Hemophilia A

A number sign (#) is used with this entry because classic hemophilia, or hemophilia A, is caused by mutation in the gene encoding coagulation factor VIII (F8; 300841) on chromosome Xq28.

Description

Hemophilia A is an X-linked recessive bleeding disorder caused by a deficiency in the activity of coagulation factor VIII. The disorder is clinically heterogeneous with variable severity, depending on the plasma levels of coagulation factor VIII: mild, with levels 6 to 30% of normal; moderate, with levels 2 to 5% of normal; and severe, with levels less than 1% of normal. Patients with mild hemophilia usually bleed excessively only after trauma or surgery, whereas those with severe hemophilia have an annual average of 20 to 30 episodes of spontaneous or excessive bleeding after minor trauma, particularly into joints and muscles. These symptoms differ substantially from those of bleeding disorders due to platelet defects or von Willebrand disease (193400), in which mucosal bleeding predominates (review by Mannucci and Tuddenham, 2001).

Nomenclature

The term 'hemophilia' is used in reference to hemophilia A (factor VIII deficiency); hemophilia B or Christmas disease (factor IX deficiency; 306900) and von Willebrand disease (von Willebrand factor deficiency; 193400). Hemophilia A and B are X-linked recessive disorders; von Willebrand disease has an autosomal dominant, or in some cases, an autosomal recessive mode of inheritance (review by Mannucci and Tuddenham, 2001).

Clinical Features

The severity and frequency of bleeding in hemophilia A is inversely related to the amount of residual factor VIII in the plasma: less than 1% factor VIII results in severe bleeding, 2 to 6% results in moderate bleeding, and 6 to 30% results in mild bleeding. The proportion of cases that are severe, moderate, and mild are about 50, 10, and 40%, respectively, The joints are frequently affected, causing swelling, pain, decreased function, and degenerative arthritis. Similarly, muscle hemorrhage can cause necrosis, contractures, and neuropathy by entrapment. Hematuria occurs occasionally and is usually painless. Intracranial hemorrhage, while uncommon, can occur after even mild head trauma and lead to severe complications. Bleeding from tongue or lip lacerations is often persistent (review by Antonarakis et al., 1995).

The clinical hallmarks of hemophilia A are joint and muscle hemorrhages, easy bruising, and prolonged hemorrhage after surgery or trauma, but no excessive bleeding after minor cuts or abrasions. Affected individuals may have little bleeding during the first year of life, but develop hemarthroses when beginning to walk. The most frequently affected joints are the knees, elbows, ankles, shoulders, and hips. Hemophilic arthropathy can be a progressive inflammatory condition which may result in limitation of motion and permanent disability (review by Hoyer, 1994).

Female Carriers

Rapaport et al. (1960) demonstrated a partial deficiency of factor VIII in heterozygous female carriers.

Most heterozygous female carriers of hemophilia A or hemophilia B (306900) have concentrations of clotting factor VIII or IX (F9; 300746) of about 50% of normal, respectively, and in most cases have mildly decreased coagulability without clinical signs. Sramek et al. (2003) followed up a cohort of 1,012 mothers of all known people with hemophilia in the Netherlands from birth to death, or the end-of-study date (41,984 person years of follow-up). Overall mortality was decreased by 22%. Deaths from ischemic heart disease were reduced by 36%. No decrease in mortality was observed for cerebral stroke (ischemic and hemorrhagic combined). Women in the cohort had an increased risk of deaths from extra cranial hemorrhage; however, the number of deaths from this cause was much lower than that for ischemic heart disease. The results were interpreted as showing that a mild decrease in coagulability has a protective effect against fatal ischemic heart disease.

In a population-based survey in the Netherlands, Plug et al. (2006) found that female carriers of hemophilia A and B bled more frequently than noncarrier women, especially after medical procedures, such as tooth extraction or tonsillectomy. Reduced clotting factor levels correlated with a mild hemophilia phenotype. Variation in clotting levels was attributed to lyonization.

Other Features

Rosendaal et al. (1990) presented evidence supporting their earlier findings that mortality due to ischemic heart disease is lower in hemophilia patients than in the general male population.

Chronic synovitis occurs in about 10% of Indian patients with severe hemophilia. Ghosh et al. (2003) reported an association between the development of chronic synovitis in patients with hemophilia and the HLA-B27 allele (142830.0001). Twenty-one (64%) of 33 patients with both disorders had HLA-B27, compared to 23 (5%) of 440 with severe hemophilia without synovitis (odds ratio of 31.6). There were 3 sib pairs with hemophilia in whom only 1 sib had synovitis; all the affected sibs had the HLA-B27 allele, whereas the unaffected sibs did not. Chronic synovitis presented as swelling of the joint with heat and redness and absence of response to treatment with factor concentrate. Ghosh et al. (2003) suggested that patients with HLA-B27 may not be able to easily downregulate inflammatory mediators after bleeding in the joints, leading to chronic synovitis.

