Chapter 120: Hereditary Qualitative Platelet Disorders

Abnormalities of platelet function manifest themselves primarily as excessive hemorrhage at mucocutaneous sites, with ecchymoses, petechiae, epistaxis, gingival hemorrhage, and menorrhagia being most common. Mild platelet function abnormalities will not cause spontaneous bleeding but may cause (excessive) hemorrhage after trauma or medical interventions. Both quantitative and qualitative platelet abnormalities can produce these symptoms, so it is necessary to exclude thrombocytopenia (Chap. 117) by performing a platelet count. Although no longer performed widely, a prolonged bleeding time in a patient with a normal platelet count is suggestive of a qualitative platelet abnormality. Some patients may have abnormalities in both platelet number and function. Chapter 121 discusses acquired qualitative platelet abnormalities and this chapter discusses hereditary qualitative platelet abnormalities.

Following injury to the blood vessel, platelets adhere to exposed subendothelium by a process that involves, among other events, the interaction of a plasma protein, von Willebrand factor (VWF), and a specific glycoprotein complex on the platelet surface, the glycoprotein (GP) Ib–IX–V complex. Adhesion is followed by recruitment of additional platelets that form clumps (aggregation), which involves binding of fibrinogen to specific platelet surface receptors, a complex comprised of integrin α_IIbβ₃ (GPIIb-IIIa). Platelet activation is required for fibrinogen binding; resting platelets do not bind fibrinogen. Activated platelets release the contents of their granules (secretion), including adenosine diphosphate (ADP) and serotonin from the dense granules, which causes the recruitment of additional platelets. Moreover, platelets play a major role in coagulation mechanisms; several key enzymatic reactions occur on the platelet membrane phospholipid surface. A number of physiologic agonists interact with platelet surface receptors to induce responses, including a change in platelet shape from discoid to spherical (shape change), aggregation, secretion, and thromboxane A₂ (TXA₂) production. The binding of agonists to their platelet receptors initiates numerous intracellular events (Chap. 112) including the production or release of several messenger molecules. One pathway leads to the hydrolysis of phosphoinositide (PI) by phospholipase C leading to the formation of diacylglycerol and inositol 1,4,5-triphosphate [IP₃]). These and other mediators induce or modulate the various platelet responses of Ca²⁺ mobilization, protein phosphorylation, aggregation, secretion, and thromboxane production. Numerous other mechanisms, such as activation of tyrosine kinases and phosphatases, are also triggered by platelet activation (Chap. 112). Inherited or acquired defects in the above and other platelet mechanisms may lead to impaired platelet function and a bleeding diathesis.

The hereditary qualitative platelet disorders can be classified according to the major locus of the defect (Table 120–1 and Fig. 120–1). Glanzmann thrombasthenia (GT) is caused by abnormalities in either integrin α_IIb (GPIIb) or β₃ (GPIIIa), resulting in loss or dysfunction of the integrin α_IIbβ₃ receptor. This results in a profound defect in platelet aggregation and secondary defects in platelet adhesion, secretion, and coagulant activity. Loss of the platelet GPIb–IX–V complex because of abnormalities in GPIbα, GPIbβ, or GPIX results in the Bernard-Soulier syndrome (BSS), which is characterized by giant platelets and thrombocytopenia. The major defect is in platelet adhesion owing to a decrease in platelet interactions with VWF. A gain-of-function defect in GPIbα (platelet-type [pseudo] von Willebrand disease [VWD]) can also produce a hemorrhagic disorder via depletion of high-molecular-weight VWF multimers. Defects in secretion of granule contents because of deficiencies in the granules or in the mechanisms that mediate secretion results in impaired platelet function. Inherited defects in agonist receptors or proteins or mediators involved in signal transduction or thromboxane synthesis may also produce hemorrhagic symptoms. Abnormalities of platelet coagulant activity, that is, the ability of platelets to facilitate thrombin generation (Chap. 113), and in cytoskeletal-linking proteins, can also lead to a hemorrhagic diathesis. Lastly, it is becoming clear that some patients may have abnormalities multiple aspects of platelet function and number related to a mutation in a hematopoietic transcription factor that regulates gene expression in megakaryocytes and platelets.

Disorders of platelet function are characterized by highly variable mucocutaneous bleeding manifestations and excessive hemorrhage following surgical procedures or trauma. These include ecchymoses, petechiae, epistaxis, gingival bleeding, and menorrhagia. Spontaneous hemarthrosis are distinctly rare, distinguishing them from the hemophilias, and deep hematomas and spontaneous central nervous system bleeding are highly unusual in these patients. Postpartum hemorrhage and postsurgical bleeding may be severe in some patients. In general, most patients with platelet function defects have mild to moderate bleeding manifestations, but may be severe in some entities, such as the GT and the BSS. Individual patients with the same functional defect or even the same genetic defect may also vary in the intensity of bleeding manifestations, suggesting one or more disease-modifying genes exist. Moreover, the severity of bleeding symptoms may vary over the lifetime of the same individual, implying that factors in addition to the platelet defect may be contributing to the bleeding risk. Sometimes, a mild platelet function defect becomes clinically manifest when the patient uses medication interfering with primary hemostasis (e.g., nonsteroidal antiinflammatory drugs [NSAIDs]).

Although normal in many of the inherited platelet function defects, the platelet count may be decreased in some entities, such as the BSS and the gray platelet syndrome (GPS) and in association with mutations in hematopoietic transcription factors. Most patients with platelet function defects, but not all, have a prolonged bleeding time, a test no longer available in most centers because of its inherent inaccuracies. In vitro platelet aggregation and secretion studies provide evidence for the dysfunction but are not generally predictive of the severity of clinical manifestations. In some patients, such as those with abnormal platelet coagulant activities, these studies may be normal. Some patients with platelet dysfunction are initially detected through abnormalities on testing with the platelet function analyzer (PFA-100).

Platelet disorders are characterized by alterations in platelet number or function or both. A general approach is shown in Fig. 120–1. A reduced platelet count can occur as an isolated platelet disorder (inherited or acquired) or with evidence of a concomitant defect in platelet function. Platelet size and examination of the blood film may provide insights into the underlying mechanism or cause. Platelet size provides clues in some entities (Chap. 117).^1,2 Decreased platelet size is characteristic of the Wiskott-Aldrich syndrome. In the Paris-Trousseau/Jacobsen syndrome, thrombocytopenia is associated with giant α granules in a subpopulation of platelets in association with mutations in the transcription factor FLI1. Transcription factor RUNX1 mutations are associated with familial thrombocytopenia, abnormal platelet function, and predisposition to leukemia. Large platelets that lack the purple granules on the peripheral smear are observed in the GPS (α-storage pool disease). The diagnosis is obtained with biochemical analysis of α-granule contents. Patients with platelet-type (pseudo-) VWD and type 2b VWD have moderate thrombocytopenia and large platelets. Studies of GPIb function and biochemistry establish the diagnosis. Patients who are hemizygous for GPIbβ because of deletion of 22q11.2, those with mutations in transcription factor GATA-1 or β₁ tubulin (R318W), and some patients who are heterozygous for defects in GPIb/IX have variable thrombocytopenia and large platelets. Mutations that activate α_IIbβ₃ are also associated with large platelets and thrombocytopenia. The platelets in BSS are truly giant; the diagnosis is confirmed with biochemical and functional analyses of the GPIb–IX–V complex.

A variety of methods have been developed to assess platelet function and new instrumentation continues to be developed.^3,4,5,6,7,8 Platelet aggregation studies performed using platelet-rich plasma can loosely separate patients into those with defects in the primary wave of platelet aggregation (dependent on either fibrinogen, VWF, their respective receptors, or agonist receptors for collagen, ADP, TXA₂) and those with defects in the secondary wave of aggregation. Enhanced ristocetin-induced platelet aggregation at low doses of ristocetin is characteristic of patients with platelet-type VWD (who have a defect in the GPIb receptor) and patients with type 2b VWD (who have gain-of-function defect in VWF) (Chap. 126). These two diseases differ in the binding of the patient’s VWF to normal platelets, or the ability of purified VWF, cryoprecipitate or asialo-VWF to aggregate patient platelets; the diagnosis of platelet-type VWD or its confirmation requires genetic analysis of GPIb.

Neither ristocetin nor the snake venom botrocetin induces platelet aggregation if the plasma lacks functional VWF, as in VWD (Chap. 126), or if the platelets lack functional GPIb–IX complexes, as in BSS. The defect in VWD, but not BSS, can be corrected by adding normal plasma or purified VWF. Direct analysis of VWF and the platelet GPIb–IX complex are used to confirm the diagnosis.

Patients whose plasma lacks fibrinogen (afibrinogenemia; Chap. 125) or whose platelets cannot bind fibrinogen because of abnormal α_IIbβ₃ receptors (GT) or inability to activate integrin α_IIbβ₃ (leukocyte adhesion deficiency [LAD]-3) as the result of a kindlin-3 abnormality will have no primary wave of platelet aggregation in response to all physiologic agonists, including ADP, epinephrine, collagen, TXA₂, and thrombin. Simple coagulation tests (prothrombin time, partial thromboplastin time, and measurement of plasma fibrinogen) and analysis of platelet integrin α_IIbβ₃ receptors, and kindlin-3 can differentiate between these two groups. Isolated defects in the primary response to collagen have been observed in patients with abnormalities in platelet integrin α₂β₁ (GPIa/IIa) or GPVI. Platelet glycoprotein analysis can separate these from each other. Because antibodies to GPVI can result in receptor depletion from circulating platelets, a search for anti-GPVI should be undertaken in patients with reduced platelet GPVI. Defects in ADP, epinephrine, or TXA₂ receptors will result in decreased platelet aggregation in response to the specific agonist. However, patients with isolated ADP and TXA₂ receptor abnormalities have impaired aggregation in response to other agonists as well because of the feedback potentiation provided by ADP and TXA₂.

A very heterogeneous group of platelet defects can result in a decreased secondary wave of platelet aggregation in response to ADP and epinephrine and diminished responses to low doses of collagen and thrombin. They can be separated into granule defects and defects in platelet secretion or the release reaction. Operationally, these two groups can be separated on the basis of their release of dense granule contents in response to high doses of thrombin. High-dose thrombin activation can overcome most or all of the release reaction (secretion) abnormalities, so platelets from patients with these disorders will release normal amounts of granule contents; in contrast, patients with reduced granule contents have abnormal granule release responses even when using high doses of thrombin. α-Granule contents and dense-body contents can be measured immunologically and biochemically; electron microscopy can establish granule defects. Specific analysis of the genes or proteins implicated in the different granule biogenesis abnormalities (Wiskott-Aldrich syndrome [WASP], Hermansky-Pudlak syndrome [HPS1–9], Chédiak-Higashi syndrome [LYST], Paris-Trousseau/Jacobson syndrome [FLI1], and inherited platelet disorder with predisposition to leukemia [RUNX1]) can establish the diagnosis. The Quebec platelet disorder is characterized by increased urokinase plasminogen activator (u-PA) in α-granules and degradation of several α-granule proteins. The diagnosis can be established by immunoblot analysis or analysis of u-PA activity and confirmed by genetic analysis of u-PA. Secretion abnormalities arise as a result of defects in mechanisms that regulate the secretion of granule contents, and may include abnormalities at various levels, including surface receptors, guanosine triphosphate (GTP)-binding proteins that link surface receptors to intracellular enzymes, phospholipase C (PLC) activation, and protein phosphorylation (protein kinase C [PKC]-θ). They also arise from defects in TXA₂ synthesis caused by deficiencies of phospholipase A₂ (PLA₂), cyclooxygenase, or thromboxane synthase. Specific studies on signal transduction mechanisms, PI metabolism, Ca²⁺ mobilization, protein phosphorylation, and thromboxane production are needed to define these defects. Because transcription factor abnormalities can affect the expression of multiple proteins involved in megakaryopoiesis and platelet function, they can simultaneously produce alterations in platelet count, structure, and function.

In the disorder of platelet coagulant activity (Scott syndrome) platelet aggregation studies are normal and the serum prothrombin time is the preferred screening assay. Other tests of platelet coagulant activity, microvesiculation, and phospholipid transfer are used to establish the diagnosis.

The introduction of microfluidic multiparameter assessments of platelet function,^7,8 coupled with advances in proteomics, RNA expression profiling, and DNA sequencing are shifting the diagnosis of platelet function disorders from a target gene approach to one in which unbiased comprehensive functional and genetic analyses are employed. These methods have identified mutations in RUNX1 and FLI-1⁹ in patients with platelet function disorders, in NBEAL2 in GPS,^10,11,12 in TMEM16 in the Scott syndrome,¹³ and in RBM8A in thrombocytopenia with absent radii (TAR) syndrome.^14,15 Many additional genetic alterations, including ones that affect multiple systems, are likely to be identified in the near future as these techniques are employed more broadly.

INTEGRIN α_IIBβ₃ (GLYCOPROTEIN IIB/IIIA; CD41/CD61)–GLANZMANN THROMBASTHENIA

Definition and History

GT is an inherited hemorrhagic disorder characterized by a severe reduction in, or absence of, platelet aggregation in response to multiple physiologic agonists as a result of qualitative or quantitative abnormalities of platelet integrin α_IIb (GPIIb; CD41) and/or integrin β₃ (GPIIIa; CD61).¹⁶

In 1918, Eduard Glanzmann, a Swiss pediatrician, described a group of patients with hemorrhagic symptoms and a defect in platelet function, namely the ability to retract clots (“weak” platelets or thrombasthenia).¹⁷ Subsequent studies demonstrated that thrombasthenic patients have prolonged bleeding times and that their platelets fail to aggregate in response to physiologic agonists^18,19,20,21 and have markedly reduced^18,20,21,22 platelet fibrinogen. In the mid-1970s, Nurden and Caen²³ and Phillips and colleagues²⁴ discovered that thrombasthenic platelets are deficient in both integrin α_IIb and β₃. Later studies demonstrated that integrin α_IIb and β₃ form a calcium-dependent complex in the platelet membrane that functions as a receptor for fibrinogen and other adhesive glycoproteins.^25,26,27,28 Cloning and sequencing of the complementary DNAs for integrin α_IIb²⁹ and β₃³⁰ identified them as separate protein subunits that are members of the integrin receptor superfamily³¹ and permitted the molecular biological characterization of patients with the disorder (see database of Glanzmann patients http://med.mssm.edu/glanzmanndb).

Etiology and Pathogenesis

GT is a rare disorder characterized by autosomal recessive inheritance with a worldwide distribution. In regions where consanguineous matings are common, groups of patients with the disorder have been identified, and in several populations founder mutations have been identified by analyzing polymorphisms in the DNA surrounding the affected mutation. These include 42 patients from South India; 39 patients from the Iraqi-Jewish population in Israel; 46 Arab patients from Israel, Jordan, and Saudi Arabia; 30 patients from Italy; a smaller number of patients from three Gypsy families; and 43 patients from Pakistan.^{22,32,33,34,35,36,37,38,39,40} Perhaps the highest frequency of a GT mutation is found in the Iraqi-Jewish population where the most common mutation causing GT was found in six of 700 individuals.³⁹

The platelet integrin α_IIbβ₃ receptor is required for platelet aggregation induced by all physiologic agonists (ADP, epinephrine, thrombin, collagen, TXA₂) (Chap. 112).⁴¹ Consequently, abnormalities in the receptor result in a failure of platelet plug formation at sites of vascular injury, and excessive bleeding.

