Each year in the United States more than 20,000 infants

Recognition and Management of Heparin-Induced Thrombocytopenia in Pediatric Cardiopulmonary Bypass Patients Lynn K. Boshkov, MD, Aileen Kirby, MD, Irving Shen, MD, and Ross M. Ungerleider, MD Departments of Pathology and Medicine, and Division of Pediatric Cardiac Surgery, Doernbecher Children s Hospital, Oregon Health and Science University, Portland, Oregon Repeated exposure to unfractionated heparin is the rule in many congenital heart disease patients. Heparin-induced thrombocytopenia occurs in 1% to 3% of adult cardiac surgeries, and carries high thrombotic morbidity (38% to 81%) and mortality (approximately 28%). Although heparin-induced thrombocytopenia appears to be infrequent in pediatric patients, particularly neonates, our evolving experience suggests postcardiopulmonary bypass congenital heart disease patients may be at increased risk. Diagnostic and therapeutic challenges include frequency of thrombocytopenia after cardiopulmonary bypass, imperfect laboratory testing, lack of established dosing of alternative anticoagulants (such as argatroban and lepirudin), and increased anticoagulantrelated bleeding in young children. (Ann Thorac Surg ) 2006 by The Society of Thoracic Surgeons Each year in the United States more than 20,000 infants with congenital heart disease (CHD) are born. About one third will have life-threatening defects requiring urgent or semiurgent cardiac surgery with cardiopulmonary bypass (CPB) to correct or palliate. In the most severe cases multiple surgeries may be needed over months to years. Use of unfractionated heparin (UFH) is ubiquitous in pediatric intensive care units and is often sustained and repetitive in CHD patients. Our clinical experience, as outlined below, suggests pediatric CHD patients may be at increased risk for heparin-induced thrombocytopenia (HIT)-related morbidity and mortality. Characteristics of Heparin-Induced Thrombocytopenia in Adult Cardiovascular Surgery Patients Pathogenesis: Laboratory Testing Heparin-induced thrombocytopenia is a clinicopathologic, immune-mediated syndrome of platelet activation that complicates 1% to 3% of UFH exposures in adult cardiovascular surgery patients (reviewed in [1, 2]) and carries a thrombotic morbidity of 38% to 81% and a thrombotic mortality of approximately 28%. Thrombosis tends to localize to sites of preexisting disease and may be arterial or venous. Thrombocytopenia typically occurs between days 5 and 10 of heparin exposure and is Presented at the Symposium on Harnessing the Effects of Neonatal Cardiopulmonary Bypass at the Fourth World Congress of Pediatric Cardiology and Cardiac Surgery, Buenos Aires, Argentina, Sept 21, 2005. Address correspondence to Dr Boshkov, Division of Hemostasis and Thrombosis, L471 Oregon Health and Science University, 3181 SW Sam Jackson Park Rd, Portland, OR 97231; e-mail: boshkovl@ohsu.edu. moderate (platelet nadir of 20 to 100 10 9 /L). Onset may be quicker or immediate with heparin reexposure within 100 days. An unexplained fall in platelets of greater than 50% in a patient on heparin even without frank thrombocytopenia should raise concern about HIT. The etiologic antibodies are formed to a neoepitope on platelet factor 4 (PF4) exposed by the binding of UFH. Platelet factor 4 is normally exteriorized from platelet granules to the platelet surface as the result of platelet stimulation. It is also released from endothelial cells in the presence of high heparin concentrations, and is abundant on transfused platelets. Postoperatively 25% to 50% of adult cardiac surgery patients will have antibodies to PF4 detectable by PF4 enzyme-linked immunosorbent assay (ELISA). Of these only 1% to 3% will have HIT syndrome, which is associated more specifically with a subset of platelet-activating immunoglobulin G antibodies found in perhaps 15% to 20% of patients and detected with greatest sensitivity and specificity by functional washed platelet assays (discussed below). These plateletactivating antibodies form immune complexes on the platelet surface, activating platelets and blebbing off procoagulant platelet microparticles with resultant thrombocytopenia and major thrombin generation, which in turn results in self-amplifying platelet activation and frequent thrombosis. Because of the major role played by thrombin in the pathogenesis of HIT, non cross-reactive direct thrombin inhibitors such as argatroban and lepirudin are preferred therapeutic agents. Even if initially recognized as thrombocytopenia uncomplicated by thrombosis, HIT in adults should be treated prophylactically with alternative anticoagulation to prevent thrombosis, which will otherwise occur in approximately 50% of patients. 2006 by The Society of Thoracic Surgeons 0003-4975/06/$32.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2006.02.075

