Ann Thorac Surg 1998;65:999-1002
© 1998 The Society of Thoracic Surgeons
Negative Reexploration for Cardiac Postoperative Bleeding: Can It Be Therapeutic?
Marc P. Pelletier, MDaa,
Susan Solymoss, MDbb,
Andrew Lee, BScaa,
Ray C.-J. Chiu, MD, PhDaa
a Division of Cardiothoracic Surgery, The Montreal General Hospital/McGill University, Montreal, Quebec, Canada
b Division of Hematology, The Montreal General Hospital, McGill University, Montreal, Quebec, Canada
Accepted for publication October 22, 1997.
Address reprint requests to Dr Chiu, Division of Cardiothoracic Surgery, The Montreal General Hospital, 1650 Cedar Ave, Rm C9.169, Montreal, Quebec, Canada, H3G 1A4
 |
Abstract
|
|---|
Background. Reexploration of the mediastinum for bleeding is required in 3% to 7% of patients after cardiac operation, with many proving to have no surgically correctable cause. In spite of a "negative exploration," the bleeding often ceases. We propose the hypothesis that such a negative exploration can be therapeutic by reducing marked fibrinolytic activity in the mediastinal cavity.
Methods. Fibrinolytic activity in shed mediastinal blood was compared with that in the system blood in 5 patients after cardiac operation by measuring fibrinogen, fibrin degradation product, plasminogen activator inhibitor-1, and 
2-antiplasmin levels.
Results. Fibrinolytic activity in mediastinal blood was markedly increased when compared with paired systemic venous blood. This was indicated by the mediastinal blood’s lower fibrinogen levels (0.47 versus 1.91 U/mL; p < 0.001), very high levels of fibrin degradation products (1,350 versus 200 ng/mL; p < 0.05), and higher levels of plasminogen activator inhibitor-1 (55.5 versus 28.1 ng/mL; p < 0.005). Decreased levels of 2-antiplasmin were also observed in the mediastinum (0.50 versus 0.61 U/mL; p < 0.05).
Conclusions. Our data confirm that fibrinolytic activity can be extremely high in the mediastinum in response to clot formation. This may explain the hemostatic effects of a negative reexploration, where irrigation and the removal of clots may reduce the fibrinolytic process; this may allow the bleeding ends of capillaries and small vessels to thrombose. Decreased levels of 2-antiplasmin observed suggest that lysine analogs, such as 
-aminocaproic acid, may have a beneficial role when locally delivered into the mediastinum.
|
Introduction
|
|---|
Postoperative hemorrhage is a serious and potentially lethal complication of cardiac operations. Initial management consists of adequate drainage to prevent cardiac tamponade, along with correction of any detectable hemostatic abnormalities. When bleeding persists, mediastinal reexploration is required; a complication that occurs in 3% to 5% of patients after cardiac operation [1–3]. For patients with previous operations or in those having valve procedures, the reexploration rate can increase to approximately 7% [4].
Although many of these patients will be found to have surgically correctable bleeding from sites, such as the side branches of the internal mammary artery or saphenous vein grafts, graft anastomoses, cannulation or aortotomy sites, in approximately one third of patients there will be no identifiable surgical bleeding sources [4]. Instead, they have generalized oozing, which is known as "nonsurgical" or "medical" bleeding. In such cases, the oozing surface is electrocoagulated as best as possible and the blood clots are evacuated. This is followed by irrigation of the mediastinum, insertion of drainage tubes, and closure of the wound. Because no surgical bleeding sites can be found, no definitive hemostatic maneuvers can be performed. Thus, the main benefit of such "negative exploration" is to definitely rule out surgical bleeding.
Surprisingly, however, after a negative exploration, as many cardiac surgeons have experienced, most patients will stop bleeding. Such an experience has been mocked as the "magic of exposure to night air." It is presumed that careful electrocoagulation of the oozing tissue, and the natural course of eventual hemostasis are the reasons for the favorable outcome. Such explanations are not always consistent with clinical observations, as illustrated in the following case.
