Platelet Antibody Testing.

A large number of assays have been developed to detect antibodies directed against or associated with platelet membrane antigens. These assays can be categorized into direct (detecting antibody associated with the patient’s platelets) or indirect (detecting antibody in the patient’s serum that binds to control platelets). Indirect tests may detect, in addition to autoantibodies, alloantibodies (e.g., HLA related), particularly in multiparous women and individuals who have received multiple transfusions, and thus give false-positive results. Autoantibodies tend to be associated with the patient’s platelets and are usually present in low concentration in serum. Therefore a direct test is the preferred test for autoimmune thrombocytopenias such as ITP.

A number of direct assays measure immunoglobulin on platelets, regardless of whether the immunoglobulin is specifically or nonspecifically bound to the platelet surface. These assays generally detect platelet-associated IgG (PAIgG), but they can also be engineered to detect IgM or IgA. PAIgG is increased in immune disorders such as ITP and in nonimmune thrombocytopenic disorders such as leukemia and myelodysplastic syndrome. Thus the specificity of these tests is limited. Moreover, the sensitivity of these tests is limited because autoantibodies, such as those responsible for ITP, represent only a small fraction of the total PAIgG. Megakaryocytes nonspecifically take up plasma proteins, including IgG and albumin, and incorporate these proteins into the alpha granules of platelets, especially in disease states associated with increased thrombopoiesis. Therefore increased PAIgG can result from elevated antibody production for any reason.

More recent assays use antigen capture techniques to detect platelet glycoprotein–specific antibodies, such as those that recognize GPIIb/IIIa and GPIa/IX/V, and have high specificity. Unfortunately, the sensitivity is generally too low for these assays to be used for the routine serologic diagnosis of most immune thrombocytopenias. A newer method is based on flow cytometric detection of autoantibodies reacting with specific platelet proteins immobilized on microbeads.

Macrophage and Platelet Fcγ Receptors.

Receptors for the Fc domain of IgG (Fcγ receptors) are found on a variety of cell types, including macrophages and platelets. These receptors have been shown to play a critical role in immune complex–mediated platelet destruction. Fcγ receptors are diverse in structure and function and fall into two major classes: those that activate effector functions, such as phagocytosis or platelet activation, and those that inhibit effector functions.

Activating Fcγ receptors on macrophages include the low-affinity receptors FcγRIIA and FcγRIIIA and the high-affinity receptor FcγRI ( Fig. 34-2 ). As discussed later, cross-linking of the activating receptor FcγRIII promotes phagocytosis of antibody-coated platelets in certain disease states, including ITP. The only Fcγ receptor expressed on platelets is the activating receptor FcγRIIA. Binding of IgG complexes to this receptor results in receptor cross-linking, which in turn initiates platelet aggregation and activation. This process is involved in heparin-induced thrombocytopenia (HIT), the most common drug-induced thrombocytopenia.

Activating Fcγ receptors on macrophages. Fcγ receptors signal phagocytosis via their phosphorylated ITAM (immunoglobulin gene–related tyrosine activation motif) domains. Receptor cross-linking stimulates src family kinases to phosphorylate tyrosine (Y) residues within the ITAM domain of FcγRIIA or within the dimerized γ subunits of FcγRI or FcγRIIIA. The tyrosine kinase syk is then recruited to the phosphorylated ITAM domains and activated. Syc mediates phagocytosis by activating phosphatidylinositol 3-kinase and phospholipase C. PI(4,5)P ₂ , phosphatidylinositol 4,5-bisphosphate.

(Modified from Aderem A, Underhill DM: Mechanisms of phagocytosis in macrophages. Annu Rev Immunol 17:595, 1999.)

Pathogenesis.

ITP is caused by autoantibodies that interact with membrane glycoproteins on the surface of platelets and megakaryocytes. These antibodies result in accelerated platelet destruction and may also impair thrombopoiesis. More than a third of adults with ITP have inadequate platelet production despite increased numbers of megakaryocytes in their bone marrow. Certain antiplatelet antibodies have been shown to inhibit megakaryocytopoiesis or egress of platelets from the marrow space. The GPIIb/IIIa complex is the autoantigen implicated most often as the cause of childhood and adult ITP. Autoantibodies directed against the GPIb/IX/V and GPIa/IIa complexes have also been reported. In rare instances platelet glycolipids have been implicated as autoantigen targets in chronic ITP. Autoantibodies diminish or disappear when platelet levels are restored to normal.

Factors that trigger platelet autoantibody formation in acute ITP are not well understood. Although various mechanisms by which viruses may induce autoimmune disease have been suggested, the link between the immune response initiated by infection (or vaccination) and the subsequent production of platelet autoantibodies has not been established. Proposed mechanisms include adsorption of virus to platelets, deposition of virus-containing immune complexes onto platelet membranes, or exposure of cryptic neoantigens on the platelet surface. There are data to both support and refute the hypothesis that acute ITP is triggered by antiviral antibodies that cross-react with platelet antigens.