Biochemical Features

Alexander and Goldstein (1953) first noted low levels of factor VIII in cases of von Willebrand disease (see 193400). This was confirmed by other workers including Nilsson et al. (1957), who studied von Willebrand's original family in the Aland Islands. Since von Willebrand disease is an autosomal disorder, these findings indicated that an autosomal locus can also cause low factor VIII levels. This overlap in phenotype between hemophilia A and von Willebrand disease is seen in families such as that of Graham et al. (1953) and Bond et al. (1962) in which the carrier females showed depression of factor VIII levels, but not as low as in hemizygous affected males and sometimes clinical hemophilia. Cooper and Wagner (1974) presented evidence that the factor VIII carrier molecule, von Willebrand factor, is normally present in the plasma of hemophilia A patients.

Using F8 antibodies, Denson (1968) and Feinstein et al. (1969) demonstrated that patients with hemophilia A have heterogeneous F8 molecules: plasma from some patients can be neutralized by the antibody, whereas plasma from other patients is not neutralized by the antibody. Denson et al. (1969) postulated that there are 2 subtypes of hemophilia A: one without any immunologically demonstrable protein and one with immunologically normal, but hemostatically defective protein. Hoyer and Breckenridge (1968) also found heterogeneity of the F8 protein in hemophilia A.

Stites et al. (1971) were able to detect F8 immunologically in all of 14 patients with hemophilia A they studied, whereas little or no F8 was identified in patients with von Willebrand disease. Zimmerman et al. (1971) found immunoreactive material in all of 22 patients with hemophilia A.

Inheritance

Hemophilia A is an X-linked recessive disorder and usually occurs in males. In familial cases, the affected boy has inherited the mutant gene from his carrier mother, but about 30% of cases arise from a spontaneous mutation (review by Mannucci and Tuddenham, 2001).

Using improved methods of carrier detection, Biggs and Rizza (1976) studied 41 mothers of presumably sporadic cases of hemophilia A and found that 39 were carriers.

Hermann (1966) reported an age effect on the mutation rate in hemophilia, but Barrai et al. (1968) concluded that there was no effect of maternal age or maternal grandfather's age.

Vogel (1977) concluded that the mutation rate causing hemophilia A is higher in males than in females; however Barrai et al. (1979) did not. Based on carrier detection tests of 21 mothers of isolated cases of severe hemophilia A, Winter et al. (1983) derived a maximum likelihood estimate of 9.6 (95% confidence limits 2.2-41.5) for the ratio of male to female mutation. Bernardi et al. (1987) found data consistent with a higher mutation rate in males than in females by using RFLP analysis in families with sporadic hemophilia A,

Rosendaal et al. (1990) collected information by mail on 462 Dutch patients with severe or moderately severe hemophilia A. Pedigree analysis on 189 of these patients who were the first hemophiliacs in their family showed, by the maximum likelihood method, that the ratio of mutation frequencies in males and females was 2.1, with a 95% confidence interval of 0.7-6.7. Rosendaal et al. (1990) performed a meta-analysis of all published studies on the sex ratio of mutation frequencies. From pooling of 6 studies, they estimated that mutations originate 3.1 times more often in males than in females (95% confidence interval 1.9-4.9). This implies that 80% of mothers of an isolated patient are expected to be hemophilia carriers. This estimate of prior risk is required for the application of Bayes theorem to probability calculations in carriership testing.

Brocker-Vriends et al. (1991) estimated that the mutation rate in males is 5.2 times that in females (95% confidence interval 1.8 to 15.1) suggesting that the probability of carriership for mothers of an isolated case amounts to 86%. Although this would imply that 14% of the mothers are not carriers in the classical sense, they may be mosaic for the mutation and, therefore, at risk of transmitting the mutation more than once.

Leuer et al. (2001) explored the hypothesis that a significant proportion of de novo mutations causing hemophilia A can be attributed to a germline or somatic mosaic originating from a mutation during early embryogenesis. They used allele-specific PCR to analyze 61 families that included members who had sporadic severe hemophilia A and known F8 gene defects. The presence of somatic mosaicism of varying degrees (0.2 to 25%) could be shown in 8 (13%) of the 61 families and was confirmed by a mutation-enrichment procedure. All mosaics were found in families with point mutations (8 of 32 families). In a subgroup of 8 families with CpG transitions, the percentage with mosaicism increased to 50% (4 of 8 families). In contrast, no mosaics were observed in 13 families with small deletions or insertions or in 16 families with intron 22 inversions. These data suggested that mosaicism may represent a fairly common event in hemophilia A. As a consequence, risk assessment in genetic counseling should include consideration of the possibility of somatic mosaicism in families with apparently de novo mutations, especially families with this subtype of point mutations.

