Cangrelor for cardiopulmonary bypass in delayed-onset heparin-induced thrombocytopenia: a case report

Abstract

Background

Anticoagulation for cardiopulmonary bypass (CPB) in cases of heparin-induced thrombocytopenia (HIT) is challenging as no convenient and proven alternative, such as heparin alone, exists. A “platelet anesthesia” concept using antiplatelet agent cangrelor with heparin has been successfully reported in this setting.

Key Clinical Question

In cases of acute HIT, is CPB with cangrelor plus heparin effective and safe?.

Clinical Approach

We report the case of a patient who developed, 2 weeks after patent foramen ovale (PFO) closure, a delayed-onset HIT complicated with carotid, popliteal, and PFO device thromboses that could not be controlled by argatroban anticoagulation and required urgent cardiac surgery. CPB for PFO occluder removal and popliteal thrombectomy were performed using cangrelor with heparin without complication. Neither a new thromboembolic event nor abnormal bleeding was noticed in the postoperative period.

Conclusion

CPB using cangrelor with heparin seems to be an effective alternative for acute HIT.

Essentials

• •Cardiopulmonary bypass (CPB) is a challenging issue in cases of acute heparin-induced thrombocytopenia (HIT).
• •No preferred anticoagulation strategy for acute HIT requiring CPB is well-established.
• •Cangrelor is a powerful antiplatelet agent with suitable properties for CPB.
• •Cangrelor with heparin is a possible anticoagulation regimen for CPB in cases of acute HIT.

1. Introduction

Heparin-induced thrombocytopenia (HIT) is a severe complication of heparin exposure caused by autoantibodies that activate platelets and cause life-threatening arterial and venous thromboses. Delayed-onset HIT is a subtype of autoimmune HIT with a more severe thrombotic phenotype where thrombocytopenia and thromboses occur at least 5 days after heparin withdrawal [1,2]. Given that heparin must be discontinued in the acute phase of HIT, cardiac surgery with cardiopulmonary bypass (CPB) appears particularly challenging since there is no alternative as safe and validated as unfractionated heparin (UFH). One strategy is to combine UFH with “platelet anesthesia” using powerful agents (eg, tirofiban, iloprost, and cangrelor). In cases of acute HIT, is CPB with cangrelor plus heparin effective and safe? We describe here the management of an urgent cardiac surgery using cangrelor in combination with UFH in a patient with delayed-onset HIT complicated by multiple arterial and venous thromboses. Cangrelor is an intravenous P2Y12 receptor antagonist that induces powerful platelet inhibition and has pharmacokinetics properties (ie, rapid onset and offset of action) suitable for use in this context.

2. Case

In July 2020, a 48-year-old man presented with a middle cerebral artery stroke for which etiological assessment revealed a patent foramen ovale (PFO) that had been closed percutaneously 8 months later, on day 0. A flush of 6000 units of UFH was administered during the procedure. The patient was discharged with aspirin and clopidogrel. On day 14, the patient was referred to the hospital after he had suffered from right limb pain and headache for 3 days. Further examinations revealed right popliteal artery thrombosis, incidental bilateral pulmonary embolisms, right internal carotid thrombosis, and bilateral centimetric thrombi linked to the PFO occluder. Delayed-onset HIT was suspected after a severe thrombocytopenia at 36 × 10⁹/L was noticed on day 15, whereas platelet count was normal 2 weeks before. Anticoagulation with danaparoid was started immediately on day 15. Qualitative rapid particle gel immunoassay (heparin-platelet factor 4 complexes Partcile Gel ImmunoAssay, Bio-Rad) for anti-PF4/heparin complex antibodies was performed and returned positive on day 16, confirmed on day 19 by a positive enzyme-linked immunosorbent assay (Asserachrom HPIA ImmunoglobulinG, Stago) with an optical density of 1.02 (positivity threshold, 0.21). Functional assay (heparin-induced platelet aggregation using washed platelets) returned positive on day 22. After 4 days of danaparoid treatment, fondaparinux was used for 1 day, and finally, continuous argatroban infusion began on day 19. Because of the absence of platelet count recovery, intravenous immunoglobulins (IVIG) at the dose of 1 g/kg were administered on days 20 and 21.