Some evidence suggests that T lymphocytes also play a role in the pathogenesis of ITP. Dysregulated helper T cells can promote the expansion of antiplatelet antibody–producing B-cell clones. Cytotoxic T cells from some ITP patients have the capacity to destroy platelets ex vivo. CD3+ T lymphocytes from ITP patients are resistant to glucocorticoid-induced apoptosis, thus suggesting that disturbed apoptosis may contribute to defective clearance of autoreactive T lymphocytes.

Genetic factors have been proposed to influence the development of ITP. Studies have failed to detect linkage of the disease to particular HLA genotypes. Polymorphisms in genes encoding phagocyte Fcγ receptors or proinflammatory cytokines may influence the development of ITP. Immunodeficiency states that have been associated with chronic ITP are discussed later.

Clinical and Laboratory Features.

The typical manifestation of acute ITP is the abrupt onset of bruising and bleeding in an otherwise healthy child. Frequently, there is a history of a viral illness in the weeks preceding the onset of bruising. Seasonal fluctuation in the diagonsis of ITP has been noted, with a peak during spring and a nadir in the autumn. Petechiae and ecchymoses are evident in most patients. Epistaxis and oral mucosal bleeding are seen in fewer than a third of patients. Hematuria, hematochezia, or melena is evident in fewer than 10%. Menorrhagia may be observed in adolescent women with ITP. Although children with ITP may have extremely low platelet counts, bleeding episodes are less severe in these patients than in those with hypoproductive thrombocytopenia. This finding has been attributed to enhanced platelet production and young, large, hemostatically effective circulating platelets. A palpable spleen is present in about 10% of reported cases of childhood ITP; this finding alone should not justify performing imaging studies on the affected child’s abdomen. Malaise, bone pain, and adenopathy are uncommon and should raise concern for another cause, such as acute leukemia.

The peak age at diagnosis is 2 to 6 years. Although acute ITP may be diagnosed in children of any age, adolescents and infants are more likely to have chronic ITP develop in combination with some other immune disorder. In children ITP is seen equally in males and females, whereas in adults ITP is seen predominantly in females by a 2-to-1 ratio ( Table 34-1 ).

Roughly 80% of children have platelet counts below 20,000/µL, often less than 10,000/µL. Leukocyte and red cell counts are usually normal, although anemia may be seen in as many as 15% of these children, especially those with a significant history of epistaxis, hematuria, or gastrointestinal bleeding. A review of the peripheral blood film is mandatory in every child suspected of having ITP; features inconsistent with a diagnosis of ITP should prompt further investigation. Bone marrow aspiration or biopsy reveals normal or increased numbers of megakaryocytes. Increased numbers of eosinophils and their precursors may be noted, although this is not predictive of outcome.

The need for bone marrow examination in children with typical features of acute ITP is a subject of debate. There is a consensus that bone marrow aspiration is not necessary if the initial management is observation alone or administration of intravenous immunoglobulin (IVIG) or anti-D. Controversy exists regarding whether bone marrow aspiration should be performed in all children before glucocorticoid therapy is started to exclude the possibility of acute leukemia (see Box 34-2 ).

Additional tests that may be considered in the initial evaluation and management of suspected ITP include blood group and Coombs test (required if anti-D therapy is contemplated), quantitative immunoglobulin levels (before IVIG therapy), and antinuclear antibody test. Retrospective studies have shown that the antinuclear antibody test is positive in approximately 30% of pediatric patients with otherwise uncomplicated ITP and may predict a subset of patients at risk for the development of further autoimmune symptoms. Tests not considered useful unless specific reasons are identified in the patient’s history and physical examination include screening coagulation tests, liver and renal function tests, serum complement levels, thyroid function tests, and testing for human immunodeficiency virus (HIV) or Helicobacter pylori infection.

Options for the Initial Management of Immune Thrombocytopenic Purpura.

Childhood acute ITP is usually a benign, self-limited disorder that requires minimal or no therapy in the majority of cases. There is no convincing evidence that medical therapy alters the natural history of the disease. Indications for treatment vary among practitioners and are a source of debate. One set of practice guidelines put forth by the American Society of Hematology (ASH) recommends that children with ITP and platelet counts less than 20,000/µL plus significant mucosal membrane bleeding or those with platelet counts less than 10,000/µL and minor purpura be treated with IVIG or a glucocorticoid. However, many pediatric hematologists take exception to this recommendation. In treatment guidelines put forth by other expert panels, a child’s condition, rather than the platelet count, steers management; children with bruising but without mucosal or more severe hemorrhage may be treated by observation alone, irrespective of the platelet count.

