70 Bleeding and haemostasis disorders

The main cause of haemostasis defects and related bleeding complications in patients with acute coronary syndromes admitted to the intensive cardiac care unit is the use of multiple antithrombotic drugs, alone or concomitantly with invasive procedures such as percutaneous coronary intervention with stent deployment and coronary artery bypass surgery. These drugs, that act upon several components of haemostasis (platelet function, coagulation, fibrinolysis), are associated with bleeding complications, particularly in elderly patients (more so in women than in men), those who are underweight, and those with comorbid conditions such as renal and liver insufficiency and diabetes. The identification of patients at higher risk of bleeding is the most important preventive strategy. Red cell and platelet transfusions, which may become necessary in patients with severe bleeding, should be used with caution, because transfused patients with acute coronary syndrome have a high rate of adverse outcomes (death, myocardial infarction, and stroke). To reduce the need of transfusion, haemostatic agents that decrease blood loss and transfusion requirements (antifibrinolytic amino acids, plasmatic prothrombin complex concentrates, recombinant factor VIIa) may be considered. However, the efficacy of these agents in the control of bleeding complications in acute coronary syndrome is not unequivocally established, and there is concern for an increased risk of re-thrombosis. A low platelet count is another cause of bleeding in the intensive cardiac care unit. The main aetiologies are drugs (unfractionated heparin and glycoprotein IIb/IIIa inhibitors), such thrombotic microangiopathies as thrombotic thrombocytopenic purpura and disseminated intravascular coagulation, that are often paradoxically associated with thrombotic manifestations. In conclusion, evidence-based recommendations for the management of bleeding in patients admitted to the intensive cardiac care unit are lacking. Accurate assessments of the risk of bleeding in the individual and prevention measures are the most valid strategies.

When the integrity of the vessel wall is altered by a perforating injury, blood loss is stopped by the intervention of the haemostasis system, ultimately leading to plug formation and vascular sealing. Haemostasis is assured by four phases that, acting in synergy, lead to the arrest of bleeding: (1) the vascular phase, (2) the platelet phase, (3) blood coagulation, and (4) fibrinolysis (FL).

The brief period of vasoconstriction that follows a vascular injury is generally unable to stop clinically significant blood losses. Consistent with the trivial relevance of this phase in securing haemostasis, congenital and acquired functional abnormalities of the vascular phase are not established causes of major bleeding.

Vascular injury exposes circulating blood to subendothelial layers, to which blood platelets tether and adhere by interacting with the adhesive glycoprotein von Willebrand factor (VWF) and collagen. The ultimate result of these early reactions is platelet aggregation and the formation of the primary haemostatic plug. Quantitative or qualitative defects of platelets lead to an inadequate or a delayed formation of the haemostatic plug and may cause bleeding, which occurs mainly in the skin and mucosal tracts (petechiae, epistaxis, menorrhagia, melaena). The most frequent defects of platelet number and function can be controlled generally by replacement therapy by means of the transfusion of allogeneic platelets.

The platelet-rich haemostatic plug is frail and unable per se to stop bleeding (particularly from large vessels), unless it is strengthened by a mesh of fibrin. Fibrin formation is the final event of blood coagulation that results from the sequential enzymatic activation of coagulation factors, mainly occurring on the platelet surface, that ultimately leads to the transformation of the soluble plasma protein fibrinogen into fibrin. Congenital or acquired defects of coagulation factors cause a bleeding tendency that is usually more clinically severe than that due to platelet defects. Haemorrhage mainly occurs in soft tissues, muscles, joints and from trauma- or surgery-induced wounds, but it may also occur in life-threatening sites such as the central nervous system (CNS), retroperitoneal space, and gastro intestinal (GI) tract. When bleeding is due to the plasma defect of single coagulation factors (as in the haemophilias), the best therapeutic approach is specific replacement of the deficient or dysfunctional coagulation factor. More often, particularly in acquired coagulation defects, more than one factor are deficient in plasma. Fresh frozen plasma is the only weapon that contains all the coagulation factors, while prothrombin complex concentrate (PCCs) fractionated from plasma contain only the coagulation factors that are depleted therapeutically by vitamin K antagonists (VKAs) (factors II, IX, X, and sometimes also VII).

The fibrinolytic system is made of enzymes that ultimately activate the proenzyme plasminogen to form the proteolytic enzyme plasmin that lyses fibrin-rich clots. The most frequent clinical conditions associated with hyperfibrinolysis are the therapeutic administration of tissue plasminogen activator (tPAs) or other fibrinolytic drugs in order to obtain thrombolysis in patients with acute coronary syndromes (ACS) or those with other arterial or venous thrombosis (particularly ischaemic stroke of recent onset).