The integrin α_IIbβ₃ receptor is also responsible for the uptake of fibrinogen from plasma into α granules,⁴² hence, patients with GT have markedly reduced platelet fibrinogen.^{18,20,21,43,44} Clot retraction requires platelets with intact integrin α_IIbβ₃ receptors,^45,46 and is, therefore, usually abnormal in GT.¹⁸

Defects in either integrin α_IIb or β₃ result in the same functional defect because both subunits are required for receptor function (Chap. 112). Biosynthetic studies indicate that integrin α_IIb and β₃ form a complex soon after protein synthesis in the rough endoplasmic reticulum^47,48,49; subsequent posttranslational processing⁵⁰ and transport to the platelet membrane require that the complex be intact (Fig. 120–2).^51,52 Complex formation protects each of the glycoproteins from proteolytic digestion,^47,48,49,50 so if either integrin α_IIb or β₃ is absent or unable to form a normal complex, the other subunit will be rapidly degraded, most likely through a proteasomal mechanism. Thus, a deficiency in either glycoprotein produces a deficiency in both. Because complex formation and vesicular transport are also required for proteolytic processing of pro-α_IIb into its constituent α_IIbα and α_IIbβ subunits,⁵⁰ if these processes do not occur normally, the very small amount of residual integrin α_IIb will be pro-α_IIb, not mature α_IIb.⁵³ Pro-α_IIb has been reported to bind to the membrane-bound endoplasmic reticulum chaperone calnexin, providing a potential mechanism for assessing whether the protein has undergone proper folding (calnexin cycle) and perhaps explaining how the receptor adopts a bent configuration.^54,55

Integrin β₃ (GPIIIa) can also combine with the integrin α_V (CD51) subunit to form the integrin α_Vβ₃ “vitronectin” receptor^30,56,57 (see Fig. 120–2; Chap. 112). This receptor can bind many of the same adhesive glycoproteins as integrin α_IIbβ₃, although there are some differences in ligand preference and binding sequences.^{57,58,59,60,61} A small number of integrin α_Vβ₃ receptors are present on platelets (50 to 100 per platelet)^60,62,63; osteoclasts, endothelial cells, macrophages, vascular smooth muscle, and uterine cells, among others, also have integrin α_Vβ₃ receptors.^64,65 In general, GT patients with defects in integrin β₃ also are deficient in integrin α_Vβ₃, whereas patients with defects in integrin α_IIb have either normal or increased numbers of platelet integrin α_Vβ₃ receptors.^{60,63,64,66,67,68} One exception to this rule is a patient with a defect in β₃ (H280P) that interferes with integrin α_IIbβ₃ biogenesis to a much greater extent than integrin α_Vβ₃ biogenesis.⁶⁹ At present, there is no evidence that patients who lack integrin α_Vβ₃ receptors in addition to lacking integrin α_IIbβ₃ receptors have a more-severe hemorrhagic diathesis or suffer from any other abnormalities, perhaps because alternative receptors containing integrin α_V associated with other β subunits can substitute for integrin α_Vβ₃.⁶³ Upregulation of integrin α₂β₁ on osteoclasts of Iraqi-Jewish patients with GT has been reported as a potential compensatory mechanism to explain the lack of bone changes despite the deficiency in osteoclast integrin α_Vβ₃.⁷⁰

The molecular biologic abnormalities in more than 100 patients with GT have been identified and they are listed in an internet database that is updated continuously⁷¹ (http://med.mssm.edu/glanzmanndb). Figure 120–3 contains information on mutations of particular interest. Of note, many of the patients with identified mutations are compound heterozygotes rather than homozygotes, indicating that a sizable number of silent carriers are present in the population. Where consanguinity is common, the disorder is more likely to be caused by a homozygous mutation arising in a founder, but even under these circumstances, more than one mutation may be present. Thus, in the Iraqi-Jewish population, in which consanguinity has been present from 586 bce to the present, two separate mutations have been identified in more than one family.³⁹ Most of the missense mutations result in decreased expression of integrin α_IIbβ₃ on the surface of platelets. This probably reflects the stringent structural requirements for proper folding and complex formation.

Mutations in Integrin α_IIbβ₃ Within the Metal Ion-Dependent Adhesion Site of Integrin β₃ and the Interface with the Integrin α_IIb β-Propeller A metal coordination site or MIDAS domain, which is highly conserved in six integrin receptor α-chain subunits and required for ligand-binding, is also present in the β-A (or I-like) domain of the integrin β₃ subunit.⁷² Mutagenesis and molecular modeling experiments suggested that a highly conserved DxSxS amino acid sequence⁷³ motif plus additional coordinating residues are brought together in the three-dimensional structure of the β₃ subunit to form a cation-binding sphere of the MIDAS domain,⁷⁴ and this was confirmed by the crystal structures of integrin α_Vβ₃ and later integrin α_IIbβ₃ (see Chap. 112, Fig. 112–111, and Fig. 120–3).^75,76 Thus, the β₃ MIDAS is composed of Asp¹¹⁹, Ser¹²¹, Ser¹²³, Glu²²⁰, and Asp²⁵¹. A region originally termed the ligand-associated metal binding site (LIMBS) in integrin α_Vβ₃,⁷⁷ but now termed the synergy metal binding site (SyMBS) in integrin α_IIbβ₃,⁷⁸ binds a Ca²⁺ ion and is required for binding of ligands to the MIDAS. It is composed of atoms from D158, N215, D217, P219, and E220. Integrin β₃ residues 214 and 216 are in close proximity with both the SyMBS residues and the interface with the α_IIb subunit. Adjacent to the MIDAS domain is a metal ion site termed the ADMIDAS (adjacent to metal ion-dependent adhesion site), in which calcium is coordinated by Ser¹²³, Asp¹²⁶, Asp¹²⁷, and Met³³⁵ in unliganded integrin α_Vβ₃ and integrin α_IIbβ₃, but Asp²⁵¹ substitutes for Met³³⁵ in the ligand-bound structures of both integrin α_Vβ₃ and integrin α_IIbβ₃. The crystal structures also demonstrated that peptide ligands containing the Arg-Gly-Asp (RGD) cell adhesion sequence interact with integrin α_IIbβ₃ and integrin α_Vβ₃ in part by coordination of the metal ion in the MIDAS by the aspartic acid in the RGD peptide.^77,79 The low-molecular-weight drugs eptifibatide and tirofiban, which block ligand binding to the α_IIb subunit, have negatively charged regions that also interact with the MIDAS cation.⁷⁶ The fibrinogen γ-chain C-terminal dodecapeptide mediates binding to integrin α_IIbβ₃ and a crystal structure of the complex demonstrates that an aspartic acid carboxyl oxygen coordinates the MIDAS cation whereas the carboxy terminal valine interacts with the nearby cation in the ADMIDAS.^76,79 A number of mutations in patients with GT have been identified within the cation-binding sphere of the MIDAS domain (see Fig. 120–3, and Chap. 112, Fig. 112–11). Two mutations D119Y (Cam variant)⁸⁰ and D119N (patient NR),⁸¹ are located within the conserved DxSxS amino acid motif and produce severe abnormalities of ligand binding to integrin α_IIbβ₃, but do not affect its surface expression. Mutations at residues R214 and R216 result in abnormal integrin α_IIbβ₃ receptors that cannot bind ligand and are very sensitive to dissociation by calcium chelation, perhaps because they are at the integrin α_IIb–β₃ interface.^32,82,83,84 Disrupting the SyMBS with a D217V mutation also leads to GT despite the expression of normal amounts of the integrin protein.⁸⁵ Further support for the importance of the MIDAS domain, SyMBS, and adjacent residues comes from studies in which the mutations D119N, R214W, D217N, E220Q, and E220K were introduced into Chinese hamster ovary CHO cells in vitro and shown to result in functional abnormalities.⁸⁶

The interface between the α_IIb β-propeller and the β₃ subunit also involves, in part, the interaction between β₃ R261, contained in a four-amino-acid 3₁₀ helix, with a number of hydrophobic residues in the α_IIb β-propeller arranged as inner and outer rings, making up a cage.⁷⁵ A β₃ subunit L262Y mutation, adjacent to R261 results in disruption of the helix and an unstable integrin α_IIbβ₃ complex that is expressed on the surface of platelets but is unable to bind fibrinogen.⁸⁷ The platelets of the patient with this mutation were able to bind fibrin and support clot retraction, suggesting different requirements for fibrinogen and fibrin binding.

Mutations in Integrin α_IIbβ₃ Within the α_IIb β–Propeller Sequence Based on their homology to another integrin α subunit, the aminoterminal 450 amino acids of integrin α_IIb and the homologous region in integrin α_V, which contain the minimal ligand-binding sequence,⁸⁸ were predicted to fold into seven repeat (blade) β-propellers, containing four cation-binding sites,⁸⁹ and this prediction was confirmed by the crystal structures of both α_V and α_IIb integrin subunits.^75,76 The upper surface of the propeller interacts with the β₃ subunit β-A (or I-like) domain to form the head of the integrin α_IIbβ₃ complex, which is the site of ligand binding. Each repeat (blade) contains four β strands that are connected by loops. The four calcium-binding sites in the α_IIb subunit, which are in β-hairpin structures, are located in loops on the undersurface of the propeller. Ligand binding in integrin α_IIb has been localized to a hydrophobic (F160, Y190, F231) and negatively charged (D224) pocket that lies adjacent to the MIDAS domain in the β₃ subunit, and is composed of contributions from the loops that link blade 2 to blade 3 (residues 144 to 171), β strand 2 to β strand 3 in blade 3 (residues 186 to 193), and blade 3 to blade 4 (residues 223 to 236). Integrin α_IIb contains a unique “cap” subdomain made up of four insertions in β propeller loops (residues 72 to 88, 111 to 126, 147 to 166, 200 to 217) that also plays a role in ligand binding.⁷⁶

GT missense mutations located within the integrin α_IIb β-propeller (see Fig. 120–3) primarily affect transport of the integrin α_IIbβ₃ complex to the cell surface,^{68,90,91,92,93} but several missense mutations and an insertion result in functionally defective receptors. Thus, Y143H affects soluble ligand binding but not adhesion or clot retraction,⁹⁴ and P145A, which has been identified in several kindreds,^32,95 and P145L, prevent ligand binding. A two-amino-acid insertion at residues 161 and 162, as well as a T176I missense mutation, also affect ligand binding.^96,97,98 A L183P mutation, which is near to, but not in the loop containing Y190, affects both receptor expression and function.⁹⁹

Mutations in Integrin α_IIbβ₃ That Affect Receptor Activation Several β₃ subunit missense mutations (C560R, V193M) result in the receptor adopting a high-affinity ligand binding state, which is paradoxical as it results in a bleeding diathesis.^100,101 A β₃ subunit S527F mutation in the third I-EGF domain was also associated with a constitutively active receptor, presumably because it prevents the receptor from assuming a bent, inactive conformation.¹⁰² The cytoplasmic domain of the β₃ subunit plays a functional role in integrin activation and the regulation of ligand binding.^103,104 Two GT mutations have been identified in this region. One is an R724X nonsense mutation (patient RM)¹⁰⁵ that results in the deletion of the carboxyterminal 39 residues of integrin β₃ and the other is a β₃ subunit S752P missense mutation (patient P or Paris I).^106,107,108 This latter patient is unusual in that he had a generally mild history of excessive hemorrhage, but he did have a prolonged bleeding time and his platelets did not aggregate in response to ADP. These mutations do not severely affect surface expression of platelet integrin α_IIbβ₃ complexes, but both mutant receptors are unresponsive to agonist stimulation. Mammalian cell expression studies of these mutations show normal adhesion to immobilized fibrinogen, but abnormal cell spreading. Cells expressing the S752P mutant receptors have reduced focal adhesion plaque formation and cells expressing the R724X mutant receptors have undetectable tyrosine phosphorylation of focal adhesion kinase, pp125^FAK. These mutations provide evidence for the role of the β₃ subunit cytoplasmic tail in inside-out signaling (i.e., platelet signals that lead to integrin α_IIbβ₃ adopting a high-affinity ligand-binding conformation) and outside-in signaling (i.e., signaling to the interior of the platelet as a result of integrin α_IIbβ₃ binding ligand; see Chap. 112, Figs. 112–3, 112–4, and 112–12).

Variants of Integrin α_IIbβ₃ in the Population The application of missense variant whole-exome and whole-genome sequencing to large numbers of individuals has provided valuable information on the frequency of missense variants in the general population and the frequency of the genetic alterations leading to GT. For the most part, the frequency of a variant in a population is a reflection of its impact on reproductive fitness and when it entered the population, with lower frequencies for variants that entered the population more recently. Thus, variants with minor allele frequencies (MAFs) of approximately 0.5 percent or less probably entered the population fewer than 2500 years ago, when the recent explosive growth in human populations began.¹⁰⁹ Data from a study^109A involving approximately 33,000 alleles from approximately 16,500 people demonstrated the presence of 114 novel missense variant affecting approximately 10 percent of the integrin α_IIb amino acids and approximately 9 percent of the β₃ subunit amino acids. Thus, approximately 1.1 percent of the population studied carried at least one missense variant. None of the known GT mutations was observed in any of the alleles studied, indicating that they have MAFs of less than 0.01 percent and thus entered the population very recently. In fact, studies of two GT populations with high intragroup marriages, Palestinian Arabs and French Manouche gypsies, estimated that the GT mutations entered the population approximately 300 to 600 and approximately 300 to 400 years ago, respectively.^110,111 Several novel missense variants identified in this study affected one of the amino acids mutated in patients with GT, and in two cases these variants were shown to profoundly affect expression of the receptor. In one case, a missense variant reduced expression by approximately 50 percent but did not alter function. A series of prediction tools indicated that somewhere between 45 and 74 percent of the 114 novel missense mutations may be deleterious. Thus, perhaps approximately 0.6 percent of individuals in the general population is a silent carrier for a GT variant that profoundly affects structure and/or function. In addition, some of the rare individuals in the healthy population with levels of integrin α_IIbβ₃ receptor expression intermediate between those of obligate GT carriers and normal individuals may reflect heterozygosity for “hypomorphic” variants that partially affect receptor expression, but not function.¹¹²

Clinical Features

Table 120–2 summarizes the clinical manifestations of 177 patients with GT obtained from two reviews.^22,33 Menorrhagia occurs in nearly all female patients. Purpura can be present immediately after birth but often is not dramatic. Petechiae of the face and subconjunctival hemorrhage associated with crying may be the first symptoms in neonates and babies. Spontaneous hemarthroses and central nervous system bleeding are very rare. The hemorrhagic diathesis in patients with GT is notable for its variability and the lack of correlation between the biochemical platelet abnormalities and clinical severity.²² Even within groups of patients such as the Iraqi Jews, most of whom share the same genetic α_IIb or β₃ subunit abnormalities, there is a wide spectrum of clinical severity.^33,39 Moreover, the severity of bleeding symptoms can vary significantly during the lifetime of individual patients. GT does not appear to protect against the development of atherosclerosis as judged by the carotid artery intima-to-media ratio.¹¹³ Carriers of GT are usually asymptomatic or only mildly symptomatic and generally have normal results in platelet function tests.^{22,33,112,114,115}

Laboratory Features

Table 120–3 provides characteristic laboratory findings in GT. Patients have normal platelet counts and morphology, prolonged bleeding times, decreased or absent clot retraction, and abnormal platelet aggregation responses to physiologic stimuli. The initial slope of high-dose ristocetin-induced aggregation is normal (or near normal), reflecting the normal plasma VWF and the normal platelet GPIb/IX content; at lower doses of ristocetin, however, where GPIb/IX–mediated activation of integrin α_IIbβ₃ (Chap. 112) normally contributes to the aggregation response, patients have decreased second wave aggregation.¹¹⁶ GT platelets undergo normal shape change in response to ADP and thrombin, demonstrating their ability to undergo metabolic and cytoskeletal changes in response to these agents. Similarly, high doses of thrombin and collagen produce normal release of dense body and α-granule contents^18,20,117; the decreased secretion observed with lower doses of these agents reflect the lack of augmentation of the release reaction normally produced by platelet aggregation.^{18,116,118,119,120}

Platelets in whole blood or platelet-rich plasma adhere to glass because fibrinogen first becomes deposited on the glass and the platelets then adhere to the immobilized fibrinogen.^121,122 Platelets from patients with GT fail to adhere to glass,^18,20,121 and this forms the basis of their abnormality in the glass bead retention assay.¹²³ Platelet coagulant activity has been variably reported as normal or abnormal.^{18,19,20,21,124,125,126} A defect in platelet microparticle formation and support of thrombin generation has been identified in some patients,^{125,126,127,128} but not in all patients.¹²⁹ Integrins α_IIbβ₃ and α_Vβ₃ bind prothrombin, probably accounting for some of the abnormalities identified.^130,131

In flow-chamber studies, thrombasthenic platelets adhere normally to deendothelialized blood vessels at low and intermediate shear rates, but do not spread normally or form platelet thrombi.^132,133,134 A defect in adhesion occurs at higher shear rates. A paradoxical increase in fibrin formation on these surfaces has been observed with thrombasthenic platelets, but the explanation for this phenomenon remains unknown.¹³⁵ In contrast to normal blood, blood from nearly all patients with GT fails to occlude a 150-μm PFA-100 aperture in collagen-coated membranes under high sheer, either in the presence of ADP or epinephrine.^136,137

Platelet integrins α_IIbβ₃ and α_Vβ₃ can be quantitated by several techniques, including, monoclonal antibody binding (using flow cytometry or radiolabeled binding), immunoblotting, and surface-labeling followed by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE). Based on such studies, GT patients are subcategorized by integrin α_IIbβ₃ content into those with less than 5 percent of normal (type I), 5 to 20 percent (type II), or 50 percent or more (variants).^22,138 In one review of 64 patients, 78 percent were type I, 14 percent were type II, and 8 percent were variants.²² This subtyping predated the identification of integrin α_IIbβ₃ abnormalities as the cause of GT and was based on functional data; this categorization provides only limited information.

