Gene therapy of the hemophilias

Development of hemophilia gene therapy depends on testing gene transfer vectors in hemophilic and nonhemophilic animals. Available animal models include factor VIII or factor IX knockout mice as well as dogs with spontaneous hemophilia A or B. Large animals (particularly dogs) more closely replicate the requirements for correction of human hemophilia than do mice. Small animals are more convenient to maintain and require significantly less vector for testing than do large animals. Nonhemophilic animals (mice or nonhuman primates), whose endogenous factor VIII and factor IX complicate analysis of the human proteins, have utility for safety testing of vectors; some assays can discriminate between human coagulation factors and the endogenous coagulation factors. Most animal models suffer the limitations imposed by the immune response to human factor VIII or IX protein. Clinical trials have failed to achieve significant factor VIII expression in hemophilia A patients, while one clinical trial in hemophilia B patients showed only transient therapeutic increments of factor IX expression. Gene therapy remains an investigational method with many obstacles to overcome before it can be widely used as treatment for hemophilia.     

Immunologic aberrations, HIV seropositivity and seroconversion rates in patients with hemophilia B

Because there have been reports that factor IX concentrate is less immunosuppressive and therefore factor IX users have less immunologic aberrations, we have studied a group of 22 patients with hemophilia B and six patients with factor VIII deficiency and high titer inhibitors with respect to lymphocyte numbers and function, human immunodeficiency virus (HIV) serology, and factor usage. This group was compared to 111 patients with hemophilia A and a group of 28 healthy male volunteer controls. When the study began in 1983, the majority of patients with hemophilia B and with higher titer factor VIII inhibitors were seronegative, 77% and 83% respectively, as compared to only 30% of patients with hemophilia A. At that time the factor IX users also had milder immune aberrations than the hemophilia A group. However, with time and increasing clotting factor concentrate usage, seroconversion and more striking abnormalities in immune function have occurred in the hemophilia B group. In a subgroup of 16 patients with hemophilia B studied twice, the incidence of seropositivity increased from 31% in 1983 to 69% in 1985. We thus conclude that factor IX concentrate in itself is not less immunosuppressive than factor VIII concentrate. Seroconversion in factor IX concentrate users appears to be lagging behind seroconversion in factor VIII concentrate users, perhaps secondary to the lower cumulative dosage of concentrate that patients with hemophilia B utilize.     

Immune complexes in hemophilia

Circulating immune complexes (ICs), assayed by the L1210 enzyme-linked immunoassay, were detected in 48% of patients with hemophilia. A, 50% of patients with von Willebrand's disease, and in none of our patients with hemophilia B. Eighty-five % of the hemophilia A and B patients had mild to moderate disease with only one patient demonstrating a circulating inhibitor. No correlation was found between IC levels and hepatitis B infection, SGOT, disease severity, total quantity of factor VIII or IX infused, time interval from list infusion, or rheumatoid factor positivity. Although the nature of the ICs is not known, the similarity of IC levels between hemophilia A and von Willebrand's disease is discussed with regard to antibodies generated to non-procoagulant portions of the factor VIII molecule.     

Inhibitor antibodies to factor VIII and factor IX: management

Inhibitor antibodies directed against factor VIII or factor IX present challenges to the clinician. Fortunately, several management options are available, although each has disadvantages as well as advantages. Alloantibodies against factor VIII (which develop in 25 to 50% of children with severe hemophilia A, as well as in a small percentage of children with mild or moderate hemophilia A) may be low titer and transient or may be high titer. Most patients with high-titer problematic inhibitors now try to eliminate the inhibitor by using one of several immune tolerance induction (ITI) regimens. For treatment of bleeding episodes in patients who have high-titer (> or = 5 Bethesda units) inhibitors, one can use a prothrombin complex concentrate (PCC) (preferably an activated PCC [APCC]), recombinant (r) factor VIIa, or porcine factor VIII. The choice of product is generally dependent on the type and severity of the patient's bleeding, degree of cross-reactivity of the patient's inhibitor with porcine factor VIII, physician familiarity with the product, product availability, and cost. In persons with hemophilia B, alloantibodies occur in only 1 to 3% of severely affected individuals. However, in roughly half of those who develop inhibitors, anaphylaxis or severe allergic reactions occur on infusion of any type of factor IX-containing product. This phenomenon usually develops after relatively few exposures to factor IX; thus it is recommended that the first 10 to 20 infusions of factor IX given to children with severe hemophilia B be given in a setting equipped for treatment of shock. For treatment of bleeding episodes in patients with severe allergic reactions, rF VIIa is the treatment of choice. ITI has been less successful in hemophilia B patients with inhibitors than in those with hemophilia A, and in a subgroup of patients with severe allergic reactions who were desensitized to factor IX and then tried on ITI, results were even poorer. Additionally, several developed nephrotic syndrome while on ITI. For hemophilia B patients with inhibitors who do not have allergic reactions to factor IX, bleeding episodes can be treated with PCC or APCC or with rF VIIa. Autoantibodies directed against factor VIII are rare but can occur in a variety of settings. They occur mainly in adults, and bleeding is often severe and life threatening. Although some factor VIII autoantibodies disappear spontaneously, most require immunosuppression. Corticosteroids and cyclophosphamide are generally recommended. For treatment of bleeding, therapeutic options include (human) factor VIII concentrates, porcine factor VIII, APCC, and rFVIIa. The choice of product is generally determined by the consulting hematologist's familiarity with the product, product availability and cost, as well as response to treatment.     