At the heart of the treatment debate is the perceived risk for ICH, a rare but potentially life-threatening complication in children with ITP. In a review of 12 case series involving 1293 children, the incidence of ICH was 0.9%. Other surveys suggest that this may be an overestimate and that the true incidence of ICH in children with ITP is between 0.1% and 0.5%. The subgroup of children with ITP who appear to be at greatest risk for ICH are those with platelet counts less than 20,000/µL and additional risk factors such as a history of head trauma, aspirin use, or arteriovenous malformation. However, the platelet count alone has never been shown to predict the severity of bleeding symptoms.

There is no evidence that medical therapy, such as administration of a glucocorticoid or IVIG, reduces the incidence of ICH. Indeed, retrospective studies have shown that ICH may occur despite previous or concomitant therapy with IVIG or a glucocorticoid. The low incidence of ICH in children with ITP precludes a randomized clinical trial to determine whether treatment reduces the risk.

The results of randomized trials for various treatment approaches have been summarized previously. Regardless of whether pharmacologic therapy is used, detailed education and careful follow-up should be provided to the patient and family. The child’s activities should be limited, and aspirin-containing medications should be avoided. Although hospitalization is appropriate for the treatment of a child with a severe bleeding episode, there is no evidence to support routine hospitalization of patients with otherwise uncomplicated ITP.

First-Line Medical Therapies.

Glucocorticoids: Glucocorticoids are presumed to act through several mechanisms, including inhibition of both phagocytosis and antibody synthesis, improved platelet production, and increased microvascular endothelial stability. The latter effect may explain why symptomatic bleeding episodes often subside before the platelet count increases.

In randomized trials prednisone therapy has been shown to induce normalization of the platelet count more promptly than placebo does. Although various doses have been used, the traditional glucocorticoid regimen is prednisone at 2 mg/kg/day (maximum of 60 to 80 mg) for approximately 21 days. A regimen of 4 mg/kg/day for 7 days and then tapered to day 21 is equally effective and appears to cause fewer side effects. Prednisone at a dose of 4 mg/kg/day orally for 4 days with no tapering is also effective. An alternative to these regimens is megadose pulse therapy (methylprednisolone, 30 mg/kg/day intravenously or orally for 3 days). Side effects of glucocorticoid therapy include cushingoid facies, weight gain, fluid retention, acne, hyperglycemia, hypertension, moodiness, pseudotumor cerebri, cataracts, growth retardation, avascular necrosis, and osteoporosis.

IVIG: Imbach and colleagues first reported the successful use of IVIG for the treatment of acute ITP in a small series of children. This was followed by many reports that documented the ability of IVIG to cause a rapid increase in the platelet count in patients with acute and chronic ITP.

IVIG slows the clearance of antibody-coated blood cells from the circulation by inhibiting the phagocytic activity of cells of the reticuloendothelial system. Studies in the early 1980s suggested that this effect is mediated primarily through the Fc portion of IgG. Specially treated preparations of intravenous IgG lacking the Fc portion of the molecule were inferior to unmodified IgG at increasing the platelet count in ITP patients. The proposed mechanism for this effect was Fc receptor blockade. Subsequent studies of FcγRIIB-deficient mice suggested that IVIG elicits its effect by activating these inhibitory receptors and not simply by Fc receptor blockade ( Fig. 34-3 ). However, recent publications cast doubt on the conclusion that IVIG acts through activation of FcγRIIB. Although some data support the notion that anti-idiotypic antibodies may also be present in commercial preparations and contribute to the action of IVIG, the clinical importance of this mechanism remains unproved.

Mechanism of action of intravenous IgG (IVIG) in immune thrombocytopenia. Binding of antibody-coated platelets to activating Fcγ receptors (FcγRIII) on macrophages results in the production of phosphatidylinositol 3,4,5-triphosphate (PI[3,4,5]P ₃ ) via the action of phosphatidylinositol 3′-kinase (PI3K). These activated macrophages phagocytose the platelet-antibody complexes. IVIG induces expression of the inhibitory Fc receptor (FcγRIIB) on macrophages. Stimulation of the inhibitory receptor results in recruitment of SHIP, a 5-phosphatase that degrades PI(3,4,5)P ₃ into phosphatidylinositol 3,4-bisphosphate PI(3,4)P ₂ .

(Modified from Lin SY, Kinet JP: Immunology: giving inhibitory receptors a boost. Science 291:446, 2001.)

The traditional dose of IVIG is 2 g/kg divided over a period of 2 to 5 days. However, several studies suggest that lower doses are effective. In a randomized trial, pediatric patients who received 0.8 g/kg IVIG had at least as rapid a response rate as those who received 2 g/kg over a 2-day period. In more recent randomized trials, favorable results have been seen with even lower doses of IVIG (250 mg/kg/day for 2 days). Among pediatric patients in whom a response is seen, IVIG produces a more rapid increase in platelet count than do traditional doses of a glucocorticoid (2 mg/kg/day of prednisone), as documented in controlled trials.

Several other general conclusions have emerged from studies on IVIG in pediatric patients. First, both splenectomized and nonsplenectomized patients may respond to IVIG. Second, responses to IVIG are generally reproducible. Third, the duration of the response is brief, approximately 2 to 4 weeks.

