Eur J Cardio-thorac

Surg (1992) 6: 18-24

Influence of 4 different membrane oxygenators on inflammation-like processes during extracorporeal circulation with pulsatile and non-pulsatile flow EDapper ‘, H. Neppl ‘33, G. Wozniak ‘, I. Strube ‘, J. Boldt 4, F. W. Hehrlein ‘, and H. Neuhof 3 of Clinical Pathophysiology and Experimental Medicine, Departments of 1 Cardiovascular Surgery. ’ Internal Medicine/Division 3 Clinical Chemistry and Pathobiochemistry, and 4 Anesthesiology and Intensive Care Medicine, Justus-Liebig-University, Giessen, FRG

Abstract. The influence of four different membrane oxygenators (HF 4000, BOS-CM 50, CML 2, Maxima) on leucocyte count, concentrations of PMN-elastase, clotting factor XII, AT-III, Cl-INH, a,-antiplasmin and C3a was registered before, during and after CPB with pulsatile and nonpulsatile flow in 80 male patients aged between 36 and 67 years. With all systems tested, there was a drop in the concentrations of clotting factor XII, AT-III, Cl-INH and ct,-antiplasmin in the early extracorporeal circulation (ECC) phase, exceeding the average hematocrit reduction accounted for by dilution. This drop was the least distinct with CML 2 systems, both with pulsatile and nonpulsatile perfusion, indicating system-inherent influences. Leucocyte count and PMN-elastase concentration rose significantly during ECC irrespective of oxygenator tested of flow type applied. The rise in leucocyte count even continued for about 4 h after ECC. During the first 40 min of ECC, these changes were paralleled by a significant rise in C3a concentration, suggesting complement activation as a main cause for PMN activation. However, there is reason to suppose involvement of further mechanisms operating in PMN activation, since the elevated C3a-concentrations began to fall off while leucocyte count and PMN-elastase concentrations were still increasing. [Eur J Cardio-thorac Surg (1992) 6: l&24] Key words: Cardiopulmonary

bypass - Complement activation ~ PMN activation - Hageman factor system

Inflammation may be considered a physiological reaction concerned with the removal of foreign, possibly harmful materials such as microorganisms and tissue debris resulting from any form of tissue damage. The inflammatory reaction is a complex event involving many plasma protein components that pertain to different, though functionally interrelated systems like the coagulation, kinin-forming, fibrinolytic and complement systems. Apart from protein factors, there is also involvement of autacoids like prostaglandines, thromboxanes and leucotriens which indicate the participation of polymorphonuclear and mononuclear cells, platelets and vascular wall components such as endothelial and smooth muscle cells [20]. The short half-life of the autacoids and the existence of plasma and cell-borne inhibitors of enzymatically active components of the above mentioned systems points to the importance of focussing the inflammatory process to a locally restricted site. The deleterious effects of the failure to locally restrain the inflammatory process, as seen Received for publication: Accepted for publication:

July 16. 1991 September 5. 1991

with sepsis, are well known to the clinician and must be avoided. In the case of extracorporeal circulation, as seen with cardiopulmonary bypass operations (CPB), hemodialysis or automated cell separators in leuco- or thrombocytopheresis, blood comes in extensive contact with foreign surfaces. These foreign surfaces of pumps, tubes and oxygenators, just like bacteria, may activate in a nonimmunological way the complement system via the alternative pathway [5, 191. Similarly the Hageman factor system, comprised of high molecular weight kininogen (HMWK), prekallikrein and the coagulation factors XII and XI may by activated via surface activation for factor XII [7, lo]. The Hageman factor system not only triggers the intrinsic pathway of coagulation but also serves as an important link to the kinin-forming and fibrinolytic system [13, 16, 301. Thus, activation of the intrinsic coagulation pathway via factor XII through foreign, preferably negatively charged surfaces, could entail activation of the other plasma systems. If the complement and Hageman factor systems were to be activated on a large scale, given the enormous ex-

19

tent of the foreign surfaces in extracorporeal systems used during CPB, systemic spread of activated factors might cause a state of “disseminated, unspecific intravascular inflammation”. Pulmonary edema, augmented intra-, and postoperative bleeding and protein-uria are but a few complications not seldom encountered with CPB and alleged consequences of inflammation-like processes [l, 3, 12, 15, 181. We compared the influence of four different membrane oxygenators on factor XII activation, polymorphonuclear leucocyte count, PMN-elastase release, plasma inhibitors AT-III, Cl-INH and cr,-antiplasmin. Differences found between the tested oxygenators, run both with pulsatile and nonpulsatile flow, will be described.

