = 7) in modei CPB. Plasma concentrations of the primary granule protein myeloperoxidase (MPO) and the secondary granule protein lactofenin (LF) were measured in radioimmunoassays, and the neutrophils were counted. After 120 min, seven to nine times baseline concentrations of LF (p < 0.OOOl) were observed with both devices. Increases of MPO were also significant, but
significantly larger (p c 0.01) with the uncoated devices. There was an equivalent reduction in neutrophil numbers in both groups. MPO did not bind to heparin-coated Sephadex particles in gel chromatography. Thus, the heparin coating most likely prevented the release of potentially harmful primary granule proteins, indicating improved biocompatibility. Adhesion of neutrophils and exocytosis of LF, which may be involved in adhesion, were unaffected. Key Words: Biocompatibility-Cardiopulmonary bypass-Heparin coating-Neutrophil activationM yeloperoxidase-Lactofemn.
Contact between the blood and various nonbiologic components of an extracorporeal circuit may lead to adverse effects, as are recognized during hernodialysis (1) and after cardiopulmonary bypass (CPB) (2). T h e “whole-body inflammatory reaction” that sometimes develops after CPB comprises increased capillary permeability with fluid accumulation in the tissues, fever, bleeding diathesis, and potentially lethal dysfunction of various organs, e.g., the lungs, heart, or kidneys. A recent study has shown that elevated concentrations of interleukin-I generated by circulating monocytes shortly after CPB are linearly correlated with the postoperative increase in body temperature (3). Activation of the complement cascade also seems to be an important trigger of the extensive systemic inflammatory response (2). The activation product C5a binds to neutrophil granulocytes via a specific receptor, and binding is followed by de-
granulation and release of granule contents into the circulation. High plasma levels of granule proteins such as myeloperoxidase (MPO), lactofemn (LF), and elastase have been observed during CPB and hernodialysis (4-9). The granule components are biologically active substances with proteolytic and cytolytic properties (10). Liberation of mediators from activated granulocytes trapped in the lungs have been proposed as a cause of pulmonary dysfunction seen after hemodialysis (1). Degranulation of neutrophils and eosinophils has also been put forward as an important pathogenetic factor in the development of adult respiratory distress syndrome (11). Breda and coworkers (12) found that reperfusion of isolated ischemic pig hearts with leukocyte-depleted blood prevented ultrastructural damage and resulted in better myocardial function than when normal blood was used. Although the clinical significance has not been fully established, it seems conceivable that a sudden massive release of granule components could be detrimental to the organism and may be partly responsible for postoperative complications
-
- Received September 1990, revised November 1990. Address correspondence and reprint requests to Dr.V. Videm at Institute for Experimental Medical Research. University of Oslo. Ullevaal Hospital, N-0407 Oslo, Norway.
REDUCED GRANULOCYTE ACTIVATION IN CPB after surgery involving CPB. Correspondingly, the liberation of elastase during hemodialysis has been used as an indicator of the biocompatibility of dialysis membranes (8). Most probably the granulocyte activation during CPB takes place at the foreign polymer surfaces (5). Heparin coating of the device improved biocompatibility in the form of better thromboresistance (13) and represents an interesting development of extracorporeal technology. Because the foreign surfaces are entirely covered with a native substance in such a device, the present study was undertaken to compare granulocyte activation in heparin-coated and uncoated devices in vitro. MATERIALS AND METHODS
