Inherited Coagulation Disorders

Jerry S. Powell

George M. Rodgers

Inherited disorders of coagulation result from the deficiency or functional abnormality of one of the plasma proteins involved in providing normal coagulation.¹ There are other proteins that regulate the kinetics of coagulation, and abnormalities in one of these proteins usually result in increased risk of thrombosis; these proteins will not be covered in this chapter. If von Willebrand disease (vWD) is included, the inherited coagulation abnormalities (Table 53.1) are common; for example, vWD may affect up to 1% of the population. The other inherited coagulation disorders occur less often.

Since the purpose of the coagulation cascade is to rapidly produce fibrin to form a clot, the clinical manifestation of any deficiency or functional abnormality in any coagulation protein will be similar, i.e., clinical bleeding, with greater bleeding associated with more severe deficiency. Therefore, the manifestations and complications of clinical bleeding will be described in this chapter in the section on hemophilia A. A recent review of treatment of hemophilia A has been published with proposed algorithms for therapy.²

TABLE 53.1 INHERITED DISORDERS OF COAGULATION

X-linked recessive traits
	Hemophilia A
	Hemophilia B (i.e., CRM⁺ and CRM^– variants; hemophilia B_m, B Leyden, etc.)
Autosomal recessive traits
	Factor XI deficiency
	Prothrombin deficiency
	Factor V deficiency
	Factor VII deficiency
	Factor X deficiency (i.e., Prower variant, Stuart variant, Friuli variant, others)
	Afibrinogenemia
	Hypofibrinogenemia
	Factor XII deficiency
	Factor XIII deficiency
Autosomal dominant traits
	von Willebrand disease
	Dysfibrinogenemias
Combined abnormalities
	Associated with factor VIII deficiency (i.e., factor V deficiency, hemophilia B, factor XI deficiency, factor VII deficiency, von Willebrand disease, dysfibrinogenemias, platelet dysfunction)
	Involving vitamin K-dependent factors (i.e., factors II, VII, IX, and X; factors IX and XII; others)
Miscellaneous
	Prekallikrein deficiency
	High-molecular-weight kininogen deficiency
	Deficiency of physiologic inhibitors (i.e., α₂-antiplasmin, abnormal α₁-antitrypsin [antithrombin Pittsburgh])
CRM, cross-reacting material.

NOMENCLATURE

Historically, the best known inherited coagulation disorder is hemophilia, which dramatically manifested itself in the royal families of Europe in the 20th century, following a spontaneous mutation that apparently arose in Queen Victoria in England. The international Roman numeral designations for these disorders are summarized in Table 18.1. Like abnormal hemoglobins, qualitatively abnormal fibrinogens are designated by the name of the city in which they were first discovered, as in fibrinogen Paris.

PRINCIPLES OF PATHOPHYSIOLOGY

With the exception of fibrinogen and prothrombin, the coagulation factors are trace proteins. Clinical laboratory measurements of their activity depend on bioassays, which detect both quantitative absence of a specific factor and a qualitative functional abnormality in a factor that is present in normal amounts. Further laboratory testing is required to determine the actual concentration of the protein, usually by enzyme-linked immunoabsorbent assay (ELISA) methodology. Thus, for example, in hemophilia A, the bioassay measures factor VIII coagulant activity (factor VIIIc) and refers to the functional property of the factor VIII molecule that corrects the coagulation defect of patients with hemophilia A. The amount of protein measures the amount of antigen by ELISA and is referred to as factor VIIIAg (Table 53.2). Many hemophilia treatment centers currently proceed to determine the genotype for individual persons with inherited coagulation disorders; such analysis is helpful to predict the severity of the clinical phenotype, to predict the risk of inhibitor development, and to aid in family counseling.

Most qualitative abnormalities in coagulation factors result in loss of function of the protein in the coagulation cascade. However, some mutations produce qualitative abnormalities that result in “gain of function.” Such factors are called abnormal or aberrant factors to distinguish them. The most clearly defined disorders of this type are the dysfibrinogenemias, in which the abnormal fibrinogen typically is not completely nonfunctional, but may also inhibit the function of normal fibrinogen. Other disorders characterized by aberrant coagulation factors include the B_m variant of hemophilia B and some of the variants of prothrombin deficiency. Aberrant procoagulant proteins also may be synthesized in acquired deficiencies of the vitamin K-dependent factors (see Chapter 54). An example of a “gain of function” mutation for factor IX was described recently, factor IX Padua, in which a leucine was substituted for arginine at position 338 (R338L) and, despite a normal concentration of factor IX protein in plasma, the factor IX clotting activity was eight times normal, resulting in clinically significant thrombosis.³

TABLE 53.2 NOMENCLATURE FOR FACTOR VIII AND VON WILLEBRAND FACTOR

Definition	International Nomenclature
Protein lacking or aberrant in hemophilia A	Factor VIII
Functional property of factor VIII that is deficient in hemophilia A and measured using coagulation assays	Factor VIIIc
Antigenic property of factor VIII that is measured by immunoassays in which homologous antibodies are used	Factor VIII_Ag
Protein required for normal platelet adhesion that is aberrant or deficient in von Willebrand disease	von Willebrand factor (vWF)
Antigenic property of vWF that is measured by immunoassays in which heterologous antibodies are used	vWF antigen (vWFAg)
Property of vWF required for platelet agglutination by ristocetin	Ristocetin cofactor activity (vWF:RCo)
Note: This table is based on recommendations of the International Committee on Thrombosis and Haemostasis.
From Marder VJ, Mannucci PM, Firkin BG, et al. Standard nomenclature for factor VIII and von Willebrand factor: a recommendation by the International Committee on Thrombosis and Haemostasis. Thromb Haemost 1985;54:871-872; and Mazurier C, Rodeghiero F. Recommended abbreviations for von Willebrand factor and its activities. Thromb Haemost 2001;86:712.

HEMOPHILIA A

Hemophilia A refers to the inherited coagulation disorder characterized by deficiency of function of the coagulation protein called factor VIII. Except for vWD, it is the most common inherited coagulation disorder, and is X-linked. As the most common bleeding disorder, historians assume that it is the severe and often fatal hemorrhagic diathesis that affected the male children of certain families and that was well recognized in antiquity, as noted in the writings of Rabbi Simon ben Gamaliel (2nd century A.D.) in the Talmud, and those of Maimonides, the Hebrew physician and philosopher, and Albucasis, the Arab (twelfth century).⁴ Complete monographs have reviewed the early literature.⁵ However, the bleeding disorder that occurred in males throughout the royal families of Europe due to a mutation that arose with Queen Victoria, long assumed to be the more common hemophilia A, turned out, after extensive medical forensic investigation, to be hemophilia B.⁶

FIGURE 53.1. The relationship of factor VIII procoagulant activity and factor VIII-related antigen (von Willebrand factor [vWF]) in the plasma of normal subjects and patients with hemophilia or von Willebrand disease. (From Hoyer LW. The factor VIII complex: structure and function. Blood 1981;58:1-13, with permission.)

Pathophysiology

The hemostatic abnormality in hemophilia A (factor VIII deficiency, classic hemophilia) is a deficiency or functional abnormality of the plasma protein factor VIII. It normally circulates in the plasma bound to a much larger molecule, vWF. The half-life of factor VIII in the absence of vWF is much shorter, as seen in individuals with the vWF Normandy mutation, resulting in loss of the protein sequences for the binding interaction site between factor VIII and vWF, and causing low levels of factor VIII and clinical bleeding that appears similar to the bleeding seen in patients with hemophilia A. The production of vWF is coded by an autosomal gene, and this protein is qualitatively normal and is present in normal or increased amounts in patients with hemophilia A (Fig. 53.1).⁷, ⁸ The function of the coagulation protein factor VIII is to serve (when activated) as a cofactor for the serine protease enzyme factor IX, and accelerate the reaction as part of the coagulation cascade. No other function for factor VIII has been clearly characterized.

Incidence

Hemophilia A has been recognized in all areas of the world and in all ethnic groups. Estimates of its incidence approximate 1 in 5,000 males, or 1 in 10,000 persons.⁹, ¹⁰ A population-based study of the southeastern United States by the Centers for Disease Control and Prevention found that 11 cases per 100,000 persons had hemophilia A and that the prevalence was similar among different racial groups.¹¹ This same survey estimated that in 1994, there were 13,320 cases of hemophilia A in the United States.¹¹

Genetics

Hemophilia A is a classic example of an X-linked recessive trait. In such a disorder, the defective gene is located on the X chromosome¹²; the factor VIII gene maps to Xq28, on the distal end of the long arm of the X chromosome. In males who lack a normal allele, the defect is manifested by clinical hemophilia (Fig. 53.2; generation I, number 1). The affected male does not transmit the disorder to his sons (generation II, numbers 4 and 5) because
his Y chromosome is normal. However, all of his daughters are obligate carriers of the trait because they inherit his X chromosome (generation II, numbers 2 and 3). Most of these women are unaffected clinically because of the presence of a normal allele from the mother. The female carrier transmits the disorder to half of her sons (generation III, numbers 6 and 7) and the carrier state to half of her daughters (generation III, numbers 8 and 9). Due to Lyonization of the X chromosome, if the normal factor VIII allele is inactivated more often by chance, then some of the carrier females may have clinical bleeding. The severity of their bleeding depends on their factor VIII activity level; and, rarely, a woman can have very low factor VIII activity and present with symptoms of moderate or even severe hemophilia.

FIGURE 53.2. The inheritance of hemophilia A and hemophilia B. The pedigree is hypothetical. Squares indicate male; circles indicate female; fully shaded squares or circles indicate affected members; half-shaded circles indicate carriers. X, normal X chromosome; x, abnormal X chromosome.

The severity of bleeding varies in different kindreds, and closely depends on the particular genetic defect. Since the same gene defect is present in the kindred, each of the affected males will have very similar clinical phenotypes.¹³ As expected, hemophilia A has been observed several times in twins.¹⁴ Occasionally, however, the coinheritance of other genetic defects can influence the clinical symptoms of hemophilia A patients. An example is the coinheritance of the factor V Leiden mutation (discussed in Chapter 55) with hemophilia A. These patients have a milder clinical phenotype than expected for the same molecular defect in factor VIII.¹⁵ The coinheritance of the factor V Leiden mutation or other prothrombotic risk factors in children with severe hemophilia A may delay the first symptomatic bleeding event.¹⁶

Intriguing, and yet to be fully explained, is the fact that in over a third of families with hemophilia A, there is no evidence or history of abnormal bleeding in other members of the family.¹⁷, ¹⁸ This percentage is consistent with the Haldane hypothesis, which predicted that maintenance of a consistent frequency of a genetic disorder in the population would require that approximately onethird of cases result from spontaneous mutation. In other instances, neonatal deaths or the passage of the trait through a succession of female carriers may explain the negative family history. For practical purposes, therefore, a negative family history is of little value in excluding the possibility of hemophilia A. Mutations that cause hemophilia A originate in males at least threefold more often than in females.¹⁹ The large size of the factor VIII gene (186 kb), the presence of hot spots (e.g., CpG dinucleotides), and the fact that the X chromosome is unpaired in males may predispose for the factor VIII gene to undergo spontaneous mutation.²⁰, ²¹, ²², ²³

Nearly 3,000 individuals with hemophilia have had their factor VIII genes analyzed; this genetic information is collected in the Hemophilia A Mutation Database (HADB, formerly the Hemophilia A Mutation, Structure, Test and Resource Site[HAMSTeRS] database).²¹ The genetic defects of hemophilia A encompass deletions, insertions, and mutations throughout the factor VIII gene.²⁴ Point mutations involving CpG dinucleotides are especially common. Approximately 5% of patients with hemophilia A have large (>50 nucleotides) deletions in the factor VIII gene.²⁵

Approximately 45% of severe hemophilia A cases result from a major inversion of a section of the tip of the long arm of the X chromosome, one breakpoint of which is situated within intron 22 of the factor VIII gene.²⁶ This common inversion is associated with severe hemophilia A, and a higher incidence of inhibitor formation. It is presumed that in the absence of homologous X chromosome pairing during male meiosis, an intrachromosomal recombination event occurs on the single X chromosome, resulting in this inversion. This event typically occurs in the father of the mother of a child with severe hemophilia A, and often leads to relief of guilt feelings in the mother, as the genetic mutation is not “her fault.” Another common inversion (intron 1) accounts for 5% of patients with severe factor VIII deficiency.²⁷ Thus, two inversions of the factor VIII gene are seen in nearly one-half of all severely affected patients with hemophilia A. In contrast, point mutations are most likely found in patients with mild to moderate hemophilia A.²⁸ The first case of hemophilia A caused by unequal homologous Alu/Alu recombination has been reported.²⁹ Because there are ˜50 Alu repeats in the factor VIII gene, it is possible that many cases of hemophilia may result from this mechanism.²⁹ More than 95% of hemophilia A patients have mutations detected.²², ³⁰, ³¹ Approximately 2% of hemophilia A patients have no detectable mutations in the coding region of the factor VIII gene.³² These results indicate that a substantial number of families with severe hemophilia A can undergo accurate gene tracking and carrier analysis. The intron 1 and 22 inversions responsible for 50% of severe hemophilia A cases can be sought using long and inverse polymerase chain reaction (PCR) techniques. Various algorithmic approaches to rapid laboratory genetic testing have been described,³⁰, ³³, ³⁴ and a recent method of performing linkage analysis of hemophilia A in families in whom mutations could not be identified has been reported.³⁵

There is a strong association between genetic mutations that lead to absence of large portions of the factor VIII protein and the development of antibodies to therapeutic factor VIII infusions. These antibodies generally develop within the first 50 infusions of factor VIII, as expected when the immune system detects a protein to which it has not been exposed previously.