A 66-year-old female patient underwent uncomplicated three-vessel coronary artery bypass grafting. At the time of closure, the wound was dry with no evidence of excessive bleeding. However, postoperatively in the intensive care unit, she continued to bleed at a rate of 300 to 500 mL/h from the mediastinal sumps, despite full correction of her coagulation parameters (international normalized ratio, 1.4; prothrombin time, 12.9 seconds; partial thromboplastin time, 27.7 seconds; platelets, 162 x 106/µL; thrombin time, 10.1 seconds [normal range, 9.3 to 13.0 seconds]; and fibrinogen 2.11 g/L [normal range, 1.5 to 3.5 g/L]) and the administration of 4 units of fresh frozen plasma, 6 units of platelets, 6 units of cryoprecipitate, and 4 units of packed red blood cells. Increases in positive end-expiratory pressure also failed to decrease the bleeding rate. After 6 hours, she was reexplored and found to have no "surgical bleeding" site. Large clots were evacuated, oozing tissue surfaces electrocoagulated, and the mediastinum irrigated with saline before closure. Despite the lack of an obvious surgical bleeding site to be corrected, there was a dramatic reduction in the rate of bleeding to less than 50 mL/h immediately after the reexploration (Fig 1). We postulate the following scenario.
View larger version (14K):
[in this window]
[in a new window]
|
Fig 1. Rate of mediastinal bleeding in an illustrative case after aortocoronary bypass grafting (CABG). Amounts listed are in milliliters after each postsurgical hour. At the seventh hour, the patient underwent reexploration and was subsequently returned to the intensive care unit at the ninth postsurgical hour.
|
|
Blood clot formation is an important mechanism to achieve hemostasis in the mediastinal wound after cardiac operation. Clot formation, however, induces counterbalancing fibrinolytic activity, which in turn may interfere with plugging of the bleeding capillaries and small vessels by thrombus formation.
No surgical bleeders are found during a negative reexploration, but the evacuation of clots, coupled with irrigation of the mediastinal wound, removes the fibrinolytic stimulus, allowing thrombi to form and remain at the divided ends of small vessels and capillaries, thus achieving hemostasis.
The following study was carried out to confirm the first part of this hypothesis, namely that there is a marked fibrinolytic activity in the shed blood from the mediastinum after cardiac operation.
|
Material and methods
|
|---|
Blood samples were collected from 5 patients during the first postoperative hour after cardiac procedures requiring cardiopulmonary bypass. Both systemic venous blood and mediastinal drainage blood samples were obtained from each patient and processed for the following hematologic tests. All patients had normal postoperative courses. Mediastinal blood samples were collected while still in liquid form from the mediastinal tubes. Both samples were collected before autotransfusion. Antibleeding agents, such as aprotinin or -aminocaproic acid (EACA), were not given to any of the patients in this study.
All samples were heparin adsorbed and double spun at 3,000 rpm for 10 minutes. The plasma of each sample was then aliquoted into marked microcentrifuge tubes and kept at -70°C until testing. The tests performed were fibrinogen levels (ACL 300; Instrumentation Laboratory), 2-antiplasmin (Chromogenic Colorimetric Assay; Diagnostica Stago, France), fibrin degradation products (Ortho Diagnostic Systems, Ontario, Canada), and plasminogen activator inhibitor-type 1 (Biopool International Inc, Ontario, Canada).
Statistical analysis
The data were compared using the software InStat. Values of p were obtained with unpaired Student’s t test; those less than 0.05 were considered significant.
|
Results
|
|---|
Some of the relevant patient characteristics are listed in Table 1. Three patients underwent aortocoronary bypass, and the other two had mitral valve repair and aortic valve replacement, respectively. All patients had an uneventful postoperative course and were discharged within 8 days.
Results of hematologic studies to determine fibrinolytic activity are shown inFigures 2 to 5. Fibrinolytic activity was markedly increased in blood drained from the mediastinum, compared with blood obtained from the venous circulation. In the shed mediastinal blood, there were lower fibrinogen levels (0.47 versus 1.91 U/mL; p < 0.001), higher levels of fibrin degradation products (1,350 versus 200 ng/mL; p < 0.05), and higher levels of plasminogen activator inhibitor (55.5 versus 28.1 ng/mL; p < 0.005). Decreased levels of 2-antiplasmin were also observed in the mediastinum (0.50 versus 0.61 U/mL; p < 0.05).
