HEPARIN-INDUCED thrombocytopenia (HIT) is an infrequent complication of heparin therapy. 1A unique problem arises in patients with HIT who need anticoagulation, especially if urgent cardiac surgery is planned. We report a case of HIT in an infant in which Argatroban (GlaxoSmithKline, Middlesex, UK) was used as antico-agulant during cardiopulmonary bypass (CPB) and the unsuccessful use of recombinant activated factor VII (rFVIIa) to reverse the anticoagulant effects postoperatively.
A 9-month-old male patient (weight, 5.5 kg) had undergone complete repair of a transposition of the great arteries, a ventricular septal defect, and a pulmonary stenosis at 7.5 months of age by means of CPB and heparin. He was readmitted to the hospital 2 weeks after discharge with congestive cardiac failure and mediastinitis. An echocardiogram showed large vegetations in the right ventricle in the vicinity of the homograft, and positive blood culture results were obtained. A diagnosis of endocarditis was made, and a Broviac catheter was inserted. His platelet count decreased from 60,000 cells/mm−3to 15,000 cells/mm−3within 2 days after receiving heparin flushes in the Broviac catheter. An HIT assay, consisting of a heparin-dependent platelet activation assay, as well as an enzyme-linked immunosorbent assay (GTI, Brookfield, WI), yielded positive results. The diagnosis of HIT was made, and all heparin was immediately stopped. An Argatroban infusion was started at 7.5 μg · kg−1· min−1after a bolus dose of 200 μg/kg. The infusion rate was adjusted to keep the activated partial thromboplastin time at 1.5–2 times the normal value. Despite antibiotic treatment, the vegetations in the right ventricle remained unchanged. It was decided to reoperate to remove the Gortex (W. L. Gore & Associates, Newark, DE) hood from the homograft and clear out any vegetation. One week after stopping heparin administration, the platelet count had recovered to 212,000 cells/mm−3.
The Argatroban infusion was stopped 4 h preoperatively to allow placement of lines and to minimize blood loss before CPB. Baseline activated clotting time (ACT) was 160 s. A total of 750 μg/kg Argatroban was administered in three divided doses over 50 min to increase the ACT to more than 400 s. Immediately on initiation of bypass, the ACT was longer than 999 s, and the Argatroban infusion, which had been started after the initial bolus, was stopped. The ACT remained at more than 999 s for the next 2 h. The total bypass time was 1 h. The patient was weaned from bypass without difficulty. At this point, hemostasis was difficult to obtain, and a total of 250 ml (45 ml/kg) fresh frozen plasma and 2 units platelets were administered over 1 h. This reduced the ACT to 394 s, but hemostasis remained inadequate. After consultation with a hematologist, two doses (90 μg/kg each) of rFVIIa (Niastase; Novo Nordisk, Bagsvärd, Denmark) were administered 20 min apart. There was no significant improvement in hemostasis clinically or in the ACT after administration of rFVIIa. (The ACT decreased to 329 s after the second dose.) An additional 20 ml/kg fresh frozen plasma was administered over the next hour. Serial ACTs, repeated every 30 min after the rFVIIa, were 329, 315, 285, and 208 s. Satisfactory hemostasis was eventually achieved 2 h after administering the rFVIIa, and the patient was transported to the intensive care unit in a stable condition.
A total of 500 ml (90 ml/kg) packed erythrocytes, 2 units platelets, and 375 ml (65 ml/kg) fresh frozen plasma were administered intraoperatively. Bleeding from the chest tube totaled 60 ml in the first 12 h postoperatively. There were no thrombi noticed in the bypass circuit postoperatively. The ACT reached a control value of 160 s at 10 h after the initial doses of Argatroban. The Argatroban infusion was restarted after 12 h and transitioned to Coumadin (Du Pont Pharma, Mississauga, Ontario, Canada) after 5 days. The patient was transferred out of the intensive care unit after 1 week. The results of an HIT assay repeated after 1 month remained positive.