Variants

Hemophilia A with autosomal dominant transmission has been reported; however, it is important to distinguish these “variant hemophilia A patients” with negative X-linked transmission from patients with variant vWD (type 2N, vWD Normandy), an autosomal vWD subtype caused by defective factor VIII binding to vWF and with a clinical picture similar to hemophilia A.³⁶

Carrier Detection

Despite the high rate of spontaneous de novo mutations in hemophilia A, detection of females who are carriers is important. When the family history of a patient suggests risk for being a carrier, then coagulation-based assays are performed, followed by DNA testing. The daughter of a man with hemophilia is an obligate carrier, whether she has symptomatic bleeding or not. Similarly, if a female has two sons with hemophilia, then she is likely a carrier; if she has one son with hemophilia and a family history of hemophilia, then she is likely a carrier; but if she has one son with hemophilia and a negative family history, then the odds are approximately 67% that she is a carrier. A number of specialized laboratories perform molecular testing for the factor VIII gene. It is most efficient and less expensive if the DNA of the relative with hemophilia is available to identify the specific genetic mutation in that kindred; then only that specific mutation is evaluated in the potential carrier.³⁷

Coagulation-based Assays

Coagulation-based assays may be useful in confirming or excluding the carrier state, although they have limitations. The regularity with which the abnormal factor VIII gene is suppressed by the normal allele in female carriers of hemophilia varies because of the phenomenon of random X chromosome inactivation (the Lyon hypothesis). Thus, although the mean concentration of factor VIIIc in the plasma of heterozygous female carriers is ˜50% of that in normal women,¹³, ³⁸ observed values scatter widely around this mean and often overlap with those found in the normal population. This is a result, in part, of the large error of assay methods and the wide range of factor VIIIc levels in normal subjects.

Although the demonstration of low levels of factor VIIIc by means of the usual assay methods strongly suggests the presence of the carrier state, the converse statement cannot be made with equal certainty—that is, the presence of normal levels of factor VIIIc does not reliably exclude the carrier state.³⁹ Furthermore, pregnancy and hormonal medications may increase the levels of factor VIIIc in female carriers.⁴⁰ Normal women and carriers with blood types A, B, or AB have higher levels of factor VIIIc and vWFAg than those with blood type O.⁴¹

The use of immunoassays for vWF has improved carrier detection in hemophilia A.⁴² These methods allow measurement of levels of vWF, which are normal⁴³ or increased in carriers of the disorder, despite mild but variable deficiencies of factor VIIIc. Results obtained when a bioassay and an immunoassay are performed on the same sample, and the ratio of VIIIc to vWF is computed, differentiate between the carrier population and the normal population with minimal overlap.⁴⁴ This ratio normally ranges from 0.74 to 2.20 and was found to be from 0.18 to 0.90 in obligatory carriers. The overall detection rate ranged from 72% to 94% in such obligatory carriers⁴⁶ and from 48% to 51% in women without hemophilic sons or fathers.⁴² In the latter group (possible carriers), 50% would be predicted to be carriers. Abnormally low ratios of VIIIc:vWF (false positives) have been encountered in an occasional normal subject,⁴³, ⁴⁵ which may be attributable to nonspecific variables such as stress.⁴⁶ Pregnancy and the use of oral contraceptives, blood type,⁴¹ or contamination of plasma samples with thrombin⁴⁶ or other proteolytic enzymes may produce falsely high ratios (false negatives) in documented obligatory carriers. Linear discriminant analysis⁴⁷ has been recommended for the expression and analysis of such laboratory data. The carrier detection rates using standardized assays with discriminant analysis exceed 90%.⁴⁸

DNA-based Assays

Molecular analyses of the factor VIII gene identify mutations in 97% of hemophilia A patients. Based on the frequency of the intron 22 inversion, severe hemophilia A patients should be initially screened for this defect. Inversion-negative patients and those with mild or moderate hemophilia A should have systematic sequencing of the factor VIII promoter, exons, and splice junctions performed. Such an approach has been reported to identify mutations in 97% of hemophilia A patients.³⁰

DNA-based assays have been adapted for prenatal diagnosis,⁴⁹ and for pre-implantation genetic diagnosis in association with in vitro fertilization. One commonly used method is to obtain chorionic villi samples during the eleventh to twelfth gestational week and perform direct genotype testing.⁵⁰ Alternatively, genetic linkage analysis of polymorphisms can be performed. The subject of genetic counseling in hemophilia has been reviewed.⁵¹

Hemophilia in the Female

Hemophilia has been well documented in human females.⁴⁴, ⁵² The most common form is that seen in a minority of heterozygous carriers, discussed previously, in whom X chromosome inactivation may occur at an unusually early stage of embryogenesis, resulting in unusually low levels of factor VIII.

A second cause of female hemophilia is a mating between an affected male and a carrier female (Fig. 53.2; generation IV, number 10).⁵³ One-half of the female offspring of such a match would inherit two abnormal X chromosomes, one from the father and one from the mother. Such homozygous female hemophilia was once thought to be lethal and to inhibit the development of the embryo. That this is not true was first suggested by the successful experimental production of hemophilia in female dogs. Homozygous hemophilia is now well authenticated in several women⁵³, ⁵⁴ and resembles the disorder seen in affected males in all respects.

Clinical Manifestations of Hemophilia

Prior to effective replacement therapy, the life expectancy for a child born with severe hemophilia was less than 20 years. These boys would die from exsanguinating hemorrhage after a trivial traumatic injury, from spontaneous internal bleeding or from spontaneous intracranial hemorrhage. It is dramatic that a boy born today with severe hemophilia A can expect to have a normal life expectancy.

However, the most characteristic bleeding manifestations in hemophilia are spontaneous bleeding in weight-bearing joints, leading to severe hemarthrosis. The frequency and severity of these joint bleeds are related to the functional activity level of factor VIII in plasma⁸, ⁵⁵ (Table 53.3). Three categories of severity have been defined by a consensus committee on the basis of FVIII activity levels (Table 53.3).⁵⁶ Severe hemophilia (factor VIII level <1 IU/dl) is manifested clinically by repeated and severe hemarthroses, resulting almost invariably in crippling arthropathy in the absence of replacement therapy; such severe cases often are called classic hemophilia. Moderate hemophilia (factor VIII level of 1 to 5 IU/dl) is associated with less frequent and less severe hemarthroses and seldom results in serious orthopedic disability. In mild hemophilia (factor VIII level of 6 to 40 IU/dl), hemarthroses and other spontaneous bleeding manifestations may be absent altogether, although serious bleeding may follow surgical procedures or traumatic injury.⁵⁷ As indicated in Table 53.3, most patients with hemophilia A have severe disease. However, one epidemiologic survey reported variable differences in the incidence of severity between individual states within the United States; for example, 13% of hemophilia cases were of moderate
severity in Massachusetts, whereas 35% of cases in Georgia were classified as moderate severity.¹¹

TABLE 53.3 PREVALENCE AND SEVERITY OF HEMOPHILIA A AND HEMOPHILIA B IN THE UNITED STATES

		Incidence (%)^b
Factor VIII or IX Level (IU/dl)	Clinical Picture^a	Hemophilia A	Hemophilia B
<1	Severe, spontaneous bleeding	70	50
1-5	Moderate bleeding with minimal trauma or surgery	15	30
6-40	Mild bleeding with major trauma or surgery	15	20
^aThe criteria for classifying hemophilia severity are taken from White GC, Rosendaal F, Aledort M, et al. Definitions in hemophilia. Thromb Haemost 2001;85:560. ^bIncidence figures are for the United States in 1989 and were provided by the National Hemophilia Foundation (Courtesy of Kathleen F. Cortes, PhD).
From Rodgers GM. Common clinical bleeding disorders. In: Boldt DH, ed. Update on hemostasis: contemporary management in internal medicine. New York: Churchill Livingstone, 1990:75-120, with permission.

Hemarthrosis

Hemarthrosis is the most common manifestation of the inherited coagulation disorders (Fig. 53.3). Joint bleeding and the consequent damage to weight-bearing joints remain the most common debilitating symptom in severe hemophilia.

Pathophysiology

Bleeding presumably originates from the synovial vessels and develops spontaneously or as the result of imperceptible or trivial trauma. Spontaneous joint bleeding in the absence of identifiable mild trauma is unusual in patients with moderate hemophilia and very rare in patients with mild hemophilia. Hemorrhage occurs into the joint cavity or into the diaphysis or epiphysis of the bone. In the acute stage, the synovial space is distended with blood. Muscular spasm further increases the intrasynovial pressure. Hemorrhage into the periarticular structures is a common complicating feature that occurs most often around small joints. The inflammatory response to joint bleeding is highly variable and not well studied. However, in many patients permanent joint damage, detectable by MRI, appears to occur after only one or two minor bleeds into a single joint.⁵⁸

The joint may regain normal function after the first few episodes of hemarthrosis. More often, however, the absorption of intraarticular blood is incomplete, and the acute bleeding leads to chronic inflammation of the synovial membrane with the development of long-term damage to the joint. Sometimes, the joint remains swollen, tender, and painful for months, and in this setting may often suffer episodes of rebleeding. Such a joint is often termed a “target joint,” and should be treated aggressively with physical therapy and appropriate factor replacement to interrupt the vicious cycle of inflammation, swollen synovium, and recurrent bleeding. Release of inflammatory cell proteolytic enzymes initiates destruction of cartilage and bone.⁵⁹

Acute hemarthroses almost invariably recur from time to time. With each recurrence, the synovium becomes progressively more thickened and vascular; folds and villi, which predispose to synovial injury during even minimal activity, may form. Proliferating synovium often fills and distends the joint, which remains swollen and enlarged in the absence of bleeding or pain (chronic proliferative synovitis).⁶⁰ Together with the weakening of the periarticular supporting structures, this process predisposes the joint to recurrent episodes of bleeding. Repeated episodes of hemarthrosis, with the associated subchondral and synovial ischemia, result in progressive loss of hyaline cartilage, particularly at the margins of the joint. Large punched-out areas of destruction are sometimes produced by subchondral hemorrhages and, in the cancellous structure of the bone, cavitation may be caused by intraosseous hemorrhage. Through disuse, diffuse demineralization of the involved bones also may occur. Subperiosteal hemorrhages are not common.

FIGURE 53.3. Hemophilic arthropathy. This figure illustrates the sequelae of recurrent joint bleeding.

The terminal stage of hemarthrosis is called chronic hemophilic arthropathy.⁶¹ It is manifested by fibrous or bony ankylosis of larger joints; complete destruction may take place in the smaller articulations because of the weaker joint structure and the thinner cortices of the smaller bones. Other permanent sequelae of hemarthrosis include atrophy and proliferation of bone, roughening of the articular surfaces with lipping and osteophyte formation, bone necrosis and cyst formation, stunted growth as the result of interference with the nutrition of the bone, and accelerated development and overgrowth of the epiphyses caused by excessive blood flow (Figs. 53.3 and 53.4). Chronic hemophilic arthropathy is less common now because of widespread use of prophylactic replacement therapy and improved availability of factor replacement therapy.

Magnetic resonance imaging is superior to standard radiography for assessment of early arthropathy.⁶² A scoring system for magnetic resonance joint evaluation has been reported⁶² (Table 53.4).

It should be emphasized that with modern prophylactic therapy of regular factor infusions to maintain the plasma factor concentration above 1% of normal, boys are now reaching adolescence and young adulthood with essentially normally functioning joints. However, such a program of regular infusions with currently available factor therapies requires intensive medical and family support.

Clinical Presentation

Most persons with severe hemophilia report a characteristic warm, tingling sensation before the onset of symptoms of joint bleeding and hemarthrosis; this is called the aura. The earliest definite symptom is pain, which in the acute form may be excruciating. Physical examination reveals muscle spasm and limited motion of the affected joint. If therapy is not initiated immediately, typically within approximately 30 minutes, then the joint may progress so that it is warm and grossly distended and discolored,
but external evidence of bleeding may be minimal or absent in chronically damaged large joints because of thickening of the articular capsule. Generally, only one joint is involved at a time, although bleeding may develop simultaneously in two or more joints. The weight-bearing joints are most commonly affected and the knees are the joints most often severely affected. However, because of the success of total knee and total hip replacement surgeries, the ankles are the most commonly affected joints that lead to chronic problems that interfere with quality of life for persons with severe hemophilia.

FIGURE 53.4. Elbow and knee joints in a patient with hemophilia A. Thickening of synovium with deposition of calcium is shown in (A) and (A1); increased intercondylar notch is shown in (B); increased density and decreased interarticular space are shown in (A), (B), and (C); and lipping along the borders of the joint surfaces is shown in (C).

TABLE 53.4 SCORING SYSTEM FOR HEMOPHILIC ARTHROPATHY USING MAGNETIC RESONANCE IMAGING

Score	Abnormalities on Imaging
O	None
I	Minimal hemosiderin
II	Large amount of hemosiderin and cartilaginous erosion
III	Cartilage destruction, bone erosion, subchondral cysts
IV	Osteoarthritis with or without ankylosis

Subcutaneous and Intramuscular Hematomas

Large ecchymoses and subcutaneous and intramuscular hematomas were common in hemophilia A prior to the use of regular infusion therapy, reflecting the substantial amounts of time patients had factor levels less than 1% of normal. With modern treatment protocols designed to keep plasma levels above 1% of normal, and immediate home infusion when a bleeding episode occurs, the frequency of large soft tissue bleeds has decreased dramatically. Such hemorrhages, when not treated promptly, characteristically spread within fascial spaces and dissect deeper structures (see Fig. 45.2). Subcutaneous bleeding may be extensive. When not treated promptly with factor infusion, the bleeding continues; and at the site of origin, the tissue is hard, indurated, raised, and purplish black. From this center, the hemorrhage extends in all directions, with each successive concentric extension less deeply colored. The point of origin of the hemorrhage may be absorbed entirely while the margin is still progressing. Intramuscular and subcutaneous hematomas may produce leukocytosis, fever, and severe pain in the absence of significant discoloration of the overlying skin.

Hematomas may produce serious consequences from the compression of vital structures. Bleeding into the tongue, throat, or neck may develop spontaneously and is especially dangerous because it may compromise the airway with surprising rapidity.⁶³ Gangrene may result from pressure on arteries; and the development of compartment syndrome and, if not promptly treated, ischemic contractures are common sequelae of hemorrhage into the calves or forearms. Peripheral nerve lesions of varying severity are common complications of untreated hemorrhage into joints or muscles. Hemophilic cysts are discussed in the section Special Aspects of Treatment.