Enthusiasm for the use of IVIG is offset by cost considerations. IVIG is considerably more expensive than glucocorticoids or anti-D. One study concluded that IVIG may be cost-effective if it reduces hospital stay or if it prevents the need for a splenectomy.

Complications associated with IVIG are common and occur in 15% to 75% of patients. Frequent side effects include flulike symptoms such as headache, nausea, lightheadedness, and fever. Some of these adverse effects can be alleviated by pretreatment with analgesics or antihistamines. Approximately 10% of patients treated with IVIG (2 g/kg) experience aseptic meningitis manifested as a severe, protracted headache and photophobia. These symptoms, which can cause considerable anxiety in patients, parents, and physicians, may spur additional diagnostic studies (e.g., computed tomography) or result in prolonged hospitalization. A rare but serious side effect of IVIG infusion is anaphylaxis, which can occur in patients who are totally deficient in IgA. Most preparations of IVIG contain small amounts of IgA, and IgE antibodies responsible for anaphylaxis form after an initial exposure to IgA in these preparations.

Anti-D: Anti-D is a plasma-derived immunoglobulin prepared from donors selected for a high titer of anti-Rh ₀ (D) antibody. Anti-D elicits a rise in platelet count, along with mild to moderate anemia, in most patients with ITP. A role for anti-D in the treatment of chronic ITP is now well established. Its utility in acute ITP has been studied less extensively, but it is also effective in this clinical setting.

Anti-D can be used to treat only Rh ₀ (D)-positive patients with ITP because Rh ₀ (D)-negative patients do not show a response. The presumed mechanism of action is phagocytic cell blockade. Patients with intact spleens are more likely to respond to anti-D than splenectomized patients are.

The generally recommended dose is 50 to 75 µg/kg as a short intravenous infusion or subcutaneous injection. Uncontrolled studies have demonstrated that anti-D treatment increases the platelet count in approximately 80% of Rh ₀ (D)-positive children. The therapeutic effect of anti-D lasts for 1 to 5 weeks. A controlled trial showed that anti-D therapy was less effective than IVIG or a glucocorticoid in terms of days required to attain a platelet count of 20,000/µL. Anti-D is less expensive than IVIG, with an estimated cost savings of 35% per episode of ITP. The shorter administration time required for anti-D therapy also makes the medication more convenient to use than IVIG.

Adverse reactions of headache, nausea, chills, dizziness, and fever have been reported in 3% of infusions, and these adverse reactions were classified as severe in only a minority of cases. Some degree of hemolysis, the main adverse reaction with anti-D, is inevitable because of binding of anti-D antibody to Rh ₀ (D)-positive erythrocytes. There is laboratory evidence of hemolysis in most patients. The average decline in hemoglobin ranges from 0.5 to 1 g/dL, and most cases of hemolysis do not require medical intervention. In a small subset of patients, more significant hemolysis has been observed, which has tempered enthusiasm for use of the drug. Postmarketing surveillance by the U.S. Food and Drug Administration documented 15 cases of hemoglobinemia or hemoglobinuria after anti-D therapy. Of these patients six required transfusion, eight experienced an onset or exacerbation of renal insufficiency, and two underwent dialysis. The incidence of intravascular hemolysis was estimated to range from 0.1% to 1.5%. Why certain patients experience severe intravascular hemolysis after anti-D therapy is unclear. Subcutaneous delivery of anti-D may be associated with a lower incidence of severe hemolytic reactions.

Relapses or Treatment Failures.

Many patients treated with a glucocorticoid, IVIG, or anti-D will become thrombocytopenic again after a few weeks. These patients are likely to respond again to the therapy used initially. Patients who require therapy and whose thrombocytopenia does not respond to initial treatment with a glucocorticoid, IVIG, or anti-D are usually treated with one of the alternative frontline therapies. No clinical studies document the response rate in these instances. Patients with persistent mild hemorrhage may benefit from low-dose glucocorticoid therapy.

Management of Chronic Immune Thrombocytopenic Purpura.

In approximately 20% of children in whom ITP is diagnosed, thrombocytopenia persists for longer than 6 months. Chronic ITP is more common in female adolescents and adults than in their male counterparts. Chronic ITP sometimes occurs in association with other autoimmune diseases or in conjunction with an underlying condition known to predispose to disorders of autoimmunity, such as lymphoma ( Box 34-3 ).

In up to a third of children with chronic ITP, spontaneous remission will occur months or years later. It is unclear why the condition resolves in these individuals. An estimated 5% of children with chronic ITP have recurrent ITP characterized by intermittent episodes of thrombocytopenia followed by lengthy periods of remission. This is presumed to reflect a chronic compensated state of ITP. During periods of remission, increased platelet production balances the increased rate of platelet destruction. During exacerbations, platelet production by the marrow is suppressed by viral infections or other factors and is unable to offset the rate of destruction.