S2356 NEONATAL CPB SYMPOSIUM BOSHKOV ET AL Ann Thorac Surg Table 1. Baseline Characteristics and Major Outcomes of our Pediatric Patients With Heparin-Induced Thrombocytopenia Case No. Age, Sex Diagnosis HIT Thrombosis HIT Death Argatroban Use C1 2 wk, F CHD: 1st operation: stage I Norwood for hypoplastic left heart variant C2 2 wk, F CHD: 1st operation: repair of tetralogy of Fallot/ pulmonary atresia C3 5 mo, M CHD: 2nd operation: bidirectional Glenn and repair anomalous pulmonary return (had stage I Norwood as neonate) C4 1 wk, F CHD: 1st operation: stage I Norwood for hypoplastic left ventricle C5 8 mo, M CHD: 1st operation: repair of tetralogy of Fallot/AV canal defect C6 7 wk, F CHD: 1st operation: repair of tetralogy of Fallot C7 9 mo, M CHD: 2nd operation: Rastelli procedure C8 5 y, M CHD: 2nd operation: Fontan (had thrombosed IVC below renals, femoral artery and vein, innominate vein, and dural sinus in HIT-related thrombosis as neonate undergoing stage I Norwood) C9 3 y, M Abdominal surgery (pullthrough for Hirshsprung s) complicated by intraabdominal sepsis C10 3.5 mo, F CHD: 1st operation: complete AV canal repair Shunt and leg Yes None None No None acute; treated with ASA. Argatroban used for catheterization at 22 months Thrombosed No pulmonary vein on ASA Pulmonary embolism and dialysis line Superior vena cava Yes No None No Infusion, ECMO Shunt No, CPB No (thrombosed fenestration and shunt after 2nd operation on UFH; laboratory testing negative for HIT) Right lobe of liver, dialysis lines No, septic death No Hemodialysis PICC thigh No Infusion therapeutic; transition to fondaparinux ASA aspirin; AV atrioventricular; CHD congenital heart disease; CPB cardiopulmonary bypass; ECMO extracorporeal membrane oxygenation; HIT heparin-induced thrombocytopenia; IVC inferior vena cava; PICC peripherally inserted central catheter; UFH unfractionated heparin. Heparin-Induced Thrombocytopenia in Neonates and Young Children With Congenital Heart Disease The Oregon Health and Science University Experience Heparin-induced thrombocytopenia is infrequently recognized in pediatric patients [3], especially neonates, and, when present, has been said not to carry major thrombotic morbidity and mortality. However, our experience at Oregon Health and Science University in pediatric CHD patients suggests this patient group may be at increased risk for HIT-related morbidity and mortality a possible association also recently noted by others [4]. Between December 2000 and May 2003 we cared for 433 neonatal and pediatric cardiac surgery patients. Of these, 5 patients exhibited clinical HIT as confirmed by functional assay (incidence, 1.2%). We have subsequently identified another 3 cases of pediatric HIT at Oregon Health and Science University, and been involved in the care of 2 additional pediatric HIT patients at another institution. These 10 patients are summarized in Table 1, and details of their treatment and outcomes are discussed in a published case series [5]. All but 1were perioperative CHD patients and 3 were neonates. Maternal HIT testing was performed in 3 neonatal cases of HIT and was negative in all 3. Two patients died of HITrelated thromboses (20% mortality), both neonates, and an additional 5 patients experienced a thrombosis in association with the diagnosis of HIT (overall thrombotic rate, 70%). We speculate the apparent high prevalence of HIT in CHD patients exposed to CPB may be related to their repeated exposure to UFH in the setting of platelet activation. Such patients are exposed to UFH to maintain catheter patency, during cardiac catheterization or angiography, during CPB, and during the use, if required, of supportive ventricular assist devices, extracorporeal membrane oxygenation, and hemodialysis. Because of their critical illness preoperatively they are normally in pediatric intensive care units where they receive UFH in lines and during cardiac catheterization. After CPB (with its associated inflammatory response and concomitant exposure to both high-dose UFH and activated platelets with PF4 on their surface), they return to the pediatric