Material

and methods

Oxygenators

and bypass circuits

Four disposable membrane oxygenators (Table 1) were included in the evaluation. Ten oxygenators of each type were used with pulsatile and ten with nonpulsatile perfusion. The bypass circuits included the oxygenator to be tested, the cardiotomy reservoire BCR 300, which like the tubing circuit Bypass 70 and the arterial filter MF 1025 C Duraflo YM were provided by American Bentley’ and a modified roller pump Model1 Multiflow from Stoeckert2. The pump was equipped with a separate trigger module PFC 2 2. which allowed easy change from pulsatile to nonpulsatile flow.

Table 1. Oxygenators Oxygenator

Membrane

Membrane

HF-4000” BOS-CM 50b CML 2” Maxima d

Hollow tibre (BOF) Hollow fibre (BIF) Flat plate Hollow tibre (BOF)

4.50 5.30 2.50 2.00

surface

m2 m2 mZ mz

William Harvey HF-4000 (Bard Cardiopulmonary International. Paris) b BOS-CM 50 (American Bentley, Irvine, Calif., USA) ’ Cobe CML 2 (Cobe Laboratories. Lakewood, USA) d Maxima (Extracorporeal, King of Prussia. USA) BOF = Blood outside fibres; BIF = Blood inside tibres

(50 ml), 10% NaCl solution (50 ml). 8.4% sodium bicarbonate (100 ml), Inzolen (10 ml/kg) and 1250 U sodium heparin. A two-stage single cannula was used for returning venous blood to the heart-lung machine (monoatrial cannulation technique), and the coronary operation was performed in “partial bypass”. After aortic cross-clamping, 15 ml/kg of ice-cold (4°C) Bretschneider’s cardioplegic solution was infused initially for 5 min by gravity, and a 200 ml reinfusion was performed 30 min later. The condition of each patient was monitored by surface electrocardiogram, systemic arterial blood pressure (radial artery), central venous pressure (internal jugular vein), pulmonary arterial and capillary pressure and cardiac output (thermodilution Swan Ganz catheter). Venous blood samples were taken at the following times: Pre anesthesia: Patient venous catheter; after sternotomy;

premeditated,

after placement

of intra-

shortly before ECC start;

ECC: At 5 min of ischemia (X-Clamp),

Patients

ECC: At 20 min of ischemia; Eighty male patients scheduled for elective aorto-coronary artery bypass grafting were selected for the study. Their ages ranged from 36 to 67 years (Table 2). All patients were staged NYHA functional class II or III. Criteria considered for exclusion from the study were a body weight exceeding 85 kg or 15% of normal, as defined by BROCA, diabetes mellitus with the need for insulin substitution, manifest renal insufficiency, coronary artery disease of one vessel only or an ejection fraction lower than 40%. Informed consent was obtained from all patients. The patients were randomly assigned one oxygenator type and perfusion mode, thus forming 8 groups consisting of 10 patients each. Anesthesia was standardized in all patients, consisting of weight-related doses of fentanyl, midazolam and pancuronium bromide. After induction of anesthesia, the patients’ lungs were ventilated mechanically with an inspired oxygen fraction of 1.O and zero end expiratory pressure.

CPB and myocardial

protection

Anticoagulation was achieved by administration of heparin 300 U/ kg before onset of extracorporeal circulation. A further 150 U/kg were added when the activated clotting time had fallen below 400 s. During the entire bypass period, a flow rate of 2.4 l/min per m2 was maintained. CPB was performed at mild hypothermia, with the lowest rectal temperature of 34” + 0.5 “C. The extracorporeal circuit prime was Ringers solution (1000 ml), 5% Glucose (1000 ml), 3.5% plasma protein solution (250 ml), 20% human albumin solution

’ Bentley Corp., Irvine, Calif., USA ’ Stoeckert, Munich, FRG

ECC: AT 40 min of ischemia; ECC: Shortly before end of ischemia; ECC: At 20 min after ischemia and shortly before end of ECC; 4 h after ECC; 28 h after ECC. Blood samples were assayed for cell-count (leucocytes, platelets, hematocrit) concentrations of PMN-elastase, C3a, clotting factor XII, and the plasma inhibitors Cl-INH, AT-III and a,-antiplasmin. Commercially available testkits were purchased from Behring 3 (AT-III, Cl-INH and a,-antiplasmin). from Merck4 (PMN-elastase) and from Progen Biotechnik 5 (C3a). Clotting factor XII concentration was determined according to Vinazzer [26] modified by H. Neppl. In brief: a prediluted plasma sample was activated with a specified volume of kaolin suspension. After 7 min of incubation, the chromogenic substrate S 2302 6 was added and followed by a further 2 min incubation period. The reaction was stopped with acetic acid. The kaolin was then removed by centrifugation and the factor XII concentration determined in the supernatant by photometry (405 nm) against a sample bank. The method is linear in the range between 25% and 150% of normal factor XII concentration. Cell-counts were determined with the flow hemocytometer Sysmex E 5000 7.