Evaluation of patients with chronic ITP should include bone marrow aspiration, if not already done; screening tests for immunodeficiency (e.g., quantitative immunoglobulin levels/subsets, specific antibody titers, lymphocyte subsets), systemic lupus erythematosus (SLE), and other autoimmune diseases (antinuclear antibody tests, direct Coombs test, antiphospholipid antibody tests, thyroid function tests); and serologic testing for HIV. Obtaining blood counts and peripheral blood smears from the parents may help exclude familial disorders.

The primary goal of treatment in patients with chronic ITP is to prevent bleeding, not cure the disease. As with acute ITP, therapy decisions should be based more on symptoms than on platelet count. In formulating a treatment plan, the physician must balance the potential side effects of treatment with the risk of bleeding episodes caused by the disease. Because the goal of treatment is a safe platelet count, not necessarily a normal platelet count, observation alone is an appropriate approach for many patients, especially those with minimal symptoms.

Splenectomy.

Splenectomy is thought to be effective because in most patients the spleen is the major site of both platelet destruction and autoantibody production. Platelet survival studies indicate improved platelet life span after splenectomy.

Splenectomy remains the most predictable intervention to achieve long-term remission in children with chronic ITP. Splenectomy should be considered in a pediatric patient with chronic ITP whose risk for hemorrhage precludes the use of observation alone. However, there is debate about where splenectomy belongs in the therapeutic decision tree for chronic ITP (see Box 34-2 ). Because remissions are often delayed in children and the risk of postsplenectomy sepsis is significant, especially in children younger than 5 years, splenectomy is deferred longer in children with ITP than in adults. ASH practice guidelines recommend that splenectomy be considered for children in whom ITP has persisted for at least 1 year and who have bleeding symptoms and a platelet count of less than 10,000/µL (ages 3 to 12) or 10,000 to 30,000/µL (ages 8 to 12 years). Practice guidelines in the United Kingdom are similar.

Laparoscopic splenectomy is preferred over open splenectomy in children with chronic ITP. Advantages of the laparoscopic procedure include less postoperative pain, earlier return of gastrointestinal function, shorter hospitalization, more rapid resumption of normal activities, and smaller incisions. To ensure adequate hemostasis, the platelet count can be raised preoperatively with a glucocorticoid, IVIG, or anti-D. Despite these interventions, many patients will still have thrombocytopenia at the time of splenectomy; however, most will not exhibit excess bleeding. Accordingly, prophylactic platelet transfusions are not warranted. Instead, platelet transfusions should be reserved for patients with intraoperative bleeding. Most experienced surgeons have become comfortable performing laparoscopic splenectomy on patients with platelet counts in the range of 50,000 to 100,000/µL. Overall perioperative mortality is less than 1%.

The majority of children with chronic ITP respond to splenectomy. In a review of 16 case series involving 271 children undergoing elective splenectomy for chronic ITP, a complete remission rate of 72% was reported. The platelet count usually rises immediately after splenectomy and reaches a maximum 1 to 2 weeks postoperatively. If the peak platelet count achieved after splenectomy is greater than 500,000/µL, permanent remission is likely. Certain patient characteristics are associated with a higher probability of response to splenectomy, including a previous response to glucocorticoid therapy. However, there is no single factor or combination of factors that will predict the response to splenectomy in all cases.

Relapse after an immediate response to splenectomy may result from transient viral suppression of thrombopoiesis. If the thrombocytopenia persists, an accessory spleen should be considered. Approximately 40% of patients with persistent thrombocytopenia will have a residual accessory spleen that can be seen with sensitive imaging techniques such as ^99m Tc sulfur colloid scans ( Fig. 34-4 ) or In platelet localization studies. Many of these patients will achieve complete remission after removal of the accessory spleen. The presence of Howell-Jolly bodies on a peripheral blood smear does not rule out an accessory spleen.

Detection of accessory spleens after splenectomy. A sulfur colloid scan shows accessory spleens (arrows) in a teenage girl with chronic immune thrombocytopenic purpura and persistent thrombocytopenia after splenectomy.

(Courtesy Dr. Robert Minkes, Louisiana State University.)

The major risk after splenectomy is fatal sepsis caused by encapsulated organisms. The risk of septicemia in splenectomized patients is estimated to be 1 per 300 to 1000 patient-years. To lessen the risk of future sepsis, patients should receive vaccinations for pneumococcus, Haemophilus influenzae type b, and meningococcus at least 2 weeks before surgery. In addition, penicillin prophylaxis is indicated for splenectomized patients younger than 5 years. The benefit of penicillin prophylaxis in older asplenic children is controversial.

Management of Life-Threatening Hemorrhage in Immune Thrombocytopenic Purpura.

