Bleeding is the commonest complication following cardiac surgery. The amount of blood loss is used to stratify patients in groups and judge their need for return to theatre, risks of postoperative complications, and need for transfusion of blood products. However, what is important for each individual patient is their demographic parameters, premorbid state, type of surgery and the risks of undertaking repeat operation.
Some patients suffer massive blood loss and surgical control may not be possible. These patients may need to be transferred to the intensive care unit with their chest open and packed with swabs. During the time spent in intensive care, the coagulation abnormalities can be largely corrected, providing better opportunities for haemostasis when returning to theatre.
An important feature of the bleeding patients in intensive care is the haemodynamic manifestation. This presents two types of problems: exsanguination and compression of structures by a clot. The intravascular blood volume lost needs merely to be replaced. However, the correct formula for which product and how much is difficult to determine for each patient. Laboratory blood tests and algorithms for how this should be done are described below. It is also important to remember that sometimes the bleeding is so fast that mechanical infuser devices and cell savers are needed. The second haemodynamic problem, compression of structures by a clot, is more challenging. A small amount of clot or fluid may be present in all cardiac surgical patients. When the amount of clot or fluid impedes filling of chambers or vessels, or causes a compression effect (tamponade, Figure 36.1), it has to be evacuated. Clinical examination, vital signs monitoring and echocardiography are used to guide decision making in these circumstances.
The bleeding patient in cardiac intensive care is also resource consuming. The patient often requires more than one nurse to help with blood sampling, requesting products, administering volume, operating cell savers, controlling temperature and frequent observations. The medical staff are often by this particular patient’s bedside. When numerous discussions on how this patient is best managed are added, the effort consumption detracts resources from other patients in the unit at the same time. Ironically, the bleeding patient is not only in a perilous state themselves, but also exposes other patients to detriment. Therefore careful early assessment and early management are essential for successful outcome.
Figure 36.1 Clinical Information System (CIS) illustration of reducing urine output, followed by reduction in blood pressure and increase in CVP and blood loss.
The Bleeding Patient
Following cardiac surgery, bleeding is one of the major expected associated problems. Therapeutic approaches include both prevention and treatment, and provide a unique opportunity to use prophylactic therapies including antifibrinolytics. In the bleeding patient, standard therapeutic approaches include blood product administration, multiple concomitant pharmacological agents, and the increasing use of both purified and recombinant haemostatic factors. It has been suggested that approximately 5–7% of patients bleed greater than 2 l of what may be reflected by chest tube drainage within the first postoperative 24 hours. When bleeding occurs, about 5% of patients require surgical re-exploration for bleeding. When cardiac surgical patients develop bleeding, this will increase the length of stay and is associated with a higher mortality.
Multiple factors contribute to the complex causes of bleeding in cardiac surgical patients. This includes surgical site bleeding, the effects of anticoagulants, including heparin, on platelet activation, contact activation, fibrinolytic and inflammatory pathway activation, dilutional changes, hypothermia, and other factors. Patients often are also on routine anticoagulants and antiplatelet agents that include oral anticoagulants (warfarin, dabigatran, rivaroxaban, apixaban, edoxaban) and platelet inhibitors (P2Y12 receptor inhibitors – clopidogrel, prasugrel or ticagrelor). Thus, bleeding and coagulopathy following cardiac surgery are due to multiple factors that can be highly problematic after cardiopulmonary bypass.
Developing a specific therapeutic plan, especially with the use of transfusion algorithms, has been shown to consistently reduce allogeneic blood administration. It is important to realise that any laboratory testing that prevents empirical blood product administration is important as part of a multimodal approach to blood conservation and reduction of allogeneic blood product use. Transfusion algorithms generally recommend the following: administration of fresh frozen plasma (FFP) when bleeding is accompanied by a PT or a PTT > 1.5 times normal value, platelet transfusions for thrombocytopenia with a platelet count <50,000–100,000, or cryoprecipitate or fibrinogen concentrates when fibrinogen levels are <200 mg/ dl (2 g/l). As will be discussed later, the critical role for fibrinogen levels continues to evolve, with most data suggesting the importance of normalising fibrinogen in a bleeding patient. With critical bleeds, and longer turnover time in standard laboratory testing, point-of-care testing, including rotational thromboelastometry (ROTEM©, Tem International, Munich, Germany), thromboelastography (TEG©, Haemonetics, Inc., Braintree, MA, USA) and/or platelet function testing, is important. In the actively bleeding patient, testing for platelet dysfunction is unreliable as most tests need a relatively normal platelet count and most of the platelet function testing may not work following dilutional changes and activation after cardiopulmonary bypass.
In the haemorrhagic patient, and from lessons learned in traumatic coagulopathy, the use of crystalloid and/or colloid administration may be helpful as first line therapy. With massive haemorrhage or uncontrollable bleeding, however, these compounds do not provide appropriate platelets or coagulation factors. The use of massive transfusion protocols, derived from lessons learned in traumatic coagulopathy, should be considered and implemented during uncontrollable bleeding in cardiac surgical patients. This routinely includes, in addition to red blood cells, FFP, platelets and cryoprecipitate (or fibrinogen concentrates). In European countries, factor concentrates (e.g., fibrinogen and prothrombin complex concentrates) are increasingly used to restore circulating levels of critical factors for haemostasis, and have been the subject of recent studies. In addition, in the patient with massive haemorrhage, hypothermia and acidosis frequently occur, requiring maintenance of normothermia and additional correction of metabolic abnormalities.