In the intensive cardiac care unit (ICCU), excessive bleeding typically occurs in patients who have multiple haemostasis defects, due to the inhibition of blood coagulation by parenteral anticoagulants (unfractionated heparin (UFH), low molecular weight heparin (LMWH), bivalirudin, fondaparinux) and oral anticoagulants (warfarin and other VKAs, and the direct oral anticoagulants dabigatran, rivaroxaban, apixaban and edoxaban). The inhibition of platelet function by antiplatelet agents such as aspirin, adenosine diphosphate (ADP) receptor (P2Y₁₂) antagonists (clopidogrel, prasugrel, ticagrelor), and glycoprotein inhibitors (GPIs) (abciximab, tirofiban, eptifibatide) is another important cause of excessive bleeding. As already mentioned, thrombolytic agents cause hyperfibrinolysis. Another typical condition associated with multiple haemostasis defects that sometimes lead to excessive blood loss is cardiac surgery and particularly repeat surgery, typically coronary artery bypass surgery. In the latter, bleeding is related to several causes: the large size of the surgical wound, a decrease of the platelet number during the circulation of blood in the extracorporeal oxygenator, hypocoagulability due to an incomplete neutralization of heparin after discontinuation of the pump, and acquired coronary by pass highlight of the year, thank you, defects of platelet function, often accompanied by hyperfibrinolysis.

With this background, the pharmacological agents that help to stop bleeding and/or reduce transfusion requirements when a major blood loss occurs will be reviewed. To validate the efficacy of these drugs, cardiac surgery will be taken as a clinical model, because haemostatic agents were extensively investigated in the context of this type of surgery, in order to reduce blood loss and transfusion requirements. Major bleeding, often associated with the use of invasive procedures in patients admitted with ACS who usually take multiple antithrombotic drugs, will also be covered. Finally, the thrombocytopenic states that sometimes occur in the ICCU and may cause not only bleeding symptoms, but also paradoxically very severe thrombotic manifestations will be discussed.

Drugs that were evaluated clinically are the antifibrinolytic amino acids aminocaproic acid and tranexamic acid, obtained by chemical synthesis; the broad-spectrum protease inhibitor aprotinin, extracted from bovine lung; the plasma-derived concentrates of the vitamin K-dependent coagulation factors (the so-called PCC) that contain factors II, IX, and X, and, in some instances, also factor VII; and the activated form of coagulation factor VII (FVIIa), produced by recombinant deoxyribonucleic acid (DNA) technology [1, 2]. Desmopressin (DDAVP), a synthetic analogue of ADH, has been used in 1980s and 1990s as a general haemostatic agent, particularly in cardiac surgery. However, early favourable results were not subsequently confirmed, so that desmopressin is currently licensed only for the treatment of bleeding episodes in patients with mild to moderate haemophilia and von Willebrand disease [1, 2].

6-aminohexanoic acid (aminocaproic acid) was the first drug of this category to be introduced and evaluated therapeutically [3]. 4-aminomethyl cyclohexanecarboxylic acid (tranexamic acid) was introduced subsequently; it is approximately ten times more potent than aminocaproic acid and has a longer half-life that allows dosing at more spaced intervals [4] (in general, every 8, instead of 4, hours). Both of these drugs inhibit fibrinolysis by saturating the lysine binding sites on the plasminogen molecule, which are essential for the binding to fibrin clots of the proenzyme of the fibrinolytic enzyme plasmin [5].

In cardiac surgery, both tranexamic and aminocaproic acid reduce the need for perioperative blood transfusion by approximately 30–40% [6]. The main perceived risk of these drugs is thrombosis, due to their inhibition of an important antithrombotic systems such as fibrinolysis. There are a number of reports on thrombotic complications in patients taking either drug, but causality is uncertain, because the clinical conditions in which these agents were used are associated per se with a high risk of thrombotic complications. Indeed, a Cochrane systematic review does not support the views that thrombotic events are more frequent in patients treated with these antifibrinolytic agents [6]. The value of tranexamic and aminocaproic acid rests on the fact that they are safer and definitely much less expensive than other haemostatic drugs such as aprotinin and factor VIIa (see next sections). For recommended dosages and schedules of administration, see Table 70.1.