Von Willebrand disease within the collective of haemophilic patients as reason for unexpected bleeding episodes

In the treatment of haemophilia A and B, recombinant coagulation factors are increasingly replacing plasma-derived factor VIII and IX concentrates. Although provided with replacement therapy, individual patients may exhibit bleeding episodes, which are difficult to control. These bleeds may be caused by von Willebrand disease (VWD) as an additional underlying coagulation disorder. We report in the present study our experience that in the collective of haemophilic patients, VWD must be anticipated at least with the same order of magnitude as it appears in a normal healthy population. Among the patients at our treatment centre, two patients (1.5%) were identified as suffering from VWD in addition to haemophilia A. In the collective of haemophilia B patients at our centre, three patients (10%) with VWD were found. Two of these patients exhibited unexpected severe bleeding episodes, which could only satisfactorily be controlled by the administration of Haemate-P or DDAVP supplementary to the recombinant coagulation factor concentrate.     

Laboratory methods for the genetic diagnosis of bleeding disorders

Accurate carrier detection and pre-natal diagnosis in haemophilia A and B and in von Willebrand's disease (VWD) can be achieved by genetic analysis. The spectrum of mutations responsible for these three disorders is described. Methods for linkage analysis using intra genic diallelic and multiallelic factor VIII and IX gene polymorphisms are mentioned, and situations where their use in carrier detection is inappropriate or fails are discussed. Linkage analysis for examination of von Willebrand factor gene inheritance in families with VWD is also described. Screening for the factor VIII gene inversion in patients with severe haemo philia A and the use of the factor VIII binding assay as a discriminant test in patients with possible mild haemophilia A or VWD are described. Point mutation screening methods AMD, CSGE, DGGE, SSCP and UHG analysis are also detailed. The variety of possible analyses available to genetically diagnose haemophilia A, B and VWD is explored.     

Cost-effectiveness in establishing hemophilia carrier detection and prenatal diagnosis services in a developing country with limited health resources

The cost-effectiveness of carrier detection and prenatal diagnosis for hemophilia at the International Hemophilia Training Center, Bangkok, Thailand was studied. From 1991 to 2002, 209 females from 124 families with hemophilia A and B were included. There were 180 hemophilia A carriers and 29 hemophilia B carriers which could be classified into 78 obligate and 131 possible carriers. The phenotypic analysis for possible carriers involved the determination of levels of factor VIII or IX clotting activity (FVIII:C, FIX:C) and the ratio of FVIII:C and von Willebrand factor antigen. The result revealed that 49 females (37.4%) were diagnosed as carriers, 65 females (49.6%) were normal and 17 females (13%) were undetermined. Additional genotypic analysis was provided to 46 families with 74 females with obligate, proven or undetermined carriers within the reproductive life. The polymorphisms associated with factor VIII and IX genes were used including Bcl I for the factor VIII gene and combined use of Mse I, Sal I, Nru I, Hha I and Dde I for the factor IX gene. The informative rate was 59.4% (44/74). Consequently, 12 prenatal diagnoses for fetus at risk were performed. Sex determination was initially determined and followed by the diagnosis of hemophilia through informative gene tracking and/or the measurement of fetal levels of FVIII:C or FIX:C. The result revealed that 3 male fetuses were affected. The total cost of carrier detection and prenatal diagnosis that the families had to pay in the government hospital was 238,600 Baht (US dollars 5,965). It was compared to the estimated cost of minimal replacement therapy using lyophilized cryoprecipitate for the survival time of 30 years in one patient with hemophilia of 1,012,500 Baht (US dollars 25,312.5). The cost of prevention was much less than the replacement therapy. In conclusion, it is cost-effective to establish the service for carrier detection and prenatal diagnosis for hemophilia especially in developing countries with limited health resources.     