ICH or another life-threatening form of bleeding mandates intervention with multiple therapeutic modalities, including IVIG (1 g/kg), anti-D (50 to 75 µg/kg), high-dose glucocorticoids (e.g., methylprednisolone, 30 mg/kg intravenously), platelet transfusions, or a combination of these interventions. Because IVIG and anti-D act to inhibit phagocytosis by distinct and potentially synergistic mechanisms (see Fig. 34-3 ), these two agents may be used together for instances of severe bleeding. TPO receptor agonists have a limited role in this setting because it takes an average of 2 weeks to see a response. Recombinant factor VIIa has also been used in the management of ICH in patients with thrombocytopenia refractory to platelet-enhancing agents. If a craniotomy is required, consideration should be given to splenic artery embolization or emergency splenectomy, although most hematologists reserve splenectomy for patients in whom ITP does not respond rapidly to therapy with glucocorticoids, IVIG, anti-D, and platelet transfusions. Optimal therapy for ICH also includes supportive care with mechanical ventilation and mannitol.

Alloantigens.

Platelet alloantigens and the platelet membrane glycoproteins on which they reside are listed in Table 34-2 . Most of the antigens were originally named on the basis of the patient’s surname. In many cases duplicate names were assigned to the same alloantigens (e.g., Pl ^A1 = Zw ^a , Pen ^a = Yuk ^b ). The HPA, or human platelet antigen, nomenclature was proposed in the 1990s to systemize platelet alloantigens. Each HPA system represents a biallelic polymorphism caused by a single amino acid substitution (see Table 33-2 ).

In white populations the most frequently implicated alloantigen in NAIT is HPA-1a (Pl ^A1 ), which accounts for 78% of serologically confirmed cases. Incompatibility for the HPA-5b alloantigen is the next most common cause. Approximately 98% of white individuals express the HPA-1a antigen on their platelets; most of the remaining 2% are homozygous for an alternative allele that encodes the HPA-1b (Pl ^A2 ) antigen. The protein products of these two alleles differ only by a single amino acid (Leu33 versus Pro33). The incidence of NAIT is approximately 1 in 2000 births, which is much lower than the predicted incidence based on distribution of the HPA-1a and HPA-1b alleles in the population. Other immune response genes appear to regulate HPA-1a alloantibody formation. Women with the HLA-DR3 antigen have about a twentyfold relative risk of forming HPA-1a alloantibodies. In Asian populations incompatibility of the HPA-4 system is the most common cause of NAIT (80%), followed by incompatibility of the HPA-3 system. HLA class I antigens are expressed on platelets, and maternal sensitization to HLA has been implicated as a cause of NAIT in a few patients. HLA alloantibodies are present in 10% to 30% of pregnant women, especially multiparous women. The relative rarity of NAIT secondary to HLA alloantibodies may result from binding of these antibodies to the placenta or other tissues. Blood group ABH alloantibodies have been found in a small number of patients with NAIT.

Clinical and Laboratory Features.

The degree of thrombocytopenia in NAIT can be severe, with platelet counts of less than 10,000/µL on the first day of life. Common hemorrhagic manifestations include petechiae (90%), hematomas (66%), and gastrointestinal bleeding (30%). The maternal platelet count is normal, an important laboratory value for distinguishing this condition from neonatal thrombocytopenia secondary to maternal ITP (discussed later). Unlike Rh disease of the newborn, NAIT can occur in both the first and subsequent pregnancies. A history of a previously affected infant provides strong supportive evidence for a diagnosis of NAIT.

Patients with NAIT are at increased risk for ICH, both prenatally and postnatally. About 15% of affected neonates have ICH, and 50% of these cases occur antenatally. In utero bleeding can result in hydrocephalus, porencephaly, seizures, or fetal loss. The pronounced bleeding diathesis associated with anti–HPA-1a alloimmune thrombocytopenia may reflect the combination of severe thrombocytopenia and antibody-mediated interference with the function of the GPIIb/IIIa complex on the few remaining platelets.

Treatment.

The diagnosis of NAIT is confirmed by serologic or genotypic testing, including immunophenotyping of maternal, paternal, and occasionally neonatal platelets; maternal or fetal serum also can be examined for the presence of antiplatelet antibody. Several sensitive assays have been developed for platelet antigen typing and detection of alloantibody. Despite the availability of sensitive tests, serologic evidence of alloantibodies is lacking in a least a third of cases identified by clinical criteria; therefore NAIT is a clinical diagnosis. Because delays may be associated with serologic testing, therapy should be initiated as soon as the diagnosis is suspected.

