Introduction

Cardiopulmonary bypass (CPB) is a prerequisite for open heart surgery, and a procedure routinely used [₁]. CPB exposes blood to artificial surfaces, to mechanical trauma from the pump, to alterations in temperature, and to dilution with fluids, whole blood, plasma products, and drugs [₂], and leads to the activation of platelets, coagulation, and fibrinolysis. In addition, activation of the complement system, increased levels of the main granulocyte components, and deposition of thrombus on the artificial surface occur during bypass [₃,₄].

CPB requires an antithrombotic agent, usually standard sodium heparin, to avoid clotting of blood in the extracorporeal circuit [₅]. Thus, open heart surgery with cardiopulmonary bypass (CPB) causes derangement of hemostasis, mainly caused by the high heparin doses [₆]. Artificial surfaces are more or less bioincompatible and thus initiate a contact activation of the intrinsic coagulation pathway, a complement activation, blood cell trauma, and reactions similar to an acute inflammation, including an acute phase reaction [₇,₈].

Several acute-phase reactants are important factors in the haemostatic mechanisms, e.g. plasminogen activator inhibitor-1 (PAI), which inhibits a tissue-plasminogen activator (tPA). Damaged endothelial cells are a major source of these proteins.

As consequence of the activation of the complement system, increased levels of neutrophil granulocyte (PMN)-components like elastase occur during bypass [₉]. Elastase represents a potent agent involved in fibrin degradation. Furthermore, in vitro-studies demonstrated that elevated plasma elastase levels are associated with a reduction in antithrombin III (AT III) [₁₀]. Since low levels of AT III lead to ineffective anticoagulation by heparin [₁₁], it is of great interest to investigate this correlation during CPB, particularly with regard to the heparin concentration and the activated clotting time (ACT).

In contrast, CPB may lead to hemorrhage caused by an increased fibrinolytic activity during and immediately after cardiac surgery [₁₂,₁₃,₁₄,₁₅].

It is obvious that there are complex interactions between coagulation/fibrinolysis and acute inflammatory reactions like "post-perfusion-syndrome" (PPS) [₁₆] or even systemic inflammatory response syndrome (SIRS) after operations with CPB. In many patients this PPS includes changes in hemostasis and represents the beginning of a severe inflammation like a SIRS. Treatment of these hemostaseologic disturbances may be a therapeutic approach to avoid or to reduce complications caused by SIRS.

Thus it was the aim of our study to investigate the cascade of coagulation and fibrinolysis during and after ECC, especially with regard to inflammatory complications by determination of specific parameters like leucocytes, plasma elastase, procalcitonin and C₁-esterase inhibitor (CEI).

In addition we wanted to investigate a potential modification by the use of prostacyclin (PGI2) and/or aprotinin. As these substances have been described to influence coagulatory and fibrinolytic reactions, their use could be a therapeutic approach for the clinical consequences of postoperative inflammatory reactions.

Materials and methods

In a randomized prospective double-blind study 40 patients undergoing aortocoronary bypass grafting (CABG) were divided into 4 groups of 10 patients each. One group served as control group (A), one received prostacyclin (PGI₂) (B), the third group was substituted with high-dosed aprotinin (C) and the last group was treated with a combination of PGI₂ and aprotinin (D).

Reoperations, emergency interventions, an ejection fraction < 35 % and severe accompanying diseases were regarded as exclusion criteria. In the preoperative course no patient was treated with intravenous heparin and the preoperative AT III-level was more than 70 % in all cases.

6 blood samples were taken of every patient perioperatively:

1. before induction of anesthesia

2. 15 minutes after start of ECC

3. end of ECC

4. 30 minutes after protamine

5. arrival on intensive care unit

6. first postoperative day

The samples were analyzed for parameters of coagulation and fibrinolysis like thrombocyte count, antithrombin III (AT III), platelet factor 4 (PF4), β-tromboglobulin (TG), thrombin-antithrombin III complex (TAT), activated clotting time (ACT), tissue-plasminogen activator (tPA), plasminogen-activator inhibitor (PAI), D-dimers, protein C, α₂-antiplasmin and heparin concentration.