Psoas and Retroperitoneal Hematomas

Spontaneous hemorrhage into internal fascial spaces and muscles of the abdomen is common in severe hemophilia A,⁶⁴ reflecting plasma factor levels less than 1% of normal. Bleeding into or around the iliopsoas muscle produces pain of progressively increasing severity and tenderness; when it occurs on the right side, it may closely simulate acute appendicitis. Femoral nerve involvement may be partial or complete, with the development of pain on the anterior surface of the thigh. The psoas sign is positive, and the hip is held in partial flexion. Paresthesias, partial or complete anesthesia, and, ultimately, weakness or paralysis of the thigh extensors with eventual muscular atrophy may ensue. Retroperitoneal hemorrhage and intraperitoneal hemorrhage also are common. Computed tomography may be helpful in the diagnosis of these hematomas.⁶⁴ Prompt factor infusion is critical to normalize coagulation and should be continued until the hematoma has resolved completely, as any residual hematoma is at higher risk for rebleeding and pseudocyst formation.

Gastrointestinal and Genitourinary Bleeding

Hemorrhage from the mouth, gums, lips, frenulum, and tongue is common and often serious. The eruption and shedding of deciduous teeth usually occur without abnormal bleeding, but may occasionally be accompanied by hemorrhage that lasts for days or weeks if not treated. Epistaxis occurs in many patients and may be of exsanguinating proportions.

Hematemesis, melena, or both are not uncommon. The source of the bleeding is usually the upper gastrointestinal tract. In most patients in whom bleeding is persistent or recurrent, it originates from a structural lesion, most commonly a peptic ulcer or gastritis. Hemorrhage may be accompanied by abdominal pain, distention, increased peristalsis, fever, and leukocytosis. Intramural bleeding into the intestinal wall may result in intussusception or obstruction.

Hematuria is more common than gastrointestinal bleeding, but it is less often the result of a demonstrable pathologic condition in the genitourinary tract. The bleeding may arise in the bladder or in one or both kidneys and may persist for days or weeks.⁶⁵ When clots form, ureteral colic may develop. Typically, the hematuria resolves after factor infusion, but if persistent, then a short course of prednisone may prove helpful to shorten the course of hematuria.⁶⁶

Traumatic Bleeding

Patients with coagulation disorders seldom bleed abnormally from small cuts such as razor nicks, reflecting the normal function of platelets, as measured by platelet function tests. All laboratory assays of platelet function are normal in patients with severe hemophilia A. After larger injuries, however, hemorrhage out of proportion to the extent of the injury is characteristic. The bleeding reflects both increased acute bleeding rates beyond what is expected after the trauma, and persistent bleeding as slow continuous oozing occurs for days, weeks, or months. Such traumatic bleeding may be massive and life-threatening unless coagulation replacement therapy is provided by immediate factor infusion.

Delayed bleeding is common. Thus, although hemostasis after an injury or a minor surgical procedure may appear to be adequate, hemorrhage, often of sudden onset and serious proportions, may develop several hours or even days later. This phenomenon apparently occurs because the processes of primary hemostasis are only temporarily effective. Delayed bleeding may occur in patients with mild hemophilia and is a significant hazard after minor surgical procedures, particularly those performed on an outpatient basis, such as tooth extraction and tonsillectomy.

Venipuncture, if skillfully performed, is without danger for the person with severe hemophilia, primarily because of the elasticity of the venous walls. If venipuncture is traumatic, digital pressure on the puncture site or a pressure dressing may prevent further complications. Subcutaneous, intracutaneous, and small intramuscular injections seldom produce hematomas if firm finger pressure is maintained for at least 5 minutes. Large intramuscular injections should be avoided. Vaccination using intramuscular injections is acceptable with minimal increase in bleeding risks.

Other Clinical Aspects

Infants usually are asymptomatic because they are insulated from trauma.⁶⁷ However, trauma during birth or afterwards may trigger
life-threatening bleeding. Infants of women who are known carriers may be at risk during delivery, and instrumentation during vaginal delivery should be avoided. Currently, the usual standard of care is to avoid testing the infant in utero due to the risks inherent in current techniques available to test the fetus, and to allow normal vaginal delivery with back-up Caesarean section available if any difficulties arise during birth. Typically, hematomas are seen first when children become active, and hemarthroses seldom develop until they begin to walk. Occasionally, evidence of the disorder is not seen until patients reach teenage years or young adult life. Spontaneous hemorrhage may be cyclic in nature. Petechiae, which are characteristic of disorders of platelets and blood vessels, are rare in patients with hemophilia but have been noted in severely affected patients during an exacerbation of bleeding. Hemorrhage from the umbilical cord or stump is unusual, but prolonged bleeding after circumcision is common and brought hemophilia to the attention of the ancient Hebrews. Pulmonary and pleural bleeding are uncommon, although mediastinal and pleural shadows have been noted radiographically and presumably originate from fresh or old hematomas. Intraocular hemorrhage is uncommon, but bleeding into the orbit and conjunctiva occurs often. Spontaneous rupture of the spleen has been reported. Intracranial bleeding is discussed in the section Special Aspects of Treatment. The website for the National Hemophilia Foundation is www.hemophilia.org and contains useful information and links to more detailed information on this bleeding disorder.

Course and Prognosis

In recent years, the prognosis in severe hemophilia has improved dramatically. With modern therapy using regular factor infusions to maintain adequate plasma factor concentrations to prevent nearly all spontaneous bleeding episodes, a person born with severe hemophilia A can expect to live an essentially normal lifespan. No other genetic disease has made such dramatic progress in the past 50 years. It is not yet clear whether hemophilia protects older hemophilia patients from thromboembolic disorders,⁶⁸ atherosclerosis,⁶⁹ or cardiovascular diseases. Despite the lack of stringent prospective clinical studies for this emerging problem of aging patients with severe hemophilia and other medical problems, reviews of current practice and expert opinion on how to treat aging patients with severe hemophilia have been published.⁷⁰, ⁷¹, ⁷², ⁷³

TABLE 53.5 LABORATORY FINDINGS IN COMMON INHERITED COAGULATION DISORDERS

Disorder	Partial Thromboplastin Time (PTT)	Prothrombin Time	Thrombin Time	Ancillary Tests
Hemophilia A	A	N	N	vWF antigen and activities are normal or increased, ratio VIIIc:vWF is low.
Hemophilia B^a	A	N^a	N	—
von Willebrand disease^a,^b	A or N	N	N	vWF antigen and VIIIc are usually low, ratio VIIIc:vWF is variable; ristocetin-induced platelet aggregation and ristocetin cofactor activity are usually diminished.
Afibrinogenemia	A	A	A	Platelet function may be abnormal.
Dysfibrinogenemia^a	vA	vA	A^c,^d	Hypofibrinogenemia,^e reptilase time is prolonged,^a fibrin(ogen) degradation products levels are increased.^a
Hypoprothrombinemia^a	A	A	N	Two-stage assay is abnormal.^f
Factor V deficiency	A	A	N	—
Factor VII deficiency^a	N	A	N	Stypven (Russell viper venom) time is normal.
Factor X deficiency^a	A	A	N	Stypven time is abnormal.
Factor XI deficiency	A	N	N	—
Factor XII deficiency	A	N	N	—
Factor XIII deficiency	N	N	N	Clot solubility tests are abnormal.
A, abnormal; N, normal; v, variable; vWF, von Willebrand factor.
Certain PTT reagents may not detect mild deficiency of factors VIII, IX, and XI.
^aFindings are significantly different in some variants. ^bCoagulation abnormalities are caused by deficiency of factor VIIIc. ^cPatient’s plasma may inhibit normal coagulation. ^dAbnormality may be corrected by increasing calcium concentration and may be magnified by diluting the thrombin solution. ^eAbnormality varies depending on technique. ^fResults of one-stage techniques may be uninterpretable.

Laboratory Diagnosis

The activity of the various coagulation factors is expressed in terms of units, defined as the activity present in 1 ml of fresh plasma from normal donors. The concentration of all coagulation factors in normal pooled plasma is thus 1 IU/ml or 100 IU/dl, or 100% activity. It should be noted that the activity levels in blood bank plasma are somewhat lower, approximately 80 IU/dl because of the dilution with anticoagulant.

Routine clinical laboratory tests (noncoagulation) are normal in patients with hemophilia A. The presence or absence of anemia or of signs of blood regeneration depends on the severity and frequency of bleeding, as in any individual patient. Neutrophilia may accompany severe hemorrhage, and the destruction of red blood cells in the hematoma may be reflected in elevated LDH, AST, or bilirubin. As in other instances of posthemorrhagic anemia, the bone marrow reflects the response to blood loss.

Screening Tests of Hemostasis and Coagulation

The partial thromboplastin time (PTT) usually is prolonged in patients with hemophilia A (Table 53.5). The biochemical conditions for most laboratory assays of PTT are set so that the PTT is abnormal if the factor VIII level is <25% of normal; however, some PTT reagents are insensitive to mild factor VIII deficiency.⁷⁴ Therefore, in most situations the abnormality of the PTT can be
normalized by mixing the patient’s plasma with normal plasma in a ratio of 1:1; and this test, called a mixing study, is routinely used to detect whether neutralizing factor VIII antibodies are present. The platelet count usually is normal, but may be elevated in the stress reaction to hemorrhage.

Factor VIIIc Assays

Assay of factor VIII is a simple technique, but requires trained and experienced technicians, as the results are sensitive to errors in laboratory technique, or to errors in proper handling of the blood specimen. In clinical practice, if the laboratory results do not seem to fit the clinical picture, then most clinicians will repeat the laboratory test, and/or discuss the results with the laboratory personnel. Two-stage methods,⁷⁵ one-stage methods,⁷⁶ and micromethods⁷⁷ are suitable for diagnosis. The one-stage techniques are used most widely because they are simple to perform. They require either a supply of plasma from a known individual with severe hemophilia, which is available commercially, or artificial substrate plasma. Two-stage assays detect approximately 20% more factor VIII than do one-stage methods,⁷⁸ and they are less subject to variables⁷⁹ such as fluctuations attributable to contaminating traces of thrombin or other proteolytic enzymes.

The World Health Organization makes international standards, but many laboratories purchase commercial secondary standards that are calibrated to the international standard. Under most circumstances, a pool of citrated plasma carefully collected from normal subjects and frozen in individual laboratories also serves as an acceptable standard.

The factor VIII assay has a large potential for error, even in expert hands. Therefore, when borderline values are obtained, the assay should always be repeated. Extensive studies of the many variables of the factor VIII assay have been reported.⁸⁰, ⁸¹, ⁸², ⁸³

In the clinical setting of delivery of a boy with a question of hemophilia A, assays for both factor VIII and vWF may be carried out with reasonable accuracy on material obtained by fetoscopy—that is, mixtures of blood and amniotic fluid⁸⁴ or unmixed fetal blood.⁸⁵

The availability of a new recombinant form of factor VIII (B domain-deleted factor VIII, Xyntha) raised concerns about accurate laboratory monitoring of patients receiving this product.⁸⁶ Chromogenic substrate assays or typical one-stage PTT assays using physiologic phospholipids are needed to quantitate factor VIII levels accurately in patients receiving B domaindeleted factor VIII.⁸⁷ Alternatively, a standard one-stage PTT assay using B domain-deleted factor VIII as the assay standard would be appropriate. The use of concentrate standards (rather than plasma standards) diluted in factor VIII-deficient plasma has been proposed as a solution to resolve discordant results using different assay methods.⁸⁸

Analysis of appropriate methods for laboratory assays of factor VIII activity levels after therapeutic infusions of the newer factor VIII products with extended half-life will be an important part of the small phase III clinical trials initiated in 2012.⁸⁹ The Scientific Subcommittee on Standards of the International Society of Thrombosis and Haemostasis is involved in defining appropriate standards for these new products.

Factor VIIIAg (cAg) Assays

Highly specific but complicated assays using immunoradiometric methods and ELISA techniques⁹⁰ are available to measure factor VIIIAg (VIIIAg). The latter methods use liquid- and solid-phase principles and both natural factor VIII antibodies and monoclonal antibodies. The results of both of these techniques demonstrate the absence of factor VIIIAg in most severely affected hemophiliacs and correlate well with factor VIIIc measurements in most cases. Results are more variable when mild hemophiliacs are tested. In most series, ˜10% of patients have detectable levels of factor VIIIAg and a low VIIIc-to-VIIIAg ratio (the CRM-positive variant). Immunoassays for vWF are discussed in the section Laboratory Diagnosis.

Differential Diagnosis

The diagnosis of hemophilia A is seldom difficult, especially in the severely affected patient with repeated and often serious hemorrhagic manifestations, including such characteristic signs as hemarthrosis. Typically the PTT is abnormal and leads to specific assays for clotting factor activity and mixing studies to rule out neutralizing inhibitors of coagulation activity. Hemarthrosis with significant orthopedic disability is rare in patients with coagulation disorders other than hemophilia A or B.

In patients with mild forms of the disorder, however, failure to recognize the existence of the disease or to make the correct diagnosis is more likely. Such patients rarely have a history of spontaneous bleeding, and the family history tends to be vague or negative. A history of abnormal bleeding after minor trauma may be difficult to establish due to variability in bleeding. Occasionally, mild hemophilia A may present with a normal PTT when the factor VIII activity assay result is greater than 25%; thus a normal PTT may be misleading. It must be emphasized that individuals with mild hemophilia are still at risk for hazardous hemorrhage after trauma or during surgical procedures.⁹¹, ⁹² A survey of hemophilia carriers found that mild reduction in factor VIII or IX levels (41 to 60 IU/dl) was associated with excessive bleeding.⁹³ A normal PTT value, therefore, cannot be relied on to exclude the possibility of hemophilia. In the mildly affected patient, specific factor assays must be performed to confirm or exclude the diagnosis of hemophilia.