The treatment of choice for severe NAIT (platelet count <30,000/µL or clinically significant bleeding) is transfusion of washed, irradiated maternal platelets ( Fig. 34-5 ). Irradiation prevents transfusion-associated graft-versus-host disease. In the interest of time, normal donor screening procedures may have to be abridged. There are often medical or practical reasons that maternal platelets cannot be collected in a timely manner. Random-donor platelet infusions can be used in the treatment of NAIT if maternal platelets are not immediately available. However, only a modest and short-lived increase in the platelet count may be observed after the infusion of random-donor platelets. IVIG (1 g/kg daily for 2 days) or a glucocorticoid (methylprednisolone, 2 mg/kg/day) can also be used as a temporary measure, although these interventions may be of limited benefit. If the offending alloantigen is known, platelets from an antigen-negative unrelated donor may be used. In white populations the use of HPA-1a– and HPA-5b–negative donors has been advocated because these two antigens are responsible for more than 95% of cases of NAIT. To provide transfusion support for fetuses and neonates with NAIT, the National Blood Service of England has established an accredited platelet donor panel of HPA-1a–negative individuals with no antibodies to HLA, HPA, or red cell antigens. The majority of these donors are HPA-5b negative.

Treatment of neonatal alloimmune thrombocytopenia with washed, irradiated maternal platelets.

(Courtesy Dr. Lori Luchtman-Jones, Washington University.)

Platelets produced by neonates with alloimmune thrombocytopenia may exhibit accelerated clearance for weeks, until antiplatelet antibody is removed from the circulation (see Fig. 34-5 ). Consequently, it is not unusual for additional transfusions of maternal (or an antigen-compatible unrelated donor) platelets to be required after the first few weeks of life.

Management of Future Pregnancies.

The likelihood is high that subsequent infants born to mothers of infants with NAIT will be affected. Moreover, the severity of antenatal and postnatal hemorrhage tends to increase in later pregnancies. Therefore detailed counseling on the risk of recurrence should be provided to the family at the time that thrombocytopenia is diagnosed in the first infant. Although the neonate’s thrombocytopenia may have resolved before completion of the full diagnostic laboratory evaluation, it is important that the diagnosis and presumed antigen incompatibility be confirmed by formal serologic and genotypic testing to guide the management of future pregnancies. Antibody titers begin to decline after delivery; thus the evaluation should ideally be initiated in the postpartum period. Because the likelihood of thrombocytopenia in subsequent infants depends on whether the father is HPA-1a homozygous or HPA-1a/1b heterozygous, performance of DNA-based genotyping is desirable.

Fetuses at risk for ICH should be monitored with serial ultrasound examinations, although fetal ICH may not always be detected by this method. In some centers prenatal diagnosis of the platelet alloantigen genotype has been performed with DNA obtained by chorionic villus sampling or amniocentesis. More commonly, percutaneous umbilical blood sampling is performed at 20 to 22 weeks of gestation to obtain a fetal platelet count and perform platelet antigen typing. However, this procedure may be associated with fetal loss because of exsanguination; thus compatible antigen-negative (e.g., maternal) platelets should be available when percutaneous umbilical blood sampling is performed and should be infused if the fetus is severely thrombocytopenic.

When the diagnosis of alloimmune thrombocytopenia is confirmed in the fetus, two antenatal treatment approaches are available. In most North American centers, the favored approach is weekly maternal infusion of IVIG. In some European centers the preferred approach for severe cases of alloimmune thrombocytopenia is repeated in utero platelet transfusions; administration of IVIG and a glucocorticoid is reserved for mildly affected fetuses.

Neonatal Autoimmune Thrombocytopenia.

This condition can occur in infants of mothers with immune thrombocytopenia as a result of ITP, SLE, or other autoimmune disorders and is caused by placental transfer of maternal autoantibodies. Antigens implicated as causes of neonatal autoimmune thrombocytopenia mirror those seen in ITP. In general, this condition is less serious than NAIT.

This diagnosis should be considered when both the neonate and the mother exhibit signs of thrombocytopenia or in infants of mothers with a previous history of immune thrombocytopenia. Neonatal thrombocytopenia has been observed in infants of mothers rendered asymptomatic for chronic ITP by splenectomy. Maternal ITP must be distinguished from the mild decrease in platelets that frequently accompanies pregnancy at term, presumably because of plasma volume expansion.

The thrombocytopenia associated with maternal immune thrombocytopenia is generally milder than NAIT. The incidence of cord platelet counts lower than 50,000/µL is only about 3%, and the incidence of ICH is only about 1%. Although clinically significant bleeding episodes are rare, newborns with this condition should be monitored closely because platelet counts often decrease in the days after birth. The threshold for medical intervention in neonates with thrombocytopenia differs from that in older children with ITP because of concern about ICH in the newborn. A platelet count of less than 40,000 to 50,000/µL in a neonate is considered by many to be a useful indicator for starting therapy. IVIG (1 g/kg daily for 2 days) elevates the platelet count in most patients. A glucocorticoid (methylprednisolone, 2 mg/kg/day) may be used in addition to or in lieu of IVIG therapy.

Perinatal management of mothers with immune thrombocytopenia, including the mode of delivery of a thrombocytopenic fetus, is controversial. Because the cord platelet count rarely is below 50,000/µL, perinatal ICH is uncommon and not necessarily related to delivery. The degree of maternal thrombocytopenia is a poor predictor of the degree of neonatal thrombocytopenia, although the clinical course of thrombocytopenia in the first sibling does predict that of the next sibling. Otherwise, there are no reliable predictors of severe thrombocytopenia in the newborn. Direct determination of the fetal platelet count by percutaneous umbilical venous sampling can be performed, but this procedure is rarely justified for this condition.