Measuring integrin α_Vβ₃ content is technically more demanding than measuring that of integrin α_IIbβ₃ because there are only approximately 50 to 100 integrin α_Vβ₃ receptors per platelet.⁶³ The integrin α_Vβ₃ level is very useful, however, in making a preliminary assessment of whether the patient has a defect in the α_IIb or β₃ subunits, because, in general, patients who lack integrin α_Vβ₃ receptors have a defect in the β₃ rather than α_IIb subunit.¹³⁹ A β₃ subunit missense mutation (H280P) that differentially affected integrin α_IIbβ₃ more than α_Vβ₃ has, however, been described.⁶⁹

Fibrinogen-binding studies assess the function of the integrin α_IIbβ₃ complex.²⁵ Early studies used radiolabeled fibrinogen to the binding of fibrinogen when the platelets are stimulated with ADP²⁵ or a similar agonist. Fibrinogen can also be labeled with a fluorescent molecule and then flow cytometry can be used to measure fibrinogen binding. These techniques are most useful in detecting qualitative abnormalities of integrin α_IIbβ₃ in patients with variant GT. The binding of a monoclonal antibody (PAC1) to platelets gives similar information because the antibody only binds to the activated form of the integrin.¹⁴⁰

Carriers of GT have essentially normal platelet function.³⁴ Their platelets, however, only contain approximately 60 percent of the normal number of integrin α_IIbβ₃ receptors; the overlap in values between normal individuals and carriers, however, doesn’t permit for unequivocal diagnosis of carriers by this technique.¹¹² Carrier detection is most accurately performed by DNA analysis.

Platelet fibrinogen is reduced to approximately 10 percent of normal in patients with marked reductions in integrin α_IIbβ₃^18,21,43,44 but is variably reduced in patients with significant amounts of integrin α_IIbβ₃.^138,141,142

Therapy and Prognosis

Therapy of GT patients is discussed in the section entitled “Management of Inherited Platelet Function Disorders.” Although GT is a severe disease, the prognosis for survival is generally good. In one series, two of 64 patients died of hemorrhage and in another series, three of 43 patients died of hemorrhage.^22,33 A nationwide survey in Japan identified 98 GT patients in 1976 and 192 in 1991.¹⁴³ The mortality rate decreased substantially during this time interval.

α_IIbβ₃: Select Macrothrombocytopenias Five heterozygous missense mutations in four different amino acids in integrin α_IIbβ₃ (α_IIb G991C, R995Q, and R995W, and β₃ L718P and D723H), as well as several different deletions, lead to variably mild reductions in both integrin α_IIbβ₃ expression and platelet aggregation, as well as constitutively active receptors, in patients with inherited aniso- and macrothrombocytopenia.^{144,145,146,147,148,149,150} The defects cluster on both sides of the transmembrane domains, and include both members of the α_IIb R995-β₃ D723 salt bridge proposed to maintain the receptor in a low affinity state.¹⁵¹ Proplatelet formation has been reported to be abnormal in several reported cases.

GLYCOPROTEIN Ib (CD42b,c)–IX (CD42a)–V: BERNARD-SOULIER SYNDROME

Definition and History

BSS is an inherited disorder of the platelet GPIb–IX–V complex characterized by thrombocytopenia, giant platelets, and a failure of the platelets to bind GPIb ligands, most importantly, VWF and thrombin.^{152,153,154,155}

In 1948 Bernard and Soulier described two children from a consanguineous family who had a severe bleeding disorder characterized by mucocutaneous hemorrhage, variable thrombocytopenia, and giant platelets.^156,157 Beginning in the early 1970s, BSS platelets were shown to have a functional defect in VWF-dependent platelet adhesion and agglutination.^158,159,160 In 1975, Nurden and Caen identified an abnormality in platelet GPIb as the cause of the functional defect.¹⁶¹ Later studies confirmed the defect in VWF–GPIb interactions^162,163,164 and identified additional defects in platelet GPV and GPIX.^165,166 Subsequent studies have identified additional ligands for the GPIb–IX complex, including thrombin,¹⁶⁷ P-selectin,¹⁶⁸ leukocyte integrin α_Mβ₂,¹⁶⁹ high-molecular-weight kininogen,¹⁷⁰ thrombospondin-1,¹⁷¹ and coagulation factors XI¹⁷² and XII¹⁷³ (Chap. 112), but the precise contributions of these interactions to the disorder are not well defined. Molecular defects in the genes for GPIbα, GPIbβ, and GPIX, but not GPV have been identified in BSS. Mouse models of BSS have been produced by gene targeting of GPIbα¹⁷⁴ and GPIbβ,¹⁷⁵ and like humans, mice deficient in GPV do not demonstrate the typical features of human BSS.^176,177

Etiology and Pathogenesis

This rare disease, with a prevalence estimated as less than one in 1,000,000, has been reported from countries around the world.^{152,154,157,165} Both autosomal recessive (“biallelic”) and autosomal dominant (“monoallelic”) forms of the disorder have been described, with the biallelic producing the most severe symptoms and the monoallelic causing macrothrombocytopenia and a mild or no bleeding syndrome. Consanguinity is common in the biallelic form, with 85 percent of the reported cases being homozygous for the causative mutation.¹⁵⁴

Six different features of BSS may contribute to the hemorrhagic diathesis: thrombocytopenia, abnormal platelet adhesive interactions with VWF, abnormal platelet interactions with thrombin, abnormal platelet coagulant activity, abnormal platelet interactions with P-selectin, and abnormal platelet interactions with leukocyte integrin α_Mβ₂.

The pathophysiology of the thrombocytopenia is uncertain. Early studies suggested a marked shortening of platelet survival, presumably from the decrease in platelet surface charge resulting from the GPIb defect.^178,179 Later studies using ¹¹¹In-oxine to label platelets reported more modest or no shortening of platelet survival, indicating that ineffective thrombopoiesis and/or decreased thrombopoiesis may contribute to the thrombocytopenia.^180,181 Morphologic abnormalities have been identified in BSS megakaryocytes and these may contribute to abnormal platelet production.¹⁸² Based on observations in other giant platelet syndromes (Chap. 117), the large size of Bernard-Soulier platelets would tend to diminish the adverse hemostatic effects of the thrombocytopenia because the platelet mass is better preserved. With only rare exceptions,¹⁸³ however, the bleeding diathesis with BSS is more severe than expected from the thrombocytopenia, reinforcing the conclusion that a qualitative platelet defect is the predominant problem.^157,184

The platelet GPIb–IX complex functions as a receptor for VWF (Chaps. 112 and 126).^152,185,186 This interaction is crucial in the adhesion of platelets to subendothelial surfaces, especially under high shear conditions, where VWF acts as a bridge between the subendothelial matrix and the platelet.^133,134 The relative roles of subendothelial VWF, plasma VWF, and platelet VWF have not been completely defined, but they probably all contribute to platelet adhesion.¹⁸⁷ The interaction of VWF with GPIb/IX initiates activation of integrin α_IIbβ₃,^188,189 which can also bind to VWF, but at a different site on the molecule.¹⁹⁰ The interaction of GPIb/IX with VWF also directly contributes to platelet–platelet interactions.^191,192,193

GPIb/IX–VWF interactions can also occur in platelet suspensions at high shear rates; this can lead to platelet activation, with subsequent aggregation mediated by integrin α_IIbβ₃.^{187,194,195,196} Whether sustained shear rates in vivo ever reach the levels required to initiate VWF binding, however, is not established.

Abnormalities of the GPIb–IX complex can be a result of genetic defects in GPIbα, GPIbβ, or GPIX, all of which are required for surface expression. BSS is the most severe form of the disease and is caused by defects in both alleles of one of the proteins as a result of a homozygous mutation, compound heterozygosity, or a combination of hemizygosity of GPIbβ because of a microdeletion and a mutation affecting the other GPIbβ allele. These abnormalities have been termed the biallelic forms.¹⁵⁴ A macrothrombocytopenic syndrome associated with a mild bleeding syndrome has been reported with heterozygous defects in GPIbα and GPIbβ.¹⁵⁴ Because obligate heterozygotes for the biallelic BSS mutations do not commonly demonstrate macrothrombocytopenia, the heterozygous defects associated with macrothrombocytopenia may exert a dominant negative effect.¹⁵⁴

The platelets of patients with BSS have a decreased response to platelet activation by thrombin, especially at limiting concentrations of thrombin.^197,198,199 BSS platelets are deficient in two different proteins that interact with thrombin, namely GPIbα, which binds thrombin,¹⁶⁷ and GPV, which is a thrombin substrate (Chap. 112). The precise nature of the interactions of thrombin with GPIbα and its biologic consequences are still unclear, but binding of thrombin to GPIbα can initiate signaling within the platelet, perhaps directly through GPIbα crosslinking or indirectly by augmenting activation of other thrombin receptors (protease-activated receptors [PARs] 1 and 4) or other thrombin-dependent events at the platelet surface.¹⁶⁷ Paradoxically, mice deficient in GPV actually have increased sensitivity to thrombin activation and variably increased thrombus formation, perhaps because GPV limits access of thrombin to GPIb.^200,201 Because thrombin is one of the major physiologic activators of platelets, the loss of thrombin binding to GPIbα may contribute to the hemorrhagic diathesis.

Platelets from patients with BSS are defective in supporting thrombin generation as judged by the serum PT,²⁰² a test performed with whole blood, but in other tests of platelet coagulant activity, BSS platelets support coagulation as well as, or better than, normal platelets.^124,203 Defects in collagen-induced coagulant activity and the association of factors V, VIII, and XI with BSS platelets have been described,²⁰³ but their significance is unclear. Similarly, GPIb/IX has been identified as a binding site for other proteins involved in coagulation, including high-molecular-weight kininogen and factor XII, but the contributions of these interactions to the coagulant abnormality are also uncertain.^170,172,173 Binding of VWF to GPIb/IX has been implicated in fibrin-dependent, but not fibrin-independent, augmentation of platelet coagulant activity and thus fibrin-dependent coagulant activity is likely to be abnormal in BSS.¹²⁶ This finding may partially explain the variability in findings between the serum PT and some of the other assays as fibrin only forms in the serum PT. Abnormal membrane lipids have also been reported.²⁰⁴

The mechanism(s) producing the giant platelets in BSS has not been identified, but since giant platelets are found in BSS variants in which GPIb/IX is present, but unable to bind ligand, it has been postulated that the abnormality is a result of the inability of GPIb/IX to bind an unknown marrow ligand.¹⁵² It cannot be because of an inability to bind VWF as, with only rare exceptions,²⁰⁵ patients lacking VWF do not have large platelets. Moreover, in a mouse model of BSS, restoring a receptor with the GPIb transmembrane and cytoplasmic domains, but not the ligand-binding domain, partially corrected both the thrombocytopenia and large platelet size.²⁰⁶ A defect in GPIb/IX–mediated signaling has also been proposed to cause the large platelets as a deficiency of PLC has also been described in BSS.^152,207 A mechanical alteration in the plasma membrane of BSS platelets has been identified by micropipette experiments, showing the plasma membrane to be more deformable than normal.²⁰⁸ Megakaryocytes in BSS have increased ploidy and volume, as well as alterations in the membrane demarcation system, granules, and microtubules.^181,182 Both the increased size and deformability may reflect the loss of the normal interaction of GPIb/IX with the cytoskeleton via actin-binding protein (filamin-1; Chap. 112).

Platelets from patients with BSS are deficient not only in GPIbα, GPIbβ, and GPIX, which are known to be associated as a complex, but also in GPV (Chap. 112).^152,166 All of these proteins share highly conserved leucine-rich regions.^152,187 One possible explanation for the loss of surface expression of all the proteins is that they need to form a complex during biosynthesis in order to be transported to the surface¹⁸⁷; evidence supports the need for GPIbα, GPIbβ, and GPIX to all be present for optimal surface expression,²⁰⁹ but data from mice deficient in GPV indicate that this glycoprotein is not required for surface expression of the GPIb–IX complex.²⁰⁰ GPV may, however, improve the efficiency of expression of the other members of the complex.²¹⁰ Moreover, data from the BSS mouse expressing a chimeric GPIbα molecule in which the leucine-rich repeat domain was replaced with the external domain of another receptor indicate that complex formation does not require the GPIbα leucine-rich domain.²⁰⁶

At the molecular level, the platelets from different patients with BSS are heterogeneous, with many having no detectable GPIb and others having variable amounts, up to 50 percent of normal.^{152,207,211,212,213,214} There also is variability in the degree of concordance in the reduction of GPIb and the other deficient proteins.^215,216

Molecular Defects The molecular biologic basis of BSS has been determined in 161 patients from 132 unrelated families¹⁵⁴ and an online registry of defects is available at http://www.bernardsoulier.org/.²¹⁷ An international consortium reported on 211 families with the recessive form of BSS, which they termed “biallelic.”¹⁵⁴ In total, 45 different mutations have been reported in GPIbα, 52 in GPIbβ, and 28 in GPIX. No defects in GPV have been identified in patients with BSS. The association with consanguineous matings was reinforced as 85 percent of the families had homozygous mutations and 13 percent were compound heterozygotes for defects in one of the genes. None of the variants were identified in several gene variant databases,¹⁵⁴ suggesting that they are all rare and likely entered the population relatively recently. A number of likely founder mutations have been identified in each of the three genes in different populations.^154,218 The ancestry of seven apparently unrelated families with a GPIbβ W89D mutation was traced to a common ancestor in 1671 in India.²¹⁸ Five mutations in GPIX account for 137 of the 184 affected GPIX alleles and GPIX N61S is found in 64 European families. A number of mutations have been identified to cause the heterozygous monoallelic form, including the GPIbα A172V mutation, which has been associated with biallelic and monoallelic forms of the disorder in 42 apparently unrelated families with macrothrombocytopenia in Italy.¹⁵⁴ Many of the defects affect the leucine-rich repeats or the conserved flanking sequences, supporting the importance of these structural elements in the biogenesis and surface expression of the GPIb–IX–V complex (see Chap. 112, Fig. 112–14, and Fig. 120–4). Three patients have been described who are homozygous for a deletion in the last two bases of codon 492 of GPIbα, resulting in a frameshift that alters the membrane spanning region and results in premature termination, and another patient has been described who is heterozygous for this deletion and a missense mutation of GPIbα.^{219,220,221,222} These defects appear to result in a poorly anchored GPIbα with GPIbα antigen present in plasma. GPIbβ mutations have affected the promoter region, at a binding site for the GATA-1 transcription factor,²²³ the signal peptide,²²⁴ and the transmembrane and intracellular domains.²²⁵ A homozygous Y88C defect in GPIbβ has been reported to cause BSS in two Japanese families and heterozygotes with this mutation have a giant platelet syndrome.^221,226 Similarly, a patient heterozygous for a GPIbβ R17C mutation also had a giant platelet syndrome.²²⁷ An N45S mutation in GPIX, affecting leucine rich repeat 1, has been reported in at least 12 different white patients, including four patients from a large Swiss family with variable clinical manifestations^228,229,230 and one Turkish patient.²³¹

BSS has been reported in seven patients in association with hemizygous deletion of GPIbβ and several neighboring genes on chromosome 22q11.2, leading to variable manifestations of the DiGeorge syndrome, including cardiac defects, dysmorphic facial features, thymic hypoplasia, and velopharyngeal insufficiency.^{154,232–239} Hemizygous mutations in the remaining GPIbβ allele have included P96S and P29L.^236,237 In other studies of patients with the 22q11.2 deletion syndrome, modest reductions in platelet count and increases in platelet volume, as well as reduced platelet agglutination to ristocetin and decreased platelet GPIb/IX expression, have been variably reported, consistent with hemizygosity for GPIbβ.^{240,241,242,243,244}

A number of monoallelic, heterozygous mutations in the genes of the GPIb–IX complex have been described as causing macrothrombocytopenia, some, but not all, of which have also been implicated in causing the biallelic form either because of homozygosity or compound heterozygosity.¹⁵⁴ These include a heterozygous mutation in the second leucine-rich repeat (L57F)²⁴⁵ in which the affected patients have moderate bleeding symptoms, moderate thrombocytopenia, and giant platelets. Additional monoallelic mutations of GPIbα that appear to produce dominant effects are N41H²⁴⁶ and Y54D.²⁴⁷

The “Bolzano” defect, which involves a mutation in the sixth leucine-rich repeat of GPIbα (A156V), results in a GPIbα molecule that has reduced ability to bind VWF, but can bind thrombin. It has been described in both biallelic and monoallelic forms. Two patients with biallelic forms have been described. In one patient, the Bolzano defect was homozygous, and the patient had a lifelong history of mucocutaneous hemorrhage in association with an approximately 50 percent reduction in GPIb surface expression and total loss of ability to bind VWF.²⁴⁸ In the other, the Bolzano mutation coexisted with a 12-amino-acid deletion and an amino acid substitution (Q181K).²⁴⁹ The monoallelic form of the Bolzano defect has been reported in more than 100 patients from 48 pedigrees, primarily from southern Italy.²⁵⁰ Most patients have no bleeding symptoms, but some have mild to moderate bleeding symptoms. Mild thrombocytopenia, increased mean platelet volume, reduced expression of GPIb/IX/V, and normal or borderline abnormal values for ristocetin-induced platelet aggregation are characteristic of this disorder.