The results of screening tests (Table 53.5) usually are sufficient to exclude the possibility of acquired hemorrhagic disorders associated with serious bleeding. Such disorders are seldom associated with a prolonged PTT and a normal prothrombin time (PT), a combination that strongly suggests an inherited disorder or an inhibitor. Among the inherited disorders characterized by this combination of findings (hemophilia A, hemophilia B, and deficiencies of factors XI, XII, prekallikrein, and high-molecularweight kininogen [HMWK]), deficiency of the latter three factors can be readily excluded because their deficiency is not associated with excessive clinical bleeding. Factor XI deficiency in males may mimic mild hemophilia, and hemophilia B is clinically identical to hemophilia A. Both factor XI deficiency and hemophilia B must be distinguished from hemophilia A in the laboratory using specific factor activity assays. The best way to accomplish this goal in evaluating patients with an isolated prolonged PTT is by performing specific assays of these three factors in the order of their statistical frequency—that is, VIII, IX, and XI. A PTT mixing study should also be performed to exclude an inhibitor. A definitive diagnosis is of great importance because specific products are used to treat each of these disorders.

Severe vWD in males may be indistinguishable from mild hemophilia A. Confirmatory tests for vWD are needed to make this distinction. Patients with the uncommon type 2N vWD may also be clinically indistinguishable from patients with hemophilia A. This diagnosis is important to establish, because therapeutic infusions of recombinant factor VIII (r-FVIII) would be ineffective due to the shortened half-life of factor VIII resulting from the absence of effective binding to von Willebrand protein. An X-linked family history of bleeding supports a diagnosis of hemophilia A in these patients, and an autosomal recessive family history of bleeding supports a diagnosis of type 2N vWD. Statistically, one would expect a rare family that might carry both genetic disorders; clearly that situation may be problematic to resolve without additional genetic testing.

Bleeding manifestations in hemophilia may simulate a variety of conditions. However, serious confusion results only when the
correct diagnosis has not been considered and appropriate laboratory studies have not been ordered. Thus, a deep hematoma may be mistaken for a suppurative condition, and surgical drainage may be attempted. Bleeding into a small joint may produce a clinical and radiologic picture suggestive of sarcoma; when larger joints are involved, findings simulate tuberculosis, arthritis, or Perthes disease. Bleeding elsewhere may suggest local causes such as kidney tumor, pulmonary disease, or peptic ulcer. For example, blood clots due to hemophilia bleeding in the renal collection system might appear as masses on a CT scan and be confused with cancer, leading potentially to inappropriate invasive procedures to obtain tissue for cancer diagnosis.

Intra-abdominal bleeding raises particularly serious diagnostic and therapeutic problems in the patient with hemophilia, even when the hemophilia condition has been accurately diagnosed. Thus, hemorrhage into the psoas, when on the right side, may simulate acute appendicitis so closely that, in the opinion of many experienced clinicians, there is no reliable clinical means to differentiate between the two diagnoses. A retroperitoneal hematoma may be mistaken for an appendiceal abscess. Intraperitoneal hemorrhage and bleeding into and around other viscera may simulate perforating peptic ulcer, bowel obstruction, or virtually any acute intra-abdominal condition. Computed tomography scanning and sonography may be particularly helpful in differentiating between intra-abdominal conditions that require surgical intervention and retroperitoneal and psoas hemorrhages. In any case, early therapeutic infusion of factor VIII will prevent further bleeding and carries no additional risk for the individual with hemophilia. Therefore, when there is a question and further diagnostic tests are planned, factor VIII should be administered first and promptly before the diagnostic tests are performed.

HEMOPHILIA B

Hemophilia B was recognized as a separate disorder from hemophilia A in 1947.⁹⁴ Hemophilia B (Christmas disease, factor IX deficiency) was established as different from hemophilia A by Aggeler et al. in 1952.⁹⁵, ⁹⁶ Hemophilia A is approximately 5 times more common than hemophilia B.¹⁸ A 1994 survey estimated that there were 3,640 cases of hemophilia B in the United States.¹¹ As expected, it appears that the severity of hemophilia B is a consequence of the plasma factor IX concentration, and that for any particular plasma concentration, the clinical severity is similar for hemophilia A and B. However, in contrast with hemophilia A, most of the genetic causes of hemophilia B are not large deletions or inversions; therefore in many patients there is some antigenic protein, and often some low level of function of the mutated protein. Consequently, the clinical phenotype is often less severe. As with hemophilia A, hemophilia B is classified as severe, moderate, or mild based on the percentage of functional factor IX in coagulation activity assays. In addition, because of the presence of low levels of factor IX protein in more patients with hemophilia B, there are fewer persons with hemophilia B that develop neutralizing inhibitors (approximately 3% versus more than 25% in hemophilia A).

Historically, when plasma from patients with hemophilia B was tested against autologous antibodies, three distinct groups were defined: A cross-reactive material (CRM)-positive variant, the most common form; a CRM-negative variant; and a third form of the disorder in which antibody neutralization was variable in extent and was approximately proportional to coagulant factor IX activity (the CRM-reduced variant).⁹⁷ These distinctions are only of historical interest now.

In a disorder called hemophilia B Leyden,⁹⁸ the clinical manifestations tend to diminish during puberty in association with a rise in the factor IX level from as low as 1 IU/dl in childhood to levels of 20 IU/dl or more in adult life. The Leyden variant is characterized as CRM-negative at birth but becomes CRM-positive or CRM-reduced with advancing age.⁹⁸ The genetic basis for factor IX Leyden is that mutations occur in the factor IX gene promoter region; this region contains an androgen response element that, with age, stimulates factor IX gene transcription and protein synthesis.⁹⁹

In another variant of hemophilia B, the PT is prolonged when performed with ox brain thromboplastin.¹⁰⁰ This disorder has been called hemophilia B_m and is characterized by the presence in the plasma of CRM that neutralizes both autologous and heterologous antibodies to factor IX.¹⁰¹ The degree of abnormality of the ox brain PT is proportional to the plasma level of CRM in both affected males and carrier females. This variant results from mutations in the carboxy terminus of the factor IX molecule, resulting in a factor IX molecule that interacts abnormally with ox brain thromboplastin.

Numerous factor IX mutations have been recognized.²², ¹⁰² Factor IX purified from the plasma of affected members of one kindred did not fragment normally when activated in vitro (hemophilia B Chapel Hill).¹⁰¹ The molecular defect was found to be a substitution of histidine for arginine at position 145, a defect that inhibits cleavage by factor XIa.¹⁰³ Studies of factor IX obtained from another family¹⁰⁴ revealed biochemical abnormalities identical to those characteristic of the descarboxy analog of this factor that is found in vitamin K deficiency or produced by coumarin drugs. Such molecules lack Ca²⁺-binding sites and do not undergo conformational changes induced by Ca²⁺. Another report concerns a unique variant with deficient Ca²⁺ binding and an abnormally high-molecular-weight (factor IX Zutphen).¹⁰⁵

Genetics

Hemophilia B is inherited as an X-linked recessive trait, but the locus on the X chromosome of the gene controlling factor IX production is remote from that involved with factor VIII biosynthesis.¹⁰⁶ Factor IX levels <10% have been documented in a few women, including some with chromosomal abnormalities.¹⁰⁷, ¹⁰⁸, ¹⁰⁹ Ten useful polymorphisms have been described that are associated with the factor IX gene.⁴⁹ An updated listing of mutations is available at www.kcl.ac.uk/ip/petergreen/haemBdatabase.html. Unlike hemophilia A, the spontaneous mutation rate is low,¹⁷ and most patients with hemophilia B have positive family histories.

Factor IX defects may be severe, moderate, or mild. Severe defects result from large gene deletions, nonsense mutations, and inversions; these defects are associated with absence of factor IX protein. Milder gene defects such as splice-site or missense mutations result in a dysfunctional protein with some residual activity. Missense mutations account for ˜80% of mutations in severe hemophilia B patients.²², ¹¹⁰, ¹¹¹, ¹¹²

Detection of Carriers

Detection of heterozygous carriers of hemophilia B involves the same principles and limitations as described for hemophilia A. Carrier detection based on coagulation assay alone usually is more reliable than is the case with hemophilia A.¹¹³ Thus, in one series of 45 obligatory carriers, the mean factor IX level in the plasma was 33 IU/dl; 40 from the group had levels <50 IU/dl, and 10 had levels <25 IU/dl. As a consequence of these low levels of factor IX, abnormal hemorrhage is not uncommon in carriers of hemophilia B.

Results of immunoassays of factor IX and the ratio of factor IX-related antigen to coagulant factor IX levels in the carrier population overlap with the normal population to a considerable degree, particularly in the CRM-positive variants.¹¹⁴, ¹¹⁵ Such studies provide a lower overall detection rate among obligatory carriers of hemophilia B than is the case with hemophilia A.¹¹⁴, ¹¹⁵ Levels of factor IX-related antigen obtained by immunoassay
are significantly increased by the use of oral contraceptives. The use of DNA probes¹¹⁶ and monoclonal immunoassays¹¹⁷ provide highly accurate methods for determining carrier status. Rapid carrier testing using allele-specific microarray methods has been described. This method permits detection of a majority of common factor IX mutations in a single assay.¹¹⁸ Prenatal diagnosis of hemophilia B has been successful in the CRM-negative variants.¹¹⁹ Modern methods for prenatal diagnosis use genetic testing and are quite reliable.¹²⁰, ¹²¹ Determination of carrier status is based on several factors, including pedigree analysis, factor IX assay results, and genotype.¹²²

Clinical Features

Severely affected patients (those with factor IX activity levels <1 IU/dl) are less common than in hemophilia A (Table 53.3), but the clinical manifestations of the two disorders are identical for the same activity levels of the coagulation factors. Specific factor assays are necessary to distinguish between hemophilia A and hemophilia B. Mild factor IX deficiency should always be considered in the differential diagnosis of patients with coagulation-type bleeding and normal routine coagulation test results (PT, PTT).¹²³ Many PTT reagents do not detect mild factor IX deficiency (factor IX levels of 20% to 30%).⁷⁴

Laboratory Diagnosis

The laboratory diagnosis of hemophilia B involves the same approach and methods as those described for the recognition of hemophilia A (Table 53.5). The screening tests reveal similar abnormalities in the two disorders, except that the PT is abnormal in the B_m variant when either ox brain or Thrombotest is used as the thromboplastin.

Hemophilia A may be easily distinguished from hemophilia B by a specific factor assay. One-stage and two-stage assays for factor IX use the same principles as those discussed for factor VIIIc. Sensitive immunoassay methods for factor IX have been developed.¹²⁴

VON WILLEBRAND DISEASE

The historical confusion that has surrounded the pathogenesis of vWD is apparent from the many names that have been applied to this disorder. These designations include angiohemophilia, vascular hemophilia, pseudohemophilia, constitutional thrombopathy, and idiopathic prolonged bleeding time. von Willebrand first recognized the disorder in a 1926 study of the inhabitants of the Åland Islands.¹²⁵ Three cardinal manifestations of this bleeding disorder are mucocutaneous hemorrhage rather than hemarthrosis and deep-muscle bleeds; autosomal dominant inheritance, rather than sex-linked as seen in hemophilia A; and prolonged bleeding time. Evaluation of one large kindred revealed variable severity of bleeding symptoms, with some obligate heterozygote carriers being asymptomatic. von Willebrand thought that the hemostatic defect resulted from combined defects in platelet function and vascular endothelium. The discovery that plasma from normal volunteers or patients with hemophilia A corrected the hemostatic defect of vWD suggested that a plasma protein distinct from factor VIII caused vWD. In 1972, von Willebrand factor (vWF) was purified. Genetic variants were correlated with various vWF structural abnormalities, and demonstration of the heterogeneous nature of vWD was furthered by development of vWF multimer analysis by gel electrophoresis.¹²⁶ In its most common form (type 1), the disorder is characterized by mild mucocutaneous hemorrhage, which is attributable to deficiency of both vWF and factor VIIIc. The disorder is not homogeneous, however, in part because of the multiple physiologic functions played by vWF. The recognition of a number of variants, together with the demonstration of multiple genetic patterns, suggested that vWD is very heterogeneous. However, all forms of vWD can be traced to insufficient amounts of vWF or defective function of this protein. Recent reviews of this area have been published,¹²⁷, ¹²⁸ and a consensus document on vWD sponsored by the National Institutes of Health reviewed all aspects of the disorder.¹²⁹, ¹³⁰

Incidence

Epidemiologic studies indicate that vWD is the most common bleeding disorder, affecting about 1% of the population.¹³¹ The high incidence of the disease is not limited to certain ethnic groups ¹³¹, ¹³²; however, only a fraction of people come to medical attention because of bleeding symptoms. This may be because of either the relatively mild nature of the disease in many affected individuals, or a lack of recognition by patients of excessive bleeding in response to either physiologic challenge (e.g., heavy menstrual bleeding) or trauma.

Nomenclature

The identification of numerous variants of vWD has led to attempts to simplify classification of this disorder. An updated classification system has been proposed (Table 53.6).¹²⁸ Quantitative defects are divided into partial deficiency (type 1) and severe deficiency with virtually complete absence of vWF (type 3). The qualitative defects (type 2) are divided into four categories according to the nature of the defect of vWF function. Type 2A refers to variants with impaired interaction between vWF and platelets, resulting from a deficiency of intermediate- and high-molecular-weight multimers of vWF. Type 2B refers to variants in which vWF exhibits increased affinity for its receptor, platelet GPIb. Paradoxically, bleeding in these patients develops as a result of clearance of larger vWF multimers and platelets from the circulation. Type 2M refers to variants with defective interaction between vWF and the platelet GPIb receptor that is not due to deficiency of high-molecular-weight multimers of vWF from plasma, but rather results from defects within the GPIb-binding domain of vWF. Finally, variants of vWD in which decreased affinity of vWF for factor VIII results in depressed plasma factor VIII levels are classified as type 2N.¹²⁸ Thus, six major categories of vWD are defined, each having distinct pathophysiology.

