Heparin-Induced Thrombocytopenia with Associated Thrombosis in Children after the Fontan Operation

Abstract

Heparin-induced thrombocytopenia is a widely recognized clinical disorder. The spectrum of disease ranges from clinically insignificant to severe thrombosis (heparin-induced thrombocytopenia with associated thrombosis). Overall, thrombosis occurs in approximately 33% of adults diagnosed with heparin-induced thrombocytopenia and has been associated with high morbidity and mortality rates. Diagnostic testing for this disorder is not standard in children with thrombocytopenia who are receiving heparin, despite the fact that children with congenital heart disease may be exposed to heparin frequently. There are few reported cases of heparin-induced thrombocytopenia with associated thrombosis in children; herein, we describe the cases of 2 children who developed this disorder after undergoing a Fontan operation. (Tex Heart Inst J 2003;30:58–61)

Heparin-induced thrombocytopenia (HIT) is a widely recognized clinical disorder that is observed in about 2% of adults who receive heparin. ^1–3 This condition has a morbidity rate of 61% and a mortality rate of 22%. ⁴ The spectrum of HIT ranges from clinically insignificant to severe thrombosis—heparin-induced thrombocytopenia with associated thrombosis (HITT). Thrombosis, usually venous, can produce devastating complications, including necrosis of the extremities, stroke, myocardial infarction, and pulmonary embolism. Thrombosis occurs in approximately 33% of adults diagnosed with HIT. Early recognition of HIT, followed by termination of heparin administration, decreases the morbidity and mortality rates.

Although heparin-induced thrombocytopenia has been recognized in adults for more than 20 years, there are few reported cases of HIT or HITT in children. ^5–9 Diagnostic testing for HIT in children with thrombocytopenia who are receiving heparin is not standard, yet children with congenital heart disease may be exposed to heparin frequently. We retrospectively reviewed records of all children who had cardiac surgery at our institution for suspected HIT from 1995 through 2000 and now describe the cases of 2 children who developed HITT after a Fontan operation.

Case Reports

Patient 1

In November 1998, a 2-year-old boy with double-inlet left ventricle, aortic stenosis, and hypoplastic aortic arch had a Fontan operation. He had previously undergone Norwood Stage I and bidirectional cavo–pulmonary anastomosis, with 4 associated heparin exposures. During the Fontan operation, he received 4,800 U of heparin. During the first 12 postoperative days, the platelet count ranged from 85–123 × 10⁹/L. The count decreased precipitously to 25 × 10⁹/L on postoperative day 14. The results of his disseminated intravascular coagulation (DIC) panel were normal, as were his antithrombin III function and protein C activity. He was given platelet transfusions, and the platelet count increased to 85 × 10⁹/L. The next day, the platelet count was again low (52 × 10⁹/L). Liver tests showed elevated function. An ultrasound of the abdomen and pelvis revealed a thrombus in the inferior vena cava. He was started on another heparin infusion. Coagulation tests were performed, including a DIC panel. Fibrinogen and fibrin split product levels were normal. On postoperative day 17, the cardiac output was diminished, and a test for HIT (heparin-induced platelet aggregation [HIPA] test) was ordered. Later that afternoon, the heparin was discontinued and angiography was performed. The angiogram showed diffuse thrombosis in the right iliac vein. The thrombus extended through the entire length of the inferior vena cava to the level of the Fontan connection. That evening, the HIPA test was reported as positive. Heparin was removed from all intravenous fluids, and no other anticoagulants were administered. The patient died in the operating room while attempts were made to remove the thrombus.

Patient 2

A 3-year-old boy with tricuspid and pulmonary valve atresia had a Fontan operation at our institution in September 1999. His most recent catheterization had been 2 weeks before the Fontan operation. During the Fontan, he had been given 4,000 U of heparin, which was his 4th exposure to heparin. Postoperatively, he had persistent DIC, with fibrin split product levels ranging from 10 to 80 μg/mL; a prothrombin time (PT) of 13–32 sec; and an activated partial thromboplastin time (aPTT) of 24–64 sec. The platelet count decreased to 51 × 10⁹/L. Coagulation factor VII and VIII levels were normal. The patient received multiple blood products, including platelets.