Unusual Inheritance Patterns

Hemophilia A can also occur in females because of inheritance of defective F8 genes from both parents or on the basis of an autosome translocation disrupting the structure of the gene (e.g., Migeon et al., 1989). Pola and Svojitka (1957) reported a homozygous affected female who was the daughter of a hemophilic man married to a double first cousin. Sie et al. (1985) reported a homozygous female. In these cases, the homozygous female was not more severely affected than the hemizygous male. Theoretically, a female can be homozygous on the basis of uniparental isodisomy. It is also possible, in some cases diagnosed as hemophilia A, that actually von Willebrand disease (see 193400) is producing the hemorrhagic diathesis with very low levels of factor VIII.

Howard et al. (1988) showed that the mother of a hemophilic boy carried a mutation in the X chromosome she received from her nonhemophilic father rather than in the X chromosome received from her mother. Thus, the father was a gonadal mosaic.

In studies of a sporadic case of hemophilia, Gitschier (1988) found that the mother had partial duplication of the F8 gene. Among her 7 children, in addition to the hemophilic male who had partial deletion of factor VIII, there were some who inherited her normal X chromosome and others who inherited her duplicated X chromosome. Possibly the duplication in the mother predisposed to deletion.

Vidaud et al. (1989) presented evidence that a case of apparent transmission of hemophilia from father to son was due to uniparental disomy. Gamete complementation, involving fertilization of a nullisomic oocyte by a disomic sperm carrying both an X and a Y, was thought to have occurred. More than 15 X-linked DNA markers indicated that the son's X chromosome was inherited from his father.

Nisen and Waber (1989) studied X-chromosome inactivation patterns, as indicated by DNA methylation, in 3 families with hemophilic daughters. One was a case of severe hemophilia B in a girl referred to in 306900. The other 2 were cases of hemophilia A. The maternal and paternal X chromosomes were distinguished by RFLPs, and then patterns of methylation of selected genes on the X-chromosome were determined using methylation-sensitive restriction endonucleases. Of the 6 X-chromosome probes tested, only the PGK (311800) and HPRT (308000) clones were informative. After digestion with HpaII or HhaI, the hybridization intensity of the RFLPs of all 3 mothers and an unaffected sister were diminished by 50%, consistent with random X-chromosome inactivation. The methylation patterns of the X chromosomes of the affected females, however, were clearly nonrandom. Depending on the probe and the patient, HPRT and PGK sequences were either completely methylated or unmethylated. Thus, nonrandom X-chromosome inactivation was the basis for severe hemophilia in these females.

Coleman et al. (1993) reported another unusual mechanism for full-blown hemophilia A in a female, namely, biased X inactivation. A female infant born to a mother with incontinentia pigmenti (IP; 308300) and a father with hemophilia A manifested both disorders. Methylation studies of peripheral blood DNA from the infant, her mother, and 2 female relatives with IP showed a highly skewed pattern of X inactivation. Random patterns were observed in the infant's 2 sisters, who did not have IP and had the usual carrier activity of factor VIII. Coleman et al. (1993) postulated that the usual negative selection against cells with the IP-bearing X chromosome as the active one had unmasked the factor VIII mutation on the infant's other X chromosome. Thus, the infant was functionally hemizygous for the F8 mutation inherited from the father.

Windsor et al. (1995) described another mechanism for severe hemophilia A in a female: the presence of 2 de novo F8 mutations, an X chromosome deletion, and a paternal F8 inversion mutation. Neither parent showed evidence of the mutation in somatic DNA.

Valleix et al. (2002) described monozygotic twin females who were heterozygous for a tyr16-to-cys mutation in the F8 gene (Y16C; 306700.0269) which most probably arose in the paternal germline. Both twins showed skewing of X inactivation toward the maternally derived normal X chromosome, the most severely affected twin exhibiting a higher percentage of inactivation of the normal X chromosome. The degree of skewing of X inactivation closely correlated with both the coagulation parameters and the clinical phenotype of the twins. Monozygotic twins may be monochorionic or dichorionic, depending on whether they develop in a single or 2 distinct chorionic sacs. Dichorionic twinning occurs prior to or around the onset of X inactivation and monochorionic twinning occurs later. A discordant X-inactivation pattern might therefore be expected to be seen more frequently in dichorionic twins. As these twins were monochorionic, Valleix et al. (2002) suggested that the twinning event in this case may have been after the onset of X inactivation.

Bicocchi et al. (2005) reported a 3-generation family in which 3 females were affected with classic hemophilia A due to a heterozygous missense mutation in the F8 gene. All 3 women showed completely skewed X inactivation with expression only of the mutant gene in all tissues analyzed, including leukocytes, skin fibroblasts, uroepithelium, and buccal mucosa. Although no mutations were identified in the XIST gene (314670), Bicocchi et al. (2005) determined that all 3 women had the same XIST allele, and suggested that an alteration within the X-inactivation center on the chromosome carrying the F8C mutation prevented it from being inactivated.