TABLE 53.6 REVISED CLASSIFICATION OF VON WILLEBRAND DISEASE

Revised Type	Features
1	Partial deficiency of vWF
2	Qualitative vWF defects
2A	vWF variants with loss of high-molecular-weight multimers and decreased vWF-dependent platelet adhesion
2B	vWF variants with loss of high-molecular-weight multimers caused by increased affinity for platelet glycoprotein Ib
2M	vWF variants with decreased vWF-dependent platelet adhesion not associated with the loss of high-molecular-weight multimers
2N	vWF variants with decreased binding affinity for factor VIII
3	Severe deficiency of vWF
vWF, von Willebrand factor.
Data from Sadler JE, Budde U, Eikenboom JC, et al. Update on the pathophysiology and classification of von Willebrand disease: a report of the Subcommittee on von Willebrand Factor. J Thromb Haemost 2006;4:2103-2114.

Genetics

A gene on chromosome 12 codes for the synthesis of the vWF macromolecule. The genetic message is composed of 52 exons covering a span of 158 kb.¹³², ¹³³ The vWF gene encodes for a protein with multiple copies of homologous motifs, including three “A,” three “B,” two “C,” and four “D” motifs (see Fig. 53.5). These motifs in turn encode protein domains that subserve the various functions of vWF. The A1 domain contains binding sites for platelet GPIb, ristocetin, and collagen,¹³⁴, ¹³⁵ the A2 domain contains a protease-sensitive domain that may have a role in regulating vWF function,¹³⁶ the A3 domain contains a second collagen-binding domain,¹³⁵ the C1 domain has an RGD sequence capable of interacting with platelet glycoprotein IIb/IIIa, and the D’ and D3 domains contain a factor VIII-binding sequence.¹³⁵

vWD appears to be inherited by multiple genetic mechanisms (Table 53.7). The most common form of the disorder (type 1 vWD) accounts for ˜70% of all cases of vWD. Diagnostic criteria for type 1 vWD have been proposed,¹²⁸ including a history of bleeding symptoms, a positive family history of bleeding symptoms, and bleeding that is attributable to quantitatively low vWF levels. Population screening suggests that the prevalence of vWD may be as high as 1% of the population, but only a minority of these patients subsequently present with clinically significant bleeding.¹³⁷ vWD is inherited as an incompletely dominant autosomal trait with variable penetrance, even among members within a single kindred. Because of variable penetrance and expression of vWD, only ˜33% of children may be affected. Figure 53.6 illustrates one of the original pedigrees described by von Willebrand.

Twin studies demonstrated that ˜60% of the variation in vWF level and ˜50% of the variation in factor VIII level is attributable to genetic factors.¹³⁸ Genetic loci outside of the vWF gene contribute to variation of vWF level. Individuals with blood group O have vWF levels that are on average 30% lower than those of people with blood group A, B, or AB,¹³⁹ and it is likely that carbohydrate groups attached to vWF play a subtle role in clearance of vWF from plasma.¹⁴⁰, ¹⁴¹ In addition, variation in expression of levels of other hemostatically active proteins, such as variation in the level of platelet adhesion receptors, may modulate the severity of symptoms conferred by vWF deficiency.¹⁴² vWF levels are under hormonal and other controls, which further complicates disease expression. Variation in levels of as much as 20% has been reported with menstrual cycle, with levels lowest in the early follicular phase (before day 7 of the cycle); and levels increase with age, rising ˜15% for each decade increase in age.¹⁴³ As will be discussed later, the criteria for diagnosing quantitative vWD deficiency, and the specificity of historic bleeding, have introduced controversy in what constitutes type 1 vWD and have led to the potential for the new diagnostic entity of “low vWF”.¹³⁷, ¹⁴⁴

Mild to moderate quantitative deficiency of vWF characterizes patients with type 1 vWD, whereas virtual absence of vWF characterizes type 3 disease. Two large studies of families with a diagnosis of type 1 vWD and genetic evaluation of patients with type 3 disease have begun to shed light on the genetics of quantitative vWF deficiency.

Very low vWF levels in the range of 5 to 20 IU/dl tend to be highly inheritable and are frequently associated with “dominant negative”-acting mutations. Mutations in these patients with type 1 vWD appear to be different than those with type 3 disease, such that type 1 disease is not simply explained by being heterozygous for a type 3 allele. Genetic defects associated with these more severe type 1 phenotypes encode amino acid changes that in turn lead to defects in protein expression. Mechanisms responsible for decreased plasma vWF level include reduced secretion related to impaired intracellular transport of vWF subunits, increased clearance from the circulation, or accelerated catabolism as a result of increased susceptibility to degradation by ADAMTS13. Changes that were most likely to cause a “dominant negative” mechanism
are often missense mutations that change the number of cysteine residues in the vWF protein. Although such changes probably affect vWF synthesis and trafficking in the cell, at least one common mutation (Tyr1584Cys) also results in enhanced degradation by ADAMTS13.¹⁴⁵, ¹⁴⁶ One non-cysteine-related mutation is seen in vWD Vicenza. In these patients, a mutation encoding for Arg1205His results in accelerated clearance of plasma vWF, with levels of vWF in the range of 15 IU/dl. With DDAVP stimulation, timed plasma sampling revealed that the released vWF had a circulation half-life that was approximately one-sixth of normal.¹⁴⁷ Several other similar mutations have recently also been described.¹⁴⁸

FIGURE 53.5. A: Structure and functional domains of von Willebrand factor (vWF). The open rectangles indicate domains of pre-pro-vWF. The first 22 amino acids of pre-pro-vWF are the signal sequence. The left arrow identifies the site of cleavage of the signal peptide, resulting in pre-vWF monomers. Pro-vWF monomers form dimers, and the vWF propeptide (vWAgII) is cleaved from mature vWF (right arrow indicates cleavage site). Dimer S-S and multimer S-S indicate locations of disulfide bonds that covalently link monomeric vWF into dimers and dimeric vWF into multimers, respectively. The red rectangle identifies exon 28, which encodes the A1 and A2 domains of mature vWF. Functional domains are identified above the open rectangles. The factor VIII binding site is located in the D’and D3 domains in the amino-terminal 272 amino acids of mature vWF. The A domains of vWF contain the binding site for the platelet glycoprotein (gp) Ib/IX complex, as well as binding sites for heparin, collagen, and sulfatides. The gpIIb/IIIa complex binding site of activated platelets is located at the carboxy-terminal portion of the C1 domain and contains the Arg-Gly-Asp sequence. The numbers below the open rectangles indicate amino acid numbers: Signal peptide, 22 amino acids; propeptide (vWA-gII), 741 amino acids; mature vWF, 2,050 amino acids. B: Mutations associated with type 2A von Willebrand disease (vWD). The open rectangle represents vWF exon 28 encoding domains A1 and A2. The arrow indicates the site of proteolysis seen in type 2A vWD. Mutations are identified using the single-letter code for amino acids. I and II identify group I and group II missense mutations that result in impaired synthesis of vWF or increased sensitivity to proteolysis in plasma, respectively. The disulfide bridge shown represents the boundary of the Cys 509-Cys 695 loop in the A1 domain. Type 2A vWD mutations associated with defective vWF multimerization are not shown in this figure; these latter mutations occur in the propeptide (vWAgII) D1 and D2 domains. C: Mutations associated with types 2B and 2M vWD. The large open rectangle represents exon 28, which encodes domains A1 and A2 of vWF. The mutations responsible for type 2B vWD are shown above the rectangle. Most type 2B mutations lie between residues 540 and 578 of mature vWF. The mutations responsible for type 2M vWD are shown below the rectangle. Asterisks identify mutations that reduce the platelet-dependent function of vWF and vWF multimeric size. δ indicates deletion of an amino acid sequence. [Modified from the Symposium Proceedings of the National Hemophilia Foundation’s 1995 Annual Meeting. Diagnosis and management of severe von Willebrand disease (types 2 and 3). Kroner PA, Montgomery RR. The molecular basis of von Willebrand disease, 1996:15-25. Published with permission of Advanstar Communications.]

TABLE 53.7 FEATURES OF COMMON VARIANTS OF VON WILLEBRAND DISEASE

Features	Type 1	Type 2A	Type 2B	Type 3	Platelet Type
Inheritance	Autosomal dominant	Autosomal dominant	Autosomal dominant	Autosomal recessive	Autosomal dominant
Factor VIIIc in plasma	Normal or reduced	Normal or reduced	Normal or reduced	Markedly reduced	Normal or reduced
vWF antigen	Normal or reduced	Normal or reduced	Normal or reduced; increased affinity for platelets	Markedly reduced	Normal or reduced; increased affinity for platelets
Ristocetin cofactor activity	Normal or reduced	Reduced	Normal or reduced	Markedly reduced	Reduced or normal
vWF multimeric analysis	Normal (plasma and platelets)	Absence of large and intermediate-sized multimers in plasma	Absence of large multimers from plasma; normal in platelets	Small multimers or absent multimers in plasma and platelets	Reduction in large multimers caused by “consumption” by platelets
Ristocetin-induced platelet aggregation	Normal or diminished	Diminished	Increased aggregation at low ristocetin concentrations	Markedly diminished	Hyperaggregation with patient’s platelets, normal plasma, and low concentration of ristocetin
vWF in platelets	Normal or reduced	Normal or absence of large and intermediatesized multimers	Normal	Absent	Normal
Ancillary findings	DDAVP usually produces significant increase in plasma VIIIc and vWF	DDAVP produces rise in factor VIIIc, but functional vWF increase is variable and may be of short duration	Variable response to DDAVP, with intravascular platelet aggregation and thrombocytopenia in some cases; ristocetin-induced platelet aggregation enhanced in presence of patient’s plasma; cryoprecipitate does not aggregate platelets in vitro unless ristocetin is added	Response to DDAVP lacking; endothelial vWF absent	Transfusion of vWF or DDAVP may produce intravascular platelet aggregation and thrombocytopenia; cryoprecipitate produces in vitro platelet aggregation
DDAVP, 1-deamino-8-d-arginine vasopressin; vWF, von Willebrand factor.

At higher levels of vWF, linkage of the vWF level to the vWF gene is found less often. The recent European type 1 vWF cohort study demonstrated this observation. Almost all families with vWF levels <30 IU/dl showed linkage to the vWF gene, but the proportion fell to only 51% with vWF levels >30 IU/dl.⁹⁸ These data were corroborated in the Canadian study.¹⁴⁶

FIGURE 53.6. Pedigree of patients described by von Willebrand. Three families are emphasized: S, E, and J. Clinical details of bleeding events in these families have been reported. Open circles indicate female nonbleeders, open squares indicate male nonbleeders. Shaded circles and squares indicate female and male bleeders, respectively. Solid symbols indicate family members who experienced severe bleeding, and crosses indicate family members who experienced hemorrhagic deaths. (Modified from von Willebrand EA. Über hereditäre pseudohämophilie. Acta Med Scand 1931;76:521-550, with permission of the publisher.)

Complete absence of vWF is responsible for the most extreme form of vWD, and this is classified as type 3 vWD.¹²⁸ As one might predict, a large variety of genetic abnormalities scattered throughout the vWF gene have been reported in families with type 3 vWD. These have included large gene deletions, small gene deletions, frame-shift mutations, splice-site mutations, nonsense mutations, and point mutations.¹⁴⁹

Type 2 vWD is characterized by production of a qualitatively defective protein, and the genetics of these variant forms is more straightforward than that of type 1 vWD. In families affected by types 2B and 2M, as well as in the majority of families with type 2A vWD, inheritance is autosomal dominant. Rare cases of type 2A disease are transmitted in an autosomal recessive fashion, as are most cases of type 2N disease. vWF gene sequencing studies combined with vWF mutation-expression analysis has revealed that single missense point mutations underlie the majority of type 2 disorders.¹²⁸ Type 3 vWD is also inherited as a recessive trait and occurs when both copies of the vWF gene are defective. In type 3 disease, gene abnormalities include large or partial gene deletions, disruption of orderly messenger RNA transcription (frame-shift mutation, splicesite, or nonsense mutation), and missense mutation. For patients with recessively inherited vWD phenotypes (type 2N, type 3, or some rare type 2A), genetic analysis may reveal either homozygous or compound heterozygous gene abnormalities. Members of the International Society of Thrombosis and Hemostasis maintain a database of vWD mutations (www.vwf.group.shef.ac.uk).

Pathophysiology

The basic defect in vWD is a deficiency or abnormality of vWF function. vWF is synthesized in endothelial cells and
megakaryocytes.¹³⁴ The primary transcription product is 2,813 amino acids in length and undergoes extensive further processing, including dimerization and polymerization, to form very-largemolecular-weight “multimers”.¹³⁴ Multimerization and intracellular trafficking of vWF through the endoplasmic reticulum and Golgi apparatus and into storage granules is directed by the vWF propeptide (vWFpp). vWFpp is cleaved off the “mature subunit” after multimer assembly. Large vWF multimers, which may exceed 20 million daltons in size, are stored in endothelial cells in Weibel-Palade bodies and in platelet α-granules.

The majority of circulating vWF is present in plasma, with a concentration of ˜10 µg/ml. The circulating half-life of plasma vWF is ˜8 to 12 hours. Approximately 15% of circulating vWF is present intracellularly within platelets. Ultralarge multimers released from endothelial cells into the plasma are further degraded to smaller multimers through the activity of a “vWF cleaving protease”.¹³⁶ The cDNA sequence identified this protein to be the thirteenth member of the ADAMTS family (a disintegrinlike and metalloproteinase with thrombospondin type 1 motifs, ADAMTS13).¹⁵⁰, ¹⁵¹ vWF is cleaved between tyrosine 1605 and methionine 1606 located in the A2 domain.¹⁵² High shear stress and tethering of vWF by platelets most likely enhance exposure of the cleavage site.¹⁵³ The physiologic importance of enzymatic degradation of vWF after release is demonstrated clinically by the association of an inherited form of thrombotic microangiopathy (Schulman-Upshaw syndrome) with congenital absence of the protein due to defects of the ADAMTS13 gene,¹⁵⁰ and the observation of plasma deficiency of the protein with presence of autoantibody in most patients with sporadic thrombotic thrombocytopenic purpura.¹⁵⁴ Endothelial cell stores of vWF can be released therapeutically with administration of desmopressin.¹⁵⁵

vWF is required for normal platelet adhesion (see Chapter 17), and also acts as a carrier of factor VIII in the plasma. When vWF is deficient or aberrant, both factor VIII deficiency and abnormalities in the early steps of primary hemostasis result. vWD is thus manifested as a multifaceted hemostatic defect. Both abnormalities— factor VIII deficiency and abnormal primary hemostasis—give rise to characteristic and distinct laboratory abnormalities, and both defects contribute to the bleeding tendency.