During this time, the patient's left lower extremity was noted to be cool and cyanotic. A catheter had been placed in the left femoral artery during surgery and subsequently removed. A vascular ultrasound demonstrated occlusion of the left femoral artery. On postoperative day 1, he was returned to the operating room for thrombectomy and primary anastomosis of the left femoral artery. An intraoperative angiogram obtained at the end of the procedure showed the left femoral artery to be patent. The patient had received 1,300 U of heparin during the thrombectomy, and heparin infusion was continued for several days. His hospital course was further complicated by multisystem organ dysfunction, including hemodynamic instability, liver dysfunction, and renal failure requiring peritoneal dialysis. Despite the thrombectomy, the patient's left foot remained cool and cyanotic and developed focal areas of dry gangrene. He was given continuing platelet transfusions due to the low platelet count, which was attributed to DIC. On postoperative day 6, the platelet count remained at 40 × 10⁹/L. Because HIT was suspected, the heparin administration was discontinued. A serotonin release assay (SRA) was ordered and reported as negative, but the results of the ELISA test measuring for heparin–platelet factor 4 complex (HPF4) antibody and the HIPA test were positive. After the diagnosis of HITT was made, heparin was removed from all intravenous fluids, and no other anticoagulants were administered. The platelet count remained low for the next week, ranging from 42–92 × 10⁹/L. On postoperative day 13, the platelet count rose to 131 × 10⁹/L. Two days later, the patient underwent a below-the-knee amputation. He was discharged from the hospital on postoperative day 39. Since then, he has done well and was last evaluated in August of 2002.

Discussion

Heparin-induced thrombocytopenia is divided into 2 types. Type I is an early-onset, mild decline in platelet count (<100 × 10⁹/L) and is reversible even with the continuation of heparin. This condition is thought to be caused by the direct platelet-aggregating effect of heparin. In contrast, HIT Type II is an immune-mediated reaction that typically occurs 2 to 5 days after the initial heparin exposure. After re-exposure to heparin, however, patients may develop HIT Type II sooner. The patients described in this report had HIT Type II.

The pathogenesis of HIT Type II is incompletely understood. The antigen in HIT Type II may be a complex of heparin and platelet factor 4 (PF4) (Fig. 1). ^10,11 Platelet factor 4 is a heparin-binding protein contained in the alpha granules of platelets and is released in large amounts in the presence of trauma, sepsis, surgical injury, and other states of platelet activation. The HIT antibodies may bind to endothelial cells, causing local cellular injury. ¹² An inflammatory state involving leukocyte activation also seems to be involved. ¹³ Immunoglobulin G antibodies bound to the heparin–PF4 complex cause platelet activation through platelet Fc-gamma-IIa receptor (FcγRIIa) binding. ¹⁴ Activated platelets release platelet microparticles that promote a hypercoagulable state and thrombin generation. ^13,14 Platelet factor 4 can also attach to endothelial glycosaminoglycans, heparan, and other polysulfated saccharides that may resemble the conformation of the heparin molecule, causing endothelial cell activation. ^10,11,15

Fig. 1 A proposed mechanism of heparin-induced thrombocytopenia depicting the immune complex of heparin with platelet factor 4.

HIT = heparin-induced thrombocytopenia; IgG = immunoglobulin G; FcγRIIa = Fc-gamma-IIa receptor; PF4 = platelet factor 4

The diagnosis of HIT is made on the basis of clinical features. Heparin-induced thrombocytopenia should be suspected in any patient who is receiving heparin and has a platelet count <100 × 10⁹/L or a 50% decrease from the baseline level. There is no platelet count that excludes the diagnosis of HIT, and there is no standardized definition of the syndrome. Typically, HIT occurs 5 or more days after heparin re-exposure, but it can occur at any time. Thrombosis has been reported within 30 minutes after heparin re-exposure. ¹⁶ Thrombosis can occur before the decline in platelet count or even when the platelet count is recovering and rising.