Ann Thorac Surg NEONATAL CPB SYMPOSIUM BOSHKOV ET AL S2357 intensive care unit where they again postoperatively are exposed to UFH to maintain catheter patency. Two Brief Illustrative Postoperative Heparin-Induced Thrombocytopenia Cases in a 2-Week-Old and a 5- Month-Old With Hypoplastic Left Heart Syndromes Case 1 (C1 on Table 1): A 2-week-old (surgery was delayed to permit treatment of neonatal meningitis) term female infant with a complex congenital heart defect (hypoplastic left ventricle variant) underwent a modified Norwood procedure with insertion of a 3.5-mm Blalock-Taussig shunt. Platelets fell from 228 to 48 10 9 /L on postoperative day (POD) 3. After sternal closure on POD 5 she had ongoing thrombocytopenia, hypoxemia, and hemodynamic instability despite sternal reopening. Angiogram revealed a large filling defect at the shunt base. Reoperation on POD 6 retrieved a 3-mm white thrombus. She received heparin in catheters and flushes, during angiograms, during CPB, and by infusion after the initial operation and reoperation. She arrested 1 hour after completion of her POD 6 surgery and died. Her left leg (the leg where the angiogram access had occurred) was noted to have turned blue and cold at the end of the POD 6 surgery. Functional washed platelet assay drawn just before death was strongly positive, confirming the diagnosis of HIT. We believe HIT was not recognized earlier in the clinical course because of the patient s age and the multiple other causes for thrombocytopenia and thrombosis present. (Technical note: the specific assay used for diagnosis in this case [6] uses the same principles as the North American gold standard functional washed platelet assay, the 14 C-serotonin release assay. It is based on release of adenosine diphosphate from platelet dense granules at therapeutic heparin concentration [0.5 U/mL], and inhibition or abrogation of adenosine diphosphate release by high heparin concentration [100 U/mL] and is done in the presence of the thrombin inhibitor d-phenylalanyl-lprolyl-l-arginine chloromethyl ketone). Case 2 (C3 on Table 1): A 5-month-old male with hypoplastic left heart syndrome status after stage I Norwood procedure as a neonate underwent a bidirectional Glenn shunt and repair of partial anomalous pulmonary venous return. Postoperatively he required extracorporeal membrane oxygenation. His cannulas were removed on POD 3, but platelets remained low. Postoperative day 5 functional HIT assay was moderately positive (unequivocal platelet activation at therapeutic but not at high heparin concentration). All heparin in catheters and flushes was stopped. There was no evidence of thrombosis, and he was treated with aspirin. Platelets rose and he developed a persistent chylothorax. On POD 20 a cardiac catheterization done using argatroban revealed thrombosis of the left pulmonary vein. He received a trial infusion of argatroban in preparation for possible surgical correction of the lesion. Unfortunately on surgical reexploration repair could not be performed. This case illustrates that HIT thrombosis can be occult, aspirin appears inadequate to prevent it, and alternative anticoagulants may be needed both for treatment and for procedures. Considerations in Diagnosing and Treating Heparin-Induced Thrombocytopenia in Pediatric Congenital Heart Disease Patients Confirmatory Assays Heparin-induced thrombocytopenia is sufficiently rare in pediatric patients, and the clinical consequences of making this diagnosis in CHD so profound, that we believe diagnosis requires confirmation by laboratory assay. As the occurrence of HIT in neonates is particularly controversial, we believe definitive diagnosis of HIT in this patient group mandates both a consistent clinical scenario and definite positivity using a functional washed platelet assay. Unfortunately laboratory tests for HIT are imperfect. Heparin-induced thrombocytopenia assays fall into two categories PF4 ELISAs and functional assays. The ELISAs have the virtue of ease of performance in the routine clinical laboratory and have high sensitivity for HIT. Unfortunately they suffer from relatively low specificity in the CPB setting. The incidence of positive PF4 ELISAs in adults undergoing reoperative cardiac surgery is approximately 50% between POD 5 and 10, and this pattern seems also to apply in young children. A published abstract [7] looked at 40 neonates (median age, 5 days) and 64 pediatric patients undergoing repeat CPB (reoperation; median age, 29 months). Patients were tested for HIT antibodies by PF4 ELISA preoperatively and on PODs 5 and 10. All tests were negative preoperatively. On POD 51of40neonates and 10 of 64 patients undergoing reoperation had detectable antibodies; this was 1 of 40 neonates and 31 of 64 patients undergoing reoperation at POD 10. Only 1 case of clinical HIT was seen (in a reoperation patient). Thus the PF4 ELISAs seem less than ideal for diagnosis of HIT after CPB, particularly in reoperative patients. Functional HIT assays can be divided into two broad types. The first is heparin-dependent platelet aggregation, normally done by adding heparin to the patient s platelet-rich plasma. This assay is neither sensitive nor specific and should not be used [8]. The other type of functional assay (the prototype is the 14 C-serotonin release) uses washed normal platelets and looks for a heparin-dependent platelet functional response in the presence of patient plasma (which contains the pathogenic platelet-activating immunoglobulin G). This response is characteristically present at therapeutic heparin concentrations, but not at high heparin concentrations. Functional washed platelet assays are considerably more specific for HIT than the PF4 ELISAs, and also have relatively high sensitivity. Although technically too demanding for most clinical laboratories to do routinely, they are presently the assays of choice for HIT, particularly in the post-cpb context. General Treatment Recommendations Optimum treatment of HIT in neonates and young children is unclear. Once the diagnosis of HIT is established it would appear prudent to follow the same basic measures applicable to treatment of HIT in adults. These include the following:

S2358 NEONATAL CPB SYMPOSIUM BOSHKOV ET AL Ann Thorac Surg 1. Imaging studies to confirm suspected thrombosis and detect occult thrombosis. 2. Discontinuing heparin in all forms (including lowmolecular-weight heparin and catheter flushes). 3. Avoidance of platelet transfusion (bleeding risk is minimal in HIT and transfused platelets have large amounts of surface PF4 and have been anecdotally reported to precipitate thrombosis). 4. Avoidance of warfarin until the platelet count is fully recovered and institution of warfarin only under alternative anticoagulant coverage with the two anticoagulants overlapped for a minimum of 5 days at full anticoagulant doses, as failure to follow these measures may precipitate venous gangrene in HIT patients. If the patient has established thrombosis, alternative anticoagulation (normally with a direct thrombin inhibitor such as argatroban or lepirudin both approved for treatment of HIT in adults) should be instituted. (Guidelines regarding argatroban infusions in neonates and young children are provided below.) In pediatric patients without established thrombosis we believe serious consideration should also be given to prophylactic administration of a direct thrombin inhibitor as follows: 1. Our experience suggests simple heparin cessation with or without acetylsalicylic acid therapy may be inadequate to prevent thrombosis (see case 2 above). 2. Adult experience suggests both lepirudin and argatroban are effective in preventing new thrombosis and extension of existing thrombosis in HIT patients; in contrast, neither agent appears particularly effective in salvaging life or limb in patients with established major thrombosis. Unresolved Issues in Diagnosing and Treating Heparin-Induced Thrombocytopenia in Neonates and Young Children in the Postcardiopulmonary Bypass Setting 1. The clinical course of HIT is not clearly delineated in neonates and young children: our patients with HIT all had clinical features consistent with an intermediate to high pretest probability as set forth in the proposed four T s scoring system for establishing pretest probability of HIT in adults [2]. This system incorporates the degree of thrombocytopenia (maximum points for a platelet nadir of 20 to 100 10 9 /L or a platelet fall 50%), the timing of the platelet fall or other sequelae (maximum points for day 5 to 10 or less than day 1 with recent heparin), the presence of thrombosis (maximum points for proven thrombosis), and the fact that no other cause for thrombocytopenia is evident (maximum points for no evident ther cause). However, reported cases of serologically confirmed HIT in neonates and young children are still too few to establish that HIT in this age group follows adult patterns. 2. Confounding causes of thrombocytopenia and thrombosis are often present: Significant thrombocytopenia ( 50% fall from baseline) is the rule after CPB in neonates and young children, and there are published data to suggest that persistent consumptive coagulopathy may continue to suppress platelet counts postoperatively in some patients along with concomitant derangement of procoagulant anticoagulant balances sufficient to precipitate thrombosis of central venous catheters [9]. Similarly sepsis in neonates and young children may present more subtly than in adults. Thus a high index of suspicion is necessary in diagnosing HIT in this patient population. 