This broad-spectrum protease inhibitor was promoted to control bleeding in cardiac surgery, because it inhibits the fibrinolytic enzyme plasmin [7, 8]. A Cochrane systematic review [6] that summarized the results of 61 trials demonstrated a reduction by 30% of the need for allogeneic transfusion of red cells, platelets, and fresh frozen plasma, and a 60% reduction in the need for reoperation following excessive peri- and post-operative bleeding. The systematic review also looked at the adverse effects of aprotinin and concluded that the drug was generally quite safe [6]. However, the numerous clinical trials of aprotinin were underpowered to assess safety. An answer on the issue of aprotinin safety stemmed from the Blood Conservation Using Antifibrinolytics in a Randomized Trial (BART) [9], carried out in 2331 patients undergoing complex and repeat cardiac operations at particularly high risk of excessive bleeding [9] and randomized to receive aprotinin, tranexamic acid, or aminocaproic acid [9]. The trial had to be prematurely interrupted, because of a higher 30-day mortality rate in patients receiving aprotinin (6.0%), compared with rates of 3.9% in tranexamic acid-treated patients and 4.0% in patients treated with aminocaproic acid [9]. The excess of deaths observed in aprotinin-treated patients were mainly due to cardiac causes. BART also demonstrated that antifibrinolytic amino acids are perhaps slightly less effective than aprotinin in reducing blood loss and transfusion requirements [6], but much safer and less expensive. As a result of these findings, aprotinin is no longer licensed for therapeutic use in some countries but is still available in others.

PCCs are extracted from human plasma and are virally inactivated with various methods. Some commercial products contain all the vitamin K-dependent coagulation factors, namely II, VII, IX, and X (for instance, Octaplex^® and Beriplex^®); others are devoid of factor VII (for instance, Bebulin^® and Profilinine SD^®). They can be used for the rapid reversal of the anticoagulant effects of VKAs such as warfarin, acenocoumarol and related drugs. A dose of 50 U/kg of PCCs, containing all vitamin K-dependent factors, is usually sufficient to normalize the prothrombin time international normalized ratio (INR), and this dose can be repeated at 12–24-hour intervals, if needed. Oral or IV vitamin K (at doses ranging from 1 to 5 mg, depending on the INR values) is another option to reverse the anticoagulant effects of VKAs, but because of its slow onset of action (particularly by the oral route) this approach is inadequate if rapid reversal is warranted such as, for instance, before emergency surgery or other invasive procedures in the frame of the ICCU.

Produced by recombinant DNA technology, this activated form of coagulation factor VII promotes haemostasis by binding to TF exposed on the damaged vessel wall and in the extravascular space, thereby generating small amounts of thrombin [10–12]. In turn, thrombin acts mainly through further generation of thrombin on the platelet surface. The reactions triggered by factor VIIa and thrombin on the platelet surface ultimately generate enough thrombin to transform plasma fibrinogen to fibrin at sites of vascular injury. Factor VIIa is also prohaemostatic through the activation by means of enhanced thrombin production of one of the principal inhibitors of fibrinolysis (i.e. thrombin-activatable fibrinolysis inhibitor, TAFI) [13].

Recombinant factor VIIa was initially developed to bypass the coagulation defect in patients with haemophilia A complicated by inhibitory anti-factor VIII antibodies, and it is licensed for the treatment of bleeding episodes in these patients. The aforementioned general haemostatic properties of factor VIIa have prompted its use in conditions other than haemophilia that are associated with major blood loss [14–17]. For the recommended dosages of recombinant factor VIIa, see Table 70.1. At the moment, recombinant factor VIIa is only licensed for indications such as haemophilia, factor VII deficiency and Glanzman thrombasthenia, so that for other indications it can only be administered on a patient-named basis for other indications. Nevertheless, it is being used with increasing frequency as a general haemostatic agent has become an attractive option in patients bleeding for various reasons. There are no randomized clinical trials showing its efficacy beyond that in haemophilia and thrombasthenia; the drug is very expensive, and the risk of thrombotic complications looms large, particularly in patients with ACS inherently at high risk.

Finally, and most importantly, when there is a situation of excessive bleeding, before considering the use of the aforementioned haemostatic agents and transfusional blood products, simple conservative measures should be considered such as pressure applied to the bleeding site, when accessible, with or without the adoption of adjunctive local measures such as fibrin glue. Table 70.1 summarizes the general strategies that can be adopted for the management of excessive bleeding.