Renault et al. (2007) described a 3-generation family segregating 2 distinct phenotypes, hemophilia A and dramatically skewed X chromosome inactivation, the convergence of which led to the expression of hemophilia A in 3 heterozygous females. All affected males and females had a proximal (type II) IVS22 inversion of the F8 gene. No female carried more than 1 inverted allele. The 3 affected females had skewed X inactivation in favor of the mutant X; 3 unaffected females also had skewed X inactivation, 2 in favor of the normal X, and the third did not carry the mutation. Renault et al. (2007) stated that known causes of skewing were not consistent with their findings in this family, suggesting that the X-chromosome inactivation ratios were genetically influenced (SXI2; 300179).

Diagnosis

Hemophilia A should be suspected whenever unusual bleeding is encountered in a male patient. Laboratory tests show normal platelet count and prothrombin time (PT), but a prolonged activated partial thromboplastin time (PTT). Since hemophilia A and B are clinically similar, specific assays of factor VIII and factor IX (F9; 300746) must be performed. Patients with von Willebrand disease (VWD; 193400) also have factor VIII deficiency secondary to a deficiency of von Willebrand factor (VWF; 613160).

Prenatal Diagnosis

Baty et al. (1986) demonstrated how DNA diagnosis can be helpful in obstetrical decisions and early care of hemophilia even though the family does not make use of the information for elective abortion. Specifically, Cesarean section was performed and the parents were psychologically prepared.

Pecorara et al. (1987) reported a relatively large experience with carrier detection and prenatal diagnosis by means of RFLP analysis.

Kogan et al. (1987) modified the procedure of PCR to use a heat-stable DNA polymerase which allowed the repeated rounds of DNA synthesis to proceed at 63 degrees C. The high sequence specificity of PCR at this temperature enabled detection of restriction-site polymorphisms, contained in PCR products derived from clinical samples to be analyzed by visual inspection of their digestion products on polyacrylamide gels. Kogan et al. (1987) used the improved method to detect carriers of hemophilia A and to diagnose hemophilia prenatally. Erlich et al. (1988) improved the PCR method using thermostable DNA polymerase from Thermus aquaticus.

Lavery (2008) described strategies for preimplantation genetic diagnosis of hemophilia, including embryo sexing, specific mutation analysis, coamplification of polymorphic markers, direct sequencing of F8, and haplotyping after multiple displacement amplification, and discussed the ethical challenges.

Mapping

Linkage studies in the early 1960s indicated that hemophilia A and B (306900) are not allelic (McKusick, 1964). The independence of the 2 loci was confirmed when Robertson and Trueman (1964) found a family in which both hemophilia A and hemophilia B were segregating; one male was deficient in both factor VIII and factor IX. From study of another family in which both hemophilia A and hemophilia B were segregating, Woodliff and Jackson (1966) concluded that the 2 loci are far apart. Direct studies of linkage between hemophilias A and B in the dog indicated that the 2 loci are at least 50 map units apart (Brinkhous et al., 1973).

Haldane and Smith (1947) concluded that there is 5-20% recombination between the color blindness (CBD; 303800) and hemophilia loci with the most probable value about 10%. However, Smith (1968) subsequently concluded that the data on which that estimate was based were heterogeneous, with some families (presumably hemophilia A) showing very close linkage and others (presumably hemophilia B) showing no linkage.

Samama et al. (1977) confirmed assignment of the hemophilia A locus to the long arm of the X chromosome by demonstration of hemophilia in a girl whose mother was a carrier and one of whose X chromosomes had partial deletion of the long arm.

In families of African descent, Boyer and Graham (1965) demonstrated close linkage of hemophilia A and the A/B polymorphism of G6PD (305900). Filippi et al. (1984) stated that 58 scorable sibs, all nonrecombinant for the linkage of HEMA and G6PD, were known by that time. From this, they inferred that the 90% upper limit of meiotic recombination between the 2 loci is less than 4%.

Harper et al. (1984) did linkage studies with the DNA probe DX13, which had been localized to band Xq28. When DNA is digested with the restriction enzyme BglII, the probe recognizes an RFLP for which 50% of females are heterozygous. No recombination was observed between the HEMA and DX13 loci. The workers concluded that the marker is useful for carrier detection and prenatal diagnosis. About 30% recombination was found between the factor VIII and IX loci.

Oberle et al. (1985) observed very close linkage of a polymorphic anonymous DNA probe called St14 (from Strasbourg, France). No recombination (theta=0) was found in 12 families (lod score 9.65). The probe was informative in more than 90% of families, and the authors suggested that it could be used in conjunction with assays of factor VIII to identify carriers with 96% confidence or better. St14 could be used for prenatal diagnosis of disorders such as hemophilia A and adrenoleukodystrophy because of close linkage (Oberle et al., 1985). Janco et al. (1987) used the more accurate intragenic F8 RFLPs to detect hemophilia A carriers.