3 4 5 6 ’

Behring, Marburg. FRG Merck, Darmstadt, FRG Progen Biotechnik, Heidelberg, FRG KabiVitrum, Mtinchen, FRG Toa Medical Electronics, Cobe. Japan

20 Table 2. Data on 10 patients Oxygenator/flow

HF-4000/p HF-4000/np BOS-CM 50/p BOS-CM 50/np CML 2/p CML 2/np Maxima/p Maxima/np p=Pulsatile

Age (years)

Surface area m2

ECC time (min)

Ischemia (min)

mean

SD

mean

SD

mean

SD

mean

SD

54.1 57.2 49.5 53.8 53.0 59.3 54.3 54.2

1.4 3.9 6.1 7.1 6.5 3.2 5.6 5.3

1.85 1.83 1.87 1.89 1.85 1.92 1.85 1.89

0.10 0.10 0.08 0.14 0.09 0.08 0.11 0.13

87.4 90.5 88.9 92.2 75.5 78.6 89.7 81.4

9.3 11.1 9.2 20.3 12.2 17.9 15.4 14.6

56.4 58.1 60.5 57.2 49.4 46.7 55.3 50.6

6.8 10.5 8.7 13.4 7.0 10.3 13.4 13.5

flow mode; np =nonpulsatile

flow mode

Statistical methods Mean and standard error was calculated at each time point. Oneway analysis of variance was applied to mean values corresponding to the same time points. The Student’s t-test was used for paired samples to compare different time points within individual groups. P values less than 0.05 were considered statistically significant.

Results

With all oxygenator systems tested, the observed alterations in Hageman factor concentration during ECC generally paralleled hematocrit changes (Figs. 1 a, 2a). Similar changes in concentration of Cl-INH, AT-III and a,-antiplasmin were noted (Figs. 1 b-d, 2b-d). However, the reduction in factor XII concentration during the early ECC phase exceeded the average hematocrit dilution by 7% - 15% with respect to the oxygenator: 7% was found with CML 2 oxygenators, both with pulsatile and nonpulsatile flow, 15% with HF-4000/pulsatile flow. BOS-CM 50 and Maxima oxygenators ranged between. Similarly, Cl -INH concentrations exceeded the average hematocrit reduction by 1% - 12%. Again, CML 2 oxygenators, irrespective of flow mode applied, showed the least and HF-4000/nonpulsatile flow the greatest fall. However, as with both factor XII and Cl-INH concentrations, the differences observed between the oxygenators were not statistically significant. Significant differences (PC 0.05) were registered with AT-III concentrations during the entire ECC phase, comparing the Maxima with the CML 2 oxygenators run in nonpulsatile flow mode (Fig. 2~); AT-III concentrations were significantly lower with Maxima oxygenators. As with clotting factor XII and Cl-INH, there was an initial fall in AT-III concentration which surpassed hematocrit reduction caused by dilution by 6% with HF-4000/nonpulsatile flow to 19% with Maxima/nonpulsatile flow. Only when employing CML 2 oxygenators did AT-III changes closely parallel those of hematocrit, both in pulsatile and nonpulsatile flow mode (Figs. 1 c, 2~). Though not statistically significant, AT-III concentrations with Maxima oxygenators were definitely lower with nonpulsatile than with pulsatile flow (Figs. 1 c, 2~).