In patients treated in the ICCU, bleeding may range from minor haemorrhage (e.g. bruises, nosebleeds and specially oozing from venous access sites) to life-threatening intracranial, GI, and retroperitoneal haemorrhages. Bleeding at the vascular access site after PCI is the most common bleeding complication in patients with ACS, even though it has been significantly reduced in the last decade, thanks to the increasing adoption of the radial access. As mentioned before, the combined use in ACS of several drugs that impair multiple phases of haemostasis is the most frequent reason for bleeding. PCI with stent haemophdeployment in patients who are taking anticoagulant and antiplatelet medications does dramatically increase the risk of bleeding, which is particularly frequent in ACS patients at high pre-procedural risk such as the elderly (more in women than in men), those who are underweight, and those with diabetes and renal and liver insufficiency and those taking dual or triple antithrombotic medications [18]. The newer antiplatelet agents prasugrel and ticagrelor, administered on top of aspirin, have a higher risk of bleeding than clopidogrel plus aspirin. A crucial issue is the clinical impact of haemorrhagic complications and their severity on the patient prognosis. Another issue is whether or not medical interventions, by means of transfusion of whole blood, red cells, or platelets (see Table 70.1), can avoid the adverse effects of bleeding complications.

There have been attempts to categorize the severity of bleeding in patients with ACS, but the definitions are not uniform. This has led to significant variations in the reported incidence of bleeding, which ranges from as little as 0.2% to as high as 11.5% [19]. Originally, bleeding was classified as major when it was intracranial or retroperitoneal, or associated with anaemia that is severe enough to warrant blood transfusion [20]. Additional criteria have been subsequently developed, in order to define more accurately the severity of bleeding in non-surgical cardiac patients [21]. The criteria adopted in the frame of the Thrombolysis In Myocardial Infarction (TIMI) clinical trials of antithrombotic therapies, based upon clinical and laboratory measurements, identify four categories of bleeding (major, minor, minimal, or none) [22]. The criteria adopted in the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO) trials also identify four categories (severe or life-threatening, moderate, mild, and none) on the basis of the need, or not, of transfusion and the presence, or not, of haemodynamic compromise [23]. The evaluation of the risk of bleeding with the TIMI or GUSTO criteria does not give identical results. Although the GUSTO criteria are thought to give more reliable and clinically relevant results [24], alternative criteria are used, adding to the existing confusion. For instance, the sensitive Acute Catheterization and Urgent Intervention Triage Strategy (ACUITY) criteria identify major bleeding on the basis of a detailed clinical assessment, changes in haemoglobin levels, haematomas of >5 cm, and the need for blood transfusion. Finally, the Bleeding Academic Research Consortium produced a consensus report containing standardized definitions of bleeding to be employed in the frame of cardiovascular clinical trials [25].

Major or moderately severe bleeding has a negative effect on prognosis, as early and firmly established by the results of meta-analysis, large registries, and clinical trials (for instance Organization for the Assessment of Strategies for Ischemic Syndromes (OASIS) registry, OASIS-2, Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE), and Global Registry of Acute Coronary Events (GRACE)). Overall, major bleeding occurring in ACS is associated with an approximately 5-fold higher incidence of death at 30 days, a 1.5-fold higher incidence of death between 30 days and 6 months, a 5-fold increase of myocardial infarction (MI) and a 3-fold increase of ischaemic stroke [26]. The 30-day risk of mortality depends on the severity of the bleeding, with a much higher mortality in patients with severe bleeding than in those with moderate or mild bleeding. Some clinicians maintain that the association between bleeding complications and excessive mortality is not causal and that excess mortality is explained by more frequent comorbidities in bleeders [27]. Another likely possibility is that bleeding frequently causes the sudden interruption of antithrombotic agents, a choice that, in these high-risk patients, may favour the recurrence of atherothrombotic events and death. Anyway, the perceived clinical relevance of these complications has prompted the addition of bleeding to the traditional triad of events (death, MI, urgent revascularization) employed to evaluate antithrombotic drugs in patients with ACS undergoing percutaneous coronary intervention (PCI) [28].

Owing to the negative prognostic role of bleeding, the early and liberal use of blood transfusion, particularly of red cells and platelets, should, in theory, benefit patients and decrease the rate of unfavourable events associated with bleeding. However, there is evidence that the clinical responses to blood transfusion are not favourable. Data from early, three large clinical trials (GUSTO IIb, the Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy (PURSUIT), and the Platelet IIb/IIIa Antagonism for the Reduction of Acute Coronary Syndrome Events in a Global Organization Network (PARAGON)), enrolling 24 112 patients with ACS, were included in a meta-analysis, in order to establish whether or not there was an association between blood transfusion and adverse outcomes in patients who had had moderate to severe bleeding complications [29]. The need for transfusion was associated with an approximately 4-fold increased risk of 30-day mortality and MI [29]. The increase in adverse outcomes was independent of the severity of anaemia and therefore seems to be due to a direct negative effect of transfused blood. A more recent meta-analysis of 17 observational studies, including more than 2.5 million patients, further confirmed a significant association of red blood cell transfusion with higher short-term and long-term all-cause mortality, as well as reinfarction rates, in patients admitted to the hospital with an ACS. Moreover, after the patients were stratified by means of the baseline haemoglobin (Hb) levels, the Authors demonstrated that blood transfusion had a neutral or beneficial effect on outcome at Hb levels below 8 g/dL, while it was harmful at Hb levels above 10 g/dL [30].