In a 9-year-old Malaysian female with de novo hemophilia A as well as a complex de novo translocation involving one X chromosome and one chromosome 17 (Muneer et al., 1986), Migeon et al. (1993) identified a breakpoint within Xq28 with deletion of the 5-prime end of the factor VIII gene, leaving the more proximal G6PD locus intact on the derivative chromosome 17. As the deleted segment included the 5-prime half of F8C as well as the subtelomeric DXYS64 locus, they concluded that F8 is oriented on the chromosome with its 5-prime region closest to the telomere.

Molecular Genetics

Ratnoff and Bennett (1973) reviewed the genetics of hereditary disorders of blood coagulation.

Gitschier et al. (1985) identified truncating mutations in the F8 gene (see, e.g., 300841.0001-300841.0003) as the basis for hemophilia A. A severe hemophiliac with no detectable factor VIIIC activity had an R2307X mutation (306700.0001). Gitschier et al. (1986) found that the same codon was converted to glutamine (R2307Q; 306700.0042) in a mild hemophiliac with 10% of normal activity. A diminished level of factor VIII Ag in the latter patient coincided with the level of clotting activity, suggesting that the abnormal factor VIII was relatively unstable.

In a study of 83 patients with hemophilia A, Youssoufian et al. (1986) identified 2 different point mutations, one in exon 18 and one in exon 22, that recurred independently in unrelated families. Each mutation produced a nonsense codon by a change of CG to TG. In the opinion of Youssoufian et al. (1986), these observations indicated that CpG dinucleotides are mutation hotspots. It had been postulated that methylated cytosines may be mutation hotspots because 5-methylcytosine can spontaneously deaminate to thymine, resulting in a C-to-T transition in DNA.

Youssoufian et al. (1987) characterized 5 different partial deletions of the F8 gene in 83 patients with hemophilia. None had developed circulating inhibitors. One of the deletions occurred de novo in a germ cell of the maternal grandmother, while a second deletion occurred in a germ cell of a maternal grandfather. The findings indicated that de novo deletions of X-linked genes can occur in either male or female gametes. Youssoufian et al. (1988) reported 6 other partial F8 gene deletions in severe hemophilia A, bringing to 12 the number of deletions among 240 patients. No association was observed between the size or location of deletions and the presence of inhibitors to factor VIII. Furthermore, no 'hotspots' for deletion breakpoints were identified.

Youssoufian et al. (1988) screened 240 patients with hemophilia A and found CG to TG transitions in an exon in 9. They identified novel missense mutations leading to severe hemophilia A and estimated that the extent of hypermutability of CpG dinucleotides is 10 to 20 times greater than the average mutation rate for hemophilia A.

Cooper and Youssoufian (1988) collated reports of single basepair mutations within gene coding regions causing human genetic disease. They found that 35% of mutations occurred within CpG dinucleotides. Over 90% of these mutations were C-to-T or G-to-A transitions, which thus occur within coding regions at a frequency 42-times higher than that predicted from random mutation. Cooper and Youssoufian (1988) believed these findings were consistent with methylation-induced deamination of 5-methylcytosine and suggested that methylation of DNA within coding regions may contribute significantly to the incidence of human genetic disease.

Higuchi et al. (1988) found deletion of about 2,000 bases spanning exon 3 and part of IVS3 of the F8 gene in a patient with severe hemophilia A. The mother was judged to be a somatic mosaic because the defective gene could be identified in only a portion of the leukocytes and cultured fibroblasts.

In a review, Antonarakis et al. (1995) collected the findings of more than 1,000 hemophilia subjects examined for F8 gene mutations. These include point mutations, inversions, deletions, and unidentified mutations which constitute 46%, 42%, 8%, 4%, and 91%, 0%, 0%, and 9%, respectively, of those with severe versus mild to moderate disease, respectively, in selected studies. The 266 point mutations described as of April, 1994 comprised missense (53%), CpG-to-TpG (16%), small deletions (12%), nonsense (9%), small inversions and splicing (3% each), and missense polymorphisms and silent mutations in exons (2% each). In addition to these point mutations 100 different larger deletions and 9 insertion mutations had been reported.