Clinical Features

Epistaxis is the most common symptom of BSS (70 percent); also common are ecchymoses (58 percent), menometrorrhagia (44 percent), gingival hemorrhage (42 percent), and gastrointestinal bleeding (22 percent).¹⁵⁷ The combination of BSS with angiodysplasia can result in particularly severe recurrent hemorrhage.^251,252,253 Hemorrhagic symptoms that occur with lower frequency include posttraumatic bleeding (13 percent), hematuria (7 percent), cerebral hemorrhage (4 percent), and retinal hemorrhage (2 percent). There is considerable variability in symptoms among patients, even among patients within a single family.^152,254 A review that includes brief descriptions of the clinical features of 55 patients, reported through 1998, has been published.¹⁵²

Laboratory Features

Thrombocytopenia is present in nearly all patients, but is variable in its severity, ranging from approximately 20 × 10⁹ platelets/L to near-normal levels. Platelets are large on smear, with more than one-third usually having diameters greater than 3.5 μm, and some being as large or larger than lymphocytes. By electron microscopy, platelets display only minor variations in vesicular structures and the open canalicular system,¹⁵⁷ but megakaryocytes have more notable abnormalities in their demarcation membranes.¹⁸² The cell membranes of platelets from patients with BSS appear to be more deformable than normal,²⁰⁸ perhaps because GPIb ordinarily interacts with the platelet cytoskeleton²⁵⁵ (Chap. 112).

Closure times of the apertures of collagen-coated membranes are markedly prolonged in the presence of ADP or epinephrine (PFA-100).¹³⁶ The hallmark findings in the BSS are the failure of platelets to aggregate in response to ristocetin¹⁵⁹ or botrocetin,^162,256 agents that require VWF–GPIb interactions. In VWD, but not BSS, this defect can be corrected by adding normal plasma (or VWF).

Although, the large size of the platelets in BSS and the thrombocytopenia make it technically difficult to perform platelet aggregation studies, in general, aggregation induced by ADP, epinephrine, or collagen is either normal or enhanced.^160,257,258 The aggregation response to thrombin is usually dose-dependent, being essentially normal in response to high doses of thrombin but characterized by a prolonged lag phase and diminished aggregation in response to low doses of thrombin.^197,259

Platelet Coagulant Activity The coagulant activity of platelets from patients with BSS has been variably reported as reduced, normal, or increased.^124,202,203 The variable presence of fibrin in the different assays used to assess platelet coagulant activity may account for these inconsistent results as GPIb–VWF interactions enhance platelet coagulant activity when fibrin is present, but not when it is absent.¹²⁶

Platelet-Thrombin Interactions Both GPIb and the seven-transmembrane domains PAR-1 and PAR-4 receptors are required for maximal response to thrombin.^167,259 Two different crystal structures of the interactions between thrombin and GPIbα have been reported; in one, two molecules of thrombin bind to each GPIbα molecule, raising the possibility that free thrombin or thrombin adherent to fibrinogen can cluster GPIb–IX–V complexes.^167,260,261 GPV, which is missing from the platelet surface in BSS, is cleaved by thrombin, but the cleavage is neither necessary nor sufficient for thrombin-induced platelet activation.^262,263 In fact, platelets of mice lacking GPV have increased responsiveness to thrombin, perhaps because GPV ordinarily limits access of thrombin to GPIbα or inhibits GPIbα crosslinking.^200,201

Ex Vivo Interaction with Subendothelial Surfaces Platelets frompatients with BSS demonstrate defective adhesion to subendothelial surfaces, especially at shear rates greater than 650 s^–1.^{133,134,158,264} The results are similar to those in patients with VWD.

Shear-Induced Platelet Aggregation Unlike normal platelets, platelets from patients with BSS are not aggregated by high shear rates.^194,195 The initial interaction in this process appears to be binding of VWF to GPIb,¹⁸⁷ with subsequent activation of integrin α_IIbβ₃. perhaps through signaling via the protein 14-3-3ζ associated with the cytoplasmic domain of GPIbα,^196,265 Fcγ receptor IIA, GPVI, and/or the Fc receptor γ chain (Chap. 112).^266,267,268 Pathologic shear stress has been reported to increase binding of α-actin to GPIb/IX as part of the signaling process.^268,269,270

Therapy

The therapy of BSS is described in the section “Management of Inherited Platelet Function Disorders” below. Splenectomy has been performed when the diagnosis of immune thrombocytopenia was mistakenly made, but this usually does not normalize the platelet count or improve the bleeding diathesis.²⁴⁹

GPIbα (CD42b,c): PLATELET-TYPE (PSEUDO-) VON WILLEBRAND DISEASE

Definition and History

A heterogeneous group of patients has been described with mild to moderate bleeding symptoms, variably enlarged platelets, variable thrombocytopenia, and diminished plasma high-molecular-weight VWF multimers.²⁷¹ The fundamental defect in these patients is thought to be an enhanced interaction between an abnormal platelet GPIb/IX receptor and normal plasma VWF.^272–281 Because these patients have some of the hallmarks of VWD, but the defect is in platelet GPIb/IX, it has been termed both pseudo-VWD and platelet-type VWD. At present, 55 patients are listed in the database of patients with this disorder (www.pt-vwd.org).^282,283

Etiology and Pathogenesis

A qualitative abnormality in GPIb is responsible for this disorder, with ongoing in vivo binding of high-molecular-weight VWF multimers to platelets causing depletion of the plasma high-molecular-weight multimers. In addition, the binding of the VWF to platelets may lead to shortened platelet survival, perhaps accounting for the variable thrombocytopenia. Inheritance is autosomal dominant.

Abnormalities in the Mr of GPIb were identified in two families,²⁷⁷ but these may have resulted from a now-recognized polymorphism in GPIb (Chap. 112) rather than being related to the functional disorder. Heterozygous point mutations in the GPIbα gene causing a variety of missense alterations (G233V, G233S, M239V, D235Y, W230L) have been found in several different families.^{271,278,284,285,286,287,288,289} The G238V mutation is the most common among the patients in the database, affecting 31 of 35 patients. All of these mutations are in the R-loop (also termed β-switch) in the carboxyterminal flanking sequence of the leucine-rich repeats, a region implicated in ligand binding (see Fig. 120–4).^{152,187,290,291,292,293} Molecular modeling suggests that the M239V substitution produces a significant conformational change in the molecule,²⁹⁴ and this was confirmed by crystallographic analysis.²⁹⁵ It has been proposed that the platelet-type VWD mutations either destabilize the compact triangular structure of the R-loop by interfering with the D235-K237 salt bridge or stabilizing the extended β-hairpin form of the R-loop, which is better able to engage the VWF A1 domain.²⁸² A mouse model of the GPIbα G233V mutation recapitulated many of the human manifestations and had an unexpected increase in bone mass.²⁹⁶ Recombinant GPIbα fragments containing the G233V and M239V mutations demonstrated enhanced interactions with VWF in several different systems, including ones under shear stress.^297,298 An increase in platelet GPIb/IX expression has also been reported.^278,281 An in-frame 27-base-pair (bp) deletion in the macroglycopeptide region of GPIbα has also been reported to cause platelet-type VWD.²⁷⁹ Because the deletion may lead to the loss of up to four glycosylation sites, it has been proposed that the glycans play a negative regulatory role in ligand binding.²⁸²

Clinical Features

Patients have variable thrombocytopenia and mild to moderate mucocutaneous hemorrhage. A study of 13 patients with six different mutations using a standardized bleeding assessment tool found a wide range of clinical severity, with approximately 40 percent having a normal bleeding score and the remainder having a wide range of abnormal scores.²⁷¹ Bleeding scores did not correlate with age or sex, but did correlate with reductions in both platelet count and ristocetin cofactor activity. All patients had macrothrombocytopenia. Of note, pregnancy may exacerbate the thrombocytopenia and bleeding symptoms.²⁷⁸

Laboratory Features and Differential Diagnosis

Mild thrombocytopenia and somewhat enlarged platelets are present in some, but not all, patients. Plasma VWF levels are variably reduced, with a disproportionate reduction in plasma high-molecular-weight multimers. Platelet VWF multimers are normal.

The most characteristic laboratory finding in platelet-type VWD is enhanced platelet aggregation in response to low concentrations of ristocetin^{272,273,274,275,276,278,288} or botrocetin.²⁹⁹ This same abnormality is present in patients with type 2b VWD, as is selective depletion of plasma high-molecular-weight VWF multimers (Chap. 126). In platelet-type VWD, however, the defect is in platelet GPIbα, whereas in type 2b VWD, the defect is in the VWF. In one study comparing platelet-type and type 2b VWD, patients with type 2b VWD had more severe bleeding, especially menorrhagia, and lower platelet counts.²⁷¹ Several assays can help differentiate between these abnormalities^{274,300,301,302}: (1) normal VWF (purified or in cryoprecipitate) will aggregate platelets from patients with platelet-type VWD, but not platelets from patients with type 2b VWD; (2) isolated platelets from patients with platelet-type VWD will bind normal VWF at lower concentrations of ristocetin than will normal platelets or platelets from patients with type 2b VWD; (3) plasma VWF from patients with type 2b VWD will bind to normal platelets at lower-than-normal concentrations of ristocetin, whereas higher-than-normal concentrations of ristocetin are required to promote the plasma VWF from patients with platelet-type VWF to bind to normal platelets³⁰¹; and (4) VWF lacking sialic acid residues (asialo-VWF) will agglutinate platelets from patients with platelet-type VWD in the presence of ethylenediaminetetraacetic acid (EDTA).³⁰³ A number of patients with platelet-type VWD were originally diagnosed as having type 2b VWD, leading to the conclusion that platelet-type VWD may be underdiagnosed, and an international registry based study supports this contention.^278,280,304

Therapy

Because normal VWF (especially the high-molecular-weight forms) can bind excessively to the platelets of patients with platelet-type VWD and potentially lead to rapid platelet clearance from the circulation, increasing the VWF level by any means (desmopressin infusion or VWF replacement with cryoprecipitate or VWF concentrates) poses a potential risk of inducing thrombocytopenia.^300,305 It may be possible to estimate this risk by assessing whether the patient’s platelets aggregate ex vivo in response to VWF (as in cryoprecipitate).²⁷³ Low-dose cryoprecipitate has successfully supported hemostasis, without inducing thrombocytopenia.^275,305,306 Currently, cryoprecipitate is generally less favored for VWF replacement therapy than plasma-derived factor VIII concentrates such as Humate-P, which is approved in the United States for the therapy of VWD, because the plasma-derived factor VIII concentrates have a reduced risk of viral infection. Consideration should also be given to platelet transfusion in appropriate circumstances. Recombinant factor VIIa infusion may be beneficial and is licensed for this indication in Europe, but this therapy is not yet approved by the FDA; it has the theoretical advantage of avoiding excessive interactions between VWF and the abnormal GPIbα receptor.^307,308

INTEGRIN α₂β₁ DEFICIENCY (GLYCOPROTEIN Ia/IIa; VLA-2; CD49B/CD29)

Integrin α₂β₁ (GPIa/IIa) can mediate platelet adhesion to collagen and platelet activation under certain conditions (Chap. 112). A female patient with excessive posttraumatic bruising and menorrhagia but no epistaxis, gum bleeding, or excessive bleeding after tonsillectomy or appendectomy was described whose platelets selectively failed to aggregate or undergo shape change in response to collagen.^309,310 The bleeding time was markedly prolonged, and the patient’s platelets failed to adhere and spread normally on subendothelial surfaces. The patient’s platelets only contained approximately 15 to 25 percent of the normal amount of integrin α₂,^309,311 and a reduction in the β₁ subunit was also apparent.³⁰⁹ It is difficult to draw conclusions about the physiologic role of integrin α₂β₁in platelet function from this patient because her α₂β₁ deficiency was incomplete, her bleeding symptoms were mild and variable, and some of the platelet function abnormalities (e.g., abnormal platelet-collagen interactions in the presence of the divalent chelating agent EDTA) are difficult to ascribe to the deficiency in α₂β₁.^309,312

Another patient with integrin α₂ deficiency has been described.³¹³ She had a history of mucocutaneous and postoperative bleeding. Her bleeding time was prolonged and platelet aggregation in response to collagen was selectively reduced, but not absent. In addition to her α₂ subunit defect, she also had little or no intact thrombospondin, and exogenous thrombospondin corrected the defect in platelet aggregation. The patient’s hemorrhagic symptoms and platelet defects disappeared when she entered menopause.

The variation in platelet integrin α₂β₁ expression in healthy individuals is very wide (10-fold) and platelet levels have been correlated with allelic variants.³¹⁴ Reduced α₂β₁ expression has been associated with alterations in megakaryocyte production and decreased mean platelet volume.³¹⁵

Mice with targeted deletion of integrin α₂β₁ do not have a hemorrhagic phenotype or prolonged tail bleeding times but they do have reduced platelet adhesion to collagen and reduced thrombus formation after vascular injury³¹⁶; mice with a conditional loss of integrin α₂β₁ in megakaryocytes and platelets have a decreased mean platelet volume.³¹⁵

CD36 (GPIV; FATTY ACYL TRANSLOCASE [FAT]; SCAVENGER RECEPTOR CLASS B, MEMBER 3 [SCARB3])

CD36 (GPIV) is a highly, but variably expressed platelet glycoprotein that is present on many cell types and documented to participate in long-chain fatty acid transport (Chap. 112). Approximately 3 percent of Japanese, 2 percent of African Americans, and 0.3 percent of whites in the United States have platelets that lack CD36 (GPIV).^317,318 Although CD36 (GPIV) has been implicated in platelet interactions with collagen, thrombospondin, advanced glycation products,³¹⁹ and myeloid-related protein (MRP)-14,^{320,321,322,323} as well as in platelet–monocyte interactions,³²⁴ individuals lacking CD36 (GPIV) do not have a hemorrhagic diathesis. Platelets from these patients can bind thrombospondin via alternative receptors³²⁵ and there are differing data on its role in adhesion to collagen.^326,327,328 A multiparameter analysis of platelet thrombus formation to different matrix proteins at differing shear rates identified a role for CD36 at low, but not high, shear rates.⁸ CD36 (GPIV) has also been implicated as a receptor for oxidized low-density lipoprotein (LDL), and the binding of very-low-density lipoprotein (VLDL) to CD36 (GPIV) has been reported to enhance collagen-induced platelet aggregation and thromboxane production.³²⁹ CD36 (GPIV) platelet expression varies widely among healthy individuals (200 to 14,000 molecules per platelet) and correlates with activation by oxidized LDL and genetic single-nucleotide variants.³³⁰

Two forms of CD36 (GPIV) deficiency have been described in Japan: type I in which both platelets and monocytes are deficient, and type II in which only platelets are deficient.^331,332,333 A P90S substitution that also leads to abnormal posttranslational modification is a common abnormality contributing to both type I and type II deficiencies. In the type I form, patients are homozygous for the abnormality, whereas in type II deficiency, patients are doubly heterozygous for the P90S abnormality and an unidentified platelet-specific expression defect.^331,334,335 Other abnormalities that have been associated with type I deficiency include a dinucleotide deletion^539,540 in exon 5, a 161-bp deletion^331–491 corresponding to loss of exon 4, a nucleotide insertion at position 1159 in codon 317 leading to a frameshift and premature stop, and splice-site mutations.^336,337,338 Other mutations have been identified in other populations.