Of all the blood product administration used clinically, data elucidating when platelet transfusion should be administered are not clear. This is in part due to the problem of measuring platelet function after cardiopulmonary bypass. In the bleeding patient, because of dilutional thrombocytopenia and acquired platelet dysfunction, platelets are often transfused, but again monitoring platelet dysfunction is problematic in this setting. This is further complicated by extensive use of antiplatelet medications as outlined above. Whole blood clotting tests, including ROTEM© and TEG©, have been used as a monitor of platelet–fibrinogen interaction based on maximal clot firmness or clot strength. With that said, these values are indirect correlates and are highly affected by fibrinogen levels. Most of the data on platelet administration are from oncology patients and we still do not know the critical platelet mass for surgical patients. Thrombocytopenia may not always correlate with abnormal bleeding.
In the patient with ongoing haemorrhage, a massive transfusion coagulopathy can develop, which is defined as 10 or more units of packed red blood cells (PRBCs) transfused in a 24 hour period, although most of this literature is from trauma patients and retrospective evaluations. Laboratory testing often lags behind blood product administration during these scenarios. As a result, transfusion protocols have been developed using fixed doses of FFP, PRBCs and platelets that are administered as a fixed 1:1:1 ratio. In cardiac surgical patients, there are few data to determine whether these therapeutic approaches improve bleeding outcomes although it may be logical to adopt massive transfusion protocols. In trauma and battlefield patients, however, fixed ratios have become the mainstay of therapy although much of these data are from retrospective analyses of large databases.
Details of the variety of blood products are discussed in Chapter 13.
Activation of the fibrinolytic system is an important mechanism of vascular homeostasis. Mechanistically, plasmin generation is the enzymatic serine protease responsible for fibrinolysis and is formed following the conversion of plasminogen to plasmin. Plasmin is an enzyme that cleaves multiple proteins, including fibrin but also fibrinogen. Following cardiopulmonary bypass and/or tissue injury that occurs with surgery or trauma, fibrinolysis is activated and represents an important cause of coagulopathy. In trauma, orthopaedic surgery, and most importantly cardiac surgery, multiple studies report the role of antifibrinolytic agent administration in order to decrease bleeding and the need for allogeneic transfusions. These synthetic antifibrinolytic agents include the lysine analogues, epsilon aminocaproic acid (EACA) and tranexamic acid (TXA) that interfere with the binding of plasminogen to fibrin, necessary for activating plasmin. Aprotinin, a broad spectrum protease inhibitor, is a direct plasmin inhibitor. The specific antifibrinolytic agents will be considered as follows.
The two antifibrinolytic agents administered clinically include EACA and TXA. As previously mentioned, EACA and TXA competitively inhibit plasminogen conversion to the active enzymatic agent plasmin. TXA can also inhibit plasmin but at higher plasma levels. Most of the studies and efficacy data with the lysine analogues are reported with TXA, but EACA continues to be extensively utilised in the USA. EACA does not consistently reduce transfusion requirements or surgical re-exploration, but it is inexpensive and tends to be frequently substituted for TXA in the USA. Although multiple studies, primarily meta-analyses of randomised controlled trials, have reported a decrease in bleeding in cardiac surgical patients, there are limited safety data about the use of antifibrinolytic agents. Most dosing studies include EACA at ~20 to 30 g per case, or TXA at doses that range from 2 to 25 g, although most TXA dosing regimens involve 2 to 8 g dosing.
One issue with TXA is the potential increased risk of seizures. The incidence of postoperative convulsive seizures in a German hospital increased from 1.3% to 3.8% after cardiac surgery, a finding that was noted to be temporally associated with high-dose TXA. They noted 24 patients who developed seizures postoperatively and had received high doses of TXA that ranged from 61 to 259 mg/kg. The mean age of the patients was ~70 years, and 21/24 had undergone valve surgery/open chamber procedures. Although the underlying mechanisms are not fully elucidated, TXA enhances neuronal excitation by antagonising inhibitory gamma-aminobutyric acid (GABA) neurotransmission. Lecker showed that TXA inhibits neural glycine receptors, while inhibition of the inhibiting neurotransmitter glycine is an established cause of seizures. Viewing the similarities in the chemical structure of TXA, GABA and glycine, it is conceivable that an interaction of TXA with the receptors of both inhibitory neurotransmitters contributes to the increase in clinical seizures observed when TXA is given. However, in view of the chemical structure of EACA and its close similarity to GABA and glycine as well, it is noteworthy that it has not been reported to produce neurological side effects.
This is a bovine serin protease inhibitor, which inhibits complement and contact factor activation via kallikrein and plasmin activation. The half-life is 5–10 hours. In 2008 Managno reported increased incidence of renal failure, myocardial infarction and heart failure in recipients. This led to the BART trial in 2008, resulting in revoking of the drug license when an excess mortality was found, despite beneficial effects on blood loss and transfusion. These findings have been disputed and aprotinin is available on a named patient basis for low risk cases. Aprotinin affects celite based tests such as some aPTT reagents and can suggest profound coagulopathy when in actual fact the patient has no excessive bleeding. Heparin monitoring is also affected.