Among the mechanisms of excess mortality after blood transfusion, impaired O₂ transport by stored blood has been implicated. Although there are alos contrasting data showing a favourable effect of blood transfusion on adverse outcomes [30], the intensive and liberal use of transfusions, in order to maintain normal levels of haemoglobin or haematocrit, should be discouraged in favour of a more restrictive approach.

Concerns on the adverse effects of blood transfusion in patients with ACS who bleed emphasize the role of prevention. General haemostatic agents, such as antifibrinolytic amino acids, and recombinant factor VIIa, have the potential to prevent or stop bleeding and to avoid transfusion. On the other hand, as mentioned earlier, they may increase the risk of recurrent or new thrombotic events in ACS patients. There is as yet no randomized study that establishes whether or not these haemostatic agents help to prevent or stop bleeding in patients with ACS and, most importantly, to avoid or at least, reduce potentially dangerous blood transfusions. Available evidence on the clinical impact of the use of haemostatic agents in ACS consists of case reports or small case series with no adequate control. Furthermore, the frequent parallel use of other agents with a potentially favourable impact on haemostasis (such as fresh frozen plasma and platelet concentrates) confuses the understanding of the efficacy of these drugs. Hence, the best strategy is to accurately evaluate the risk of bleeding in each patient and then to tailor the choice and dosages of antithrombotic drugs on the risk magnitude.

Evaluation of the patient’s individual risk is based upon the knowledge that there are comorbid conditions and/or patient features that increase the risk of bleeding during antithrombotic therapy for ACS. As mentioned before, increasing age (particularly in women older than 75 years), low body weight, diabetes, and kidney and liver diseases are independent and strong risk factors. Several studies also report an association between bleeding and severe hypertension, cerebrovascular disease and alcoholism. Malignancy, previous GI bleeding, and the recent administration of anticoagulants (warfarin, heparins, fondaparinux, bivalirudin, and the direct oral agents; dabigatran, rivaroxaban, apixaban and edoxaban) are other strong predictors of bleeding. There have been efforts to combine several risk factors, in an attempt to develop bleeding risk scores aimed to facilitate the choice and dosages of antithrombotic drugs for the medical treatment of ACS with or without PCI. For instance, a bleeding score validated in a large number of patients with ACS is proposed by Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the ACC/AHA guidelines (CRUSADE) [31], which includes eight parameters (see Table 70.2). According to CRUSADE, the risk of in-hospital bleeding ranks from very low (score <20, predicted rate of major bleeding of 3.1%), low (between 21 and 30, bleeding rate of 5.5%), moderate (31–40, 8.6%), to high (41–50, 11.2%) and very high (>50, 19.5%) [31]. Bleeding risk should be also reassessed during the in-hospital phase and before discharge in order to predict the out-of-hospital risk of bleeding and to define the best long-term antithrombotic treatment for any individual patient. The PRECISE-DAPT score has been recently proposed for this purpose: it can support clinical decision making for antiplatelet treatment selection and duration in the context of a comprehensive clinical evaluation process. The score takes into consideration 5 variables (age, creatinine clearance, haemoglobin, white-blood-cell count and previous spontaneous bleeding) and has been included into the algorithm related to DAPT duration decision making of the 2017 ESC focused update guidelines on dual antiplatelet therapy : score values ≥25 discourage the use of a long-lasting DAPT and the selection of prasugrel as P2Y12 inhibitor, while for this same bleeding risk ticagrelor, as an alternative to clopidogrel, is admitted only in ACS patients treated with PCI and stent implantation or with CABG [33,34].

When patients who are taking antithrombotic drugs need surgery or other invasive procedures in the frame of acute care settings, bleeding is the principal concern. On the other hand, these procedures increase the risk of atherothrombotic complications in these high-risk patients, making it impossible to implement safely the simple choice of stopping drugs affecting haemostasis. The most frequent and typical situations in ICCUs include: (1) patients on anticoagulant therapy for AF who develop an ACS and (2) patients with coronary stents on single or dual antiplatelet therapy who develop a new ACS that warrants surgery or other invasive procedures. The forthcoming recommendations on how to handle these complex clinical situations are based on little evidence other than expert opinions [32]. In general, in order to handle optimally these situations, it is important for clinicians to know the elimination half-life of the major antithrombotic drugs used in the ICCU (see Table 70.3).