3. Sites of localization of HIT thrombosis may differ from the adult setting. Our series suggests that thrombosis will localize to sites of preexisting disease, and particular attention should be paid in CHD patients to sites of recent surgical anastomosis and to sites of central venous catheters. 4. Doses of alternative anticoagulants such as the direct thrombin inhibitors argatroban and lepirudin, both approved for use in adult HIT (reviewed in [10, 11]), are not completely established in pediatrics and the risk benefit of anticoagulation in HIT in neonates and young children is unclear. Although our experience suggests in the setting of CHD the natural history of HIT, even in very young children, is unfavorable, the risk of anticoagulation (and particularly the risk of intracranial hemorrhage) may be greater in neonates and young children than in adults, particularly if the diagnosis of HIT has not been definitively established. Existing studies suggest there is also more bleeding in adults anticoagulated with direct thrombin inhibitors than there is with heparin anticoagulation, possibly because both these direct thrombin inhibitors are small molecules and, unlike heparin, can inactivate both clot-bound and fluid-phase thrombin. Neither drug is reversible, although both have fairly short half-lives (approximately 1 hour). Lepirudin is renally cleared, and clearance can be markedly impaired in patients with renal insufficiency. Argatroban is hepatically cleared, and doses must be reduced in hepatic insufficiency. There are also other anticoagulants that could provide rational therapy for HIT. These include danaparoid (long half-life and unavailable in the United States), the anti-xa pentasaccharide fondaparinux, and another direct thrombin inhibitor, bivalirudin. (For pediatric use of danaparoid, see [12, 13]). Both fondaparinux and bivalirudin have been used in treatment of adult HIT. However, there is far less experience with these agents in the treatment of acute HIT than there is with lepirudin and argatroban, and pediatric experience with these agents is virtually nonexistent, whereas there at least is a small experience with both lepirudin and argatroban in pediatric patients [4, 5, 14 16].

Ann Thorac Surg NEONATAL CPB SYMPOSIUM BOSHKOV ET AL S2359 The Oregon Health and Science University Experience With Argatroban for Treatment of Heparin-Induced Thrombocytopenia in Neonates and Young Children: Status of an Ongoing United States Pediatric Argatroban Trial, Infusion Dosing Recommendations A multicenter US pediatric argatroban trial (SKF 105043/ 013, GlaxoSmithKline) was initiated in 2004 and is currently in progress. As a result of study interim data analysis, US Food and Drug Administration approval for use of argatroban in pediatrics is currently being sought (January 2006). This study is enrolling pediatric patients (including those with suspected HIT), who in the opinion of the treating physician need alternative anticoagulation to UFH. Pharmacokinetic data and outcome data are being gathered, and, in the case of suspected HIT, baseline serotonin release assay is being performed at a reference laboratory. Argatroban is being used in this study both for infusion and for procedures (CPB, extracorporeal membrane oxygenation, cardiac catheterization, and so forth). As a prelude to establishment of a formal study protocol (which was followed by Oregon Health and Science University s formal Institutional Review Board approved participation in this study in 2004), as well as during our participation in this study, we have treated a number of our neonatal and pediatric HIT patients at Oregon Health and Science University with argatroban after documentation of parental informed consent. We have also reported our evolving experience with this agent [5, 15, 16]. We have now successfully used argatroban anticoagulation in 10 neonatal and pediatric patients in a variety of clinical contexts. These have included prophylaxis and treatment of thrombosis (n 7), and for procedures: cardiac catheterization (n 6), extracorporeal membrane oxygenation (n 2), CPB (n 2) and hemodialysis (n 1). Activated clotting times, partial thromboplastin times, and plasma argatroban generally changed in concert during times of clinical stability. Significant bleeding only occurred with overanticoagulation (ie, activated clotting time 999 during CPB) and was successfully treated with recombinant factor