Comparing a,-antiplasmin concentrations (Figs. 1 d, 2d), significant differences (PC 0.05) were only found between CML 2/nonpulsatile flow and BOS CM 5O/pulsatile flow, the former combination showing signiticantly lower values. However, this difference was observed not only during the ECC phase but already before ECC start. With all oxygenators run in either flow mode, the concentrations of clotting factor XII, Cl-INH, AT-III and a,antiplasmin had again reached pre-surgical values 28 h after ECC. Compared with the parameters described above, quite different changes were observed with leucocyte count (Figs. 1 e, 2e) and PMN-elastase concentration (Figs. 1 f, 2f) during and after the ECC phase. After a drop caused by dilution, leucocyte count continuously increased, reaching at ECC end about twice the count measured before ECC start. There was a further increase 4 h after ECC: 28 h after ECC the leucocyte count had decreased but still had not returned to pre-surgical values. Statistically significant differences were not observed between the oxygenators run with pulsatile or nonpulsatile flow. During ECC, the changes in leucocyte count were paralleled by those in PMN-elastase concentration. Four hours after ECC, PMN-elastase concentrations were still at the level observed at ECC end with CML 2 and HF4000 oxygenators: 28 h after ECC, PMN-elastase concentrations had declined with all oxygenators but were still above pre ECC level. Though not statistically significant, the rise in PMN-elastase concentration was definitely higher with CML 2 than with BOS-CM 50 or Maxima with both flow types (Figs. 1 f, 2f). BOS-CM 50 and Maxima showed quantitatively similar increases, HT4000 oxygenators being intermediate between CML 2 and BOS-CM 50 or Maxima. Due to the high costs of the assay and the rather difficult sampling procedure, C3a concentrations were determined only with 2 patients in each group at selected time points. There was a significant (PcO.01) rise in the concentrations of C3a, reaching a peak after 40 min of ischemia and decreasing towards the end of ECC. This time course was observed irrespective of the perfusion mode selected (results not shown as graphs).

21 Hct 50 120 Factor Wol WI 1 (XI

XII

30. 10

a0

11

lo-

20.

20

0

tic1 50

,“.a,.,

lo-

bo-

02

(Hyc~~0~100-crp-Anllpl~rmln W

40

30

30.

80.

20

20.

40.

. 10.

10

30. 20. 10.

do-

CO

Fig. 1 a-f.

o-

Pulsatile flow mode

Discussion With all oxygenators tested - irrespective of flow type the fall in factor XII concentration at the beginning of ECC exceeded the average hematocrit reduction by 7% 15%. This finding is suggestive of factor XII consumption. Reduction of hematocrit and, to an identical extent, of all other blood constituents at this time is due to dilution through the priming solution when the patient is being connected to the ECC system. Since there was already a slight decrease of factor XII concentration between anesthesia and ECC begin, i.e., when the patient undergoes surgery in order to be connected to the ECC

system, the factor XII consumption might be considered a consequence of surgical manipulation and would thus not be attributed to contact activation with foreign surfaces of the ECC systems. This view is supported by the finding that, later on, factor XII concentrations remained stable until ECC end and showed only a slight increase thereafter. Comparing changes in the concentrations of factor XII, Cl-INH and AT-III between the different oxygenator systems, however, there is strong indication of factor XII consumption due to activation from foreign surfaces of the ECC system. In this respect it is important to point out the strikingly low degree of factor XII consumption

10

20

a0 J

0i

10

20 10

co

0

I

MT

i t.eococytes

II ~OOOI

14. 12. 10.

Fig. 2 a-f.

Nonpulsatile

flow mode

seen with the CML 2 oxygenators. In accordance with this finding, it was observed that Cl-INH and AT-III concentrations with this oxygenator only slightly exceeded the decrease due to dilution. Cl-INH and AT-III are both inhibitors of activated clotting factor XII. These striking, though statistically not significant differences between the CML 2 and the other oxygenators point towards a dependency on the oxygenator used for ECC. This would indicate that not only surgical manipulation but also the oxygenator used contributes to factor XII activation. Considering leucocyte count and PMN-elastase concentration, there is strong evidence that ECC systems

cause PMN activation. There was a slight rise in leucocyte count and PMN-elastase concentration in the time period between anesthesia and ECC begin, suggesting PMN activation due to surgical manipulation as already considered partially responsible for factor XII activation. This pre-ECC rise in leucocyte count and PMN-elastase concentration was followed by a sharp drop at ECC begin due to dilution, but unlike factor XII, Cl-INH, ATIII and cr,-antiplasmin, whose concentrations remained stable until ECC end, leucocyte count and PMN-elastase concentrations continuously increased. A fall in leucocyte number in the early ECC phase, exceeding the reduction accounted for by dilution, was not observed with any