View larger version (13K):
[in this window]
[in a new window]
|
Fig 2. Comparison of fibrinogen levels in the systemic blood (black bar) versus mediastinal blood (white bar). Values are in units per milliliter.
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Fig 3. Comparison of fibrin degradation products (XDP) in the systemic blood (black bar) versus mediastinal blood (white bar). Values are in nanograms per milliliter.
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Fig 4. Comparison of plasminogen activator inhibitor-1 (PAI-1) levels in the systemic blood (black bar) versus mediastinal blood (white bar). Values are in nanograms per milliliter.
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Fig 5. Comparison of 2-antiplasmin (A2AP) levels in the systemic blood (black bar) versus mediastinal blood (white bar). Values are in units per milliliter.
|
|
|
Comment
|
|---|
The importance of a negative reexploration for excessive postoperative hemorrhage after cardiac operation is generally thought to be the exclusion of surgical bleeding, a cause that can be readily controlled by surgical maneuvers such as ligation and suturing. Various hematologic abnormalities, which may cause continued bleeding, are sought in the systemic circulation. Our hypothesis states that such abnormalities can exist independently in the mediastinal blood and may be removed during reexploration.
The blood interacts with its environment in a complex way to prevent both undesirable bleeding and the equally detrimental complication of intravascular thrombosis. In the systemic circulation, a delicate balance between the blood coagulation cascade and the fibrinolytic system evolved to achieve this goal. However, in the nonphysiologic circumstance of blood collecting in the mediastinum, these two systems can be activated, but the delicate balance may not be maintained. Fortunately for our patients, most of the postoperative bleeding stops in a manner similar to the 5 patients studied here, although even in such cases, the mediastinal blood obtained from the sump drain showed marked fibrinolytic activity. In some patients, the imbalance may be even more pronounced, and bleeding continues conspicuously as in our patient, who responded dramatically to reexploration presumably because this intervention reversed the balance. Thus, we suggest two categories of medical bleeding. One denotes the well-known group of patients with an abnormal hemostatic profile of the systemic blood. The other represents those in whom an abnormality exists locally in the mediastinal blood.
The complex interactions of fibrinolytic activators and inhibitors have been described by Collen and Lijnen [5]. Our data indicate clearly that there is marked activation of this system in the mediastinal blood, compared with that in the systemic circulation. It is well known to surgeons that pericardial blood often does not clot readily, and is usually partially defibrinated. It had been postulated that the latter could be attributable to consumption or mechanical agitation associated with heart beats. Our results, particularly those indicating markedly elevated fibrin degradation products, confirmed that fibrinolysis is a main mechanism for nonclotting of such blood [6]. Our finding of lower mediastinal levels of 2-antiplasmin is also interesting, as this molecule inhibits fibrinolysis by competing for the lysine-binding site of plasminogen, reducing the number of plasminogen molecules available for binding to fibrin. It is for this reason that analogs of lysine, such as EACA, are used as antifibrinolytic agents. By occupying free lysine-binding sites, these compounds prevent plasminogen from binding to fibrin, blocking its subsequent activation on fibrin to plasmin [7], which can then split fibrin. Although some reports suggest that EACA may reduce cardiac postoperative bleeding [8–14], it is not clear whether this is mediated by its effect on blood within the systemic circulation, or by the delivery of EACA to the mediastinum and acting locally. Thus, the local administration of EACA in the mediastinum may have merit, given the evidence of greater fibrinolytic activity there compared with that seen in the systemic blood, as noted in our study. However, appropriate clinical trials are needed to confirm the efficacy of this approach.
We tend to ascribe cessation of bleeding after exploration without a surgical bleeding site to additional electrocoagulation of the oozing wound. Another perception is that the bleeding has run its course during reexploration. But the abrupt reduction in the rate of bleeding after reexploration, as shown in our illustrated case, and experienced frequently by many surgeons, is not consistent with such an explanation. If the pathophysiology we described is valid, the negative reexploration may not be so negative after all, and indeed can be therapeutic.
|
Acknowledgments
|
|---|
We thank Jenny Kwan, BSc, RT, for her technical assistance and support.
|
References
|
|---|
- Mehta S.M., Pae W.E., Jr Complications of cardiac surgery. In: Edmunds L.H., Jr, ed. Cardiac surgery in the adult, 1 ed. New York: McGraw-Hill, 1997:392-393.