When a patient with atrial fibrillation (AF) on long-term anticoagulant therapy develops an ACS that warrants surgical revascularization, this therapy cannot be continued at the time of the procedure, due to the certainty of excessive bleeding, irrespective of the anticoagulant drug used vitamin K antagonist (VKAs), such as warfarin, or the inhibitors of activated coagulation factors dabigatran, rivaroxaban, apixaban, and edoxaban. On the other hand, the high risk of thrombosis, related to the underlying ACS, plus the procoagulant effects of major surgery demand that anticoagulation is interrupted for as briefly as possible in the perioperative period and restarted as soon as possible after surgery is completed and the risk of bleeding returns to being relatively low.

If warfarin or related VKAs are used as anticoagulants, drug intake should be stopped at least 4–5 days before the procedure, because of the relatively long elimination time of the effect of these drugs (see Table 70.3). However, the risk of thrombosis demands that the so-called bridging anticoagulant therapy is started at the time of stopping VKAs, employing therapeutic doses of such short-acting drugs such as UFH or LMWH, that are cleared relatively rapidly from the circulation, provided the renal function is not severely impaired (see Table 70.3). These parenteral anticoagulants should, in turn, be discontinued no earlier than 12 hours before surgery if LMWHs are used for bridging, or 6 hours with UFH that has a shorter half-life. It is advisable to measure the prothrombin, and partial thromboplastin and thrombin times before starting surgery, to be certain that normal values of these simple tests indicate blood coagulation has returned to normal, so that the risk of excessive bleeding during surgery is small. Heparins should be started again 12–24 hours after the completion of surgery, while also resuming at the same time the dose of warfarin that the patient was taking before the surgical procedure. Heparins should then be stopped when INR testing confirms that the desired level of anticoagulation is reached again.

Bridging therapy is planned quite differently, in terms of the start and stop times when patients with AF are treated with direct oral anticoagulants, which differ from VKAs for a much shorter half-life (and hence a more rapid elimination of the anticoagulant effects) (see Table 70.3) and at the same time for an earlier return to full anticoagulation when these drugs are started again. Accordingly, they can be stopped at a time much closer to surgery than warfarin (1–2, instead of 4–5, days, at least in patients with normal renal function), and full anticoagulation is reached again more rapidly (2–3 hours) after resuming them in the post-operative period (12–24 hours after the end of the operation). As above, the prothrombin, time and the partial thromboplastin and thrombin times should be carried out to be certain that normal coagulation times indicate that the direct anticoagulants no longer exert any significant inhibitory effect.

Coronary stenting, used in approximately 85% of patients with ATS treated with PCI requires dual antiplatelet therapy (DAPT) for a minimum of 1 month up to 12 months, depending on the type of stent used subsequently asprin alone is used for at least 12 months in carriers of bare metal stent, whereas drug eluting stents DES warrant dual theorapy for at least 3–6 months with a P2Y₁₂ antagonist used on top of asprin (clopidogrel, prasugrel, or ticagrelor). When patients carrying stents develop a new ACS that warrants relatively urgent surgical revascularization, a strategy tailored to reduce the risk of perioperative bleeding, and at the same time to avoid atherothrombotic episodes, should be planned in these high-risk patients. Handling this situation differs if the patient is on aspirin alone or on DAPT at the time of surgery.

In patients on aspirin alone, we suggest to continue this drug during the whole perioperative period, because discontinuation carries a risk of rebound platelet hyperactivity and an at least 3-fold increased risk of major cardiac adverse events (AEs). The surgeons are used to operating on ‘aspirinated’ patients without mishaps, although the amount of blood loss is noticeably increased, particularly if large doses of aspirin are used (>300 mg daily).

In patients on DAPT, P2Y₁₂ inhibitors should be stopped before surgery, in order to avoid the excessive bleeding associated with their continuation at the time of major cardiac surgery, whereas aspirin should be continued. P2Y₁₂ inhibitors, such as clopidogrel and prasugrel, should be discontinued long before the procedure (5–7 days), because they irreversibly inhibit platelet function, so that this time lag is needed for the production of fresh non-inhibited platelets, and thus the return to normal of haemostasis. On the other hand, the use of the reversible inhibitor ticagrelor demands its interruption much closer to surgery (3–5 days), because the inhibitory effect of this antiplatelet agent is more short-lasting than that of clopidogrel and prasugrel. P2Y₁₂ antagonists should be resumed 12–24 hours after completion of surgery, using the loading dose that is routinely recommended in ACS. Bridging therapy, with short-acting, reversible parenteral antiplatelet agents, such as tirofiban and eptifibatide, to be stopped 2–3 hours before sugery, can be considered in patients at particular high risk of atherothrombosis (see Table 70.4).