In a study of 147 sporadic cases of severe hemophilia A, Becker et al. (1996) were able to identify the causative defect in the F8 gene in 126 patients (85.7%). An inversion of the gene was found in 55 patients (37.4%), a point mutation in 47 (32%), a small deletion in 14 (9.5%), a large deletion in 8 (5.4%), and a small insertion in 2 (1.4%). In 4 (2.7%), mutations were localized but not yet sequenced. No mutation was identified in 17 patients (11.6%). The identified mutations occurred in the B domain in 16 (10.9%); 4 of these were located in an adenosine nucleotide stretch at codon 1192, indicating a mutation hotspot. Somatic mosaicism was detected in 3 (3.9%) of 76 patients' mothers, comprising 3 of 16 de novo mutations in the patients' mothers. Investigation of family relatives allowed detection of a de novo mutation in 16 of 76 2-generation and 28 of 34 3-generation families. On the basis of these data, Becker et al. (1996) estimated the male:female ratio of mutation frequencies (k) to be 3.6. By use of the quotients of mutation origin in maternal grandfather to patients' mother or to maternal grandmother, k values were directly estimated as 15 and 7.5, respectively. Considering each mutation type separately, they found a mutation type-specific sex ratio of mutation frequencies. Point mutations showed a 5-to-10-fold-higher and inversions a more than 10-fold-higher mutation rate in male germ cells, whereas deletions showed a more than 5-fold-higher mutation rate in female germ cells. Consequently, and in accordance with the data of other disorders such as Duchenne muscular dystrophy, the results indicated to Becker et al. (1996) that at least for X-chromosomal disorders the male:female mutation rate is determined by its proportion of the different mutation types.

The molecular diagnosis of hemophilia A is challenging because of the high number of different causative mutations that are distributed through the large F8 gene. The putative role of the novel mutations, especially missense mutations, may be difficult to interpret as causing hemophilia A. Guillet et al. (2006) identified 95 novel mutations out of 180 different mutations found among 515 patients with hemophilia A from 406 unrelated families followed up at a single hemophilia treatment center in a Paris hospital. The 95 novel mutations comprised 55 missense mutations, 12 nonsense mutations, 11 splice site mutations, and 17 small insertions/deletions. They used a strategy in interpreting the causality of novel F8 mutations based on a combination of the familial segregation of the mutation, the resulting biologic and clinical hemophilia A phenotype, and the molecular consequences of the amino acid substitution. For the latter, they studied the putative biochemical modifications: its conservation status with cross-species factor VIII and homologous proteins, its putative location in known factor VIII functional regions, and its spatial position in the available factor VIII 3D structures.

Among 1,410 Italian patients with hemophilia A, Santacroce et al. (2008) identified 382 different mutations in the F8 gene, 217 (57%) of which had not previously been reported. Mutations leading to a null allele accounted for 82%, 15%, and less than 1% of severe, moderate, or mild hemophilia, respectively. Missense mutations were identified in 16%, 68%, and 81% of severe, moderate, or mild hemophilia, respectively, yielding a good genotype/phenotype correlation useful for treatment and genetic counseling.

In order to establish a national database of F8 mutations, Green et al. (2008) identified and cataloged multiple mutations in approximately one-third of the U.K. hemophilia A population. The risk of developing inhibitors for patients with nonsense mutations was greater when the stop codon was in the 3-prime half of the mRNA. The most common change was the intron 22 inversion (306700.0067), which accounted for 16.6% of all mutations and for 38% of those causing severe disease.

Inversion Mutations in Intron 22 of the F8 Gene

Intron 22 of the human F8 gene is hypomethylated on the active X and methylated on the inactive X. Inaba et al. (1990) described an MspI RFLP in intron 22 of the F8 gene. Japanese showed 45% heterozygosity and Asian Indians showed 13%; polymorphism was not found in American blacks or Caucasians.

Naylor et al. (1992) found an unusual cluster of mutations involving regions of intron 22 not examined earlier and leading to defective joining of exons 22 and 23 in the mRNA (300841.0067) as the cause of hemophilia A in 10 of 24 severely affected UK patients. These results confirmed predictions about the efficacy of the mRNA-based method suggested by Naylor et al. (1991), and also excluded hypotheses proposing that mutations outside the F8 gene are responsible for a large proportion of severe hemophilia A.

Of the 28 patients reported by Naylor et al. (1993), 5 had mild or moderate disease and all had a missense mutation. The other 23 patients were severely affected; unexpectedly, intron 22 seemed to be the target of approximately 40% of the mutations causing severe hemophilia A. Naylor et al. (1993) found that the basis of the unique F8 mRNA defect that prevented PCR amplification across the boundary between exons 22 and 23 was an abnormality in the internal regions of intron 22. They showed that exons 1-22 of the F8 mRNA had become part of a hybrid message containing new multi-exonic sequences expressed in normal cells. The novel sequences were not located in a YAC containing the whole F8 gene. Southern blots from patients probed by novel sequences and clones covering intron 22 showed no obvious abnormalities. Naylor et al. (1993) also suggested that inversions involving intron 22 repeated sequences are the basis of the mRNA defect. These mutations in severely affected patients occur at the surprising rate of approximately 4 x 10(-6) per gene per gamete per generation. Furthermore, it has been shown that these de novo inversions occur more frequently in males than females with a ratio of 302:1 estimated in male:female germ cells.