Specific parameters of inflammation, such as leucocyte count, plasma elastase (PE), procalcitonin (PCT) and C₁-esterase inhibitor (CEI), were also determined in all blood samples.

Furthermore we particularly analyzed those patients who suffered from an inflammatory reaction in the postoperative period like PPS or SIRS.

PPS in patients after extracorporeal circulation was defined as a temporarily limited need for catecholamines despite sufficient substitution of volume accompanied by a low systemic vascular resistance (SVR). The diagnosis "SIRS" was made according to the criteria of the Consensus Conference [₁₇]. The group of patients with inflammatory complications was compared to patients with an uncomplicated postoperative course. All data were compared by t-test, Mann-Whithney-test or analysis of variance and p < 0.05 was considered to indicate significance.

Using Mann-Whitney-test for tables 1-8 we defined the following significant differences between the 4 groups (A-D):

Figure 1

Figure 2

Table 1: Platelets (n x 103/µl)

Figure 3

Table 2: Tissue Plasminogen activator - tPA (ng/ml)

Figure 5

Table 4: PAI-activity (IU/ml)

Figure 6

Table 4: PAI-activity (IU/ml)

Figure 7

Table 5: a2-antiplasmin-activity (%)

Figure 8

Table 6: Plasminogen-activity (%)

Figure 9

Table 7: Elastase (µg/l)

Figure 10

Table 8 : Leucocytes (x nl-1)

Figure 11

Table 9: ACT and heparin concentration (n=40)

Figure 12

Figure 1: Elastase (µg/l) (PPS-SIRS +/ - )

Results

Under PGI₂ (group B) the mean number of platelets at the end of ECC was higher than in group A (A:132.9 ± 14.9/nl; B:166.5 ± 12.5/nl) (p<0.05). Apart from a significantly lower concentration of tPA (A:68 ± 17; B:29 ± 6 ng/ml) (p<0.05) under PGI₂ no major influence of PGI₂ on coagulation, fibrinolysis or its inhibition could be found.

Under aprotinin (group C) the number of platelets was also significantly higher during and after ECC compared to group A (A:132.9 ± 14.9/nl; C:217.3 ± 17.0/nl) (p<0.05). During ECC significantly lower TAT-complexes were found (A:143 ± 13 µg/l; C:102 ± 12 µg/l) (p<0.05). Furthermore an inhibition of the fibrinolysis occurred during ECC due to an increased activity of PAI as well as of α₂-antiplasmin (A:PAI:17.1 ± 3.6 U/ml, antiplasmin:91 ± 10 %; C:PAI:97.5 ± 13.8 U/ml, antiplasmin:242 ± 10 %) (p<0.05). tPA during ECC was measured in lower concentration under aprotinin (A:38 ± 11 ng/ml; C:21 ± 5 ng/ml) (p<0.05). Accordingly a significantly lower value was measured for plasminogen activity (A:125 ± 15 %; C:76 ± 18 %) (p<0.05) and for the D-dimers (A:2755 ± 439 ng/ml; C:448 ± 60 ng/ml) (p<0.05). In group D we could not find further changes compared to aprotinin alone.

PF4 and β-TG were significantly decreased during ECC only in group D patients (PGI₂ and aprotinin) (A: PF4: 67 ± 11 IU/ml, β-TG: 196 ± 33 IU/ml; D: PF4: 39 ± 3 IU/ml, β-TG: 127 ± 12 IU/ml) (p<0.05). For plasminogen, fibrinogen and protein C we could not find significant differences between the groups at all sample time points.

In the tables 1-6 the exact values for platelets (1), tPA (2), D-dimers (3), PAI- (4), α₂-antiplasmin- (5) and plasminogen-activity (6) are shown.

Levels of plasma elastase increased significantly in all intra- and postoperative blood samples compared to the preoperative baseline values (< 30 µg/l) (p<0.05). The elastase release was even more pronounced in the control- and in the aprotinin-group (170 ± 23 µg/l; 175 ± 14 µg/l during ECC) compared to patients who received prostacyclin (132 ± 21 µg/l) (p<0.05) (table 7). The leucocyte count rose significantly in each group during CPB without significant differences between the groups (table 8).