While on single antiplatelet therapy (aspirin) and effectively anticoagulated with argatroban, the patient experienced a transient ischemic attack on day 21. On day 22, considering a recovery of platelet count at 101 × 10⁹/L and the growing cardiac thrombus showed by transesophageal echocardiography despite adequate anticoagulation, cardiac surgery was scheduled on that day (Figure). Argatroban was stopped 2 hours before leaving for the operating room, where a first activated clotting time (ACT) (Medtronic), a whole blood coagulation test, was measured and served as reference for heparin dosing (normal range, 70-120 seconds; over 450 seconds during CPB). A preprocedural thromboelastography test (citrated TEG 6s, Haemonetics) showed a normal coagulation profile. A 30-μg/kg cangrelor bolus was administered, followed by 4 μg/kg/min continuous infusion, first through a dedicated central venous line and then, as soon as needed, directly through the CPB circuit for the safety of the administration. Five minutes later, a bolus of 25,000 units of UFH (300 UI/kg) was administered, and then, UFH was continuously infused and managed according to ACT measurement every 30 minutes. Tranexamic acid was used with a routine dose of 15 mg/kg bolus followed by a 2 mg/kg/h infusion, and started before surgical incision and stopped after chest closure. Surgery consisted of right and left atrial thrombectomy, PFO occluder removal, and surgical PFO closure. CPB and aortic clamp times lasted 61 and 37 minutes, respectively. At the end of CPB, cangrelor infusion was stopped 5 minutes after the heparin infusion and 10 minutes before the protamine sulfate bolus administration (23,000 units). In order to avoid heparin exposure, 5 minutes after the end of the cangrelor infusion, CPB blood was returned to the cell saver and not to the patient. Thirty minutes after stopping cangrelor, that is, 5 elimination half-lives, assessment of platelet function using thromboelastography (platelet mapping TEG 6s, Haemonetics) still showed 99% Adenosine DiPhosphate inhibition. After CPB removal, a right popliteal thrombectomy was performed without heparin. A second protamine bolus of 2000 units was administered 40 minutes after the initial one, according to ACT rebound. The postoperative course showed neither new thrombosis nor relevant bleeding. Argatroban continuous infusion was resumed 2 hours after the surgery at the dose of 2 μg/kg/min.

Figure.

Cardiopulmonary bypass (CPB) using heparin and cangrelor: timeline with timing of drugs. Red arrows and numbers indicate the time between injections of the different drugs (heparin, cangrelor, and protamine). The level of heparinization during surgery is indicated by the values of the different activated clotting times (ACTs). ’, minutes; G/L, giga per liter; Hb, hemoglobin; s, seconds; UFH, unfractionated heparin.

Chest tubes were removed on the first postoperative day (POD) after having removed 160 mL of blood. No blood products were transfused during or after the surgery. After the patient was discharged from the critical care unit, where he spent 4 days, an asymptomatic pericardial effusion was drained percutaneously on POD 13. As the platelet count remained below 100 × 10⁹/L under effective argatroban regimen, with a nadir of 61 × 10⁹/L on POD 5, corticosteroids and a new course of IVIG (1 g/kg for 2 days) were administered on POD 14, which allowed platelet count normalization (124 × 10⁹/L on POD 19 and 194 × 10⁹/L on POD 22). Warfarin was initiated on POD 26, before stopping argatroban on POD 40. After 7 months of warfarin with aspirin administration, the patient had fully recovered and HIT antibodies disappeared (a enzyme-linked immunosorbent assay optical density of 0.09), allowing for warfarin discontinuation.