CD36 (GPIV) deficiency can result in refractoriness to platelet transfusions because of isoimmunization and has been implicated in posttransfusion purpura (Chap. 117),³³⁹ as well as thrombocytopenia caused by the passive transfer of anti-CD36 antibodies.³⁴⁰

GLYCOPROTEIN VI DEFICIENCY

GPVI can mediate platelet adhesion to collagen and is important in collagen-induced signal transduction (Chap. 112). Twelve patients with mild to moderate bleeding disorders and variable deficiencies of platelet GPVI or signaling have been described; one had concomitant GPS (α-granule deficiency).^341–350 The others had selective abnormalities in platelet–collagen interactions. Platelet GPVI deficiency associated with an autoantibody to GPVI has been described in several patients, one of whom also had systemic lupus erythematosus and another of whom had coexisting antibodies to integrin α_IIbβ₃.^351,352,353 Antibody to GPVI may be detectable in eluates prepared from patient platelets even when it is not detectable in patient plasma.³⁵⁴ Acquired forms of GPVI-specific signal transduction have also been described in patients with myelodysplastic syndromes and chronic lymphocytic leukemia.³⁴⁷ Studies in mice and primates demonstrated that antibodies to GPVI can result in loss of GPVI from the platelet surface through either proteolytic shedding or a cyclic adenosine monophosphate (cAMP)-mediated internalization mechanism, even though the platelets continue to circulate.^346,355 Thus, it is likely that the deficiency of GPVI most commonly results from the autoantibodies.³⁵³ It is unclear whether the patients reported earlier as having GPVI deficiency might also have had an immune basis for their GPVI deficiency.

Three different inherited forms of GPVI deficiency have been described. One patient with a lifelong history of “mild” mucocutaneous, posttraumatic, and postsurgery bleeding had a marked deficiency of platelet membrane GPVI resulting from both a 16-base out-of-frame deletion and a S175N missense mutation.³⁵⁰ The patient’s platelets failed to respond to collagen, convulxin, or collagen-related peptide. Of note, FcRγ expression was normal. Another patient with easy bruising, a prolonged bleeding time, abnormal PFA-100 closure time to collagen/epinephrine, and no platelet aggregation in response to collagen had a combination of a R38C mutation in one allele and a five-nucleotide duplication insertion in the other, leading to a nonsense codon.³⁵⁶ The R38C mutation led to misfolding of the protein, reduced surface expression, and a qualitative defect in collagen binding. Five subjects from four apparently unrelated families in Chile with variable histories of easy bruising, epistaxis, excessive bleeding after minor trauma, and gingival bleeding were found to have no platelet aggregation in response to collagen, convulxin, or collagen-related peptide.³⁵⁷ DNA analysis showed a homozygous adenine insertion between bases 711 and 712. Heterozygotes were asymptomatic and had nearly normal platelet function.

Platelets contain at least three types of granules: dense or δ granules containing ADP, ATP, calcium, serotonin, and pyrophosphate; α granules containing a variety of proteins, some derived from plasma, others synthesized by the megakaryocyte; and lysosomes containing acid hydrolases. Following platelet activation, the contents of the α and dense granules are secreted. Inability to release these granule contents by virtue of a deficiency of the granules or their contents or in the cellular mechanisms governing the secretory process is associated with impaired platelet function. A heterogeneous group of disorders involving platelet granules has been described. They are broadly categorized into defects affecting dense granules (δ-storage pool deficiency [SPD]), α granules (α-SPD, or GPS), or both dense bodies and α granules (αδ-SPD). Another disorder affecting granules is the Quebec platelet disorder, which affects the α-granules.

δ-STORAGE POOL DEFICIENCY

Definition and History

In 1969, a family with impaired platelet aggregation was described whose platelets displayed decreased levels of ADP.³⁵⁸ Holmsen and Weiss subsequently established that the defect was a deficiency in the nonmetabolic pool or “storage pool” of ADP present in the dense granules.³⁵⁹ δ-SPD is a heterogeneous disorder characterized by a bleeding tendency, abnormalities in the second wave of platelet aggregation, and variable deficiencies of platelet dense granule contents.

Etiology and Pathogenesis

Normal platelets contain two pools of adenine nucleotides that exchange very slowly.³⁶⁰ One pool is a metabolic nongranule pool in which the ratio of ATP to ADP content is 8 to 10:1. The second pool is the “storage pool” present in the dense granules and contains 65 percent of the platelet adenine nucleotides with an ATP-to-ADP ratio of 2:3. It is this storage pool that is deficient in δ-SPD. Dense granules also contain serotonin, which is taken up from plasma at a ratio of approximately 1000:1 via a pH-dependent amine trapping mechanism. Platelet serotonin levels are also decreased in δ-SPD.

δ-SPD can be a primary, inherited platelet disorder or a component of a multisystem (syndromic) disorder, such as the Hermansky-Pudlak syndrome (HPS)^{361,362,363,364,365} (variable oculocutaneous albinism, hemorrhagic disorder, and neurologic manifestations), the Chédiak-Higashi syndrome^361,365,366 (partial oculocutaneous albinism, giant lysosomal granules, and frequent pyogenic infections), and the Wiskott-Aldrich syndrome^361,367,368 (see below and Chap. 80). Other diseases associated with δ-SPD are Ehlers-Danlos syndrome,³⁶⁹ osteogenesis imperfecta,³⁷⁰ and TAR syndrome.^361,371 The mode of inheritance in δ-SPD is not well defined, but an autosomal dominant pattern for the primary form has been identified in some patients.³⁷² The inheritance pattern of the syndromic forms follows the autosomal recessive and X-linked patterns characteristic of those disorders. Essential criteria for the diagnosis of HPS are the tyrosinase-positive oculocutaneous albinism and the δ-SPD in platelets.³⁷³ The hallmarks of albinism are diffuse hypopigmentation of skin, iris, hair and retina, although there is phenotypic heterogeneity.

Studies from animal models and, particularly, in patients with the syndromic variants indicate that defects in biogenesis of lysosome-related organelles form the basis of the disorders.^361,374 These organelles share features with lysosomes but have distinct morphology, composition, and functions, and include melanosomes in melanocytes, platelet δ-granules, and Weibel-Palade bodies in endothelial cells.^361,374 In δ-SPD associated with HPS, there may be a total failure of δ-granule formation as judged by electron microscopy of platelets and megakaryocytes³⁷⁵ and the absence of CD63 (granulophysin; ME491; LIMP-1; LAMP-3), a lysosomal and dense granule membrane protein of Mr 40,000 that is also found in melanosomes.^{361,363,376,377} The defect in melanosomes accounts for the oculocutaneous albinism. The defect in lysosomes results in accumulation of ceroid lipofuscin, a lipid-protein complex leading to granulomatous colitis and pulmonary fibrosis, which are variably manifest in these patients. Abnormalities of nine genes have been implicated in causing HPS (Fig. 120–5). Ultrastructural studies in patients with a variety of types of HPS indicate that α granules and other organelles are unaffected.³⁷⁸ Mutations in HPS-associated genes result in defects in intracellular protein trafficking and in the biogenesis of lysosomes and lysosomes-related organelles.³⁶¹ The HPS gene products operate in distinct complexes termed biogenesis of lysosome-related organelles complexes (BLOCs), which consist of multiple proteins.³⁶¹ HPS is unusually common in patients from northwest Puerto Rico (frequency: 1:1800), and linkage analysis of these patients led to the identification of the abnormality in HPS gene (HPS1). The gene encodes a 700-amino-acid protein that participates in the complex of proteins called BLOC-3.^361,363 The mutation in the Puerto Rican kindreds is a 16-bp duplication in HPS1 exon 15; other mutations of the same gene have been identified in patients from this region and in other ethnic groups.^363,379,380 HPS2 is caused by mutations in the gene AP3B1 (HPS2), which codes for the β₃A subunit of the heterotetrameric adaptor complex-3 (AP-3), which, in turn, facilitates the formation of vesicles of lysosomal lineage from membranes of the trans-Golgi network or late endosomes. Patients with mutations in this protein tend to also have neutropenia and childhood infections.^373,381

Defects in the HPS3 gene cause a relatively mild form of HPS and pulmonary involvement is usually minimal.³⁸² HPS4 encodes a protein that interacts with the HPS1 protein in the BLOC-3 complex.³⁸³ Patients with mutations in this gene tend to have severe disease, and like patients with defects in HPS1, pulmonary involvement is common. The gene products of HPS5 and HPS6 interact with the HPS3 gene product to form BLOC-2.^384,385 Proteins implicated in HPS7 (encoded by DTNBP1) and HPS8 (encoded by BLOC1S3) are components of BLOC-1.^386,387 Improper trafficking of melanocyte-specific proteins, including tyrosinase, has been found in the melanosomes of patients with HPS5.³⁸⁸ HPS is also associated with mutations in pallidin (BLOC1S6/PLDN), a protein of BLOC-1 and labeled as HPS9.³⁸⁹

Not all patients with δ-SPD have HPS. Some of the patients described to have δ-SPD have been subsequently shown to have mutations in RUNX1.^{358,372,390,391}

In other forms of δ-SPD, data obtained with uranaffin, a dye that specifically stains amine-containing granules, indicate that dense granule membranes are formed but are not properly filled.^392,393,394 The defects in the different substances contained in dense granules are also heterogeneous, with some patients able to secrete significant amounts of calcium and pyrophosphate even when adenine nucleotide secretion is nearly absent.³⁹²

Chédiak-Higashi syndrome results from mutation of the LYST gene, which encodes a protein of estimated molecular mass of 429 kDa, predicted from domain analysis to participate in vesicle transport and to interact with microtubules; an HPS1-like region is also present.^361,366

Numerous animal models of human δ-SPD and HPS have been reported and several represent specific counterparts of the human disease. Thus, more than 20 separate inherited mouse defects have been reported to include dense granule deficiencies; of these, pale ear (ep) is linked to the mouse equivalent of the HPS1 gene; pearl (pe) to the mouse equivalent of HSP2 (β₃A subunit of AP-3 complex); cocoa to HPS3; light ear (le) to HPS4; ruby-eye-2 (ru2) to HPS5; ruby-eye (ru) to HPS6; sandy (sdy) to HPS7; and pallid to HPS9 (PLDN).^361,395 The beige mouse and rat serve as models for Chédiak-Higashi syndrome.³⁶¹

Clinical Features

Patients with δ-SPD as part of the HPS may have severe, or even lethal, hemorrhage.³⁶³ For all other forms of the disorder, the bleeding tendency is variable, and mild to moderate.

Laboratory Features

When tested with the PFA-100 instrument, both prolonged and normal closure times have been reported, akin to prior findings with bleeding times.^{396,397,398,399} Interestingly, flow-dependent thrombus formation, assessed using a multisurface and multiparameter flow assay, is reported to be decreased.⁸ In platelet aggregation studies, ADP and epinephrine induce a normal primary wave of aggregation, but the secondary wave is variably abnormal. The abnormal platelet response is more easily discernible at low collagen doses than high doses. Secretion of ATP from platelets can be measured by luminescence simultaneously with platelet aggregation using a specially designed instrument.⁴⁰⁰ ATP release in response to activation is absent or decreased in δ-SPD patients. Thrombin at high doses causes maximal release of platelet dense body contents, even in patients with secretion abnormalities unrelated to granule deficiency, and, therefore, this reagent may distinguish between δ-SPD (diminished release) and abnormalities of platelet secretory mechanisms, wherein the release may be normal or near-normal.

Measurement of platelet granule contents can establish further the platelet abnormality. The total platelet content of adenine nucleotides is reduced in δ-SPD, and the ratio of total platelet ATP to ADP is increased because it more closely reflects the ratio in the cytoplasmic, “metabolic” pool of adenine nucleotides (approximately 8:1) than in the “storage” pool in dense granules (approximately 2:3).^360,401 Platelet serotonin is variably reduced.⁴⁰² Serotonin is taken up by platelets of patients with δ-SPD, but because it cannot be stored in dense granules, it is rapidly catabolized.⁴⁰² Abnormalities in platelet secretion and arachidonic acid metabolism have been identified, but are quite variable.⁴⁰³ Reduced plasma and platelet VWF activity in association with a decrease in plasma high-molecular-weight multimers has been reported in HPS,^404,405 and may reflect abnormalities involving the endothelial Weibel-Palade bodies.

The decrease or absence of platelet dense bodies can be confirmed by electron microscopy, using either whole mounts or thin sections of platelets fixed in the presence of calcium, although it requires expertise to interpret the results.^5,406 Some patients have abnormal granules. Uranaffin and osmium may help to identify dense granules. The fluorescent amine mepacrine can be used to quantify dense bodies by fluorescent microscopy or by flow cytometry.^407,408 Immunoblot analysis of skin fibroblasts in HPS may facilitate identification of the protein responsible for the defect.⁴⁰⁹

Therapy, Course, and Prognosis

The general principles of patient management are similar to those described for all patients with platelet function defects. Patients with HPS suffer from a number of specific problems related to their albinism, colitis, and pulmonary fibrosis.³⁷³ In particular, they should avoid sun exposure. An antifibrotic agent, pirfenidone, demonstrated modest benefit in the initial study.⁴¹⁰ A second study⁴¹¹ failed to confirm this and was terminated because of futility.

GRAY PLATELET SYNDROME (α-STORAGE POOL DEFICIENCY)

Definition and History

The GPS is a markedly heterogeneous bleeding disorder characterized by selective deficiency of platelet α-granules and their contents, in combination with thrombocytopenia and large platelet size.^412,413 The name derives from the initial observation by Raccuglia, in 1971,⁴¹⁴ of the gray appearance of platelets with paucity of granules on peripheral blood films from a patient with a life-long bleeding disorder.

Etiology and Pathogenesis

Normal platelets contain approximately 50 spherical or elongated α granules that contain a large number of proteins, some relatively specific for platelets (platelet factor 4 [PF4], β-thromboglobulin [βTG]), others that are also found in plasma, and yet others whose role in platelets is poorly understood.^412,415,416 Plasma proteins present in α granules include fibrinogen, VWF, albumin, coagulation factor V, immunoglobulin (Ig) G, fibronectin, and several protease inhibitors. Some of these proteins, such as VWF are synthesized by megakaryocytes, whereas others, such as albumin and IgG, are incorporated into platelets by endocytosis. The α-granule membrane contains several proteins present in the platelet plasma membrane (integrin α_IIbβ₃, GPIb-IX-V) and some specific for α granules (P-selectin and osteonectin).