10 of the 40 patients suffered from a PPS or a SIRS in the postoperative period; in these patients intraoperative elastase levels were significantly higher (188 ± 26 µg/l compared to 138 ± 22 µg/l) (p<0.05), whereas postoperative values did not differ significantly (figure 1). The levels of PCT (figure 2) and CEI did not change significantly during and after ECC.

Figure 13

Figure 2: Procalcitonin (ng/ml) (PPS-SIRS +/ -)

When levels of elastase were rising during ECC, we could find that values of antithrombin III decreased significantly at the same time (preoperative: 84 ± 5 %; end of CPB: 53 ± 4 %) (p<0.05). This could be directly correlated to the rise in elastase; the coefficient was -0.99. Figure 3 shows the course of the mean values of elastase and AT III in all 40 patients at all sample time points.

Figure 14

Figure 3: Mean values of elastase and AT III

Figure 15

Figure 4: Mean values of ACT and heparin concentration

There was also a significant reduction of the ACT with 821 ± 49 seconds when starting CPB compared to 450 ± 30 seconds at the end of CPB (p<0.05), although the heparin concentration did not differ significantly. The exact values for the 40 patients are shown in table 9, figure 4 shows the course of ACT and heparin concentration in the perioperative period (samples 1-6).

{image:15}

Discussion

The defective haemostasis observed during cardiac surgery with CPB may be related to a number of factors such as loss of vascular integrity, defects in platelet function, a hyper-fibrinolytic state, thrombocytopenia and coagulation defects [₁₈,₁₉,₂₀]. The contact of blood with foreign surfaces, such as those encountered with passage through the heart-lung machine during CPB, results in activation of both coagulation and fibrinolysis, producing adverse changes in blood constituents, blood cells and specific coagulation proteins. A number of these changes are controlled by amplification cascades of proteolytic enzymes [₂₁]. The result is a whole-body inflammatory response, termed post-perfusion syndrome, which manifests as bleeding tendencies, pulmonary dysfunction, renal dysfunction and increased susceptibility to infection [₈,₂₂].

The intrinsic coagulation pathway is initiated when blood comes into contact with a foreign surface and factor XII is converted into its active form [₂₃]. Activated factor XII, kininogen and kallikrein catalyse the activation of multiple biological systems, including stimulation of the coagulation system, synthesis of the inflammatory mediator bradykinin and activation of C₁ of the complement system.

The effects on fibrinolysis concern all routes of plasminogen activation. Bradykinin is capable of releasing t-PA from endothelium, and kallikrein is capable of activating both plasma pro-urokinase and the factor XII-dependent plasminogen proactiavtor. Fibrinolytic activity is increased during CPB, as evidenced by increased levels of fibrin and fibrin breakdown products [₂₄,₂₅,₂₆,₂₇]. This activation of fibrinolysis may also contribute to a bleeding tendency in patients undergoing cardiac surgery with CPB.

In addition to the effects on the coagulation and fibrinolytic systems, an acquired defect in platelet formation plays an important role in the haemostatic dysfunction associated with extracorporeal circulation. During CPB the contact of blood with artificial surfaces induces platelet activation or damage, resulting in platelet adhesion to the foreign surfaces and platelet aggregation accompanied by the release of different substances [₂₈]. Platelets do not only secret platelet-specific proteins such as PF 4 and β-TG [₂₉,₃₀], but also synthesize Thromboxan (TX) A₂ [₃₁,₃₂], a pro-aggregatory vasoconstrictor, which may induce further deleterious effects on the circulation. The release reaction may be aggravated by heparin. On the other hand it has been reported that prostacyclin (PGI₂), anti-aggregatory vasodilator is also produced from the vessel wall during CPB, but not in sufficient amounts to counteract the effects of TXA₂, resulting in an imbalance between thromboxane and prostacyclin [₃₁,₃₂].