The F8A gene (305423) is contained entirely within intron 22 of the F8 gene and is transcript in the reverse orientation from the F8 gene (Levinson et al., 1990). Lakich et al. (1993) proposed that many of the previously unidentified mutations resulting in severe hemophilia A are based on recombination between the homologous F8A sequences within intron 22 and upstream of the F8 gene. Such a recombination would lead to an inversion of all intervening DNA and a disruption of the gene. Lakich et al. (1993) presented evidence to support this model and described a Southern blot assay that detects the inversion. They suggested that this assay should permit genetic prediction of hemophilia A in approximately 45% of families with severe disease.

Inversion mutations resulting from recombinations between DNA sequences in the A gene in intron 22 of the F8 gene and 1 of 2 other A genes upstream to F8 have been shown to cause a large portion of cases. From data on more than 2,000 samples, Antonarakis et al. (1995) concluded that the common inversion mutations are found in 42% of all severe hemophilia A subjects. Whereas 98% of the mothers of those with inversions were carriers of the inversion, only about 1 de novo inversion was found in maternal cells for every 25 mothers of sporadic cases. When the maternal grandparental origin of inversions was examined the ratio of de novo occurrences in male:female germ cells was 69:1.

Brinke et al. (1996) reported the presence of a novel inversion in 2 hemophilic monozygotic twins. These patients showed an inversion that affects the first intron of the F8 gene, displacing the most telomeric exon (exon 1) of F8 further towards the telomere and close to the C6.1A gene (BRCC3; 300617). Brinke et al. (1996) noted that this novel inversion creates 2 hybrid transcription units. One of these is formed by the promoter and first exon of F8 and widely expressed sequences that map telomeric to the C6.1A sequence. The other hybrid transcription unit contains the CpG island and all of the known sequence of C6.1A and the 3-prime section of most of the F8 gene.

It is hypothesized that the inversion mutations occur almost exclusively in germ cells during meiotic cell division by an intrachromosomal recombination between a 9.6-kb sequence within intron 22 and 1 of 2 almost identical copies located about 300 kb distal to the F8 gene at the telomeric end of the X chromosome. Most inversion mutations originate in male germ cells, where the lack of bivalent formation may facilitate flipping of the telomeric end of the single X chromosome. Oldenburg et al. (2000) reported the first instance of intron 22 inversion presenting as somatic mosaicism in a female, affecting only about 50% of lymphocyte and fibroblast cells of the proposita. Supposing a postzygotic de novo mutation as the usual cause of somatic mosaicism, the finding implies that the intron 22 inversion mutation is not restricted to meiotic cell divisions but can also occur during mitotic cell divisions, either in germ cell precursors or in somatic cells.

Development of Factor VIII Inhibitors

Approximately 10 to 20% of patients with severe hemophilia A develop antibodies, known as inhibitors, to factor VIII following treatment with exogenous factor VIII. Most of these patients have nonsense mutations or deletions in the F8 gene (Antonarakis et al., 1995).

Antonarakis et al. (1985) identified several molecular defects in families with hemophilia A. One family had a deletion of about 80 kb in the F8 gene, whereas another had a single nucleotide change in the coding region of the gene, resulting in a nonsense codon and premature termination. In addition, they used 2 common polymorphic sites in the F8 gene to differentiate the normal gene from the defective gene in 4 of 6 obligate carriers from families with patients in whom inhibitors did not develop. In both the family with a large deletion and the family with premature termination, affected persons developed inhibitors.

A variety of F8 gene mutations have been found in patients with hemophilia A due to inhibitors. Among 30 such cases, Antonarakis et al. (1995) found that 87% and 13% had different nonsense and missense mutations, respectively. F8 gene inversions do not seem to be a major predisposing factor for the development of inhibitors. Among severe hemophilia A cases, 16% of those without inversions and 20% of those with inversions developed inhibitors.

Schwaab et al. (1995) found that the probability of developing factor VIII inhibitors is greater in patients with large deletions in the F8 gene.

Viel et al. (2009) sequenced the F8 gene in 78 black patients with hemophilia to identify the causative mutations and background haplotypes, which the authors designated H1 to H5. They found that 24% of the patients had an H3 or H4 haplotype, and that the prevalence of inhibitors was higher among patients with either of those haplotypes than among patients with haplotypes H1 or H2 (odds ratio, 3.6; p = 0.04), despite a similar spectrum of hemophilic mutations and degree of severity of illness in the 2 subgroups. Noting that Caucasians carry only the H1 or H2 haplotypes and that most blood donors are Caucasian, Viel et al. (2009) suggested that mismatched factor VIII replacement therapy might be a risk factor for the development of anti-factor VIII alloantibodies.