Abnormalities of Primary Hemostasis

Convincing evidence exists that the platelets, although intrinsically normal, do not adhere to subendothelium normally in the setting of deficient or defective vWF.¹⁵⁶ This abnormality is well demonstrated by methods that measure platelet adhesion to subendothelial surfaces under flow conditions (the Baumgartner technique). The physiologic stimulus for interaction of vWF with platelets is not completely defined, but tethering of vWF on exposed subendothelial cell matrix or shear stress may cause vWF to undergo conformational change. Results of experimental studies suggest that in the absence of vWF, platelet adhesion to collagen in a flowing stream of blood may be particularly deficient at high rates of shear, when only a short time is available for formation of a platelet-collagen bond.¹⁵⁷ Deficient platelet plug formation has been observed directly in experimental wounds.¹⁵⁸ In vitro, the interaction of vWF with the platelet GPIb receptor can be induced by addition of the antibiotic ristocetin, by the snake venom protein botrocetin, or by subjecting platelets to high shear stress in the presence of vWF. For the effective support of platelet adhesion, vWF is multimerized into long chains (strings), which can exceed 2 µm in length. Plasma vWF demonstrates a spectrum of sizes; however, the longest vWF multimers are the most efficient at supporting platelet adhesion.¹³⁴

Abnormalities of Secondary Hemostasis

A second function of vWF is as a molecular chaperone for factor VIII, increasing the plasma half-life of factor VIII approximately fivefold.¹⁵⁹ This effect may occur through protection of factor VIII from activated protein C-mediated degradation.¹⁶⁰ Furthermore, the binding of vWF to subendothelium at sites of vascular injury could serve to co-localize factor VIII, where it can participate in its role in the regulation of hemostasis on the activated platelet membrane.¹³⁴ It is generally thought that whereas vWF is synthesized in endothelial cells and megakaryocytes, factor VIII is synthesized by hepatocytes. When the drug desmopressin (DDAVP, 1-deamino-8-D-arginine vasopressin) is administered to normal subjects or individuals with mild quantitative deficiency of vWF or factor VIII, levels of both molecules rise. Studies of hemophilia patients and patients with type 3 vWD on replacement therapies suggest that both vWF and factor VIII must be endogenously synthesized in order for there to be coordinated release after DDAVP. The cell responsible for that coordinated synthesis remains controversial.

Response to Transfusion

When blood products containing the vWF-factor VIII complex are infused into patients with hemophilia A, peak levels of factor VIIIc are present immediately after the infusion; this activity then declines rapidly, with an overall half-life of 8 to 10 hours (Fig. 53.7). Twenty-four hours later, factor VIIIc activity is minimal. This response is highly predictable and is discussed in a later section concerning replacement therapy for hemophilia A. In most patients with vWD, the infusion of normal vWF-factor VIII complex produces an initial rise in factor VIIIc that is predicted from the preinfusion level and the amount of factor infused. This is followed by a sustained but variable rise in factor VIIIc activity that reaches a plateau about 24 hours later and may persist for 48 to 72 hours. This phenomenon is highly variable and irregular; in some patients, a rapid initial fall of factor VIIIc is followed by a secondary rise in activity. This disproportionate response to transfused factor VIII has been called the secondary transfusion response, a feature that is apparently unique to vWD. Various hypotheses have been advanced to explain the disproportionate response to transfusion described above. The leading hypothesis is that the infused vWF results in stimulation of increased factor VIII release into the plasma with the attachment of the newly released factor VIII onto infused vWF.¹⁶¹

FIGURE 53.7. The phenomenon of new factor VIII synthesis in patients with von Willebrand disease (vWD) who receive plasma transfusion. Open circles indicate changes in the plasma factor VIII levels of a patient with vWD after infusion of plasma from a patient with severe hemophilia A. A significant and sustained increase in the factor VIII levels of the recipient was observed even though no active factor VIII was present in the infused plasma. Solid circles indicate effects of infusing plasma from a patient with vWD into a patient with severe hemophilia A. The factor VIII level in the infused plasma was 15% of normal. Note the slight and transitory effects. (From Shulman NR, Cowan DH, Libre EP, et al. The physiologic basis for therapy of classic hemophilia (factor VIII deficiency) and related disorders. Ann Intern Med 1967;67:856-882, with permission.)

Transfusions of blood products containing vWF shorten the bleeding time to a variable degree in patients with vWD. This corrective effect seldom persists for more than a few hours, even after massive transfusions that raise vWF to high levels. The correction of the bleeding time apparently requires the large molecular forms of vWF that are present in cryoprecipitate and select intermediate-purity factor VIII concentrates,¹⁶² but that are completely absent from monoclonally purified or recombinant concentrates of factor VIII.¹⁶³ Failure of the skin template bleeding time of patients with vWD to show sustained correction with appropriate replacement therapy has been noted, and does not necessarily predict defective surgical hemostasis. This apparent paradox may be a result of sensitivity of the bleeding time test to platelet α-granule vWF, which is not replenished during vWF replacement therapies.¹⁶⁴

Clinical Manifestations

The bleeding manifestations in vWD are heterogeneous but consistent with the dual roles of vWF in supporting both primary and secondary hemostasis. Thus, in the mild forms of vWD, the clinical picture is dominated by cutaneous and mucosal bleeding, which appears to be mainly the result of disordered primary hemostasis. In the most severe forms of the disorder, in which factor VIII levels are low, hemarthroses and dissecting intramuscular hematomas may develop. As in mild classic hemophilia, serious hemorrhage resulting from traumatic injuries or after surgical procedures is a significant hazard in severe vWD. Petechiae are rare, but hematoma formation and the extent of bruising are excessive compared with the inciting trauma.

The bleeding manifestations in the usual patient with type 1 vWD are mild, however, and many patients are virtually asymptomatic. The disorder may be symptomatic at any age, but it is not always recognized. Easy bruising is common in vWD, but it is not specific. Mucosal bleeding is particularly common. Childhood epistaxis, a life-long history of easy bruising, bleeding with dental extraction, heavy menstrual bleeding or anemia attributed to excessive menstrual blood losses, or postpartum hemorrhage are all included in the spectrum of vWD symptomatology. Systemic bleeding disorders are often not considered in women with menorrhagia,¹⁶⁵ but clinical studies reveal that the prevalence if vWD is significant in individuals who present with this complaint.¹⁶⁶ Angiodysplasia of various vascular beds, particularly those in the gut, has been demonstrated in some patients with vWD and may be an important contributory factor in chronic gastrointestinal bleeding.¹⁶⁷ In type 1 vWD, symptoms usually become milder during pregnancy or with estrogen therapy, when the vWF and factor VIIIc levels rise significantly.⁴⁰ The disorder may also decrease in severity with advancing age. Bleeding manifestations tend to be more prominent in patients with more severe quantitative deficiency and in patients with qualitative (type 2) defects.

While eliciting the history of bleeding, one should bear in mind that women are often inaccurate assessors of whether their menstrual flow is normal or excessive. Supplemental questioning regarding the frequency of changes of menstrual protection, and questions regarding a history of iron deficiency anemia and transfusion, may be informative. Retrospective studies of women with vWD show that abnormal bleeding can often be traced to menarche.¹⁶⁸ Family history also may provide important clues to a diagnosis. The physician should specifically seek a history of the response to hemostatic challenge, such as dental extraction, tonsillectomy, surgical procedures, menstruation, peripartum hemorrhage, and transfusion. A model questionnaire with a scoring system has been published.¹⁶⁹ Coexistent conditions or drug effects can strongly influence the severity of bleeding symptoms. Risk of hemorrhage may increase in patients with concomitant liver disease, uremia, gastrointestinal ulcer, or angiodysplasia. Excessive bleeding in response to challenge with aspirin may also point toward a bleeding disorder, but this history is not specific for vWD. Valproic acid use can result in lower vWF activity levels, as may hypothyroidism and valvular heart disease. Conversely, oral contraception frequently ameliorates menorrhagia.¹²⁷

Laboratory Diagnosis

Criteria for the laboratory diagnosis of vWD are imperfect. One goal of applying a series of laboratory evaluations is to exclude other causes of bleeding. When specifically considering vWD, the laboratory evaluation may be complicated. Even with the more elaborate confirmatory tests now available, the diagnosis of this disorder may be difficult and may require repeated observations over a period of time. Both vWF and factor VIII are “acute-phase reactants,” increasing with stress, trauma, estrogen therapy, or pregnancy, such that levels may fluctuate from time to time. Evaluation of a patient at a time remote from acute infection, pregnancy, or strenuous activity is preferable. However, pregnancy and acute-phase reaction are unlikely to obscure diagnosis in patients with more severe quantitative deficiency or qualitative defects.

Diagnosis of vWD is complicated, because of both the breadth of vWF functions and the spectrum of disorders that constitutes vWD. Consequently, a battery of studies should be considered to evaluate a patient conclusively for vWD.¹³⁴ Screening tests such as bleeding time, Platelet Function Analyzer-100 (PFA-100), and PTT lack sensitivity; specific tests such as those for vWF antigen, vWF activity, and factor VIII level are preferred. Correlating the values of these various quantitative assays with each other, or obtaining qualitative tests such as vWF multimer analysis or lowdose ristocetin-induced platelet aggregation, may provide important clues toward a diagnosis of type 2 vWD (Table 53.7).

Bleeding Time and Platelet Function Analyzer-100

The template bleeding time, the Ivy bleeding time, or technical modifications thereof may be useful in detecting the abnormality characteristic of vWD. However, the bleeding time may be normal in many patients with vWD.¹⁷⁰ Consequently, the bleeding time adds to the diagnosis of vWD if it is prolonged, but otherwise may have little clinical utility in the diagnosis and management of patients with vWD. The bleeding time may not improve following replacement therapy with vWF, and monitoring is generally not recommended,¹⁷¹ because surgical bleeding can generally be prevented by replacement of plasma vWF.

Standardization issues, inconvenience, and residual scarring of patients are some of the issues that have fostered development of in vitro tests to replace the bleeding time. The PFA-100 evaluates platelet adhesion and aggregation to collagen in a whole-blood assay under high shear conditions.¹⁷² Citrate-anticoagulated whole blood is aspirated through an aperture in a collagen-coated membrane, and the device measures the time from first flow until flow ceases as the “closure time.” PFA-100 closure times are sensitive to multiple factors, including vWF function, platelet count, platelet function, and hematocrit. Initial studies using the PFA-100 in patients with diagnosed vWD indicated that the assay was sensitive to most forms of vWD. However, when the closure time test was applied in clinical settings where patients were referred for evaluation of bleeding symptoms, the sensitivity of the PFA-100 for diagnosis of vWD ranged from 62% to 80%, with specificity ranging from 84% to 89%.¹⁷³, ¹⁷⁴ In addition, the closure time may remain normal in acquired vWD despite a very reduced plasma vWF level, because the platelet vWF level may remain normal. The PFA-100 closure time may also be normal in patients with type 2N vWD, in which the major defect is low factor VIII levels. As with the bleeding time, correction of plasma vWF level in the absence of concomitant correction of the platelet vWF may be insufficient to correct the closure time, so this test may not be very helpful for perioperative monitoring of patients on replacement therapy.¹⁷⁵, ¹⁷⁶

Partial Thromboplastin Time and Factor VIIIc Assay

Among the simple screening tests (Table 53.5), the PTT may be abnormal, reflecting reduced levels of factor VIIIc. Plasma levels of factor VIIIc vary greatly in this disorder and range from as low as 3 IU/dl in patients with type 3 vWD to normal levels in patients with mild type 1 vWD. Because factor VIIIc deficiency is often mild in patients with vWD, a diagnosis of vWD may be missed by the routine screening PTT. Consequently, a specific assay is required to reliably detect factor VIIIc deficiency in patients with vWD.

von Willebrand Factor Immunoassay

The immunoassay of vWF is a quantitative measure of vWF protein that is one of the most sensitive methods available for the diagnosis of vWD.¹⁷⁷ A variety of methods have been used, all of which depend on antibodies specific to vWF. The electroimmunodiffusion method of Laurell was initially widely used because it was simple to perform, but ELISA or automated assays using related methodologies have now largely replaced this technique. Average levels of vWF antigen (vWF:Ag) obtained by different laboratories will vary, but a level of ˜45 to 50 IU/dl is the lower limit of normal reported by many laboratories. The ratio of factor VIII to vWF also varies by laboratory but normally ranges from ˜0.7 to 2.2 (Fig. 53.1). Plasma levels of vWF:Ag are regulated to some extent by carbohydrate residues attached to the protein; ˜25% lower vWF antigenic levels have been noted in patients with blood type O compared to patients with other blood types.¹³⁹ The contribution of blood group O is more marked in patients with mild quantitative vWF deficiency, where the low vWF level is less likely to be related to abnormalities of the vWF gene.¹⁴⁵, ¹⁴⁶ Whether these patients actually have vWD type I is subject to debate, and the role of blood group in the diagnosis of vWF deficiency remains to be resolved.¹²⁸ Patients with type 2 vWD may have normal vWF:Ag levels, emphasizing the need to use more than just a vWF antigen assay to evaluate a patient for vWD.

von Willebrand Factor Functional Assays

vWF has multiple functions, and assays have been devised to specifically evaluate many of these. The function most often clinically assessed is its ability to interact with platelets. An assay that evaluates the ability of vWF to interact with collagen has also been proposed as a functional assay. The relevance of collagen-binding activity analysis is questioned for routine screening, and collagenbinding activity assay has not gained widespread acceptance in the United States.¹⁷⁸ Finally, vWF binding of factor VIII can be assessed in special circumstances when questioning whether mild factor VIII deficiency is due to failure of vWF chaperone function. vWF functional assay generally correlates well with vWF:Ag in patients with quantitative disorders of vWF. Observing a discrepancy between vWF function and vWF:Ag provides a useful clue for qualitative defects, provided that the testing laboratory uses the same control plasma for determining both assays.¹⁷⁹

Ristocetin Cofactor Activity and Ristocetin-induced Platelet Aggregation

The assay for ristocetin cofactor activity (vWF:RCo) is a quantitative technique for estimating the ability of vWF in patient plasma to bind target platelets that are specially prepared or preserved for use in the assay.¹⁸⁰ Ristocetin induces an activated conformation of vWF such that it will then bind platelets via the platelet GPIb receptor, mimicking the in vivo sequence of events that allows interaction between subendothelial tissue-bound vWF and platelets. Quantitation of vWF:RCo is held to be the single most sensitive and specific assay for vWD by multiple authorities.¹⁸¹, ¹⁸², ¹⁸³ vWF:RCo activity is adversely affected by loss of the larger-molecular-weight vWF multimers. The discrepancy between vWF:RCo and vWF:Ag that is observed in qualitative defects characterized by loss of both the intermediate- and higher-molecular-weight multimers (such as type 2A vWD) is thought to result from this. Ristocetin cofactor activity is also generally utilized to follow a vWD patient’s response to therapeutic interventions. However, obtaining typical ristocetin cofactor activity is a time-consuming procedure, and assay standardization remains somewhat problematic. Automation of the vWF:RCo assay has resulted in a clinically more available, but less sensitive technique, which requires modification to detect levels below ˜10 to 20 IU/dl. Simpler ELISA-based methods using monoclonal antibody to the GPIb-binding domain of vWF have been explored for estimating vWF activity.¹⁸⁴ Unfortunately, commercial functional epitope assays have not performed sufficiently well to allow them to replace vWF:RCo assay,¹⁸⁵, ¹⁸⁶ and these assays do not measure the increased functionality of the larger vWF multimers directly.