Nagaoka [₃₃] reported therapeutic interventions in the platelet release reaction during CPB including use of antiaggregatory vasodilator prostacyclin [₃₄,₃₅], and use of the platelet activation inhibitor aprotinin [₃₆]. The use of PGI₂-infusion during CPB remains controversial. Some investigators [₃₄,₃₅] have suggested that PGI₂-treated patients had less platelet loss, β-TG and PF 4 production than those of untreated patients, indicating the protective effects of prostacyclin infusion on platelets in open heart surgery. In contrast, others [₃₆] have not only failed to find any beneficial effect of prostacyclin on platelet count or function, but have also found some adverse results.

Our data show that ECC causes significant alterations of hemostatic variables with the potential risk of thromboembolic and bleeding complications. In all patients we could find an activation of fibrinolysis, as a result of the increased activation of plasminogen and the reduction of plasmin inhibitors. Apart from a significantly lower concentration of t-PA under prostacyclin during ECC no major influence of PGI₂ on coagulation, fibrinolysis or its inhibition could be found. The effect of PGI₂ is achieved via the known thrombocytic mechanism [₃₇].

The effect of aprotinin in our patients consisted of three interrelated mechanisms: inhibition of thrombin activation, inhibition of fibrinolysis and a protecting effect on thrombocytes, expressed by a significantly higher number of platelets during and after CPB.

The combined application of PGI₂ and aprotinin shows an additive effect on platelets, coagulation and inhibition of fibrinolysis, but without significant advantages compared with the application of aprotinin alone.

Former investigators could show a common mechanism to be responsible for the postoperative inflammatory reaction in cardiac surgery patients, involving the inflammatory mediators of the complement system and cytokines [₃₈,₃₉]. Cytokines are potent intercellular signaling molecules known to participate in the regulation of cellular growth, function, and differentiation. Interleukin-6 and interleukin-8 (Il-6, Il-8) levels have been shown to be elevated during and after CPB and are associated with cardiac and pulmonary dysfunction after bypass [₄₀]. The role of neutrophil granulocytes (PMN) in this inflammatory process is quite clear and may be expressed by levels of plasma elastase as a neutrophil product.

To initiate an effective therapy in patients with inflammatory reactions, it would be of importance to recognize a PPS immediately, if possible even in the intraoperative period. Our results show that CPB initiates an elastase release that can be suppressed by prostacyclin. Increased intraoperative elastase levels in patients with PPS or SIRS in the postoperative course show that elastase may be an indicator of ongoing systemic inflammation. It can be speculated that early PGI2-infusion could be a therapeutic option in inflammatory diseases caused by ECC.
As described above disturbances of haemostasis can be found in patients with inflammatory reactions, particularly cases of hypercoagulability such as coronary thrombosis.

With our data we could show that elevated elastase levels during CPB - either due to mechanical alteration or inflammatory activation of PMN - can be correlated to a reduction of AT III. This may be one reason for perioperative thromboembolic complications in patients with open heart surgery and for a hypercoagulatory state in patients suffering from an inflammatory reaction. We could confirm this by significantly reduced levels of ACT despite comparable heparin concentrations in patients with elevated plasma elastase. The administration of AT III can be considered to be useful in selected patients to avoid microvascular thrombosis, endothelial injury, systemic inflammation, organ failure and shock.

Summarizing our results we conclude:

• During CPB the activation of fibrinolysis, as a result of the increased activation of plasminogen and the reduction of plasmin inhibitors, seems to be of importance for postoperative bleeding complications.

• Prostacyclin only acts on thrombocytes directly.

• The application of aprotinin is followed by inhibition of thrombin activation, inhibition of fibrinolysis and a protecting effect on thrombocytes.

• CPB initiates elastase release.

• Elastase release can be suppressed by prostacyclin

• Elastase may be an indicator of postoperative inflammatory reactions

• Elastase release could be correlated to a reduction of AT III

• The use of AT III seems to be useful in selected patients.

According to our results we think that PGI₂-infusion as therapeutic option in inflammatory diseases should be investigated in further studies, particularly with regard to the predictive value of an increased elastase level during operations with CPB.

Correspondence to

Dr. Jens Litmathe Department of Thoracic and Cardiovascular Surgery Heinrich Heine University Moorenstraβe 5 D-40225 Düsseldorf Germany Tel.: + 49 211 811 8331 Fax: + 49 211 811 8333 e-mail: litmathe@med.uni-duesseldorf.de