Genotype/Phenotype Correlations

In a Japanese family with mild to moderately severe hemophilia A, Young et al. (1997) found a deletion of a single nucleotide T within an A(8)TA(2) sequence of exon 14 of the F8 gene. The severity of the clinical phenotype did not correspond to that expected of a frameshift mutation. A small amount of functional factor VIII protein was detected in the patient's plasma. Analysis of DNA and RNA molecules from normal and affected individuals and in vitro transcription/translation suggested a partial correction of the molecular defect, because of the following: (i) DNA replication/RNA transcription errors resulted in restoration of the reading frame and/or (ii) 'ribosomal frameshifting' resulted in the production of normal factor VIII polypeptide and, thus, in a milder-than-expected hemophilia A. All of these mechanisms probably were promoted by the longer run of adenines, A(10) instead of A(8)TA(2), after the deleted T. Young et al. (1997) concluded that errors in the complex steps of gene expression therefore may partially correct a severe frameshift defect and ameliorate an expected severe phenotype.

Cutler et al. (2002) identified 81 mutations in the F8C gene in 96 unrelated patients, all of whom had previously typed negative for the common IVS22 inversion mutation (306700.0067). Forty-one of these mutations were not recorded in F8C gene mutation databases. Analysis of these 41 mutations with regard to location, possible cross-species conservation, and type of substitution, in correlation with the clinical severity of the disease, supported the view that the phenotypic result of a mutation in the F8C gene correlates more with the position of the amino acid change within the 3-dimensional structure of the protein than with the actual nature of the alteration.

Clinical Management

The mainstay of routine treatment for hemophilia A is infusion of factor VIII using amounts that are required to restore the factor VIII activity to therapeutic levels. Desmopressin (dDAVP), a synthetic analog of the neurohypophyseal nonapeptide arginine vasopressin (AVP; 192340), has been approved for treatment of mild hemophilia A and von Willebrand disease. Following dDAVP in some cases concentrations of factor VIII and von Willebrand factor are transiently increased to levels that allow minor surgery (Richardson and Robinson, 1985; review by Hoyer, 1994).

Lewis et al. (1985) reported that a hemophiliac who received a liver transplant from a normal donor had nearly normal levels of factor VIII coagulant activity in the postoperative period.

Nilsson et al. (1988) used combined cyclophosphamide, intravenous IgG, and factor VIII therapy to induce immune tolerance to factor VIII infusions in patients with antibodies to factor VIII. Factor VIII coagulant antibodies disappeared in 9 of 11 patients so treated; the other 2 patients did not respond. Earlier treatment with either factor VIII and cyclophosphamide or factor VIII and IgG had been ineffective, suggesting that all 3 components of the protocol are necessary for the successful induction of tolerance. Pignone et al. (1992) had success within induction of immune tolerance in patients with hemophilia A and factor VIII inhibitors: combined treatment with gammaglobulin, cyclophosphamide, and factor VIII.

Schwartz et al. (1990) used antihemophilic factor produced by recombinant DNA methods in the successful treatment of hemophilia A in 107 subjects. The half-lives equaled or exceeded those of plasma-derived factor VIII, and immunogenicity appeared to be no greater. This represented a major advance because of the opportunity to avoid exposure to transfusion-associated viral diseases. F8 was one of the largest genes cloned to that time and, with the study of Schwartz et al. (1990), became the largest cloned protein to be used in clinical trials.

Through a suppressive effect on premature termination codons, aminoglycoside antibiotics such as gentamicin have been used for therapeutic benefit in a number of conditions including cystic fibrosis (602421) and Duchenne muscular dystrophy (300377). James et al. (2005) evaluated the effect of gentamicin on the factor VIII and factor IX levels of severe hemophiliacs with known nonsense mutations. They concluded that gentamicin was unlikely to be an effective treatment for severe hemophilia because of its potential toxicity and the minimal response observed.

In hemophilic SCID mice, Aronovich et al. (2006) reported successful treatment of hemophilia by transplantation of fetal pig spleen harvested at embryonic day 42 before the appearance of mature T cells. The transplanted tissue exhibited good growth and subsequent expression of factor VIII, leading to complete alleviation of hemophilia within 2 to 3 months after transplant. The results provided proof of principle that transplantation of fetal spleen can correct hemophilia while avoiding graft-versus-host disease (GVHD; see 614395).

Development of Factor VIII Inhibitors

An acquired disorder resembling hemophilia A can be caused by the development by autoantibodies against factor VIII (Nilsson and Lamme, 1980; Zimmerman et al., 1971). Approximately 10 to 20% of patients with severe hemophilia A develop antibodies, known as inhibitors, to factor VIII following treatment with exogenous factor VIII. Most of these patients have nonsense mutations or deletions in the F8 gene (Antonarakis et al., 1995).

Frommel et al. (1977) studied 10 sibships of hemophilia A, each of which included 1 or 2 hemophilic brothers with antibody to factor VIII. Their results suggested linkage of the MHC (142800