Search

A monoclonal antibody to human antihemophilic factor (AHF, factor VIII) was derived from BALB/c mouse spleen cells fused with P3x63Ag8 mouse plasmacytoma cells. This antibody, harvested from culture medium or ascites fluid, reacted with purified AHF and with plasmas with normal subjects or classic hemophiliacs, as measured by enzyme-linked immunosorbent assay (ELISA), but not with plasmas from patients with severe von Willebrand's disease. The antibody possessed only IgG, heavy chains and kappa light chains. It blocked ristocetin-induced platelet agglutination and, to a lesser degree, platelet retention by glass bead columns, but it did not inhibit the procoagulant activity of AHF significantly. An amount of rabbit antiserum against AHF that provided equivalent inhibition of ristocetin-induced platelet agglutination inhibited glass bead retention much more effectively than the mouse monoclonal antibody. This difference was exaggerated in studies of the corresponding Fab fragments. These data suggest that the site or sites on the AHF complex molecule that are associated with ristocetin-induced platelet agglutination differ quantitatively or qualitatively from those associated with enhancement of platelet retention by glass beads. ELISA titers of immunoreactive AHF, using the monoclonal antibody, were closely correlated to those using rabbit antiserum against AHF in normal, hemophilic, and most von Willebrand's disease plasma.

This study examined the hypothesis that non-inhibitor haemophilic plasma contains antibodies which are specific for sites other than the active procoagulatn site on factor VIII, and that some of them might be sufficiently close to the active site that pre-incubation of such plasma with factor VIII would block the subsequent binding of inhibitor antibody. Among the 26 non-inhibitor plasmas examined, none was found to contain such blocking antibody. This result does not eliminate the possibility that antibody is present in such non-inhibitor plasmas which is neither specific for the active enzyme site of factor VIII nor capable of blocking the binding of antibody which does have that specificity.

To define further the factor VIII abnormality in acquired von Willebrand's disease, we performed immunoelectrophoresis of factor VIII antigen, as well as quantitative measurements of the antigen, factor VIII procoagulant activity and von Willebrand factor activity on plasma from an affected 57-year-old man who also had a poorly differentiated lymphocytic lymphoma. No evidence for an inhibitor against factor VIII procoagulant activity or von Willebrand factor activity was detected, but immunoelectrophoresis showed none of the less anodic forms of factor VIII antigen. There were concomitant decreases in total antigen (0.19 U per milliliter) and von Willebrand factor levels (0.12 U per milliliter). Factor VIII-procoagulant activity was borderline low (0.45 U per milliliter). Correction of both the abnormal immunoelectrophoresis pattern and the quantitative abnormalities followed radiotherapy of the lymphoma. The factor VIII abnormalities might have resulted from binding or destruction of theless anodic forms of factor VIII antigen by the malignant lymphocytes.

A 1-year follow-up study of the T-cell subset abnormalities was carried out in 16 severe haemophilia A patients, treated "on demand" with an average amount of 500 U/kg/yr of factor VIII concentrate (group A) and in 15 mild haemophiliacs or von Willebrand patients treated only sporadically with less than 3000 U of factor VIII and no longer exposed to any other blood component in the 2 years preceding the beginning of the study (group B). In group A, 50% and 70% of patients showed a reduced or inverted T 4/T 8 ratio, respectively, at the beginning and at the end of follow-up. These values were of 30% and 20% in patients of group B, suggesting a long-lasting effect of concentrate therapy on T-cell subsets. The low T 4/T 8 ratio was mainly due to an increase of suppressor cells. None of the patients was found positive for anti HTLV-I, whereas 3 patients, all belonging to the group A, showed antibodies against HTLV-III. Thus, in these patients, HTLV-III seems not to be the only cause of low T 4/T 8 ratio.

A patient with clinical and laboratory evidence of von Willebrand syndrome is described in association with an IgG-kappa immunoglobulin and Bence-Jones proteinuria due to a probable lymphoproliferative disorder. He had a persistently prolonged bleeding time of greater than 20 minutes, factor VIII related antigen (VIII:R.Ag), factor VIII procoagulant activity (VIII:C) and factor VIII ristocetin co-factor (VIIIR:Rcof) below 10%. Following cryoprecipitate or high purity factor VIII concentrate infusion, he had the expected immediate rise in VIII:C, VIII:R.Ag, and VIIIR:Rcof, but there was a rapid decline in all three components within two hours. The larger forms of VIII:R.Ag were preferentially removed from the plasma, and this paralleled the fall in plasma VIIIR:Rcof level. However, no inhibitory activity could be demonstrated in vitro using the patient's plasma or IgG. Using protein A it was possible to demonstrate that his plasma or IgG bound factor VIII and that this complex retained its biological activity in vitro. It is postulated that the monoclonal IgG forms complexes with factor VIII in vivo and these are rapidly removed by the reticuloendothelial system (RES).