This is the first reported case in which Argatroban has been used as an anticoagulant during cardiac surgery in an infant with HIT, a recognized complication of heparin exposure that is rarely reported in children. 1,2The cornerstone of treating a patient with HIT is the discontinuation of all heparin. Heparin cessation alone may not be effective in preventing thromboembolic events, 3and alternative anticoagulants should be considered because thromboembolism often occurs when the platelet count rebounds. An Argatroban infusion was started in our patient because he was considered to be in the acute phase of HIT and at high risk of a thromboembolic event. However, discontinuation of heparin poses another challenging problem in patients with HIT who need subsequent cardiac surgery. Alternative anticoagulants include danaparoid sodium, lepirudin, ancrod, Argatroban, 4and recently also bivalirudin. 5Routine coagulation tests cannot be used to monitor the anticoagulant effect of danaparoid, but the ecarin clotting time does reliably monitor the anticoagulant effect of lepirudin and bivalirudin. The pharmacokinetic profile of bivalirudin in particular makes it an attractive alternative in a situation in which heparin is contraindicated. 5
Argatroban is a synthetic small-molecule direct thrombin inhibitor derived from l-arginine. 6It inhibits free and clot-bound thrombin, the interaction with thrombin being reversible. As with danaparoid and other direct thrombin inhibitors, Argatroban has no specific antidote, but it nevertheless has potential advantages. It has a relatively short half-life (40–50 min). It does not require antithrombin III as cofactor, and it undergoes hepatobiliary excretion, making its use safe in renal failure. In addition, its activity can be measured with routine coagulation tests, such as activated partial thromboplastin time and ACT. 7
Argatroban has been used in vascular surgery, including left heart assist, 8in adult and pediatric extracorporeal membrane oxygenation, 9,10and also in off-pump coronary artery bypass surgery. 11Argatroban has also been used successfully as an anticoagulant in CPB studies in a dog model. 12In addition, there are case reports of its successful use as the anticoagulant in CPB in adult patients with antithrombin III deficiency 13and with HIT. 14Furukawa et al. 13recommend an ACT of more than 400 s when using Argatroban for cardiac surgery. They used a bolus of 0.1 mg/kg, followed by a continuous infusion of 5–10 μg · kg−1· min−1. In our case, we needed to administer a total of 750 μg/kg to increase the ACT to more than 400 s. Interestingly however, immediately on initiating CPB, the ACT in our patient was more than 999 s. This probably reflected abrupt hemodilution on the bypass circuit; the commonly seen increase in ACT on initiating bypass 15is perhaps exacerbated when Argatroban is used.
Although rFVIIa is not indicated for the reversal of Argatroban, there are successful reports of the use of rFVIIa to treat coagulopathy and bleeding after CPB in infants and children. 16,17There was no significant improvement clinically in our patient after administration of rFVIIa or in the ACT values taken in the next 2 h. The reduction in ACT over the 2-h period probably largely represented metabolism of the drug as well as additional fresh frozen plasma administration. There is no documented clinical experience with the reversal of direct thrombin inhibitors with rFVIIa, but our observation in this case corresponds to laboratory data in which rFVIIa at very high doses failed to fully restore hemostasis or produce a significant reduction in blood loss in anesthetized rats treated with a direct thrombin inhibitor, melagatran. 18Among the possibilities are that rFVIIa will not work or that the dose was inadequate given what seems to have been an excessive anticoagulation effect of Argatroban. The only clinically proven efficacy of rFVIIa is for hemophilia, but when diagnosis and treatment of specific coagulation defects fail to correct coagulopathy, it may be reasonable to use rFVIIa as rescue therapy, as was the case in our patient. 19Clinical trials are needed to address the potential reversal of direct thrombin inhibitors with rFVIIa.
Our patient experienced major bleeding during surgery and needed infusion of large volumes of fresh frozen plasma, packed erythrocytes, and platelets. This case illustrates the potential risks of using new anticoagulants for cardiac surgery. Although it was not an option in our patient because the HIT assay results remained positive after 1 month, repeated use of unfractionated heparin remains the preferred approach to anticoagulation in patients with HIT who need repeat cardiac surgery, provided that HIT antibodies are no longer detectable. 20,21
We conclude that Argatroban is a potential alternative anticoagulant for use during CPB, although formal pharmacokinetic studies are needed in infants and small children to establish a safe and optimal dosing regimen. Careful titration of Argatroban to a therapeutic ACT, activated partial thromboplastin time, or possibly ecarin clotting time is advised, bearing in mind a possible unpredictable ACT response due to hemodilution on initiating CPB in infants and small children. rFVIIa also did not reverse the anticoagulant effect of Argatroban in our patient.