Ristocetin-induced platelet aggregation (RIPA) is a distinct test from ristocetin cofactor activity. RIPA is a qualitative test that involves evaluating the rate or extent of agglutination of patientderived platelet-rich plasma in response to the addition of the antibiotic ristocetin. The concentration of vWF and platelets present in the test plasma, as well as the amount of ristocetin added, affect RIPA. The assay is relatively insensitive to mild deficiency of vWF, but decreased aggregation response is observed in both more severe deficiency and in patients with Bernard-Soulier syndrome (in which platelets are missing the GPIb receptor). Titrating down the amount of ristocetin added in the assay may be of specific interest in defining variants of vWD in which there are “gain of function” mutations (type 2B and platelet-type [pseudo] vWD). These two conditions are characterized by increased sensitivity of patient vWF or platelets to the effect of ristocetin.¹²⁸ However, insensitivity of patient samples to low-dose ristocetin aggregation is seen in occasional patients with type 2B vWD in whom there is marked reduction in the amount of high- and intermediate-molecular-weight multimers, and genetic analysis may be required to distinguish some cases of type 2B from type 2A vWD. In addition, low-dose ristocetin aggregation of patient-derived platelet-rich plasma cannot differentiate type 2B vWD from the rarer platelettype (pseudo) vWD, and mixing studies using patient plasma and donor platelets have been devised to help make that distinction at the phenotypic level.¹⁸⁷

von Willebrand Collagen-binding Activity

vWF contains collagen-binding sites in both the A1 and A3 domains of the vWF protein, and collagen-binding activity (vWF:CB) may be an important function that allows vWF to adhere to platelets at sites of vascular injury. Like ristocetin cofactor activity, vWF:CB is dependent on multimer size, with the larger multimers binding collagen more avidly. The assay is generally done in an ELISA format, in which type I and/or type III collagen is plated into microtiter wells, and the amount of vWF captured is assayed.¹⁸⁸ vWF:CB correlates well with vWF:Ag in normal individuals and those with quantitative deficiency of vWF, and discrepant results are indicative of vWD subtypes that are characterized by deficiency of larger vWF multimers. A single case of vWD characterized by mutation of the vWF A3 domain with isolated loss of collagen-binding function despite the presence of a normal multimer distribution has been described.¹⁸⁹ Although some studies suggest that vWF:CB testing complements assays of vWF:Ag and vWF:RCo in distinguishing type 1 vWD from type 2A and 2B variants, the role of vWF:CB testing in routine clinical practice remains to be determined.¹²⁷, ¹⁷⁸, ¹⁹⁰

Assay of von Willebrand Factor’s Ability to Bind Factor VIII

Factor VIII binding by vWF (vWF:FVIIIB) can be measured in a few reference coagulation laboratories, and is indicated for the evaluation of patients suspected of having factor VIII deficiency on the basis of defective vWF chaperone function (type 2N vWD).¹⁹¹ In this ELISA format assay, ELISA plate wells are coated with antibody to vWF in order to capture vWF from patient plasma. After washing patient-derived factor VIII from the well, the ability of patient-derived vWF to capture r-FVIII is assessed using a chromogenic or immunoassay for factor VIII, and the quantity of patient-derived vWF:Ag in the well is quantified by standard ELISA technique. The calculated ratio of captured r-FVIII to vWF:Ag is expected to be in the normal range when testing plasma from a patient with hemophilia A, whereas it will be abnormally low in the rare cases of type 2N vWD.¹⁹²

Multimeric Analysis of von Willebrand Factor

Normal vWF in plasma consists of a complex series of multimers, ranging in size from about 500 to 20,000 kDa. vWD initially diagnosed by abnormal assays for factor VIIIc, vWF antigen, or ristocetin cofactor activity should be further characterized by multimeric analysis. In this test, the spectrum of vWF molecular sizes of the patient’s vWF is assessed by size-based separation of vWF multimers via agarose gel electrophoresis, and vWF multimers are imaged by immunologic methods coupled to autoradiography or other detection methods.¹⁹³ Multimer analysis is characterized as either “low-resolution” (which differentiates the largest multimer forms from the intermediate and smaller forms), or “high-resolution” (which resolves each multimer band into three to five satellite bands), which inform on the activity of ADAMTS13 on plasma vWF. Multimer analysis will confirm the presence of the entire spectrum of vWF multimers in normal individuals and those with type 1 vWD, and allows identification of the loss of intermediate- or high-molecular-weight multimers of vWF as is characteristic of the type 2A and type 2B qualitative defects.¹⁹³ Unfortunately, the distinction between type 2A and type 2B vWD is not always possible using multimer analysis alone, necessitating further testing. Multimeric analysis is a very labor-intensive test and is usually performed only in coagulation reference laboratories. The multimeric composition of plasma from normal subjects and patients with vWD is shown in Figure 53.8. This methodology has been reviewed.¹⁹⁴

Genetic Testing in von Willebrand Disease

Direct mutational analysis by vWF gene sequencing, allele-specific PCR, or restriction enzyme analysis is useful in confirming a diagnosis of type 2 variant vWD.¹⁹⁵ The functional domain structure of the vWF gene results in clustering of the various type 2 vWD defects in limited areas of the vWF sequence.¹²⁸ vWF exon 28 encodes for the A1 loop of vWF responsible for the interaction of vWF with platelet GPIb, and vWD variants characterized by defects of this interaction (types 2B and 2M) are encoded by genetic mutations clustered in this exon.¹⁹⁶ Exon 28 also encodes the A2 domain, which contains the ADAMTS13-sensitive bond tyrosine1605-methionine 1606. Mutations that increase the susceptibility of this bond to cleavage by ADAMTS13 underlie a high fraction of type 2A vWD (previously classified as 2A, subtype 2). Less commonly, type 2A vWD results from a defect of the multimerization process. The carboxy terminus of the vWF peptide (encoded by exons 51 to 52) is involved in initial vWF dimer formation in the rough endoplasmic reticulum. The vWFpp (encoded by exons 3 to 17) directs further multimerization via cross-linkage of the amino-terminal D’ and D3 of the mature vWF protein (encoded by exons 18 to 28). As one might expect, type 2A vWD on the basis of failure of multimerization (previously classified as 2A, subtype 1) has been attributed to genetic abnormalities in each of these regions of the genome. Mutations responsible for defective interaction of vWF with factor VIII (type 2N vWD) are predominantly clustered in exons 18 to 20, which encode for the factor VIII-binding domain of vWF; however, rare cases have been attributed to mutations in exons 24 and 25. An online database of defects found in vWD patients can be accessed at www.vwf.group. shef.ac.uk. Figure 53.5 shows the location of these various genetic abnormalities schematically. Rapid, complete vWF gene sequencing has been reported.¹⁹⁷

FIGURE 53.8. Multimeric composition of von Willebrand factor from normal plasma and that from types I, IIA, IIB, IIC, IID, and IIE von Willebrand disease (vWD), as well as from a type IIC heterozygote (IICh). Using the revised classification scheme, these types would correspond to 1, 2A, 2B, 2A, 2A, and 2A, respectively. The three brackets include the smallest normal oligomers, with arrows pointing to the central predominant band in each. In type 1 vWD, the relative concentration of large multimers varies. (From Zimmerman TS, Ruggeri ZM. von Willebrand disease. Hum Pathol 1987;18:140-152, with permission.)

Genetics of Type 1 and Type 3 vWD

Mild to moderate quantitative deficiency of vWF characterizes patients with type 1 vWD, whereas virtual absence of vWF characterizes type 3 disease. Genetic analysis is not generally required to make these diagnoses, but may provide useful additional information concerning pathogenesis of a patient’s condition, responsiveness to DDAVP, or risk for alloantibody formation. In distinction from type 2 vWD, genetic defects underlying quantitative vWF deficiency are distributed throughout the vWF gene, complicating evaluation. In addition, for families with patients who demonstrate normal multimer structure and higher vWF levels, the chances of finding linkage to the vWF gene are reduced,¹⁴⁵, ¹⁴⁶ and family members are usually asymptomatic.¹³⁷ Type 3 patients with large gene deletions are particularly prone to develop anti-vWF alloantibodies after treatment with vWF-containing concentrates.

Type 1 von Willebrand Disease and “Low von Willebrand Factor”

Patients with bleeding attributed to decreased production of qualitatively normal vWF are classified as having type 1 vWD. Most patients with vWD (70% to 80%) fall into this category.¹³⁴ The majority of patients have a mildly symptomatic disorder, but bleeding may increase with physical trauma, surgery, or during menstruation. vWF levels are low, with concordant reduction of vWF:Ag and vWF:RCo (and vWF:CB, if that is measured). Factor VIII levels are generally equal to or higher than vWF levels, and all vWF multimers should be present if that analysis is performed as part of the patient evaluation. The diagnosis of type 1 vWD may be simpler in an individual with mucosal bleeding symptoms, a strong family history of similar symptoms, and quantitative vWF level (vWF:Ag and/or vWF:RCo) <30 IU/dl. Although not required for a diagnosis of type 1 vWD, the probability of finding linkage of the phenotype to the vWF gene in such an individual would be high, such that both therapeutic choices and genetic counseling would be straightforward.

Unfortunately, many factors make the diagnosis of type 1 vWD difficult.¹⁴⁴ Mild bleeding symptoms are common in hemostatically normal people (see Chapter 45). In addition, the relative risk of bleeding for individuals with only modest reductions in vWF level appears to be minimal. Finally, because of the strong influence of blood group on vWF levels, there is a high prevalence of blood group O among individuals with mildly depressed vWF levels. These observations have led to the recent proposal of the concept of “low vWF”.¹³⁷,¹⁴⁴ The term “low vWF” could be applied to patients with vWF levels below the normal reference range but above some lower limit. A recently convened “expert group” brought together by the National Heart, Lung and Blood Institute is considering this proposal.¹²⁹ Studies of families of patients with a diagnosis of type 1 vWD suggest that linkage to the vWF gene is less often seen when the vWF level rises above 30 IU/dl, suggesting that this might be a candidate for the lower limit of the “low vWF” category.¹⁴⁵, ¹⁹⁸ Empiric therapy to raise vWF levels in patients with “low vWF” at times of hemostatic challenge would be reasonable for patients with a history of bleeding, but the genetic counseling issues would clearly be different.¹⁴⁴

On the other hand, two population-based studies have shown that “low vWF” causes increased bleeding.¹⁹⁹, ²⁰⁰ These latter studies would suggest that abnormally low ristocetin cofactor activity, and not the presence of a mutation, should primarily be used to identify bleeders.