Kinetic aspects of the removal of IgG and inhibitors in hemophiliacs using protein A immunoadsorption.

Seven patients with hemophilia A and B and 3 patients with acquired hemophilia were treated on 13 and 4 occasions, respectively, with protein A immunoadsorption to reduce anti-factor VIII or IX antibodies. Usually the patients were treated on 2 consecutive days. On each treatment day an average of 3 (range: 1.02-5.83) plasma volumes were processed in the congenital patients and 1.5 (range: 1.02-2.90) in those with acquired hemophilia. Plasma levels of IgG decreased from 18.9 +/- 1.9 to 3.1 +/- 1.2 g/l in the congenital group, and from 11.5 +/- 2.3 to 2.3 +/- 0.6 g/l in the acquired group. In the congenital hemophiliacs a corresponding reduction in inhibitor level of 70-95% was regularly seen; in 1 exceptional patient the inhibitor was reduced from 4,350 to 12 Bethesda Units/ml (BU/ml) during 5 days of treatment. In the congenital hemophiliacs immunoadsorption was followed by factor infusion to peak levels between 8 and 215 IU/dl. In the patients with acquired hemophilia a satisfactory reduction in inhibitor levels was obtained in 2 of the 4 treatments, which were followed by DDAVP or factor infusion. Some recommendations for the use of protein A immunoadsorption in the treatment of hemophilic patients will be given.     

Oral treatment of hemophilia A using traditional kanpo medicine, Huang-lien-chieh-tu-tang (plant extract)

Huang-lien-chieh-tu-tang (HLCT) extract was given orally to a patient with severe hemophilia A. The plasma concentration of factor VIII rose from less than 1 to 41% 1 h after oral administration of HLCT extract. APTT also decreased from 102.0 to 70.9 s. In vitro experiments, HLCT showed factor VIII, IX and X biological activity at concentrations around 0.08%. There was no factor-VIII-antigen-related activity in HLCT.     

Danazol fails to increase factor VIII or IX levels in a double-blind crossover study of patients with haemophilia A and B

Twenty-one patients, 14 with haemophilia A and seven with haemophilia B, completed a double-blind crossover study to evaluate the effects of danazol on factor VIII and factor IX levels. Clotting and immunoradiometric assays were used to measure factor levels at baseline, 2 weeks and 8 weeks on both danazol and placebo. Fibrinogen, plasminogen and activated partial thromboplastin time were measured on all patients during placebo and danazol treatment. Although plasminogen levels rose significantly (P less than 0.01) and fibrinogen decreased (P less than 0.01), factor VIII and IX levels did not change. While on danazol, three patients had increased bleeding and shortened euglobulin lysis times compared to their baseline levels. We conclude that danazol does not raise factor VIII or IX levels and increases bleeding in some patients.     

von Willebrand's disease: an important cause of dysfunctional uterine bleeding.

We assessed the prevalence of von Willebrand's disease (VWD) in patients with objectively confirmed dysfunctional uterine bleeding. A case-control study was designed to include 38 patients with objectively confirmed dysfunctional uterine bleeding and 38 age-matched controls with normal menstrual blood loss (MBL). Menorrhagia was defined as a mean MBL of greater than 80 ml on three consecutive menses as measured by the alkali haematin method. von Willebrand factor antigen, von Willebrand factor activity (VWF:Ac) and factor VIII:C were measured on three serial venous blood samples 1 week apart. VWD was diagnosed in five of 38 (13%) patients with menorrhagia and one of 38 (2.6%) patients with normal menstrual blood loss. The mean VWF:Ac value was significantly reduced in patients with menorrhagia (mean +/- standard deviation, 84.5 +/- 26.7 IU/dl versus 103.9 +/- 34.5 IU/dl; P < 0.01) and this effect persisted after exclusion of patients diagnosed with VWD. Failure to investigate patients for VWD will limit the potential benefits of medical therapies such as tranexamic acid or nasal desmopressin [1-desamino-8-D-arginine vasopressin, (DDAVP)] and, in addition, will lead to an increased risk associated with surgical intervention in patients with undiagnosed VWD.     