The molecular mechanisms leading to α-granule deficiency in GPS and the combined αδ-SPD are also heterogeneous; thus, they have been attributed to failure of α-granule maturation during MK differentiation, transport or targeting of proteins to α granules, and/or synthesis of granule membranes.^412,415 Proteomic studies in a GPS patient suggested a failure to incorporate endogenously synthesized MK proteins into α granules.⁴¹⁵ Some GPS patients have elevated plasma PF4⁴¹² suggesting that PF4 synthesis was normal and the primary defect was impaired granule biogenesis with leakage of PF4. Some patients with decreased α-granule contents have a mutation in the gene for the transcription factor RUNX1^372,390 and PF4 is a transcriptional target of RUNX1.⁴¹⁷ This suggests that decreased platelet PF4 levels in these patients represents a defect in transcriptional regulation and synthesis of PF4. GPS has been reported in association with a X-linked thrombocytopenia, thalassemia, and an R216N mutation in GATA-1, a major regulator of megakaryopoiesis.⁴¹⁸ Autosomal dominant GPS resulting from mutations in transcription factor gene GFI1B has been reported in two kindreds.^419,420

Three groups^10,11,12 have reported mutations in NBEAL2 in patients with GPS, a gene that encodes for a protein linked to vesicle transport in neuronal cells. These studies implicate defective vesicle transport as a major mechanism in many forms of GPS. Studies in Nbeal2-deficient mice provide evidence^421,422 that Nbeal2 is required for α-granule biogenesis, platelet function, and MK survival, development and platelet production—key evidence linking GPS and Nbeal2. These mice were also found to have abnormalities in arterial thrombosis, inflammation, and wound repair. A mutation in the gene encoding the VPS33B protein (a member of the Sec1/Munc18 protein family) involved in vesicle trafficking has also been associated with human α-granule deficiency in the arthrogryposis multiplex congenital, renal dysfunction, and cholestasis (ARC) syndrome.⁴²³ A mutation in VPS16B gene has also been linked to α-granule deficiency in the ARC syndrome, and it appears that VPS16B and VPS33B interact with each other.⁴²⁴ Thus, multiple mechanisms can lead to GPS. The inheritance of α-granule deficiency has been autosomal recessive in most reports, but autosomal dominant and sex-linked cases have also been described.^{372,390,412,418}

Clinical Features

GPS patients have a lifelong mild to moderate bleeding diathesis, which is variable in its manifestations.⁴¹²

Laboratory Features

Platelets appear as larger-than-normal, pale, ghost-like, oval forms on blood films and often difficult to identify (Fig. 120–6). Thrombocytopenia is variable but can be moderately severe, with counts below 50,000/μL. Platelet aggregation abnormalities vary considerably.^{343,412,413,425,426,427} ADP- and epinephrine-induced aggregation is normal or nearly normal. Collagen- and thrombin-induced aggregation tend to be more abnormal, but this is not a consistent finding. Concomitant GPVI deficiency was reported in one patient and if the association is more widespread, it may explain the variable abnormalities in collagen-induced aggregation.³⁴³ Studies in one patient showed flow-dependent thrombus formation to be decreased.⁸ Additional abnormalities in PI metabolism, protein phosphorylation, calcium mobilization, platelet factor Va, and platelet secretion have been described^428,429,430 and may contribute to the platelet aggregation abnormalities and clinical symptoms. The failure of exogenous α-granule proteins to fully correct the aggregation defects suggests that these abnormalities may be important in the overall platelet dysfunction in GPS.⁴²⁵

Under the electron microscope platelets and megakaryocytes from GPS patients reveal absent or markedly decreased α granules (Fig. 120–6).⁴¹² The platelets are deficient in α-granule proteins: PF4, βTG, VWF, thrombospondin, fibronectin, factor V, high-molecular-weight kininogen, transforming growth factor (TGF)-β₁, and platelet-derived growth factor (PDGF); albumin and IgG may also be decreased. There is increased reticulin in the marrow of some patients with GPS,^431,432,433 which may be associated with splenomegaly and evidence of extramedullary hematopoiesis.^431,434 The marrow fibrosis has been attributed to elevated marrow levels of PDGF and TGF-β₁ leaked from megakaryocytes.⁴¹² Megakaryocytes show emperipolesis and the capture of neutrophils. P-selectin is a protein present in the α-granule membranes and translocates to platelet surface on platelet activation. The content and surface expression of P-selectin has been reported as normal^412,435 or decreased,⁴³⁶ which underscores the heterogeneity of the GPS.

α,δ-STORAGE POOL DEFICIENCY

This disorder is characterized by moderate to severe defects in both α and δ granules, with heterogeneous expression in the few patients in whom it has been reported.³⁷² Clinical and laboratory features are similar to those of δ-SPD. In general, the functional consequences of the defect in dense granules is more severe than the defect in α granules.

QUEBEC PLATELET DISORDER

Quebec platelet disorder (QPD) is an autosomal dominant bleeding disorder associated with reduced platelet counts and decreased α-granule proteins as a result of increased plasmin-mediated degradation of the α-granule proteins.^437,438 Originally described as factor V Quebec, the early description of this disorder included severe bleeding after trauma, mild thrombocytopenia, decreased functional platelet factor V, and normal plasma factor V.^437,439 The bleeding time is normal to mildly prolonged and epinephrine-induced platelet aggregation is selectively decreased. Aggregation response to other agonists is variable. Platelets from patients with QPD display reduced levels of several α-granule proteins (factor V, fibrinogen, VWF, fibronectin, thrombospondin, multimerin, and osteonectin) as a consequence of enhanced proteolysis.^437,440 The excessive plasmin generation inducing the degradation of α-granule proteins results from increased megakaryocyte expression of u-PA because of a tandem duplication mutation of the cis regulatory elements of the u-PA gene (PLAU).^441,442 Plasma tests of systemic fibrinolysis (fibrinogen, D-dimer, plasminogen, plasmin-α₂ antiplasmin complexes, and u-PA), are normal in these patients. Genetic testing for the PLAU mutation provides a definitive diagnosis. Treatment with fibrinolytic inhibitors appears to be effective in controlling bleeding.⁴³⁷

A sizable percentage of patients with variably severe mucocutaneous bleeding manifestations, mostly mild, have defects in platelet aggregation and secretion. In most of these patients the underlying platelet molecular mechanisms are unknown. The most common pattern on laboratory studies is blunted platelet aggregation and absence of the second wave of aggregation on exposure to ADP, epinephrine, collagen, or U46619, and decreased dense granule secretion. Such patients have been lumped together, more out of convenience than because of an understanding of the mechanisms, categorized as primary secretion defects, activation defects, or signal transduction defects.^{443,444,445,446} Simplistically, platelet activation is a complex process involving agonist binding to surface receptors; signal transduction through G-protein–coupled receptors and other types of receptors; phosphoinositol metabolism resulting in calcium mobilization and phosphorylation of target proteins; arachidonic acid metabolism leading to TXA₂ production; activation of the integrin α_IIbβ₃ receptor; and release of granule contents (Chap. 112). Defects involving these and other processes can result in impaired platelet function.

DEFECTS IN PLATELET AGONIST RECEPTORS OR AGONIST-SPECIFIC SIGNAL TRANSDUCTION

Thromboxane A₂ Receptor Defect

Platelets contain two different isoforms of the TXA₂ receptor. Both forms activate PLC, but they differ in their effects on adenylyl cyclase, with one stimulating and the other inhibiting this enzyme.⁴⁴⁷ A mutation in the first cytoplasmic loop of the TXA₂ receptor (R601L) has been described as causing an inherited bleeding disorder in several families from Japan.^448,449 The platelets of these patients do not aggregate in response to TXA₂ mimetics. The aggregation defect also extends to other agonists, such as ADP, in which TXA₂ made by activated platelets and released into the surrounding medium, augments the response. The defect appears to be in signal initiation rather than ligand binding. TXA₂-induced activation of PLC (measured as Ca²⁺ mobilization, inositol trisphosphate, and phosphatidic acid formation) is impaired while PLA₂ activation and TXA₂ production is normal. Of note, the mutation appears to inhibit PLC activation by both receptor isoforms and impairs adenylyl cyclase stimulation by one of the isoforms; it does not, however, affect the inhibition of adenylyl cyclase produced by the other isoform. Both dominant and recessive inheritance patterns have been reported. The abnormal aggregation responses in heterozygous family members suggests a dominant negative effect of the mutation.⁴⁴⁹ Another report⁴⁵⁰ describes a heterozygous D304N substitution in the seventh transmembrane region of the TXA₂ receptor associated with a bleeding history, a 50 percent reduction in ligand binding and loss of receptor function. A heterozygous TXA₂R mutation (V2416) in the third intracellular loop has been reported⁷ in a subject without any bleeding symptoms. This subject had impaired aggregation and Ca²⁺ mobilization in response to U46619, normal platelet receptor levels and had aspirin resistance in microfluidic experiments assessing platelet deposition under flow.

Adenosine Diphosphate Receptor Defects (P2Y₁₂ and P2X₁)

Multiple receptors (P2Y₁₂, P2Y₁, and P2X₁) mediate ADP interactions with platelets (Chap. 112).⁴⁵¹ P2Y₁ receptors induce PLC activation, intracellular Ca²⁺ mobilization, and shape change, while P2Y₁₂ receptors mediate inhibition of cAMP formation by adenylyl cyclase. ADP-induced platelet aggregation requires activation of both P2Y₁ and P2Y₁₂ receptors. P2X₁ receptors function as an ATP- and ADP-gated cation channel (Chap. 112). Patients with P2Y₁₂ receptor abnormalities have blunted ADP-induced platelet aggregation responses, impaired suppression of prostaglandin E₁ (PGE₁)-induced elevations in cAMP, and normal ADP-stimulated shape change.^{452,453,454,455,456} Bleeding symptoms have been variable, with some demonstrating moderately severe hemorrhage in association with surgery and trauma. Because ADP released from platelets potentiates the responses to other agonists, such as collagen and TXA₂, platelet aggregation in response to these agonists is also abnormal in these patients. Platelet binding of ADP or the ADP analogue 2-methylthio-ADP^{452,453,454,456} was decreased in all but one patient studied.⁴⁵⁷ Decreased platelet 2-methylthio-ADP binding has also been reported in other patients with impaired aggregation and secretion in response to several agonists, including ADP.⁴⁵⁹

The genetic defects have been defined in some of these patients. In three patients, homozygous deletions have been demonstrated in the P2Y₁₂ gene, resulting in premature termination and a lack of P2Y₁₂ protein.^445,452,456 A homozygous missense mutation in the translation initiation codon was described in another patient,⁴⁵⁴ and another patient was reported to have a two-nucleotide deletion (at amino acid 240) in one P2Y₁₂ gene allele, resulting in a frameshift and a premature stop codon.^453,460 Although this last patient had one P2Y₁₂ allele with a normal coding region, the patient’s platelets lacked P2Y₁₂ receptors, suggesting repression of the normal allele or an unrelated abnormality in its transcriptional regulation. In contrast, platelets from the patient’s daughter had an intermediate number of ADP-binding sites, a normal platelet response to ADP, and one frame-shifted allele and one normal allele, suggesting that the mutant allele does not act in a dominant negative manner.⁴⁵³ Studies in yet another patient with abnormal ADP-induced aggregation revealed a compound heterozygous state with one allele containing an R256N substitution in the sixth transmembrane domain, and the other allele containing an R265W substitution in the third extracellular loop of the receptor.⁴⁵⁷ Platelet binding of ³³P-2MeS ADP was normal; neither mutation affected the translocation of the P2Y₁₂ receptor to the cell surface, but ADP-induced inhibition of adenylyl cyclase was partially reduced, indicating a functionally abnormal receptor. A heterozygous mutation (K174E) in the second extracellular loop of P2Y₁₂ was identified in one patient⁴⁵⁵; this was associated with decreased 2-methylthio-ADP binding. Another heterozygous mutation, P258T in the third extracellular loop has been described in association with a bleeding diathesis.⁴⁶¹ Interestingly, a heterozygous mutation in P2Y₁₂ (P341A) has been shown to induce altered interaction with Rab guanosine triphosphatases (GTPases) and endosomal trafficking of the receptor leading to its decreased surface expression.⁴⁶²

A defect in the P2X₁ purinergic receptor has been described in a 6-year-old patient with bleeding manifestations.⁴⁶³ The patient had isolated impairment of ADP-induced platelet aggregation and was heterozygous for a deletion of a single leucine in a stretch of four leucine residues^{351,352,353,354} in the second transmembrane domain of P2X₁. The mutant protein apparently caused a dominant negative effect on P2X₁-mediated calcium channel activity.

Epinephrine Receptor Defects

Abnormalities of α-adrenergic receptors or α-adrenergic–specific signal transduction have been described in several patients,^464,465,466 but the relationship to bleeding manifestations remains unclear, particularly because responses to epinephrine are blunted even in some otherwise normal individuals. Responses to epinephrine are blunted in the QPD.⁴³⁷

Platelet-Activating Factor Receptor Defect

A defect in the platelet-activating factor receptor or platelet-activating factor-specific signal transduction has been reported.⁴⁶⁷

GUANOSINE TRIPHOSPHATE-BINDING PROTEIN DEFECTS

GTP-binding proteins are a heterotrimeric class of proteins (consisting of α, β, and γ subunits) that link surface receptors and intracellular enzymes (Chap. 112). Abnormalities involving Gαq, Gαi₁, and Gαs proteins have been described.

Gαq Deficiency

Gαq plays a major role in mediating platelet responses to activation of G-protein–coupled receptors. One patient has been described with a selective platelet Gαq deficiency in association with a mild bleeding disorder, abnormal platelet aggregation and secretion in response to a number of agonists, and diminished GTPase activity (a reflection of Gα-subunit dysfunction) in response to platelet activation.^468,469 The downstream events from Gαq, including Ca²⁺ mobilization, release of arachidonic acid from phospholipids, and activation of integrin α_IIbβ₃ receptors, were impaired. The Gαq coding sequence in this patient was normal, but Gαq mRNA levels were decreased in platelets, suggesting a potential defect in transcriptional regulation of the gene. This abnormality appeared to be selective for platelets as the patient’s neutrophils had normal Gαq protein.⁴⁷⁰

Gαs Hyperfunction and Genetic Variation in Extra Large Gαs

Two unrelated families have been described with inducible hyperactivity of Gαs.⁴⁷¹ These patients had a bleeding diathesis, prolonged bleeding times, variable mental retardation, and mild skeletal malformations. Platelet aggregation responses to physiologic agonists were normal, but the platelets showed increased sensitivity to inhibition by agents (PGE₁, prostacyclin [PGI₂]) that elevate cAMP. Platelet Gαs, which when activated increases platelet cAMP levels and inhibits platelet aggregation and secretion, was increased in these patients. The Gαs gene (GNAS1) has multiple alternative promoters and isoforms as a result of alternative splicing, including extralarge Gαs (XLαs). XLαs is imprinted and thus normally only expressed from the paternal allele. A heterozygous 36-bp insertion and a 2-bp substitution were identified in exon 1 of the paternal XLαs gene in these patients. Because XLαs is not activated by the usual platelet Gαs-coupled receptors, the mechanisms leading to increased cAMP levels and enhanced expression of Gαs protein is unclear. Of note, 2.2 percent of control subjects also had the same polymorphism, but only those individuals inheriting it from their father had inducible Gαs hyperfunction and increased platelet Gαs protein.

Platelet Gαs deficiency has also been described in a patient with pseudohypoparathyroidism Ib in association with disturbed imprinting and altered methylation in the GNAS1 gene cluster that encompasses the four GNAS1 splice variants, including the Gαs subunit.⁴⁷² The Gαs coding sequence was normal. As expected from the deficiency in Gαs protein, there was decreased platelet cAMP formation upon activation of receptors linked to Gαs. The authors did not indicate whether the patient had a bleeding diathesis.

Gαi1 Deficiency

Platelet Gαi1 deficiency has been reported in association with a bleeding disorder and abnormalities in integrin α_IIbβ₃ activation, platelet aggregation, and dense granule secretion upon activation with one or more agonists.⁴⁷³ In keeping with the known function of Gαi in inhibiting the activation of adenylyl cyclase and the subsequent increases in cAMP levels, the patient’s platelets failed to inhibit forskolin-stimulated cAMP levels on activation. Platelet Gαi1 protein was decreased by 75 percent, whereas other members of the Gαi family (Gαi2, Gαi3, Gαiz) and Gαq were normal. Although a large subset of patients has been considered as having a defect in Gi signaling,⁴⁷⁴ this is based on abnormal responses on platelet aggregation and secretion studies to ADP and epinephrine, and direct evidence at the molecular level to support this conclusion has not been provided.