Type 3 von Willebrand Disease

Type 3 vWD is the most severe form of vWD, because it results from complete failure of vWF synthesis. This bleeding disorder is generally diagnosed during infancy. Hematoma formation is common, epistaxis may be life-threatening, and hemarthrosis may occur as a result of the low factor VIII levels that are seen in this condition. Plasma from patients with type 3 vWD contains virtually no detectable vWF, and vWF is not present in either platelets or endothelial cells. Factor VIII level is generally in the range of 1 to 10 IU/dl, similar to that seen in moderate to mild hemophilia A. Multimeric analysis of the little vWF present yields variable results, in some cases revealing only small multimers.¹⁹³ Type 3 vWD is rare, with an estimated frequency of 1 in a million people.²⁰¹ Genetic analysis reveals either homozygous or compound (double) heterozygous defects of the vWF gene, with gene deletions, frame-shift mutations, missense mutations, or nonsense mutations. Careful laboratory evaluation of the parents of patients with type 3 disease may reveal mild quantitative deficiency of vWF. In a review of 117 obligate heterozygotes, the mean plasma vWF level was 45 IU/dl, with a range from 5 to 130 IU/dl.¹⁴⁴ Parents are frequently asymptomatic, consistent with the impression that type 3 disease is inherited as a recessive trait. Consanguinity is common in kindreds with this variant. In some patients with type 3 vWD, large gene deletions have been identified on one or both vWF alleles.¹²⁸,²⁰² Such patients are at increased risk of developing inhibitory alloantibodies to vWF after transfusion therapy.²⁰²

Because these patients lack vWF in their endothelial cells and platelets, there is no rise in vWF level in response to DDAVP. For replacement therapy, these patients require vWF-containing concentrates. This is usually provided in the form of specific intermediate-purity factor VIII concentrates that have been documented to contain intact vWF.¹⁵⁵ Because these patients possess a normal factor VIII gene but are missing the vWF chaperone, posttransfusion factor VIII recovery and survival may be longer in a patient with type 3 vWD than in a patient with hemophilia A, owing to endogenous factor VIII production. Although life-threatening bleeds require immediate replacement of both vWF and factor VIII, replacement with vWF alone may be sufficient if therapy is begun 12 to 24 hours before elective surgery.²⁰³

Qualitative Defects of von Willebrand Factor

Several subsets of patients with vWD differ substantially from those with the common quantitative deficiency states (Table 53.6).¹²⁸ Classified as type 2 variants, these qualitative defects of vWF are less common, but they may represent up to 20% to 25% of cases of vWD. Type 2 forms of vWD are suspected when the severity of the patient’s symptoms seems in excess of the observed vWF and factor VIII levels, when there is discordant reduction between vWF antigen and vWF functional activity or factor VIII assay, or when there is concomitant vWF deficiency and thrombocytopenia. Discrimination between qualitatively normal and abnormal vWF can be difficult; however, when the ratio of vWF:RCo to vWF:Ag or vWF:CB to vWF:Ag is found to be <0.6, dysfunctional vWF should be considered. The sensitivity and specificity of such ratio screening depends to a large extent on the expertise and precision of the laboratory performing the assays, but the predictive value of ratios decreases when the level of vWF antigen is <20 IU/dl. Examining the structure of patient vWF may provide further definition of the nature of the qualitative defect. This can be demonstrated by vWF multimer analysis. If required, further information is provided by supplemental studies of vWF function, or via genetic analysis of the vWF gene (Fig. 53.5).¹²⁶,¹²⁸,¹⁹³ Type 2 defects were initially classified using a Roman numeral system, but further understanding of the genetics of vWD led to the consolidation and reclassification into types 2A, 2B, 2M, and 2N,¹²⁸ based on the nature of the vWF functional defect. In addition to the well-defined variants described in this chapter, other forms of vWD continue to be described.

Type 2A von Willebrand Disease

vWD with defective interaction of vWF with platelets due to deficiency of intermediate- and high-molecular-weight forms of vWF is classified as type 2A vWD.¹²⁸ This form of the disorder is generally inherited as an autosomal dominant trait,²⁰⁴ and it accounts for the majority of patients with type 2 defects. Levels of factor VIIIc and vWF:Ag in the plasma may be normal or reduced. The vWF activity as assayed by either vWF:RCo or vWF:CB is significantly lower than vWF:Ag because of the absence of the larger multimers (which are more potent in their ability to interact with platelet GPIb and collagen). Analysis of vWF multimers reveals a relative reduction in intermediate- and high-molecularweight species.¹²⁷ Protein studies followed by genetic analysis of the basis of type 2A vWD revealed multiple mechanisms for the generation of this disorder.²⁰⁵, ²⁰⁶, ²⁰⁷ In some patients, there is failure to synthesize full-length multimers (type 2A, subgroup 1). Mutations associated with impaired vWF multimerization have been localized to the vWFpp (formerly called type IIC), the region of the mature vWF protein associated with the amino terminus area of multimer formation (formerly called type IIE), and the carboxy terminus region that is involved in initial dimer formation (formerly called type IID). Subtle abnormalities uncovered by high-resolution multimer analysis or genetic evaluation can be used to distinguish these subtypes of type 2A vWD. The spectrum of multimers released into plasma after DDAVP administration would not be expected to improve significantly in these patients whose defects prevent normal multimer assembly.

A second group of type 2A patients has been described in whom there is excessive catabolism of fully multimerized vWF after vWF is released into the plasma. These patients demonstrate mutations in exon 28, encoding the A1 and A2 regions of vWF. These mutations allow increased proteolysis of multimerized vWF by ADAMTS13. Again, high-resolution multimer gel studies may be informative by demonstrating increased “satellite” band intensity, which accumulates as a result of excessive ADAMTS13 activity. Because intracellular vWF is protected from cleavage by ADAMTS13, these patients may demonstrate very transient
improvement in both vWF:RCo and multimer spectrum after DDAVP administration.

Inheritance of type 2A vWD is generally autosomal dominant, with the majority of cases caused by mutations clustered in the region of exon 28, which encodes the vWF A2 homologous repeat (Fig. 53.5).¹²⁸ Rare recessive forms of type 2A vWD have been reported with some multimerization defects (formerly called types IIC and IID).

Because of the underlying structural abnormality of the vWF produced in these patients, neither stress nor pregnancy significantly increases the functional amount of vWF protein in the plasma. As noted above, the administration of DDAVP does not result in a significant rise in vWF:RCo in patients with type 2A vWD, in which the underlying mechanism is defective multimerization or transport, but a very short-lived correction may be observed in patients with mutations associated with increased vWF cleavage by ADAMTS13.

Type 2B von Willebrand Disease

Type 2B vWD is a paradoxical bleeding disorder, characterized by increased interaction of patient vWF with platelets which is demonstrated in vitro in the presence of low doses of ristocetin. Increased in vivo interaction of the larger multimers of type 2B vWF with platelets is thought to result from mutations that either allow increased access of platelets to the A1 loop of vWF or that stabilize that interaction.¹²⁷ This may in turn allow increased generation of vWF-platelet complexes, which are subsequently cleared from circulation, resulting in thrombocytopenia.

This bleeding disorder is characterized by deficient vWF function attributable to mild reduction of vWF:Ag, a somewhat more marked deficiency of vWF:RCo activity (and vWF:CB, if measured). Multimeric analysis reveals a deficiency of the highestmolecular-weight vWF multimers. Measurements of factor VIIIc are variable. Many patients with the type 2B variant have mild persistent thrombocytopenia.²⁰⁸ The platelet count may fall further during physiologic stresses, pregnancy, in association with surgical procedures, or after the administration of DDAVP.²⁰⁹

Additional laboratory evaluation is required to support a diagnosis of type 2B vWD. In most cases, multimeric analysis reveals absence of higher-molecular-weight forms of plasma vWF, but distinction of type 2B from type 2A vWD is not possible by multimeric analysis alone. To show the distinct “gain in function” in type 2B disease, RIPA studies are done to reveal enhanced interaction of patient vWF and platelets with low doses of the drug. Finally, to differentiate the more common type 2B vWD from the rare platelet-type (pseudo) vWD, one should prove that the defect resides in vWF. This is done either through performing mixing studies of patient plasma with donor platelets¹⁸⁷ or by analysis of the patient’s vWF gene.

Type 2B vWD is inherited as an autosomal dominant trait. A small cluster of mutations within the portion of vWF exon 28 that encodes the vWF A1 domain that interacts with platelet GPIb accounts for all the cases of type 2B vWD that have been reported (Fig. 53.5).¹²⁸

Recent studies evaluating a large number of patients with type 2B vWD mutations revealed diversity in laboratory findings.²¹⁰ Not all patients had reduced vWF:RCo activity or abnormal multimeric analysis, and not all patients had thrombocytopenia.²¹⁰ Type 2B patients with normal multimers did not develop thrombocytopenia.

Type 2M von Willebrand Disease

The type 2M vWD variant is defined by diminished platelet-dependent vWF function that is not attributed to deficiency of vWF multimers.¹²⁸ Thus, these patients have an initial laboratory profile that in many ways is similar to that of patients with type 2A vWD, revealing variable deficiency of vWF:Ag but disproportionately decreased interaction of vWF with platelets in the presence of ristocetin as measured by vWF:RCo assay. Factor VIII level is proportionate to the vWF:Ag, and the platelet counts are normal. What differentiates type 2M vWD from the type 2A patients is that the vWF multimeric analysis is normal (and, if measured, the vWF:CB is usually similar to the vWF:Ag). Type 2M vWD is inherited as an autosomal dominant trait, and, when investigated, mutations have been found in the region of exon 28 that encodes the A1 domain of vWF. Unlike the type 2B mutations, 2M mutations cause impairment of the binding of vWF to the GP1b receptor and have been shown to be localized to an alternative area in the A1 domain (Fig. 53.5).¹⁹⁶

One case classified as a type 2M defect was described in which the vWF interaction with collagen was defective. In that case, a mutation in the A3 domain that contributes to the interaction of vWF with collagen was found.¹⁸⁹

Type 2N von Willebrand Disease

Mutations affecting the association of vWF with factor VIII can result in “autosomal hemophilia” in which factor VIII levels are significantly reduced relative to vWF.¹²⁸ Affected patients have factor VIII levels in the range of 5 to 30 IU/dl, attributed to mutations at the factor VIII-binding site near the amino terminus of the vWF subunit.²¹¹ Indeed, genetic studies indicate that the majority of mutations underlying type 2N vWD involve vWF exons 18, 19, and 20, which encode the bulk of the factor VIII-binding domain of vWF. The other functions of vWF in patients with type 2N vWD are qualitatively normal, and thus vWF laboratory parameters (vWF:Ag, ristocetin cofactor activity, etc.) are usually normal (unless there is coinheritance of type 1 vWD). This variant was originally named vWD Normandy but has been renamed type 2N vWD in the revised classification.¹²⁸ The factor VIII-binding defect in these patients is inherited in an autosomal recessive manner, and thus affected patients must inherit 2N alleles from each of their two parents, or inherit type 1 vWD from one parent while inheriting a 2N allele from the other. Patients with factor VIII deficiency and a bleeding disorder that is not clearly transmitted as an X-linked disorder or who demonstrate an unexpectedly short in vivo survival of infused factor VIII should be evaluated for type 2N vWD.²¹² Diagnosis requires performance of an assay that assesses the interaction of factor VIII with vWF,¹⁹¹, ¹⁹² or genetic studies of the vWF gene. A survey of almost 400 unrelated patients with either hemophilia A or type 1 vWD indicated a prevalence of type 2N vWD in these patient populations of 3% and 1.5%, respectively.²¹³ Consideration of this diagnosis is important because both therapy and genetic counseling are distinct from that of mild hemophilia A.

Platelet-type (Pseudo) von Willebrand Disease

Platelet-type (pseudo) vWD resembles type 2B vWD in most respects, except that the basis for platelet-type vWD is a structural defect in platelet GP1b²¹⁴ rather than a defect of vWF (thus platelet-type [pseudo] vWD is a form of platelet dysfunction). However, as with type 2B vWD, a genetic defect results in increased interaction between GP1b and vWF. Phenotypically, platelet-type (pseudo) vWD is manifested by mild thrombocytopenia, a prolonged bleeding time, and variable deficiency of plasma vWF and factor VIIIc. The reductions in plasma vWF and factor VIIIc may be a result of the attachment of these proteins to platelets that are subsequently removed from the circulation. This “consumption” of vWF and platelets results in preferential loss of high-molecular-weight multimers from patient plasma.²¹⁴ In vitro platelet aggregation studies reveal platelet agglutination at unusually low concentrations of ristocetin. Spontaneous intravascular platelet clumping may also occur. Platelet-type (pseudo) vWF is inherited as an autosomal dominant trait in most families.²¹⁴, ²¹⁵

Distinction of platelet-type vWD from type 2B vWD is based on showing that the enhanced interaction of platelets with vWF is either platelet-based or resides in the plasma phase (type 2B vWD).²¹⁶ This can sometimes be demonstrated through low-dose ristocetin-based agglutination assays using mixtures of normal donor and patient samples, or genetic analysis can be used. To date, all cases of platelet-type vWD have been traced to genetic variations of platelet GPIbα.²¹⁷ These include missense mutations that alter the amino acid sequence from Gly233 to Met 239. This region of the protein adopts a β-sheet conformation after interaction with vWF, and these mutations may function to stabilize this conformation.²¹⁸ In addition, a 27-bp gene deletion that encodes for deletion of proline 421 through serine 429 has been shown to underlie platelet-type vWD in one kindred. Thus, genetic analysis of the GPIbα gene could confirm a diagnosis of platelet-type vWD in cases in which phenotypic studies were unsuccessful.

CLINICAL DISORDERS OF THE FIBRINOGEN MOLECULE

The clinical disorders of fibrinogen are complex, and depending on the types of mutation may present with bleeding or with thrombotic symptoms, or both. These disorders can be divided based on whether the defect is predominantly a quantitative or qualitative abnormality of the fibrinogen molecule.²¹⁹, ²²⁰ Quantitative abnormalities are disorders associated with the complete absence of fibrinogen (afibrinogenemia) or with low levels of fibrinogen (hypofibrinogenemia). Qualitative abnormalities are disorders resulting from synthesis of an abnormal fibrinogen molecule (dysfibrinogenemia) with altered functional properties.²¹⁹, ²²⁰, ²²¹, ²²² Dysfibrinogenemia may be asymptomatic and identified only through laboratory studies, or it may be associated with either hemorrhagic or thrombotic consequences.

Fibrinogen is a 340-kDa plasma protein that circulates at a concentration of 1.5 to 3.5 mg/ml. It is a symmetric disulfidelinked dimer with a central E domain linked via “coiled coil” peptide chains to outer D domains.²²³ Each half-molecule consists of a set of three different peptide chains, termed Aα, Bβ, and γ, which are linked at their amino termini by disulfide bonds to form the E domain. The D domains are formed by disulfide linkages near the carboxy termini of the peptides. Fibrinogen is synthesized in hepatocytes²²⁴ by coordinated expression of three separate genes on chromosome 4.²²⁵ The rate-limiting step in fibrinogen synthesis is transcription of the Bβ gene. Fibrinogen is an acute-phase reactant, and levels may rise considerably with inflammation. The circulation half-life of plasma fibrinogen is ˜4 days. Fibrinogen participates in multiple physiologic processes, including fibrin clot formation mediated by the enzymatic activities of thrombin and factor XIIIa, and cohesion of activated platelets through interaction with the GP IIb/IIIa receptor. Fibrinogen also acts as a plasma carrier for factor XIII. Fibrinogen physiology has been reviewed.²²³

Only gold members can continue reading. Log In or Register to continue