23

oxygenator. This finding does not correspond with observations made by others, who describe a strong, transient decrease independent of dilution [I I]. The rise in PMN-elastase concentration indicates PMN activation and possibly PMN damage. The most likely candidates considered for PMN activation may be breakdown products of complement components [2, 8, 11,14,22]. As demonstrated only at random in our study, there was a signiticant increase of C3a concentration with all oxygenators employed, both with pulsatile and nonpulsatile flow. Many investigators have demonstrated complement activation via the alternative pathway through the foreign surfaces of ECC systems [4, 5, 191. According to Kazatchkine and Nydegger [17], complement activating surfaces provide a microenvironment that protects autocatalytically formed C3b from inactivation and C3bBb from active decay-dissociation by the regulatory proteins. Thus, by allowing the formation and by protecting the function of the surface bound amplification convertase C3bBb, activating surfaces advance the alternative pathway from a slow fluid phase turnover to a surface directed amplification phase of C3 cleavage. The complement fragment C5a is known to be a potent activator of PMNs [25,29]. Though very difficult to assay because of its rapid binding to C5a receptors on PMNs [6] and various other cells, C5a was shown to be elevated during ECC [5]. A further indication for the formation of C5a comes from the demonstration of elevated concentrations of the terminal complement complex during ECC [24]. C5a is essential for the formation of the terminal complement complex. Being aware of the known tissue damaging effects of activated PMNs through reactive oxygen species [23], and the many different lysosomal proteinases [9, 211 of which elastase is but one of the best studied, it becomes imperative to use ECC systems which lead to no or at most very low PMN activation. Since PMN activation can be triggered by fragments of complement components, plasma kallikrein [27] and activated clotting factor XII [28], greater efforts must be made to develope materials that do not provide an environment for C3b or C3bBb protection and activation of the Hageman factor system. Certainly, the enormous number of successfully performed interventions with ECC systems confirms the high standard and the systems seem to be perfectly adequate for daily routine. However to give but one example, the not so rare individuals with an innate deficiency in cl,-PI may perhaps suffer damage from the unobstructed action of massive PMN-elastase released, especially when the inhibitor deficiency is further augmented through blood dilution as seen in ECC. The concentrations of elastase measured in our study may be considered inactivated because the assay system is based on measuring complexes made up of elastase and a,-PI formed in the circulation. Thus, even under conditions of hemodilution, the inhibitory potential of normal individuals seems sufficient; whether this holds equally true with innate a,-PI deficiency can not be concluded from this study. Acknowledgements. We thank C. Ruhl for technical assistance and P. Miiller for graphic illustration.

1. Abel RM, Buckley MJ, Austen WG, Bamett GO, Beck CH, Fischer JE (1976) Etiology, incidence, and prognosis of renal failure following cardiac operations. Results of a prospective analysis of 500 consecutive patients. J Thorac Cardiovasc Surg 71: 323-333 2. Antonsen S, Brandslund I, Clemensen S, Sofeldt S, Madsen T, Alstrup P (1987) Neutrophil lysosomal enzyme release and complement activation during cardiopulmonary bypass. Stand J Thorac Cardiovasc Surg 21:47-52 3. Bick RL (1985) Hemostasis defects associated with cardiac surgery, prosthetic devices, and other extracorporeal circuits. Semin Thromb Hemost 11: 249 -280 4. Cavarocchi NC, Pluth JR, Schaff HV, Orszulak TA, Homburger HA, Solis E, Kaye MP, Clancy MS, Kolff J, Deeb GM, Brounstein L, Schnell WA (1986) Complement activation during cardiopulmonary bypass. J Thorac Cardiovasc Surg 91:252-258 5. Chenoweth DE, Cooper SW, Hugli TE, Stewart RW, Blackstone EH, Kirklin JW (1981) Complement activation during cardiopulmonary bypass. Evidence for generation of C3a and C5a anaphylatoxins. N Engl J Med 304497-503 6. Chenoweth DE, Hugh TE (1978) Demonstration of specific C5a receptor on intact human polymorphonuclear leucocytes. Proc Nat1 Acad Sci USA 75: 3943-3947 7. Cochrane CG, Griffin JH (1979) Molecular assembly in the contact phase of the Hageman factor system. Am J Med 67: 657-664 8. Craddock PR, Hammerschmidt DE, White JG, Dalmasso AP, Jacob HS (1977) Complement (CSa)-induced granulocyte aggregation in vitro. A possible mechanism of complement-mediated leucostasis and leucopenia. J Clin Invest 60: 260-264 9. Fritz H, Jochum M (1984) Granulocyte proteinases as mediators of unspecific proteolysis in inflammation: a review. In: Goldberg DM, Werner M (eds) Selected topics in clinical enzymology 2. de Gruyter, Berlin New York, pp 305-327 10. Griffin JH (1978) Role of surface in surface-dependent activation of Hageman factor (blood coagulation factor XII). Proc Nat1 Acad Sci USA 75: 1998-2002 11. Hammerschmidt DE, Stroncek DF, Bowers TK (1981) Complement activation and neutropenia occurring during cardiopulmonary bypass. J Thorac Cardiovasc Surg 81: 370-377 12. Harker LA, Malpass TW, Branson HE, Hessel EA, Slichter SJ (1980) Mechanism of abnormal bleeding in patients undergoing cardiopulmonary bypass: acquired transient platelet dysfunction associated with selective granule release. Blood 56: 824834 13. Heimark RL, Kurachi K, Fujikawa K, Davie EN (1980) Surface activation of blood coagulation, fibrinolysis, and kinin formation. Nature (London) 286: 456-460 14. Jacob HS, Craddock PR, Hammerschmidt DE, Moldow CF (1980) Complement-induced granulocyte aggregation. An unsuspected mechanism of disease. N Engl J Med 302: 789-794 15. Kalter RD, Saul CM, Wetstein L, Soriano C, Reiss RF (1979) Cardiopulmonary bypass. Associated hemostatic abnormalities. J Thorac Cardiovasc Surg 77: 427-435 16. Kaplan AP (1983) Hageman factor-dependent pathways: mechanism of initiation and bradykinin formation. Fed Proc 42:3123-3127 17. Kazatchkine MD, Nydegger UE (1982) The human alternative complement pathway: biology and immunopathology of activation and regulation. Prog Allergy 30: 193-234 18. Kirklin JK, Westaby S, Blackstone EH, Kirklin JW, Chenoweth DE, Pacific0 AD (1983) Complement and the damaging effects of cardiopulmonary bypass. J Thorac Cardiovasc Surg 86: 845857 19. Kutsal A, Ersoy U, Ersoy F, Yeniay I, Bakkaloglu A, Bozer Y (1989) Complement activation during cardiopulmonary bypass. J Cardiovasc Surg 30: 359-363