As mentioned earlier, these should be avoided as much as possible not only to stop antithrombotic agents in case of bleeding, but also to implement measures meant to reverse their action, because the risk of re-thrombosis looms large. However, there are situations in which life-threatening bleeding (for instance, in the central nervous system (CNS), retroperitoneal space, and GI tract) demands the use of antidotes, i.e. medications that stop rapidly the effect of antithrombotic drugs and restore haemostasis. The simple approach to stop the actual antithrombotic drug is particularly valid for those with a relatively short half-life such as heparin, thrombolytic agents, and direct oral anticoagulants such as dabigatran, rivaroxaban, apixaban and edoxaban. For agents with a longer half-life, such as VKAs, reversal of the anticoagulant action with the use of PCCs is featured in detail in a previous section (see Prothrombin complex concentrates section). For other long-acting drugs, such as antiplatelet agents, platelet transfusion should be considered if bleeding is truly life-threatening, albeit with caution according to the recommendations previously given in Table 70.1.

Low platelet counts are sometime related to the use of drugs [33] but can also develop as a consequence of massive transfusion after major blood loss. Some thrombocytopenias may be associated with a bleeding tendency such as those occurring within the first 24 hours after the use of inhibitors of platelet GP IIb/IIIa. Among them, the most frequently implicated is abciximab (incidence 0.5–1.0%), which lowers the platelet count through the formation of antibodies directed against the murine component of the monoclonal antibody, that leads to the subsequent removal of antibody-loaded platelets from the circulating blood. Other GPIs, such as tirofiban and eptifibatide, are less frequent causes of thrombocytopenia (0.2–0.5%), that develops due to the generation of neoantigens and antiplatelet antibodies through the binding of these drugs to their target, i.e. the platelet GP IIb/IIIa. The rate of occurrence of thrombocytopenia is much higher after drug re-exposure. The clinical manifestations of thrombocytopenia ensuing after the intake of these drugs are seldom severe and usually do not warrant replacement therapy with platelet concentrates, nor the intake of corticosteroids to block the production and action of antibodies.

Heparin induced thrombocytopenia (HIT) typically develops 5–10 days after the onset of heparin therapy and is due to autoantibodies reacting with complexes that form on the platelet surface between heparin and platelet factor 4. UFH is much more immunogenic than LMWHs and fondaparinux [34–36]. However, the risk associated with the latter two is not absent, because they both share with UFH the pentasaccharide structure. The direct anticoagulants dabigatran, rivaroxaban, apixaban, and edoxaban are free of risk of HIT. Complexes that form between autoantibodies, heparin, and factor 4 activate platelets and enhance thrombin generation, leading to a strong hypercoagulable state. Hence, the most serious clinical consequence of HIT is the occurrence, in up to 75% of patients, of venous or arterial thrombosis [36], often developing in the same vascular district that initially made necessary the antithrombotic use of heparin. Mortality from HIT can be as high as 20%, and major complications, such as stroke or limb loss due to ischaemia, are not rare. Serological tests detect antiplatelet antibodies, but they are useful only for their negative predictive value, because positivity may occur in heparin-treated patients, in spite of the fact that they are, and will remain, symptom-free [36–38]. The most important therapeutic measure in HIT is to stop UFH and to treat any thrombotic manifestation with anticoagulants that are structurally unrelated to heparin such as the direct thrombin inhibitors (DTIs) lepirudin, bivalirudin or argatroban, and the heparinoid danaparoid. VKAs, like warfarin, should not be used in the acute phase of life-threatening HIT-associated thrombosis, because the onset of anticoagulation induced by these drugs is delayed, and early treatment of thrombotic manifestations is compelling. This drawback should be avoided by the use of the direct anticoagulants that inhibit directly thrombin (dabigatran) or activated factor X (rivaroxaban, apixaban, edoxaban) and are active 2–3 hours after intake. However, there is only very limited experience on their use in HIT. Table 70.5 lists the cornerstones of the management of HIT in patients with or without thrombosis admitted to the ICCU.