3. Discussion

HIT is associated with thrombotic complications and significant morbidity and mortality if not treated [3]. It is caused by ImmunoglobulinG antibodies directed toward heparin–platelet factor 4 complexes, leading to platelet aggregation and thrombocytopenia in the presence of heparin. Although its overall incidence could be up to 5%, it is around 0.5% for patients undergoing vascular surgery and transaortic valve replacement—procedures similar to PFO closure [4,5]. Even if the risk increases with the dose, it has been established that HIT is possible regardless of the type and dose of heparin. This is true even for a single dose, as epitomized by our case where the patient experienced a dramatic HIT after a single bolus of 6000 units of UFH [1,6,7].

Delayed-onset HIT is defined by thrombocytopenia and thrombosis occurring more than 5 days after heparin cessation, and is associated with a more severe presentation than in other HIT subtypes [1,8]. In a case series of 12 patients with delayed-onset HIT, all had 1 or more thromboses and a more severe thrombocytopenia than in classical HIT. Also, serologic assays showed a high titer of antibodies associated with strongly positive heparin-dependent and heparin-independent platelet activation. A highly reactive and elevated titer of HIT antibodies could exhibit heparin-independent platelet activation when reacting with endothelial glycosaminoglycans such as heparan sulfate, thus explaining activation without heparin [2].

HIT is a challenging issue when CPB is required. Indeed, heparin avoidance is the cornerstone of the treatment and heparin is the anticoagulant of choice for CPB owing to its properties: low cost, pharmacokinetics, easy monitoring, reversibility with protamine, and vast use experience. While CPB with heparin is possible for remote or subacute B HIT for its low probability of recurrence [9], the recommendations are different in the case of acute HIT and advocate for the following alternatives: intraoperative anticoagulation with bivalirudin, intraoperative heparin after treatment with plasma exchange, or intraoperative heparin in combination with a potent antiplatelet agent [10]. Although long recommended as a first-line treatment [11], bivalirudin has never undergone wide acceptance for CPB because of its unfavorable properties: no easy monitoring, no antagonist, drug cleavage in stagnant blood (CPB reservoir and cell saver), and accumulation in case of renal failure. Consequently, an approach called “platelet anesthesia” combining heparin and a powerful antiplatelet agent (tirofiban, iloprost, and cangrelor) has emerged. Cangrelor is a P2Y12 receptor antagonist providing powerful platelet inhibition with pharmacokinetics suitable for CPB: a rapid onset of action and a short half-life (2 and 5 minutes, respectively), a clearance that is unaffected by renal or liver functions, and a possible point-of-care monitoring. Therefore, cangrelor has been successfully used in combination with UFH in subacute HIT in recent case series [[12], [13], [14], [15], [16]]. However, to our knowledge, our case is only the second report of CPB with cangrelor performed in such complicated HIT [17].

The management of CPB with cangrelor in this highly prothrombotic situation was successfully performed without any perioperative bleeding or postoperative thrombosis. We believe that several points were decisive for the favorable outcome. Timing of drug infusion and caution at the end of CPB are crucial to avoid any contact between still-heparinized blood from CPB and the patient’s blood. Cell saver usage, a citrated one in our case, seems possible and safe to transfuse CPB circuit blood. Point-of-care tests, such as ACT and platelet mapping TEG, are useful to monitor more precisely the pharmacokinetics of antihemostatic drugs that are unlikely to be predicted in such patients with severe disease. IVIG probably had in our case a positive effect by preventing platelet reactivation, particularly during the time without any anticoagulation in the immediate postoperative period between the end of cangrelor activity and anticoagulation resumption. Therefore, some authors recommend to use them systematically with cangrelor [16]. Finally, as any postoperative platelet transfusion could carry a high thrombotic risk, as evoked by Godier et al. [17], the fact that we did not transfuse any blood product probably had a protective effect on the occurrence of thrombosis.

In conclusion, cangrelor combined with UFH seems a possible alternative in CPB in case of complicated and acute HIT. The management of acute HIT in this context still remains challenging.

The patient gave his informed consent to the writing of this case report.