- Moulton M.J., Creswell L.L., Mackey M.E., et al. Reexploration for bleeding is a risk factor for adverse outcomes after cardiac operations. J Thorac Card Surg 1996;111:1037-1046.[Abstract/Free Full Text]
- Spiess B.D., Gillies B.S., Chandler W., Verrier E. Changes in transfusion therapy and reexploration rate after institution of a blood management program in cardiac surgical patients. J Cardio Vasc Anest 1995;9:168-173.
- Unsworth-White M.J., Herriot A., Valencia O., et al. Resternotomy for bleeding after cardiac operation: a marker for increased morbidity and mortality. Ann Thorac Surg 1995;59:664-667.[Abstract/Free Full Text]
- Collen D., Lijnen H.R. Molecular and cellular basis of fibrinolysis. In: Hoffman R., Benz E.J., Jr, Shattil S.J., Furie B., Cohen H.J., eds. Hematology: basic principles and practice. New York: Churchill Livingstone Inc, 1991:1232.
- Tabuchi N., de Haan J., Boonstra P.W., van Oeveren W. Activation of fibrinolysis in the pericardial cavity during cardiopulmonary bypass. J Thorac Card Surg 1993;106:828-833.[Abstract]
- Rapaport S.I. Introduction to hematology, 2 ed. Philadelphia: Lippincott, 1987:464.
- Vander Salm T.J., Kaur S., Lancy R.A., et al. Reduction of bleeding after heart operations through the prophylactic use of epsilon-aminocaproic acid. J Thorac Card Surg 1996;112:1098-1107.[Abstract/Free Full Text]
- Levy J.H., Pifarré R., Schaff H.V., et al. A multicenter, double-blind, placebo-controlled trial of aprotinin for reducing blood loss and the requirement for donor-blood transfusion in patients undergoing repeat coronary artery bypass grafting. Circulation 1995;92:2236-2244.[Abstract/Free Full Text]
- Gans H., Krivit W. Problems in hemostasis during open-heart surgery. III: Epsilon amino caproic acid as an inhibitor of plasminogen activator activity. Ann Surg 1962;155:268-276.[Medline]
- DelRossi A.J., Cernaianu A.C., Botros S., et al. Prophylactic treatment of postperfusion bleeding using EACA. Chest 1989;96:27-30.[Abstract/Free Full Text]
- Arom K.V., Emery R.W. Decreased postoperative drainage with addition of -aminocaproic acid before cardiopulmonary bypass. Ann Thorac Surg 1994;57:1108-1112.[Abstract]
- Daily P.O., Lamphere J.A., Dembitsky W.P., et al. Effect of prophylactic epsilon-aminocaproic acid on blood loss and transfusion requirements in patients undergoing first-time coronary artery bypass grafting: a randomized, prospective, double-blind study. J Thorac Cardiovasc Surg 1994;108:99-106.[Abstract/Free Full Text]
- Vander Salm T.J., Ansell J.E., Okike O.N., et al. The role of epsilon-aminocaproic acid in reducing bleeding after cardiac operation: a double-blind randomized study. J Thorac Cardiovasc Surg 1988;95:538-540.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
R. Salenger, J. S. Gammie, and T. J. Vander Salm
Postoperative Care of Cardiac Surgical Patients
Card. Surg. Adult,
January 1, 2003;
2(2003):
439 - 469.
[Full Text]
|
|
|
|
|
|
|
G. J. Despotis and L. T. Goodnough
Management approaches to platelet-related microvascular bleeding in cardiothoracic surgery
Ann. Thorac. Surg.,
August 1, 2000;
70(2):
S20 - 32.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
F. Forestier, S. Belisle, D. Robitaille, R. Martineau, L. P. Perrault, and J.-F. Hardy
Low-dose aprotinin is ineffective to treat excessive bleeding after cardiopulmonary bypass
Ann. Thorac. Surg.,
February 1, 2000;
69(2):
452 - 456.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. B. Ross, A. McIntyre, and C. MacDougall
Reoperation for cardiac postoperative bleeding
Ann. Thorac. Surg.,
October 1, 1998;
66(4):
1472 - 1472.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