Experimental procedure Fresh whole human blood from informed volunteer donors was collected in bags (Laboratoires Travenol, La Ch$tre, France) to which 4,000 IU of heparin had been added. Fourteen Maxima hollowfiber oxygenators (Medtronic, Anaheim, California, U.S.A.) were set up with tubings and a cardiotomy reservoir in a standardized manner to form an extracorporeal circuit. Seven sets were delivered with the entire blood-contact surface coated with endpoint attached active heparin, Carrneda Bio-Active Surface (13) (Carmeda a s , Oslo, Norway). In brief, this method consists of surface amination with polyethyleneimine followed by incubation with partially degraded heparin, NaBN,CN, and NaC1, giving covalent bonding of terminal aldehyde groups of the heparin fragments to the amino groups on the surface. In a previous study, the resulting heparinized surface completely inhibited thrombin activity in defibrogenated plasma when tested with a synthetic chromogenic peptide assay (13). Fresh, normal human blood did not coagulate when exposed to the surface, and the platelet count in the exposed blood did not change significantly. The remaining seven sets were uncoated. Coated and uncoated sets were used in random order. Each set was primed with 125 ml dextran 60 mgl ml in saline, 125 ml Ringer’s acetate, 125 ml glucose 50 mg/ml, 125 ml mannitol 150 mg/ml, and 50 ml sodium hydrogen carbonate 500 mmol/L. A baseline blood sample was drawn from the bag before 1 U of blood was filled into the circuit. The temperature was maintained at 30°C by means of a heat exchanger and the bloodprimer mixture was circulated through the setup at 4 L/min for 2 h, using a nonocclusive roller pump (Gambro, Horten, Norway). Test samples were obtained at 0, 15, 30, 60,
91
and 120 min. Bloodprimer controls (n = 14) were drawn into glass tubes from the circuit at 0 min and were kept in a 30°C water bath. Control samples were drawn from these tubes at 60 and 120 min. All test and control samples were drawn in glass tubes containing EDTA and were kept on ice until analysis or centrifugation shortly thereafter. Plasma was stored at -70°C. Because MPO is a polycation and heparin is a polyanion, additional experiments were performed to investigate possible binding of MPO to the heparin coating. Sephadex G15 gel (Pharmacia Fine Chemicals, Uppsala, Sweden) was coated with h e p arin employing the same method as used in the CPB circuits (13). Pasteur pipettes were then packed with 5 cm of heparin-coated Sephadex G15. Pipettes packed with untreated Sephadex G 15 or aminated Sephadex G15 were used as controls. Three different solutions with increasing concentrations of MPO were prepared. Sample 1 contained 1.5 ml of pooled human serum and 450 p1 of a buffer solution (phosphate buffer 0.5 M 100 ml, EDTA 0.1 M 100 ml, NaN3 2% 10 ml, bovine serum albumin 10% 20 ml, NaCl 9% 50 ml, Tween 5 ml, H20millipore ad 1,OOO ml) with 0.2% N-cetyl-N,N,N-trimethyl ammonium bromide. Sample 2 contained 1.5 ml of the pooled serum and 450 p,1 of a solution with 1,600 pg/L, of purified MPO (14). Sample 3 contained 1.5 ml of the pooled serum and 450 pI of a solution with 3,200 pg/L of purified MPO. Five hundred microliters of each sample was passed through each type of gel. Test samples for measurement of MPO concentrations were obtained before and after the gel filtration. Analysis of samples The numbers of leukocytes and the hemoglobin concentration (Hgb) in the samples were determined in an electronic counter (Coulter Electronics, Harkenden, England). Neutrophil counts were calculated from May/Griinwald/Giemsa-stained blood smears. MPO and LF were quaniitated by radioimmunological technique (15,16). The results were corrected for hemodilution by multiplication with the factor: initial Hgbhample Hgb. StatiStiCS
The Friedman test (17) was used for analyses of time-dependent changes within each group. The Wilcoxon rank sum test and the Kruskal-Wallis test (17) were applied for intergroup comparisons. Results are presented as median based on Walsh numbers with the 95% nonparametric confidence interval in parentheses. Arr$ Orgons, Vol. 15, No.2. 1991
V . VIDEM ET AL.