VIIa. (Optimal treatment of bleeding on argatroban is unclear.) Dose requirements for argatroban varied widely even in the same patient, however, and our sense was that dosing requirements paralleled clinical status (lower doses if sicker ). Interim pharmacokinetic data analysis in the first 11 subjects (age range, 6 months to 16 years; 10 patients deemed critically ill) from the pediatric argatroban trial (letter to study investigators July 11, 2005) suggests the following dosing guidelines for continuous infusion therapy: ages 6 to 16: 1.0 to 2.0 g kg 1 min 1 ; ages 6 months to 6 years: 0.5 to 1.0 g kg 1 min 1 ; age younger than 6 months: 0.5 to 1.0 g kg 1 min 1. These doses would be expected to achieve average partial thromboplastin times of 1.7 to 2.7 times baseline. Dosages should be adjusted as clinically indicated to achieve a target PTT of 1.5 to 3 times the baseline value. Transition to warfarin can be a problem as argatroban prolongs the baseline prothrombin international normalized ratio. Our approach to this issue is to monitor warfarin anticoagulation using chromogenic factor X assays, with factor X levels less than 0.3 considered therapeutic. Neonates and young children on argatroban requiring transition to an alternative agent for longer term anticoagulation in whom warfarin is impracticable can be transitioned to fondaparinux, which is given every day subcutaneously, and we have done this in 1 patient [17]. However, adult experience with fondaparinux in HIT is limited, and there is almost no pediatric experience. Again, fondaparinux is not approved for pediatric uses, therapeutic dose ranges are not established in either adults or children, and monitoring requires generation of a special fondaparinux curve (low-molecular-weight heparin curves cannot be used). References 1. Warkentin TE, Greinacher A. Heparin-induced thrombocytopenia and cardiac surgery. Ann Thorac Surg 2003;76:2121 31. 2. Warkentin TE. Review: heparin-induced thrombocytopenia: pathogenesis and management. Br J Haematol 2003;121:535 55. 3. Severin T, Sutor AH. Heparin-induced thrombocytopenia in pediatrics. Semin Thromb Hemost 2001;27:293 9. 4. Klenner AF, Lubenow N, Raschke R, Greinacher A. Heparininduced thrombocytopenia in children: 12 new cases and review of the literature. Thomb Haemost 2004;91:719 24. 5. Alsoufi B, Boshkov LK, Kirby A, et al. Heparin-induced thrombocytopenia (HIT) in pediatric cardiac surgery: an emerging cause of morbidity and mortality. Semin Thorac Cardiovasc Surg Pediatr Cardiac Surg Annu 2004;7:155 71. 6. Stewart MW, Etches WS, Boshkov LK, Gordon PA. Heparininduced thrombocytopenia: an improved method of detection based on lumi-aggregometry. Br J Haematol 1995:91: 173 7. 7. Mullen MP, Thomas KC, McGowan FX, et al. Heparin-induced thrombocytopenia in pediatric patients undergoing cardiopulmonary bypass. Circulation 2000;102:468 (Abstract 12). 8. Warkentin TE, Heddle NM. Laboratory diagnosis of immune heparin-induced thrombocytopenia. Curr Hematol Rep 2003;2:148 57. 9. Petaja J, Peltola K, Sairanen H, et al. Fibrinolysis, antithrombin III and protein C in neonates during cardiac operations. J Thorac Cardiovasc Surg 1996:112:665 71. 10. Lewis BE, Wallis DE. Leya F, et al. Argatroban-915 investigators. Argatroban anticoagulation in patients with heparininduced thrombocytopenia. Arch Int Med 2003;163:1849 56. 11. Lubenow N, Greinacher A. Hirudin in heparin-induced thrombocytopenia. Semin Thromb Hemost 2002;28:431 8. 12. Saxon BR, Black MD, Edgell D, et al. Pediatric heparininduced thrombocytopenia: management with danaparoid (Orgaran). Ann Thorac Surg 1999;68:1076 8. 13. Zohrer B, Zenz W, Rettenbacher A, et al. Danaparoid sodium (Orgaran) in four children with heparin-induced thrombocytopenia type II. Acta Paediatr 2002;90:765 71. 14. Deitcher SR, Toupoulos AP, Bartholomew JR, et al. Lepirudin anticoagulation for heparin-induced thrombocytopenia. J Pediatr 2002;140:264 6. 15. Mejak B, Giacomuzzi C, Heller E, et al. Argatroban usage for ECMO on a past-cardiac patient with heparin-induced thrombocytopenia. J Extra Corpor Technol 2004;36:178 81. 16. Mejak B, Giacomuzzi C, Shen I, et al. Cardiopulmonary bypass using argatroban as an anticoagulant for a 6.0 kg pediatric patient. J Extra Corpor Technol 2005;37:303 5. 17. Boshkov LK, Kirby A, Heuschkel M. Pharmacokinetics of fondaparinux by anti-xa levels and clinical response to anticoagulation in a 4 month old congenital cardiac patient with heparin-induced thrombocytopenia (HIT) and established venous thrombosis transitioned from argatroban to fondaparinux. Blood 2004;104: Abstract 4072.