Heparin-induced thrombocytopenia (HIT) is an immune-mediated disease where antibodies against a neo-antigen of heparin bound to a plasma protein (mostly platelet factor 4 (PF4)) are generated. The antigen/antibody complexes bind to the Fc receptor of platelets thus causing platelet activation, with further release of PF4, and massive thrombin generation. The clinical correlate of this self-amplifying cascade is a rapid decrease of the platelet count, which in some patients is associated with massive thromboembolic events: thrombotic HIT (T) [4].

Typically, HIT antibodies generate after a prolonged therapy with heparins (5–14 days) and disappear 40–100 days after stop of heparin therapy. In case of re-exposure to heparin (even a single bolus), it usually takes 5–7 days until new HIT antibodies are generated.

Clinical scoring systems, such as the “4 T Score,” are helpful to assess the probability of HIT [4, 5]. However, the diagnosis is finally confirmed using a specific ELISA test which detects the HIT/PF4 IgG antibodies or by a platelet activation test in which patient serum and defined amounts of heparin lead to the aggregation of washed donor platelets [5].

If HIT is diagnosed or even suspected, heparin must be stopped and therapeutic anticoagulation started with an alternative anticoagulant [5]. Argatroban, a direct thrombin inhibitor, is approved for therapy and prophylaxis of HIT in the USA and most countries in Europe.

The incidence of HIT varies largely among different patient populations. While in standard cardiac surgical patients the incidence of HIT is 0.5–1%, in MCS patients a rate of approximately 10% has been described [6, 7]. If HIT occurs in this special patient population, the rate of thromboembolic complications is high and patient outcome worse, if alternative anticoagulation strategies are not initiated immediately [6, 7].

Management of a patient with HIT during cardiac surgery, particularly when CPB must be used, is difficult as no alternative anticoagulation strategies are approved. However, the actual 9th ACCP Guidelines provide useful information for management of this complex condition [5].

The highest premise in the management of a patient diagnosed with HIT(T) is that the underlying risks of the alternative anticoagulation procedure should not exceed the HIT-associated risks, which, however, is hard to be defined prospectively. The risk of the alternative anticoagulation approach should be clearly defined in the light of the status of the patient, the issues of the planned surgery, and the experience of the team involved with the use of such strategies. In select cases, it should be discussed if modification of the surgical strategy may help to reduce the intraoperative risk. For example, the implantation of a LVAD on ECLS and interventional implantation of a closure device may be preferred to implantation of the LVAD and surgical closure of a persistent foramen ovale with use of CPB. Only after such a process of “risk stratification” the optimal case-sensitive decision can be made.

The HIT antibodies are transient, and after re-exposure to heparin, it takes days until new antibodies are generated. Therefore, when a negative HIT antibody status is proven by laboratory assays, surgery can be safely performed using heparin anticoagulation. However, special care must be taken that any further exposure to heparin in the pre- and postoperative period is strictly avoided [5].

In line with this recommendation, surgery in patients with a positive HIT antibody status should be postponed whenever possible until antibodies have been proven to have disappeared [5]. Although this strategy might not be applicable for most of the urgent candidates for VAD implantation, it might be useful in selected patients which can be recompensated by further medical therapy.

In patients with acute HIT who need urgent surgery, bivalirudin anticoagulation should be preferred to other alternative agents or strategies using heparin and a short-acting antiplatelet agent [5].

Bivalirudin, a short-acting (elimination half-life approximately 25 min) direct thrombin inhibitor, is the only agent which has been prospectively studied in cardiac surgical patients with and without HIT. Bivalirudin binds bivalent to thrombin and thereby inhibits its anticoagulant action. However, the bonded bivalirudin molecule is cleaved by thrombin itself, so that the thrombin molecule achieves its anticoagulant action again. While approximately 80% of the elimination of bivalirudin is performed via this enzymatic mechanism, only 20% are eliminated via the renal pathway. An antidote is not available. The administration of bivalirudin usually follows a strict protocol with an initial bolus and a defined constant infusion to maintain drug concentrations [5] (◘ Table 18.2). As hemofiltration impacts bivalirudin concentrations, this procedure is not recommended during the course of CPB, particularly as the ACT is not a specific assay for monitoring bivalirudin concentrations.

The fact that bivalirudin has a short elimination half-life and an elimination mechanism largely independent of the function of special organ systems makes it an optimal agent for short and effective anticoagulation. However, what at the first sight appears promising, particularly when a VAD implantation is performed, turns out to increase the complexity of the procedure. In spaces where there is no flow and blood stagnation and therapeutic bivalirudin concentrations cannot be maintained by continuous systemic infusion, “clot formation” will occur. Under the special condition of VAD implantation using CPB, this will occur in the pericardial and pleural space, the CPB reservoir, and flushed grafts, cannulas, or the chamber of flushed devices. Therefore, cardiotomy suction should be replaced by cell salvage and flow in reservoirs maintained by using shunting lines and continuous stirring of the volume. When a VAD system is implanted, modification of the surgical strategy may be necessary to reduce the risk of “clot formation.” Grafts and cannulas of a VAD system should remain clamped or periodically flushed until all anastomoses are performed. Only after having completed this the chamber of the device can be deaired, and the action of the VAD should be started immediately. Successful use of bivalirudin anticoagulation for total artificial heart and LVAD implantation on CPB and LVAD implantation on an ECLS, using a lower-dose bivalirudin protocol, has been described [8–10].

A case series using argatroban anticoagulation during ECLS support has been published recently [11]. After having experienced massive intraventricular thrombus formation in one patient, changing the implantation strategy, and increasing argatroban dosing (target activated partial thromboplastin time 70–80 s), four of six patients needed re-exploration due to bleeding complications. Therefore, argatroban may not be recommended in this indication.