CalDAG-GEFI Deficiency

Three siblings from first cousin parents and severe mucocutaneous bleeding manifestations, had prolonged bleeding times and reduced platelet aggregation in response to ADP or epinephrine, and to low doses, but not high doses, of a thrombin receptor–activating peptide and collagen.⁴⁷⁵ Clot retraction was normal. Whole-exome analysis revealed a homozygous G248W mutation in RAS guanyl-releasing protein-2 (RASGRP2), the gene for the protein calcium- and diacylglycerol (DAG)-regulated guanine exchange factor-1 (CalDAG-GEFI), affecting the CDC25 catalytic domain critical for interacting with GTPases. The platelets demonstrated decreased Rap1 activation and fibrinogen binding and abnormal adhesion and spreading on immobilized fibrinogen and collagen. These results support a model of platelet activation in which CalDAG-GEFI stimulates GTP loading of Rap1 and Rac1 in response to an increase in intracellular Ca²⁺ and the activated Rap1 leads to integrin α_IIbβ₃ activation and the activated Rac1 enhances platelet spreading. The heterozygotes were asymptomatic but their platelets had a defective spreading.⁴⁷⁵

Phospholipase C–β₂ Deficiency and Defects in Phospholipase C Activation

Several investigators have described patients with relatively mild bleeding diatheses and impaired platelet aggregation and dense granule secretion, despite normal granule stores and the ability to synthesize TXA₂.^{445,446,476,477,478} An early event after stimulating several platelet G-protein–coupled receptors is activation of PLC-β, leading to formation of the intracellular mediators IP₃ and DAG (Chap. 112); the former mediates Ca²⁺ mobilization and the latter for PKC-induced protein phosphorylation. Defects in one or more of these responses has been documented in several patients. In one study of eight patients with abnormal platelet aggregation and secretion in response to several different receptor-mediated agonists, Ca²⁺ mobilization and/or pleckstrin phosphorylation was abnormal in seven patients, suggesting that the impaired secretion and aggregation resulted from upstream abnormalities in early signaling events.⁴⁷⁸ Specific defects at the level of PLC-β₂,^479,480 Gαq,⁴⁶⁸ and PKC-θ⁴⁸¹ were identified in these eight patients. In another study, eight patients were described who had decreased initial rates and extents of platelet aggregation in response to ADP, epinephrine, and the TXA₂ mimetic U44069⁴⁷⁶; subsequent studies in one patient demonstrated impaired phosphatidylinositol hydrolysis, phosphatidic acid formation, and pleckstrin phosphorylation.^482,483

In two related patients described with PLC-β₂ deficiency, platelet aggregation and secretion were impaired in association with impaired IP₃ and DAG formation, calcium mobilization, and pleckstrin phosphorylation following activation with ADP, collagen, platelet-activating factor, or thrombin, indicating a defect in PLC activation.⁴⁷⁹ These patients had a mild bleeding disorder. Human platelets contain at least seven PLC isozymes and a selective decrease was observed in only the PLC-β₂ isozyme.⁴⁸⁰ The decreased platelet PLC-β₂ protein levels were associated with a normal gene coding sequence but with diminished PLC-β₂ mRNA levels in platelets, but not neutrophils, suggesting a hematopoietic lineage-specific defect in PLC-β₂ gene regulation.⁴⁸⁴ Defects in phosphatidylinositol metabolism and protein phosphorylation have been described in other such patients, although the primary protein abnormalities were not defined.^{482,483,485,486,487,488}

Defects in Protein Phosphorylation: Protein Kinase C-θ Deficiency

PKC isozymes, a family of serine- and threonine-specific protein kinases, phosphorylate a wide array of proteins involved in signal transduction. PKC enzymes regulate several aspects of platelet function, including activation of integrin α_IIbβ₃ receptors, platelet aggregation and secretion, and platelet production (Chap. 112). Deficiency of a human platelet PKC isozyme (PKC-θ) has been described in a patient with lifelong mucocutaneous bleeding manifestations, mild thrombocytopenia, and markedly abnormal platelet aggregation (including primary wave) and dense granule secretion in response to multiple agonists.^481,489 Agonist-induced phosphorylation of pleckstrin and myosin light chain were diminished in the patient’s platelets. This subject was subsequently shown to have a heterozygous mutation in a transcription factor, RUNX1 (also termed core-binding factor A2, CBFA2, or AML1), which has been linked to a familial platelet function defect, associated with thrombocytopenia and predisposition to acute leukemia^481,490 (see “Transcription Factor Mutations and Associated Platelet Dysfunction” below). Platelet expression of myosin light chain (MYL9) was also decreased in this patient.⁴⁹¹

DEFECTS IN ARACHIDONIC ACID METABOLISM AND THROMBOXANE PRODUCTION

Defects in Arachidonic Acid Release from Phospholipids

Release of free arachidonic acid from phospholipids, mediated by cytosolic PLA₂, is the initial and rate-limiting step in thromboxane synthesis upon platelet activation. Several patients have been described with abnormalities in the release of arachidonic acid.^{469,485,492,493,494} In general, their platelets aggregated normally in response to arachidonic acid but not to ADP, epinephrine, and/or collagen. In one of these patients this defect was related to an upstream abnormality in Gαq (see “Gαq Deficiency” above).⁴⁶⁸ Another patient had the HPS with δ-SPD and abnormal PLA₂ activity.⁴⁹² An inherited deficiency in cytosolic PLA₂ has been reported in a patient with recurrent small intestinal ulceration, markedly decreased eicosanoid synthesis (including in thromboxane, 12-hydroxyeicosatetraenoic acid [12-HETE], and leukotriene B₄) and impaired aggregation with ADP and collagen but normal with arachidonic acid.⁴⁹³ This patient had two heterozygous single-base-pair mutations in the PLA₂ coding region, leading to S111P and R485H substitutions. Another report documents twins with a history of gastrointestinal ulcers, associated with similarly impaired platelet function associated with a homozygous D575H mutation in PLA₂ (PLA2G4A).⁴⁹⁴ These patients had mildly decreased plasma factor XI as well.

Cyclooxygenase (Prostaglandin H₂ Synthase-1) Deficiency

Deficient platelet cyclooxygenase (prostaglandin H₂ synthase-1) activity leading to impaired platelet function and a mild bleeding disorder has been identified in a number of patients.^495–502 Platelets from such patients cannot make thromboxane from arachidonic acid but can make it from cyclic endoperoxides (prostaglandin G₂ and prostaglandin H₂). While some patients have had decreased platelet cyclooxygenase protein, others have had evidence of a dysfunctional molecule.^501,502

Thromboxane Synthase Deficiency

Presumed platelet thromboxane synthase deficiencies have been identified in two families based on the failure of cyclic endoperoxides to be converted into TXA₂.^503,504

DEFINITION AND HISTORY

Activated platelets play an essential role in providing the membrane surface on which specific blood coagulation reactions occur leading to thrombin generation.^505,506 Patients whose platelets fail to facilitate thrombin generation are defined as having defects in platelet coagulant activity (PCA) (Chap. 112). Only a few patients have been described with isolated defects in PCA and normal aggregation and secretion responses.^505–513 Defects in PCA may also be secondary to abnormalities in platelet aggregation, such as in patients with SPD and thrombasthenia.⁵⁰⁵ Patients with isolated abnormalities in PCA are referred to as having the Scott syndrome after the first patient described in 1979 by Weiss and colleagues^{505,507,508,509,510}

ETIOLOGY AND PATHOGENESIS

The main functional abnormality in the Scott syndrome is the impaired ability of activated platelets to promote coagulation reactions; a second abnormality in these patients is a defect in release of microvesicles on cell activation.^514,515 In resting platelets, membrane phospholipids are asymmetrically distributed, with the aminophospholipids phosphatidylserine (PS) and phosphatidylethanolamine (PE) concentrated in the inner membrane leaflet, and phosphatidylcholine (PC) and sphingomyelin concentrated in the outer leaflet. Cell activation induces phospholipid translocation, with PS moving to the outer leaflet. This process is regulated by several proteins that are descriptively identified by their functions. They include a “flippase” (i.e., an aminophospholipid translocase identified as a P4 adenosine triphosphatase [ATPase]), which promotes inward transport of lipids; a “floppase” that regulates outward phospholipid transport (encoded by gene ABCC1); and one or more “scramblase” that promotes bidirectional movement of lipids between the two layers.⁵⁰⁶ Surface expression of PS is essential for platelets to accelerate coagulation reactions, in particular, activation of the tenase complex leading to the activation of factor X to Xa and of the prothrombinase complex that converts prothrombin to thrombin (Chaps. 113 and 114). Scott syndrome platelets have a defect in PS translocation resulting in decreased binding of factors Va-Xa and VIIIa-IXa,^505,506 thus leading to impaired blood coagulation. Erythrocytes and lymphocytes also demonstrate similar defects in both microvesicle formation and coagulant activity.^505,512 In a French family with Scott syndrome,⁵⁰⁸ the propositus’ platelets were found to have a defect in protein tyrosine phosphorylation, suggesting an additional defect in signal transduction.⁵¹⁰

Apart from the above patients with the Scott syndrome, four patients from three unrelated families have been reported to have abnormal PCA, a bleeding disorder, impaired serum prothrombin consumption, and reduced microparticle formation.⁵¹⁶ However, in contrast to the Scott syndrome, the prothrombinase activity was normal.

The inheritance pattern in the Scott syndrome appears to be autosomal recessive.^505,508 A heterozygous missense mutation has been reported in one patient in the gene ABCA1, which encodes the ATP-binding cassette transporter protein implicated in PS translocation; the significance of this remains unclear.⁵¹⁷ Mutations in TMEM16F have been identified in two patients with Scott syndrome. The patient Scott was found to have a homozygous mutation at the splice acceptor site for intron 12 of TMEM16F, resulting in a frameshift and premature termination of protein translation.⁵¹⁸ A second patient had compound heterozygosity for a mutation at the donor splice site of intron 6 and a single nucleotide insertion in exon 12, causing a frameshift and premature termination of translation.¹³ While these findings constitute strong evidence linking TMEM16F mutations to Scott syndrome, they leave open whether TMEM16F is itself the membrane scramblase or a protein that regulates the scramblase.

CLINICAL AND LABORATORY FEATURES

The bleeding symptoms in patients with the Scott syndrome are similar to those with other platelet defects. In contrast to other qualitative platelet abnormalities, the bleeding time in Scott syndrome patients is normal.^508,509,514 The serum PT, which reflects the completeness of clotting of whole blood and consumption of prothrombin, is abnormally, indicating incomplete coagulation.^507,508,509 More specific assays of “platelet factor 3,” the phenomenologic designation of the platelet contribution to accelerating clot formation, are also abnormal.⁵¹⁹

Patients with the Scott syndrome have normal platelet aggregation and secretion in response to the usually used agonists.⁵⁰⁵ The patient Scott⁵⁰⁵ also had normal platelet phospholipid content, normal to enhanced platelet adhesion to subendothelium with diminished thrombus formation, diminished factor Va binding to platelets and platelet microparticles, and diminished platelet acceleration of both factor X activation and prothrombin activation. Abnormalities in exposure of negatively charged phospholipids and shedding of microparticles have been consistent findings in all patients described.^{506,508,509,510,511,512}

TREATMENT

Platelet or whole blood transfusions have been effective as prophylaxis and as therapy for bleeding episodes.^{505,507,508,509} Prothrombin complex concentrates were effective in the patient Scott,³⁹² but these preparations may be associated with thrombotic complications.

Megakaryocytes and platelets express primarily and selectively the β₁ isoform of tubulin. A heterozygous β₁-tubulin Q43P polymorphism was identified in a group of patients with a recessive form of macrothrombocytopenia; the polymorphism could not account fully for the macrothrombocytopenia because of the difference in inheritance and its presence in approximately 11 percent of the normal population.⁵²⁰ Individuals heterozygous for the polymorphism had normal platelet counts, relatively high mean platelet volume values, abnormally rounded platelets with abnormal marginal bands of microtubules, and mild abnormalities of platelet aggregation, secretion, and adhesion to collagen. One study found the polymorphism was associated with decreased collagen-induced platelet aggregation and increased risk of intracerebral hemorrhage in men.⁵²¹

Subsequently, two patients with macrothrombocytopenia from a single kindred were reported to be heterozygous for a R318W β₁ tubulin mutation.⁵²² The mutation is strategically located at the α–β tubulin interface. Of note, the Q43P polymorphism and an R207H substitution were also found in this family, but neither was judged to be responsible for the macrothrombocytopenia.

Filamins are large dimeric actin-binding proteins that stabilize actin filament networks. Filamin A is the predominant platelet filamin. Several patients have been reported with dominant mutations of X-linked Filamin A (FLNA) gene associated with thrombocytopenia, and abnormalities in platelet aggregation, secretion, GPVI signaling and thrombus growth on collagen.^522A

WISKOTT-ALDRICH SYNDROME PROTEIN

The Wiskott-Aldrich syndrome (WAS) is an X-chromosome–linked inherited disorder characterized by small platelets, thrombocytopenia, recurrent infections, eczema, and an increased incidence of autoimmunity and malignancies.^523,524 In addition, a variety of immunologic abnormalities affecting T-lymphocyte function, Ig levels, cellular immunity, and responsiveness to polysaccharide antigens are commonly present. Death from infection, hemorrhage, or malignancy is common before adulthood. Some patients with WAS mutations may have only thrombocytopenia (X-linked thrombocytopenia [XLT]) without other features. WAS is caused by mutations in the WAS gene which encodes the WAS protein (WASP), a multidomain protein that relays signals from the cell surface to the actin skeleton and modulates the latter’s reorganization. In platelets, WASP is localized to the cytoskeleton.^524,525 It is phosphorylated on platelet activation by several different protein kinases, including Btk, Grb2, PLC-γ₂, PKC-θ, and SGK1.⁵²⁴

Microthrombocytopenia is a consistent feature of WAS and XLT and the major contributor to the bleeding diathesis, which may be life-threatening in some patients because of gastrointestinal hemorrhage or intracranial hemorrhage. Marrow megakaryocytes are normal in number, but platelet formation is abnormal and platelet survival is decreased.^523,524

Abnormalities in platelet surface glycoprotein sialophorin (CD43, gp115, leukosialin), GPIb, platelet integrin α₂, integrin α_IIbβ₃, and GPIV have been reported in some WAS patients.^526,527 Platelets from patients with WAS also have qualitative defects, including SPD^367,368,528 and impaired energy metabolism.^528,529

Platelet aggregation in WAS has been reported to be reduced, normal, or enhanced.^{524,530,531,532,533} Interpreting these studies is confounded by the low platelet count, methodological differences, and timing in relation to splenectomy.⁵²⁴ Despite the role of WASP in cytoskeletal reorganization, shape change and actin polymerization are normal in WAS platelets.^531,534,535

WASP has been implicated in regulating responses dependent on integrin α_IIbβ₃ outside-in signaling.⁵³⁶ Although Pac-1 binding is normal in WAS platelets, there is decreased spreading on fibrinogen and decreased clot retraction associated with enhanced PS exposure.⁵³⁶

Splenectomy usually improves the thrombocytopenia.^523,524 Hematopoietic stem cell transplantation is the accepted curative approach for WAS and can correct all aspects of the disease provided reconstitution is achieved.^523,524 Autologous gene-modified hematopoietic stem cell transplantation is an emerging therapy for WAS patients.^523,524

DEFINITION AND HISTORY/ETIOLOGY AND PATHOGENESIS

A syndrome with the features of both mild LAD-1 and GT was first described in 1997⁵³⁷ and termed LAD-1 variant or LAD-3. Since then more than 10 families have been reported, with several from Turkey.^{538,539,540,541} The etiology is a deficiency or defect in the cytoskeletal linking protein kindlin-3 (FERMTS3). Kindlin-3 is a protein expressed exclusively in hematopoietic cells with homology to talin that also binds to the cytoplasmic domain of the β₃ subunit of integrin α_IIbβ₃ (Chap. 112). It has been implicated in the inside-out activation of integrin α_IIbβ₃ in mice.⁵⁴² It also participates in the activation of leukocyte integrins, which accounts for the defects in immune function. It may also affect red blood cell structure. Defects in CALDAGGEF1, the gene for another exchange factor, also cause abnormalities in platelet function and are discussed in the section on Guanosine Triphosphate-Binding Protein Defects.^538,543