Pathogenesis.

Several mechanisms have been implicated in drug-induced immune thrombocytopenia.

Clinical Features.

Severe thrombocytopenia generally develops within 1 day of ingestion of quinine in a previously sensitized individual. A minimum of 1 week is required for initiation of the primary immune response in individuals taking this drug for the first time. Prodromal symptoms such as fever and chills may be observed. Bleeding may be severe, and its onset is abrupt. Hemorrhagic vesicles are common in the oral mucosa. The presence of a platelet antibody may be verified by a variety of techniques. Readministration of a suspected drug in an attempt to confirm an etiologic relationship is not recommended as a routine diagnostic measure.

Treatment.

Discontinuation of treatment with the offending drug is the main therapy. With quinine the thrombocytopenia usually begins to abate in 1 day if no additional drug is taken, and a normal platelet count is restored within 1 week. In patients with severe drug-induced thrombocytopenia, IVIG or a glucocorticoid has been used to ameliorate bleeding. However, there is little evidence that these medications are effective in this clinical setting. For life-threatening hemorrhage, platelet transfusion may be used, although this may be of limited value because of the persistence of slowly dissociating drug-antibody complexes. Because a sensitized individual is presumed to remain at lifelong risk for additional episodes of drug-induced thrombocytopenia, further use of the offending drug is contraindicated.

Pathogenesis.

The antibodies responsible for HIT recognize the complex of heparin and platelet factor 4 (PF4), an alpha granule constituent. After treatment with heparin, PF4 is released into the circulation and binds with heparin to form a reactive antigen on the platelet surface. The heparin-PF4 complex is immunogenic in some patients. IgG against this complex triggers platelet activation by binding to FcγRIIA receptors on the surface membrane. The activated platelets release substances that stimulate other platelets and result in thrombin generation. PF4 bound to glycosaminoglycans on the surface of endothelial cells may also serve as a target for antibody binding and lead to antibody-mediated injury to the endothelium and a predisposition to thrombosis.

Approximately 50% of patients exposed to heparin during cardiac catheterization and subsequently during cardiac surgery form antibodies against heparin-PF4 complexes, but HITTS develops in only a small percentage. It is unclear why heparin-PF4 complexes, which are composed of two substances normally present in the body, are so immunogenic. Moreover, it is not known why thrombocytopenia or thrombosis develops in some patients and not in others.

Diagnosis.

The diagnosis of HIT is based on the interpretation of clinical findings together with laboratory confirmation of heparin-PF4 antibodies. A scoring system has been proposed to assess the pretest probability of HIT, although this system has not been validated for pediatric patients. The gold standard laboratory test for diagnosing HIT is a functional test, the platelet [ ¹⁴ C]serotonin release assay. Unfortunately, the complexity and availability of this assay limit its usefulness. An enzyme-linked immunosorbent assay (ELISA) for detection of antibodies against heparin-PF4 complexes is available, and results obtained with the serotonin release assay and ELISA are generally in agreement in patients with a clinical diagnosis of HIT. However, ELISA has a more limited specificity and therefore produces false-positive results. A negative ELISA result generally excludes HIT. If the ELISA result is positive, a serotonin release assay may be performed to confirm the diagnosis or to invalidate a false-positive ELISA result.

Treatment.

It is standard practice to discontinue heparin in patients in whom HIT is suspected, but many of these patients will have thrombotic complications over the subsequent days. Therefore ongoing antithrombotic therapy is indicated. Patients with HIT should not be given warfarin as a substitute for heparin during the acute episode because this can cause an abrupt decrease in the level of protein C and put the patient at greater risk for thromboembolic complications such as warfarin-induced limb gangrene. LMWH is not recommended for the treatment of HIT because unfractionated heparin and LMWH preparations display significant platelet aggregation cross-reactivity in vitro and in vivo. Hirudin, a compound originally purified from the leech salivary gland, inactivates thrombin by forming a 1 : 1 complex. A recombinant hirudin, lepirudin, is effective in the treatment of HIT or HITTS. In children lepirudin should be started as a continuous infusion at 0.1 mg/kg/h without a preceding bolus, with a target activated partial thromboplastin time (APTT) of 1.5 to 2.5 times the patient’s baseline value (first assessed 4 hours after infusion). Other agents that are effective in this clinical setting are the synthetic antithrombin argatroban and the synthetic pentasaccharide fondaparinux. Danaparoid, a low-molecular-weight heparinoid consisting of a mixture of heparan sulfate, dermatan sulfate, and small amounts of chondroitin sulfate, has been used for more than a decade to treat patients with HIT. Danaparoid was voluntarily withdrawn from the United States market with the introduction of fondaparinux but is still available in other countries.