92
NC diff
RESULTS Neutrophil counts are shown in Fig. 1. The initial values for the coated and uncoated groups were not significantly different. During the experiment there was a significant reduction of neutrophil counts in both groups. The reduction was equivalent for uncoated and coated devices (Fig. 2). MPO concentrations increased significantly in both groups (Fig. 3). In the controls there was a slight but significant rise after 120 min. The increases from 0 to 60 min and from 0 to 120 min were significantly lower in the coated than in the uncoated group (Fig. 4). LF concentrations rose significantly during the experiment in both oxygenator groups and in the controls (Fig. 5 ) . There were no signifkant differences between the uncoated and coated groups (Fig. 6), but the levels in both oxygenator groups were significantly higher than in the controls after 60 and 120 min. When serum containing various known amounts of MPO was passed through columns of heparincoated or uncoated Sephadex particles, the recovery of MPO was close to 100% (Table 1). In contrast, the aminated Sephadex particles bound approximately 90% of the MPO present in each
2
0
uncoated heparin coated
1
0 0-60
0-120 min
FIG.2. Reduction in neutrophil counts (NCdiw (median with 95% confidence intervals) after circulation for 60 and 120 min during in vitro cardiopulmonary bypass. Differences between coated and uncoated groups are not statistically significant.
DISCUSSION
6-
0 uncoated heparincoated
7 -
6 -
4 1
I
I
T
I
I
I
1
1
0
15
30
60
120
mln
FIG. 1. Neutrophil counts (NC)(median with 95% confidence intervals) versus circulation time through heparin-coated and uncoated cardiopulmonary bypass circuits in vitro. Asterisks denote statistically significant changes compared with initial value: 'p < 0.05, "p < 0.01, ""p < 0.0001.
Arfi/Organs. bol I S . N o 2 . 1991
In the present study there was an equivalent reduction in granulocyte counts in CPB circuits coated with functionally active heparin and in uncoated circuits. The plasma MPO concentrations were approximately five times higher in the uncoated group, whereas there were no significant differences in LF concentrations. One possible explanation for the reduced circulating levels of MPO in the heparin-coated group was binding of MPO to the heparin. However, no binding affinity to'heparin was found in the control experiments, in which MPO-enriched serum was passed through heparincoated Sephadex gel. The validity of the control experiments is supported by the fact that MPO showed a high affinity to the aminated gel. Thus, the authors infer that MPO liberation from the granulocytes was significantly lower after heparin coating. Most investigators find a transient drop in the number of circulating neutrophils during CPB (18,19). The initial fall during perfusion may be explained by adhesion to polymer surfaces and filters or by trapping of granulocytes in the lung circulation. Because there was no filter in the CPB circuit
REDUCED GRANULOCYTE ACTIVATION IN CPB
93
3058
700
600
FIG. 3. Plasma concentrations of myeloperoxidase (MPO) (median with 95% confidence intervals) versus circulation time through heparincoated and uncoated circuits during in vitro cardiopulmonary bypass and in control samples. Asterisks denote statistically significant changes compared with initial value: ‘p < 0.05, ”p < 0.01, **-p < o.Ooo1.
400
300
0
15
30
60
120
min
in the present study, the decline in neutrophil numbers was most probably caused by adhesion to the polymer surfaces. The similar magnitude of the reduction with coated and uncoated devices indicates that the differences in surface properties did not influence neutrophil adherence. Boxer et al. (20) observed neutropenia and margination of neutrophils in the microvasculature of the hamster cheek pouch after infusion of LF, providing evidence that L F is involved in neutrophil adherence to the endothelial cell. Based on the equivalent L F increases in the two oxygenator groups, the authors postulate that LF may be involved in the neutrophil adherence to polymer surfaces, whether they are coated with heparin or not. The mechanisms of neutrophil activation during extracorporeal circulation are not fully understood. Craddock and co-workers (1) allege that activated complement is responsible. Other studies show that activation of complement and granulocytes are not always parallel phenomena. Horl et al. (21) found considerable granulocyte activation accompanied by only little complement activation during hemodialysis, which points at a complement-independent
mechanism. Hdlgren et al. ( 5 ) found no granulocyte-activating properties of serum from dialyzed patients and assumed that degranulation in this setting took place locally from cells attached to the membranes. The authors found that the liberation of L F from secondary granules was not affected by heparin coating. On the other hand, the release of MPO from primary granules was substantially reduced. According to Gallin (lo), exocytosis of secondary granules occurs with milder stimuli such as nonspecific surface adhesion, whereas ‘exocytosis of primary granules generally requires more powerful stimuli. The heparin-coated surface seems to permit neutrophil adhesion and liberation of L F but not further activation with release of the more potent primary granule components such as MPO. The polymer surface is thus rendered more “endothelium-like.” The mechanism for this effect of the heparin coating is not clear. CONCLUSION It is concluded that modification of the polymer surfaces in a model CPB circuit by heparin coating
Arrf Organs. Vol.