The disorder is characterized by a hemorrhagic diathesis in combination with a variable predisposition to infections and inflammation without pus formation, poor wound healing, delayed umbilical cord stump detachment, and variable osteopetrosis. Intracerebral hemorrhage at birth or soon thereafter has been reported in several of the patients, as well as relatively severe mucosal and gastrointestinal bleeding. Thus, the bleeding diathesis is more severe than is found in patients with GT, perhaps because of additional abnormalities in blood vessels. The need for red blood cell transfusions in infancy has been reported in several patients as a result of blood loss from mucosal surfaces and perhaps red blood cell abnormalities. Leukocytosis, as is found in other LAD syndromes is a constant finding. Normal platelet counts are the usual finding, but thrombocytopenia has been reported. Platelet aggregation studies demonstrate defects similar to those observed in GT.^541,542,543 Hematopoietic stem cell transplantation has been successful in restoring normal hematopoietic function in patients with life threatening hemorrhagic and infectious complications of the disease.⁵⁴¹

Transcription factors, and the cis-regulatory sequences to which they bind, regulate lineage-specific gene expression. Transcription factors RUNX1, FLI1 (a member of the ETS [E-twenty-six] family), GATA-1, and GFI1B (growth factor independent 1B) are important regulators of hematopoietic lineage differentiation, megakaryopoiesis, and platelet production.^419,420,544 They interact in a combinatorial manner in regulating megakaryocytic genes.⁵⁴⁴ A single transcription factor mutation may alter expression of numerous genes, affect diverse cellular mechanisms, and lead to defects in both platelet number and function. Until recently the pursuit of the molecular mechanisms in patients with platelet dysfunction has focused on delineating mutations in the genes encoding postulated candidate proteins. Therefore, the increasing spotlight on transcription factor mutations to explain platelet dysfunction is a paradigm shift.^9,545,545A

RUNX1 (FAMILIAL PLATELET DISORDER WITH PREDISPOSITION TO ACUTE MYELOGENOUS LEUKEMIA)

An association between inherited platelet dysfunction, thrombocytopenia, and a predisposition to acute myeloid leukemia has been reported in several families in which the platelet abnormalities antedated the leukemia.^{390,490,546,547,548,549,550,551} Inherited mutations in RUNX1 (AML1, CBFA2) is the basis for this constellation, which is inherited as an autosomal dominant trait, because of haploinsufficiency.⁴⁹⁰ Patients generally display mild thrombocytopenia from birth and a bleeding disorder disproportionate to the thrombocytopenia. Approximately one-third of patients develop leukemia, with a median age of onset of 33 years.⁵⁵²

Platelet abnormalities reported in patients with RUNX1 mutations include decreased aggregation, secretion, protein phosphorylation (myosin light chain and pleckstrin), production of 12-hydroxyeicosapentaenoic acid; decreased integrin α_IIbβ₃ activation upon platelet activation; δ and/or α-granule SPD; and a selective decrease in one PKC isoform (PKC-θ).^{372,390,553,554,555} Of note, several patients described earlier as having SPD (δ or α granules) have been subsequently shown to harbor RUNX1 mutations.^{358,372,390,391} In one patient, platelet albumin and IgG were diminished, suggesting a defect in the uptake and packaging of these proteins into α granules.^481,489

Most mutations of RUNX1 have been in the conserved Runt domain,³⁹⁰ although a mutation in the transactivating domain (Y260X) has been reported.³⁹⁰ Platelet transcript expression profiling in a patient with RUNX1 haploinsufficiency revealed downregulation of numerous genes involved in platelet structure and function, including MYL9 (myosin light chain), ALOX12 (12-lipoxygenase), PF4, and PRKCQ (PKC-θ).^{491,554,555,556} ALOX12, PRKCQ, PF4, and MYL9^554,555,556 are direct transcriptional targets of RUNX1. Patients with RUNX1 haploinsufficiency also have impaired megakaryopoiesis⁴⁹⁰ and decreased platelet thrombopoietin receptors (Mpl).⁵⁵¹ Targeted correction of RUNX1 mutation in induced pluripotent stem cells (iPSCs) developed from skin fibroblasts from two patients resulted in normalization of the defect in megakaryopoiesis.³⁹¹ These studies raise the potential for targeted gene therapy for this disorder. Using next-generation sequencing, Stockley and colleagues⁹ identified mutations in RUNX1 and FLI1, in six of 13 patients with excessive bleeding, and impaired aggregation and platelet dense granule secretion in response to multiple agonists. Thus, transcription factor mutations are an important mechanism for inherited platelet dysfunction.^545,545A

GATA-1

GATA-1 is a critical regulator of both megakaryocyte and erythroid development. GATA-1 mutations have been associated with an X-linked syndrome consisting of dyserythropoiesis, anemia, thrombocytopenia, and large platelets²; selectively impaired responses to collagen and ristocetin related to abnormalities in GPIbβ^557,558; diminished platelet Gαs protein and mRNA⁵⁵⁷; and a form of GPS (with R216N mutation).⁴¹⁸

FLI-1 (DIMORPHIC DYSMORPHIC PLATELETS WITH GIANT α-GRANULES AND THROMBOCYTOPENIA (PARIS-TROUSSEAU/JACOBSEN SYNDROME])

The Paris-Trousseau syndrome, a variant of Jacobsen syndrome, is a rare autosomal dominant disorder^{559,560,561,562} characterized by mental retardation, congenital macrothrombocytopenia, giant α granules (1 to 2 μm in diameter) in a subpopulation (1 to 5 percent) of circulating platelets, and marrow dysmegakaryopoiesis in association with deletion of the distal part of either the maternally or paternally derived chromosome 11 (11q23.3–24). Among the genes deleted is the transcription factor FLI1, which is important in megakaryocyte development via its effects on expression of several genes, including ITGA2, GPIX, GPIbα, and c-MPL.² Although platelet survival is normal, there is dramatic expansion of marrow megakaryocytes resulting from arrested megakaryocyte development. Thrombin-induced platelet release of α-granule contents is impaired.⁵⁵⁹ Both the inheritance pattern and the dimorphic population of normal and dysmorphic giant α granules are explained by the observation that during a period in early megakaryocyte development, only one of the two FLI1 alleles appears to be expressed in any single megakaryocyte precursor.^561,562

GFI1B

Two studies^419,420 implicate autosomal dominant mutations in GFI1B with a bleeding disorder and multiple alterations in platelet number and function, and red cell anisopoikilocytosis. In one study,⁴²⁰ the affected family members had a single nucleotide insertion in exon 7 of GFI1B leading to a frameshift mutation associated with macrothrombocytopenia, impaired platelet aggregation responses, α-granule deficiency, and decreased platelet P-selectin, fibrinogen, GPIbα, and integrin β₃. The second study identified⁴¹⁹ a dominant-negative truncating mutation (c.859C to T) in the zinc finger 5 region of GFI1B in a family originally reported in 1968 with macrothrombocytopenia and platelet dysfunction.⁵⁶³ The family members had decreased platelet α granules, PF4, βTG, and GP1bα, and deficiency of platelet factor 3. There was myelofibrosis and emperipolesis in the marrow.

The TAR syndrome is characterized by a reduction in platelet counts, absence of the radius bone in the forearm, skeletal abnormalities, and decreased marrow megakaryocytes. Dense granule SPD and impaired platelet aggregation and secretion have also been reported in TAR syndrome.³⁷¹ Early studies reported that majority of these patients have a deletion on chromosome 1q21.1,¹⁵ and later studies showed that these patients have both a rare null allele of RBM8A along with one of two low-frequency single nucleotide polymorphisms (SNPs) in the gene’s regulatory regions.¹⁴ RBM8A encodes for the Y14 subunit of the exon-junction complex (EJC), which plays an essential in RNA processing.

Platelet function abnormalities have also been reported in inherited connective tissue disorders such as osteogenesis imperfecta, the Ehlers-Danlos syndrome, and the Marfan syndrome^{369,370,564,565}; bleeding manifestations are more likely caused by the underlying connective tissue defect than by the platelet dysfunction. Abnormalities in platelet responses and or granules have been reported in patients with hexokinase deficiency,⁵⁶⁶ glucose-6 phosphatase deficiency (glycogen storage disease, type I),^567,568 and the Down syndrome.^{569,570,571,572,573} In glucose-6 phosphatase deficiency the platelet abnormalities were reversed following total parenteral nutrition for 10 to 12 days^567,568 indicating that the platelets may be intrinsically normal. The MYH9-related disorders (May-Hegglin anomaly) are characterized by giant platelets, thrombocytopenia, and basophilic granulocyte inclusions; some patients with this anomaly have platelet function and ultrastructural abnormalities.^8,574,575 Despite the large platelet size, the surface membrane glycoproteins appear to be normal.⁵⁷⁶ Markedly impaired platelet responses to multiple agonists have been reported with partial trisomy 18p associated with three copies of the PACAP (pituitary adenylate cyclase-activating polypeptide) gene and elevated plasma levels of PACAP, which induces increased platelet cAMP levels via stimulation of Gαs.⁵⁷⁷

Familial hemophagocytic lymphohistiocytosis (FHLH) is a genetic disorder of lymphocyte cytotoxicity associated with mutations in the gene encoding perforin or proteins important for vesicular trafficking and exocytosis. Flow cytometric analyses of the platelets of FHLH type 5 patients, who have MUNC18–2 (STXBP2) mutations, revealed that thrombin-induced secretion from both α and δ granules is impaired.^578,579 Platelets from an FHLH type 4 patient with a mutation in syntaxin-11 (STX11) also had a defect in agonist-induced secretion associated with normal cargo levels.⁵⁸⁰

Management of patients with inherited platelet function disorders needs to be individualized because of the wide variation in clinical manifestations, even in patients with the same defect. A general approach is described here; additional features specific to some individual entities are provided in their respective descriptions. Management of these patients involves preventive measures and treatment of specific bleeding episodes.^581,582,583 Dental hygiene is important in minimizing gingival hemorrhage. Antiplatelet agents should be avoided as they increase the bleeding manifestations. Iron and folate supplementation may be needed in patients with chronic hemorrhage. Hepatitis B vaccine should be administered early in life.

PLATELET TRANSFUSIONS AND GENERAL APPROACHES

Transfusion of platelets (Chap. 139) is a time-tested therapy for serious bleeding and as prophylaxis prior to surgery or invasive procedures. In addition to the usual risks associated with transfusions (transmission of infections, allergic reactions, Rh-immunization in Rh-negative individuals, and rarely hemolytic reactions) patients with GT and the BSS may develop specific antibodies against the missing glycoproteins, which may seriously compromise efficacy of future platelet transfusions.^581,582 This occurs particularly in patients whose platelets have no detectable integrin α_IIbβ₃.⁵⁸⁴ Therefore, platelet transfusions should be kept to a minimum. Transfusions of both platelets and red blood cells should be given with leukocyte depletion filters to decrease the risk of alloimmunization and cytomegalovirus transmission. It is reasonable to use human leukocyte antigen (HLA)-matched and ABO-matched platelets to minimize the risk of alloimmunization and side effects.^581,582,583

Treatment with 1-deamino-8-d-arginine vasopressin (DDAVP; desmopressin) may shorten the bleeding time and/or improve hemostasis in some, but not all, patients with platelet function defects.^{585,586,587,588} Responses to DDAVP appear dependent on the cause of the platelet dysfunction.^585,587,588 Most patients with thrombasthenia do not respond to DDAVP with a shortening of the bleeding time^{585,587,588,589} with exceptions,⁵⁹⁰ but it is unknown whether DDAVP improves hemostasis in these patients despite a lack of shortening of the bleeding time. Responses to DDAVP in SPD patients have been variable with a shortening of the bleeding time in some patients^588,591 but not others.^585,587 In uncontrolled studies it has been feasible to manage selected patients with inherited platelet defects undergoing surgical procedures with DDAVP alone.^585,587 However, this approach needs to be individualized based on the nature and location of the surgery and the intensity of the patient’s bleeding symptoms, and platelet need to be readily available for transfusion in the event of excess hemorrhage. The abnormal in vitro platelet aggregation or secretion responses in patients with platelet defects are usually not corrected by DDAVP.^587,592 It has been proposed that one mechanism by which DDAVP improves platelet function is via increased formation of procoagulant “COAT” platelets induced by combined activation by collagen and thrombin.⁵⁹²

Investigators have reported the successful use of recombinant factor VIIa (rFVIIa) in the management of bleeding events in patients with inherited platelet defects, including GT (including patients with antibodies to integrin α_IIbβ₃), the BSS and SPD.^{581,583,593,594,595,596} rFVIIa is thought to increase thrombin generation through both tissue factor dependent and independent mechanisms.

The antifibrinolytic agents epsilon-aminocaproic acid (EACA) and tranexamic acid,^581,582,583 which may be given orally, intravenously, or topically have been successfully used in patients with coagulation disorders and platelet function abnormalities.^586,597 Antifibrinolytic agents are useful in patients with gingival bleeding, epistaxis, and menorrhagia, and those undergoing dental extractions. A tranexamic acid mouthwash (10 mL of a 5 percent solution used four times daily) has been found effective in controlling gum bleeding and bleeding after tooth extractions.⁵⁸² A short 3- to 4-day course of prednisone (20 to 50 mg)⁵⁹⁸ has also been used.

Topical agents can also help arrest bleeding and have been used in GT patients. Gelfoam (a form of resolvable, oxidized, regenerated cellulose) soaked in either tranexamic acid or topical thrombin may be effective.⁵⁹⁹ Fibrin sealants prepared from a source of fibrinogen and a source of thrombin (exogenous or from the patient’s own plasma), with or without antifibrinolytic agents or other components⁵⁹⁹ have been used successfully in patients with GT.⁶⁰⁰ Some preparations of bovine thrombin induced antibody formation to itself and contaminated factor V and factor XI and have been associated with serious hemorrhage; recombinant human thrombin is now available and appears to have low immunogenicity.⁶⁰¹

Allogeneic marrow transplantation has successfully in treated patients with GT, BSS,⁵⁸¹ LAD-3,⁵⁴¹ and WAS.^523,524 Progress has been made in gene therapy approaches to correcting the genetic defect in GT in megakaryocytes.^602,603,604 Several GT animal models are available, including the integrin subunits α_IIb and β₃-null mouse models,⁶⁰⁵ and dog models involving mutations in the α_IIb subunit.^604,606,607 With respect to BSS, in mouse models, lentiviral transduction of hematopoietic stem cells with human GPIbα under the control of the integrin α_IIb promoter has been effective in improving hemostasis when transplanted into animals.²⁸⁹ Thus, as methods of marrow transplantation and gene transfer therapy improve, it will be important to reassess the risk-to-benefit ratios of these therapies for individual patients with GT.

BLEEDING AT SPECIFIC SITES

Control of epistaxis can be particularly difficult in some patients.⁶⁰⁸ When topical measures fail, platelet transfusions or factor VIIa should be considered. Nosebleeds occur primarily along the anterior nasal septum at the Kiesselbach area,⁶⁰⁸ posterior nosebleeds can occur either along the septum or the lateral nasal wall. Self-administered home therapy for anterior hemorrhage consists of pinching the outer aspect of the nose against the septum for 15 minutes to tamponade the septal vessels.⁶⁰⁸ If this fails, topical application by medical personnel of an anesthetic such as lidocaine in combination with a vasoconstrictor such as phenylephrine or oxymetazoline is commonly effective. Electrical cautery sometimes is effective when chemical cauterization fails. In many cases anterior or posterior packing may be needed for persistent severe epistaxis.

Menarche may be associated with the severe bleeding manifestations and require transfusions in some patients. Antifibrinolytics have been used for menorrhagia; hormonal therapy with progesterone alone or combined progesterone-estrogen is effective in those with persistent hemorrhage.^581,583

PREGNANCY

Management during pregnancy and delivery requires close interaction between the hematologist and the obstetrician. Most patients with severe bleeding symptoms, particularly those with GT and the BSS, will need platelet transfusions during childbirth and this need may continue for several days after delivery.^581,582,583 Postpartum bleeding may occur 2 to 4 weeks after delivery in some patients. rFVIIa has been used successfully in women with GT with persistent bleeding despite platelet transfusions, and those with integrin α_IIbβ₃ antibodies.^{581,582,583,609,610} Of note, fetal thrombocytopenia and intracranial hemorrhage may occur because of transplacental passage of the antibodies.

SURGERY

Management during surgical procedures needs to be individualized and depends on the bleeding history of the patient, the nature of the surgery, and on information such as alloimmunization and refractoriness to prior transfusions. Therapeutic options include DDAVP, platelet transfusions, rFVIIa and ancillary measures, such as antifibrinolytics, which may be continued for several days postsurgery.^581,582,583