Thrombotic thrombocytopenic purpura (TTP) is a rare disease that may occur in patients with ACS treated with antiplatelet agents of the thienopyridine family (ticlopidine and, less frequently, clopidogrel). It is also a rare complication of cardiac surgery. TTP is associated with a decrease in platelet count, accompanied by haemolytic anaemia, due to the in vivo fragmentation of red cells [37–39]. A low platelet count is due to in vivo platelet consumption following the disseminated intravascular formation of platelet-rich thrombi in the microcirculation, leading, in turn, to ischaemia in multiple organs (more often the brain, but also the kidney, myocardium, and GI tract). Platelet-rich thrombi are due to in vivo platelet aggregation induced by highly thrombogenic forms of VWF that are not cleaved to the less thrombogenic physiological forms of this adhesive protein, due to the deficiency or dysfunction of the plasmatic VWF-cleaving metalloprotease ADAMTS13 [37, 38]. The thienopyridine-associated forms of TTP, which usually develop 10–15 days after the intake of ticlopidine, or should be treated with daily sessions of plasma exchange (4–6 L of plasma), together with the administration of large doses of corticosteroids (1.0–1.5 mg/kg, tapered off over 30–40 days). The outcome of TTP after thienopyridine drug intake is usually favourable, provided this complication is promptly diagnosed, the implicated drugs are stopped, and plasma exchange is initiated early and continued for an average of 7–10 days. ADAMTS13 assays are useful, but not essential, for diagnosis, because there are TTP cases with normal levels of the VWF-cleaving protease [39].

DIC may develop in ICCUs for two main reasons: bacterial infections (often related to infected catheters and other devices used for venous access) and post-haemorrhagic hypovolaemic shock. Bleeding symptoms may be prominent (oozing from venous lines, GI bleeding, soft tissue haematomas), but ischaemic symptoms due to microvascular thrombosis may also occur, due to impairment of the microcirculation of several organs (mainly the kidney) by intravascular thrombi [40]. Thrombocytopenia varies from moderate to severe, and there are signs of heightened activation of the coagulation system, the most typical being the presence of marked elevations of the fibrin degradation product D-dimer, which is usually much in excess of 1000–2000 micrograms/mL [40, 41]. More rarely, there are laboratory signs of consumption coagulopathy, with prolongations of the prothrombin, partial thromboplastin, and low plasma levels of fibrinogen and other coagulation factors and inhibitors (factors V and VIII, antithrombin, protein C).

The mechanistic approach to the treatment of DIC is principally based upon the removal of the underlying condition, such as bacterial infections and haemorrhagic shock, and the correction of the often-present acidosis (see Table 70.6). Replacement of the deficient factors and platelets, by means of the infusion of fresh frozen plasma or platelet concentrates, is usually advisable when bleeding is associated with abnormal coagulation tests, even though there is some concern that the supply of exogenous factors and platelets may help to maintain the DIC process. Replacement therapy with naturally occurring anticoagulants, such as the administration of large doses of antithrombin concentrates (80–100 U/kg), has been advocated, but evidence of its clinical efficacy is weak [41].

The combined use of multiple antithrombotic drugs has dramatically improved the prognosis of ACS. However, these drugs induce multiple abnormalities of haemostasis and thereby may cause bleeding, even though the balance is definitely in favour of their use. The most important approach is prevention, which must be realized through the scrutiny of the pattern of bleeding risk factors in the individual. It is not recommended to skip the antithrombotic strategies and invasive procedures that are of proven efficacy in ACS patients at high risk of bleeding (the elderly, those who are underweight, and those with renal insufficiency). Yet, the combination of drugs and their dosages may be tailored and modified, according to each individual patient and his/her category of risk (for examples, see Personal perspective section).

If bleeding is severe, blood transfusion should be considered. However, the concept that patients with coronary artery disease (CAD) should be aggressively transfused, due to the putative adverse effects of anaemia on their heart disease, is challenged by the accumulating evidence that intense transfusional regimens may increase mortality and cardiovascular events. It remains to be established whether or not haemostatic drugs are efficacious, when bleeding occurs in patients with ACS treated with antithrombotic drugs. Successful results in this field stem from case reports and case series, and they cannot be taken as solid evidence of efficacy. The question should be answered by means of controlled clinical trials. Antifibrinolytic amino acids should be preferred, not only to aprotinin, but also to recombinant factor VIIa, for their higher safety and lower costs.

Tables 70.7 and 70.8 summarize the sequential steps that I recommend to consider for the general management of bleeding in patients with ACS admitted to ICCU and treated with multiple antithrombotic drugs that lead to complex defects of haemostasis.