IS, No. 2 , 1991
V . VIDEM ET AL.
94
0 uncoated
700
600
1
0 uncoated I3 heparin coated
c) heparin coated
w control
control 1426
2938
1618
1000
-
800
-
500
600
400 -
400
1
200
T
C
0-60
0-120, min
0-60
2006 1037
flG. 4. Increase in myeloperoxidase concentratiqns (MPO diff) (median with 95% confidence intervals) after circulation for 60 and 120 rnin during in vitro cardiopulmonary bypass and in control samples. Statistically significant differences after So rnin: uncoated vs. coated p < 0.001, uncoated vs. control p < 0.001, coated vs. control p < 0.05; after 720 min: uncoated vs. coated p < 0.01. uncoated vs. control p < O.OOO1, coated vs. control p < 0.01.
0-120 min
FIG. 6. Increase in plasma lactoferrin concentrations (LF diff) (median with 95% confidence intervals) after circulation for 60 and 120 rnin during in vitro cardiopulmonary bypass and in control samples. Statistically significant differences after 60 min: uncoated vs. coated not significant (NS),uncoated vs. control p < 0.001, coated vs. control p < 0.01; after 720 min: uncoated vs. coated NS, uncoated vs. control p < 0.0001, coated vs. control p < O.OOO1.
LF
(PS-J 1)
2209
lo00
heperincoated
900Boo700
FIG. 5. Plasma concentrations of lactoferrin (LF)(median with 95% confidehce intervals) versus circulation time through heparin-coated and uncoated circuits during in vitro cardiopulmonary bypass and in control samples. Asterisks denote statistically significant changes compared with initial value: 'p < 0.05, "p < 0.01, 49-p < 0.0001.
-
boo-
I
0
15
30
60
120
mln
ArrfOrgons. V o l . I S . N o . 2. 1991
REDUCED GRANULOCYTE ACTIVATION IN CPB TABLE 1. Concentration of myeloperoxidase (micrograms per liter) in three human serum samples before and after filtration through three Sephadex gel preparations Before filtration After Sephadex GI5 After aminated Sephadex G15 After heparin-coated Sephadex GI5
Sample 1
Sample 2
Sample 3
280 252
643 63 1
1,250 1,300
34
85
165
385
686
1,370
leads to alteration of the granulocyte response. Adhesion of neutrophils was unaffected, whereas further activation was significantly reduced. The authors interpret these findings as an indication of improved biocompatibility, which may prove clinically advantageous in cardiac surgery. Acknowledgment: We are grateful for excellent technical assistance from Mr. Hans Nielsen and the staff of the Surgical Department Laboratory and the Blood Bank,U1levaal Hospital, and Ms. Kerstin Lindblad, Department of CIinical Chemistry, University Hospital, Uppsala. The study was supported by Anders Jahres Fund for the Promotion of Science, the Norwegian Council on Cardiovascular Research, and the Swedish Medical Research Council.