A high-potency, single-donor cryoprecipitate of known factor VIII content dispensed in vials

Current factor VIII products expose recipients to many donors and hence to a high risk of acquiring blood-borne infections. Plasma-exchange donation of cryoprecipitate can reduce donor exposure by repeatedly obtaining large yields of factor VIII from individual donors. In this study, donor factor VIII levels were stimulated with desmopressin before donation. Mean yield per donation increased from 1399 +/- 425 IU in controls to 3818 +/- 1350 IU in stimulated donations (p less than 0.001), and mean factor VIII concentration in the cryoprecipitate increased from 8.2 +/- 3 IU/mL to 24 +/- 12 IU/mL (p less than 0.001). A new packaging system dispenses assayed aliquots of stimulated cryoprecipitate in plastic vials. The direct cost of production for this material is $.065 per unit. The cryoprecipitate is hemostatically active and convenient to use, and the aggregate yields from sequential donations by stimulated persons are high enough to allow long-term, single-donor support of many adults with hemophilia.     

Cold storage of citrated whole blood induces drastic time-dependent losses in factor VIII and von Willebrand factor: potential for misdiagnosis of haemophilia and von Willebrand disease.

This study investigates the effect of pre-analytic storage conditions on the laboratory evaluation of von Willebrand disease (VWD) and haemophilia. Samples from healthy controls and patients with VWD were stored as whole blood and as separated plasma, both at room temperature and on crushed ice, for two different time periods (3 or 6 h). In samples from healthy individuals (n=10) and in patients with suspected type 1 VWD (n=10), storage of whole blood on ice caused a drastic time-dependent decrease in von Willebrand factor (VWF):ristocetin cofactor activity, in VWF:antigen activity and factor VIII activity (mean+/-SD) to 35+/-18, 55+/-23 and 53+/-15% of baseline levels after 6 h storage, respectively. Patients with type 2 VWD and non-detectable VWF:ristocetin cofactor activity did not demonstrate such drastic cold-induced losses in VWF and factor VIII levels. Storage of plasma caused only minor changes in VWF levels. The cold-induced loss in VWF might thus depend on the presence of platelets and of high molecular weight VWF. Chilling of platelets induces a clustering of the glycoprotein Ib subunit. We therefore hypothesize that cold-induced loss in VWF might be due to a cold-promoted binding of VWF to glycoprotein VWF receptor Ib alpha. These results suggest a serious potential for misdiagnosis of haemophilia or VWD due to inappropriate pre-analytical handling of blood.     

The effect of near-normoglycemic control on plasma factor VIII/von Willebrand factor and fibrin degradation products in insulin-dependent diabetic patients

Diabetic patients have elevated plasma levels of factor VIII/von Willebrand factor (F VIII/vWF), and such elevations have been linked to vascular endothelial injury. In a prospective study we investigated the effect of metabolic regulation on the plasma levels of F VIII/vWF and cross-linked fibrin degradation products (XL-FDP), an indicator of intravascular coagulation, in 15 insulin-dependent diabetic patients who had no demonstrable vascular abnormalities. Eight patients had newly diagnosed diabetes, and 7 had been diabetic for an average of 12 yr. The patients were tested before and 1, 2, 4, and 8 weeks after the start of a structured diabetes education and care program, including introduction of a basal-bolus form of insulin treatment. Treatment for 8 weeks resulted in a highly significant improvement of metabolic control [hemoglobin Aic, 11.1 +/- 1.3% (+/- SD) vs. 6.8 +/- 1.0%; plasma fructosamine, 4.8 +/- 1.0 vs. 2.9 +/- 0.7 mmol/L; plasma glucose, 13.5 +/- 4.2 vs. 6.3 +/- 2.2 mmol/L; P less than 0.0001, respectively]. Compared to age- and sex-matched normal subjects, plasma activity of factor VIII (F VIII:C) was significantly elevated in the diabetic patients initially (1.5 +/- 0.6 vs. 1.0 +/- 0.1 x 10(3) U/L; P less than 0.01). After 2 weeks of intensified therapy it was 1.1 +/- 0.4 x 10(3) U/L. The mean plasma vWF value also was significantly elevated initially [vWF antigen, 1.8 +/- 0.7; normal group, 0.9 +/- 0.1 x 10(3) U/L (P less than 0.01); vWF ristocetin cofactor activity, 1.9 +/- 0.9; normal group, 1.0 +/- 0.3 x 10(3) U/L (P less than 0.001)] and decreased significantly after only 1 week of therapy. In the following 7-week period plasma vWF remained near normal. Plasma XL-FDP levels were elevated in all patients initially (190 +/- 150; normal group, 35 +/- 30 micrograms/L): the value was most abnormal in the patients with newly diagnosed disease (300 +/- 150 micrograms/L), indicating intravascular fibrin formation. The mean XL-FDP level declined significantly in the patients with newly diagnosed diabetes after 1 week of therapy; in the other patients, however, XL-FDP levels remained slightly elevated. In all 15 patients the plasma F VIII:C and XL-FDP levels were correlated significantly at all times. The plasma vWF and XL-FDP levels were correlated after 1, 2, 4, and 8 weeks of treatment as were the plasma vWF levels and glucose concentrations before and 1 and 2 weeks after the start of treatment program.(ABSTRACT TRUNCATED AT 400 WORDS)     