24 20. Neuhof H (1984) Zur pathogenetischen Bedeutung der klassischen Kaskadensysteme und des Arachidonsaure-Metabolismus. Med Welt 35: 1457- 1463 21. Neuhof H, Hoffmann Ch, Seeger W, Suttorp N, Fritz H (1989) Proteases as mediators of pulmonary vascular permeability. In: Schlag G, Red1 H, (eds) Second Vienna shock forum. Liss, New York, pp 305-314 22. O’Flaherty JT, Craddock PR, Jacobs HS (1978) Effect of intravascular complement activation on granulocyte adhesiveness and distribution. Blood 51: 731-739 23. Sacks T, Moldow CF, Craddock PR, Bowers TK, Jacob HS (1978) Oxygen radicals mediate endothelial cell damage by complement-stimulated granulocytes. An in vitro model of immune vascular damage. J Clin Invest 61: 1161- 1167 24. Salama A, Hugo F, Heinrich D, Hijge R, Mtiller R, Kiefel V, Mueller-Eckhardt C, Bhakdi S (1988) Deposition of terminal CSb-9 complement complexes on erythrocytes and leukocytes during cardiopulmonary bypass. N Engl J Med 318408-414 25. Till GO, Ward PA (1986) Systemic complement activation and acute lung injury. Fed Proc 45: 13-18 26. Vinazzer H (1979) Assay of total factor XII and of activator factor XII in plasma with a chromogenic substrate. Thromb Res 14:155-166

27. Wachtfogel YT, Kucich U, James HL (1983) Human plasma kallikrein releases neutrophil elastase during blood coagulation. J Clin Invest 72: 1672- 3677 28. Wachtfogel YT, Pixley RA, Kucich U, Abrams W, Weinbaum G, Schapira M, Colman RW (1986) Purified plasma factor XII a aggregates human neutrophils and causes degranulation. Blood 67: 1731-1737 29. Webster RO, Hong SR, Johnston RB, Henson PM (1980) Biological effects of the human complements fragments C5a and C5a on neutrophil function. Immunopharmacology 2: 201-219 30. Wiggins RC (1983) Kinin release from high molecular weight kininogen by the action of Hageman factor in the absence of kallikrein. J Biol Chem 258: 896338970

Dr. F. Dapper Abteilung fur Herzgefigchirurgie Justus-Liebig-Universitlt KlinikstraBe 29 W-6300 GieBen Federal Republic of Germany