Laboratory assays are useful to help confirm a clinical diagnosis of HIT. A number of assays have been developed for the laboratory diagnosis of HIT, including platelet aggregometry (HIPA), the SRA, and the ELISA for measuring antibody to the heparin–PF4 complex. The HIPA test is a platelet activation test in which the patient's serum is mixed with donor platelets in the presence of heparin. Aggregation of the donor platelets indicates the presence of antibodies to the heparin–PF4 complex. The SRA is another platelet activation test in which a patient's serum is mixed with washed donor platelets that have been incubated previously with ¹⁴C-serotonin. Platelet activation, measured as the release of ¹⁴C-serotonin, indicates the presence of heparin antibody. The SRA is considered to have the best sensitivity and specificity (reported as high as 80% and 99%, respectively) for the detection of heparin-induced antibodies. ^17,18 However, the SRA test is technically demanding and cannot be performed by most laboratories. The HIPA test has a high specificity, but it generally has a low sensitivity for HIT (∼40%). ¹⁸ Recently, ELISA assays to measure heparin-induced antiplatelet antibody titers have been developed in an attempt to provide a simpler and more sensitive method. These assays have shown a sensitivity of 60% to 70% and a specificity of 80% to 90%. However, some patients with clinical manifestations of HIT do not show an elevated antibody titer. ¹⁸ Conversely, positive titers have been observed in patients who have no clinical symptoms of HIT. ^10,19 Clearly, there is currently no assay for HIT with optimum sensitivity and specificity. ^18,20

Reports of HITT in children are rare. We report 2 cases of HITT seen at our institution within a 5-year span. In the 1st patient, the HIPA test was positive. The patient had the clinical signs of HITT (extensive inferior vena cava thrombosis associated with thrombocytopenia during heparin administration). In the 2nd patient, both the HIPA and the HPF4 ELISA test results were positive, and the SRA test result was negative. Because the SRA has a false negative rate of approximately 15%, ¹⁷ the positive HPF4 ELISA result is indicative of the greater sensitivity of the ELISA test. The 2nd patient's thrombocytopenia may also have resulted from postoperative DIC. Even as the coagulopathy resolved, however, the platelet count remained severely low. The platelet count did not begin to recover until 7 days after the heparin was discontinued. Another unusual aspect of this case is the rapidity of the thrombosis (only a few hours) after heparin re-exposure. The patient's previous heparin exposure had been just 2 weeks before the Fontan procedure. Antibodies may have developed after the catheterization, which would explain the rapid thrombosis after heparin re-exposure.

The hemodynamics of the Fontan circuit may contribute to HITT. Fontan patients with prolonged hospital courses may require continuous heparin infusions. The sluggish flow in the cavo–pulmonary connection, coupled with elevated venous filling pressure, may create an environment conducive to thrombosis. Most of the children who require surgery for congenital heart disease do not have lengthy hospital stays and seldom require prolonged exposure to heparin, which may explain why HITT in children has been reported infrequently.

The site of trauma to the endothelium is a likely nidus for clot formation. A number of studies ^21,22 have suggested that thrombotic events are more likely to occur at sites of pre-existing injury, including sites of plaque formation. Singer and colleagues ²¹ treated 4 patients with intravascular catheters who developed HITT after cardiopulmonary bypass and required limb amputation. In another series, ²² 19 of 25 patients who developed HITT had had a catheter in the affected extremity. In summary, HIT with or without thrombosis does occur in children, and pediatricians should be prepared to discontinue heparin infusion and provide appropriate treatment in children who develop this condition.