Adeno-associated virus-mediated gene transfer for hemophilia B

Hemophilia is the bleeding diathesis caused by mutations in the gene encoding factor VIII (hemophilia A) or factor IX (hemophilia B). Currently, the disease is treated by intravenous infusion of the missing purified clotting factor. The goal of gene transfer for treating hemophilia is to achieve sustained expression of factor VIII or factor IX at levels high enough to improve the symptoms of the disease. Hemophilia has proven to be an attractive model for those interested in gene transfer, and multiple gene-transfer strategies are currently being investigated for the hemophilias. The most promising preclinical studies have been with adeno-associated viral vectors (AAV); introduction of AAV vectors expressing factor IX into skeletal muscle or liver in hemophilic dogs has resulted in the long-term expression of factor IX at levels that are adequate to improve disease symptoms. Efforts to translate these findings into the clinical arena have proceeded slowly because of the lack of prior clinical experience with parenteral administration of AAV. In a staged approach, AAV-factor IX (AAV-F.IX) was first administered at doses of up to 1.8 x 10(12) vector genomes/kg (vg/kg) into the skeletal muscles of men with hemophilia B. This trial established the safety of parenteral administration and also showed that general characteristics of AAV transduction were similar in mice, dogs, and humans. In an ongoing trial, AAV-F.IX is being administered into the hepatic circulation of men with severe hemophilia B. The goal of these studies is to identify a safe dose that reliably yields circulating levels of factor IX >2% of normal levels in all subjects. This goal has already been achieved in the hemophilia B dog model; the ongoing study will determine whether a similar result can be achieved in humans with hemophilia B.     

Aminoglycoside suppression of nonsense mutations in severe hemophilia

Aminoglycoside antibiotics exhibit their bactericidal effect by interfering with normal ribosomal activity. In this pilot study, we have evaluated the effect of the aminoglycoside antibiotic gentamicin on the factor VIII (FVIII) and IX levels of severe hemophiliacs with known nonsense mutations. Five patients were enrolled and each patient was given 3 consecutive days of gentamicin at a dose of 7 mg/kg intravenously every 24 hours. Two patients (patient no. 1: hemophilia A, Ser1395Stop; and patient no. 5: hemophilia B, Arg333Stop) showed a decrease in their activated partial thromboplastin time (aPTT), an increase in their FVIII (0.016 IU/mL, 1.6%) or FIX (0.02 IU/mL, 2%) levels, and an increase in thrombin generation. The remaining 3 patients (patient no. 2: hemophilia B, Arg252Stop; patient no. 3: hemophilia A, Arg2116Stop; and patient no. 4: hemophilia A, Arg427Stop) showed no response in the aPTTs or factor levels, but one (patient no. 2: hemophilia B, Arg252Stop) showed an increase in the factor IX antigen level (2%-5.5%) that persisted throughout the period of the study and was concordant with an increase in thrombin generation. Gentamicin is unlikely to be an effective treatment for severe hemophilia due to its potential toxicities and the minimal response documented in this report. This study, however, does provide a proof of principle, suggesting that ribosomal interference with a less toxic agent may be a potential therapeutic mechanism for severe hemophilia patients with nonsense mutations.     

A randomized placebo-controlled double-blind study of danazol in hemophilia A.