REFERENCES I . Craddock PR, Fehr J, Brigham KL, Kronenberg RS, Jacob HS. Complement and leukocyte-mediated pulmonary dysfunction in hemodialysis. New Engl J Med 1977;2%:769-74. 2. Kirklin JK, Westaby S, Blackstone EH. Kirklin JW, Chenoweth DE, Pacific0 AD. Complement and the damaging effects of cardiopulmonary bypass. J Thorac Cardiovasc SUrg 1983;86:845--57. 3. Haeffner-Cavaillon N, Roussellier N, Ponzio 0, Carreno MP, Laude M, Carpentier A, Kazatchkine MD. Induction of interleukin-1 production in patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1989;98:1100-6. 4. Nilsson L, Brunnkvist S, Nilsson U, NysMm SO, Tydtn H, Venge P, Aberg T. Activation of inflammatory systems during cardiopulmonary bypass. Scand J Thorac Cardiovasc Surg 1988;22:51-3. 5 . Hallgren R, Venge P, Wikstr8m B. Hemodialysis-induced increase in serum lactofemn and serum eosinophil cationic
95
protein as signs of local neutrophil and eosinophil degranulation. Nephron 1981;29:233-8. 6. Wachtfogel YT, Kurich U, Greenplate J, Gluszko P, Abrams W. Weinbaum G, Wenger RK, Rucinski B. Niewiarowski S, Edmunds LH, Colman RW. Human neutrophil degranulation during extracorporeal circulation. Blood 1987;69:32430. 7. Antonsen S, Brandslund J. Clemensen S , Soefeldt S , Madsen T, Alstrup P. Neutrophil lysosomal enzyme release and complement activation during cardiopulmonary bypass. Scand J Thorac Cardiovasc Surg 1987;21:41-52. 8. Knudsen F, Nielsen AH, Pedersen JO, Jersild C. On the kinetics of the complement activation, leukopenia and gmnulocyte-elastase release induced by haemodialysis. Scand J CIin Lob Invest 1985;45:75M. 9. Riegel W, Spillner B, Schlosser V, Hdrl WH. Plasma levels of main granulocyte components during cardiopulmonary bypass. J Thorac Cardiovasc Surg 1988;95:1014-9. 10. Gallin JI. Neutrophil specific granules: a fuse that ignites the inflammatory response. Clin Res 19&;32:320-8. 1 1 . HUgren R, Borg T, Venge P, Modig J. Signs of neutrophil and eosinophil activation in adult respiratory distress syndrome. Crir Care Med 1984;12:14-8. 12. Breda MA, Drinkwater DC,Laks H, Bhuta S, Como AF, Davtyan HG, Chang P. Prevention of reperfusion injury in the neonatal heart with leukocyte-depleted blood. J Thorac Cardiovasc Surg 1989;654-65. 13. Larm 0.Larsson R, Olsson P. A new non-thrombogenic surface prepared by selective covalent binding of heparin via a modified reducing terminal residue. Biomater Med Devices Arrif Organs 1983;1 1:161-73. 14. Olsson I, Venge P. Cationic proteins of human granulocytes. 11. Separation of the cationic proteins of the granules of leucemic myeloicf cells. Blood 1974;44:235-46. IS. Olofsson T, Olsson J, Venge P. Myeloperoxidase and lactofemn of blood neutrophils and plasma in chronic granulocytic leukemia. Scand J Haemafol 1977;18:113-20. 16. Olofsson T, Olsson I, Venge P, Elgefors B. Serum myeloperoxidase and lactofemn in neutropenia. Scand J Haemato1 1977;1 8 : 7 M . 17. Conover WJ. Practical nonparmerric statistics, 2nd ed. New York: Wiley, 1980:229-37, 299-300. 18. Fosse E, Mollnes TE, Ingvaldsen B. Complement activation during major operations with or without cardiopulmonary bypass. J Thorac Cardiovasc Surg 1987;93:860-6. 19. van Oeveren W, Kazatchkine MD,Dcscamps-Latscha B, Maillet F, Fischer E, Carpentier A, Wildevuur CRH. Deleterious effects of cardiopulmonary bypass. J Thorac Cardiovasc Surg 1985;89888-99. 20. Boxer LA, Bjarksttn B, BjBrk J, Yang HH, Allen JM, Baehner RL. Neutropenia induced by systemic infusion of lactofenin. J Lub CIin Med 1982;99:866-72. 21. Hbrl WH, Riegel W, Schollmeyer P, Rautenberg W, Neumann S. Different complement and granulocyte activation in patients dialyzed with PMMA dialyzers. Clin Nephrol 1986;25:30&7.
Artf Organs. Vol. 15. No. 2, 1991