A randomized double-blind placebo-controlled crossover trial of danazol was carried out in 19 cases of hemophilia A. Danazol was given for 3 months at a dose of 150 mg/day to patients under 15 years of age, and 300 mg/day to older patients. The basal factor VIII:C level was 8.3 +/- 5.6% (mean +/- SD), and after 3 months of danazol treatment was 15.3 +/- 11.0% (p = 0.02). Six patients (basal factor VIII:C 2-22%) showed a 1.36- to 2.87-fold elevation of factor VIII:C levels after danazol. 0/2, 1/4 and 5/13 cases of severe, moderate and mild disease, respectively, responded. Decreases in the number of bleeding episodes and cryoprecipitate requirement were seen in the responders. No adverse reactions to danazol were encountered. Danazol appears to raise the factor VIII:C levels in selected cases of hemophilia A.     

Determinants of plasma factor VIIa levels in humans

Several enzymes can activate factor VII in vitro, but the protease responsible for generating factor VIIa in vivo has not been determined. Using recombinant tissue factor that has undergone a COOH-terminal truncation, a sensitive functional assay has been established for measuring plasma factor VIIa levels. To evaluate the mechanism responsible for the generation of factor VIIa in vivo, we measured the levels of this enzyme after administering purified concentrates of factor IX and factor VIII to patients with severe deficiencies of these clotting factors. In patients with hemophilia B, factor VIIa levels were initially reduced to 0.5 +/- 0.1 ng/mL and gradually increased to normal after infusing 100 U/kg of body weight (BW) of factor IX. Despite these increases, there were no significant changes in the generation of factor Xa or thrombin. In patients with hemophilia A, only a slight reduction in factor VIIa levels (2.5 +/- 1.3 ng/mL) was observed as compared with controls (3.3 +/- 1.1 ng/mL) and no significant changes were observed after factor VIII levels were normalized. The administration of recombinant factor VIIa (10 micrograms/kg BW) to patients with factor VII deficiency increased the mean circulating level of the enzyme to 118 ng/mL, but this only resulted in normalization of the levels of the activation peptides of factor IX and factor X. The above data indicate that factor IXa is primarily responsible for the basal levels of free factor VIIa generated in vivo (ie, in the absence of thrombosis or provocative stimuli) and that changes in the plasma concentrations of free factor VIIa in the blood do not necessarily lead to alterations in the extent of factor X activation.     

Diagnosis and management of adult patients with von Willebrand disease in South Australia

We have analyzed the databases for von Willebrand disease (VWD) from the hemophilia center for adult patients with bleeding disorders in South Australia. We define the prevalence of types of VWD to determine the proportion of who would be treated with factor (F) VIII/von Willebrand factor (VWF) concentrate to prevent or control hemorrhage. In severe or moderately severe patients, we use plasma-derived FVIII/VWF concentrate, and for mild to severe cases, we use desmopressin plus tranexamic acid. There are 103 patients with VWF ristocetin (RCo) ≤50 IU/dL: 38 (37%) severe (VWF:RCo <10 IU/dL), 28 (27%) moderate (VWF:RCo 10 to 29 IU/dL), and 37 (36%) mild (VWF:RCo 30 to 50 IU/dL). Hence in 66 (64%), FVIII/VWF concentrate is the mainstay of treatment. The prevalence of VWD in our region according to data from our center is ~1 per 12,000. A total of 52% of patients are type 1, 44% type 2, and 5% type 3. In our experience, type 2M (45% of type 2) is much more common than types 2A and 2B (each 9% of type 2). Mutation detection is useful for identifying some subtypes of VWD.     

Severe factor VIII and factor IX deficiency in females.

A survey of 11 hemophilia centers produced data concerning 28 females with extremely low levels of factor VIII or IX coagulant activity. Ten of the 28 have hemophilia A, six have hemophilia B, and 12 have severe von Willebrand's disease. The 16 females who have severe factor VIII or factor IX deficiency as an isolated defect exemplify several of the possible genetic explanations for the occurrence of hemophilia in females. All 16 bruise excessively, and several have had recurrent hemarthroses. Three of these girls, ages five, 10 and 23 years, have evidence of chronic hemophilic arthropathy. The 12 females with severe von Willebrand's disease are either homozygous for von Willebrand's disease or severely affected heterozygotes. All 12 have mucous membrane bleeding. In addition, five of the 12 have recurrent hemarthroses and three have evidence of chronic joint disease. However, the major problem in the adult females with von Willebrand's disease has been extreme menorrhagia. One of the seven adults underwent irradiation sterilization and another had a hysterectomy because of menorrhagia. The others have been managed with anovulatory drugs or plasma infusions and EACA. Despite menorrhagia, five pregnancies and deliveries have been uneventful in three of these women.