Disturbances of haemostasis caused immunologically and non-immunologically were observed after transfusion of blood and blood derivatives. Transfusion of heparin blood increased the bleeding susceptibility only in case of pre-existing high-degree defects of haemostasis or if they were performed as massive or exchange transfusions. Massive transfusions with blood stored for a long time will induce complex defects. Under intensive substitution therapy of haemophilia A the so-called paradoxical bleeding will occur in spite of a high factor VIII level. These bleedings are supposed to be disturbances of the thrombocyte function and are caused by fibrin(ogen) derivatives. Post-transfusional thrombocytopenias may be brought to remission by repeated plasmapheresis. Factor specific inhibitory bodies will appear after substitution in a small percentage of haemophilic patients. 5 to 7 days after the onset of therapy an anamnestic reaction can be observed as a titre increase by leaps. Usually, the inhibitory titre will decrease to a mostly low basal value in the course of three to five months. The therapy with cyclophosphamide simultaneously started with the substitution will more frequently prevent the anamnestic reaction or reduce it. Titres with more than 5 units cannot be overcome at the beginning even by higher concentrations of preparations. The substitution therapy should be preceded by exchange transfusions or plasmapheresis of up to 25 units. With still higher titres only procedures of inhibitor-bypassing are possible with factor VIII preparations of animal origin or better with activated prothrombin complex preparations, such as FEIBA. Recent reports give evidence that permanent substitution with factor VIII concentrates at a highest dosage can eliminate the production of inhibitors completely.

A patient with acquired von Willebrand disease associated with multiple myeloma (IgG-lambda) is described. Mixture of his plasma or IgG fraction with washed control platelets resulted in the inhibition of aggregation with ristocetin, but mixture of control plasma or IgG fraction with washed patient platelets showed no inhibition of ristocetin-induced aggregation. Although his vWF: Ag, RCo, and factor VIII coagulant activity were all normal, inactivation of RCo was induced in normal plasma by incubation with patient plasma. Crossed immunoelectrophoretic analysis showed that vWF:Ag was composed of much more anodic component. A marked increase of Factor VIII and a rapid return of RCo to the baseline after 1-deamino-8-arginine vasopressin (DDAVP) infusion were observed. A transient increase in vWF:Ag after the infusion of DDAVP showed with less anodic forms and in the relative proportion as in normal. Treatment of the underlying disease also led to a correction of the bleeding time, improvement of platelet adhesion and ristocetin-induced aggregation, and normalization of crossed immunoelectrophoresis of vWF:Ag. The present study showed that myeloma-associated IgG interacted specifically with the antigenic sites on the von Willebrand portion of the Factor VIII complex.

Antiserum specific for that part of the factor VIII complex designated ristocetin aggregation factor (VIIIRAF) was prepared by immunizing rabbits with VIIIRAF derived from the plasma of a hemophilic patient with circulating antibody to factor VIII procoagulant activity (VIIIcoag). The rabbit antiserum prevented ristocetin-induced platelet aggregation and platelet retention by glass-bead columns. Although the antiserum also inactivated VIIIcoag in normal plasma, it did not inactivate VIIIcoag which had been dissociated from VIIIRAF by chromatography in 1 M NaCl, thus establishing the antigenic specificity of these two factors. When the VIIIRAF antibody was conjugated with fluorescein isothiocyanate and used to examine platelets from normal subjects and patients with von Willebrand's disease, the normal platelets showed a granular fluorescence similar to that observed with antifibrinogen serum whole von Willebrand platelets showed no fluorescent staining. The normal platelets retained the VIIIRAF granules during 18 hours incubation and 5 washings in artificial medium while the von Willebrand platelets failed to acquire granules after 18 hours in normal plasma. When the unstained platelets from patients with von Willebrand's disease were suspended in normal plasma and then aggregated by the addition of ristocetin, the aggregates not only stained brightly for VIIRAF, but fluorescent granules could be seen on individual platelets in the aggregates. These observations suggest that ristocetin causes the binding of VIIIRAF to platelets as well as platelet-to-platelet adhesion.

We have examined nine patients with presumed von Willebrand's disease who present the spectrum of that disorder. Two had findings that would be accepted generally as diagnostic of von Willebrand's disease, and seven had variations of the usual pattern. The commonest variation was the combination of borderline and variable levels of coagulant Factor VIII, commensurate levels of Factor VIII-related antigen, and low levels of ristocetin-Willebrand factor.

Heterogeneous factor VIII related antigen isolated from endothelial cell postculture medium was characterized. On crossed immunoelectrophoresis, slow moving less anodal populations of factor VIII related antigen molecules were more prominent in endothelial cell postculture medium than in plasma. The protein synthesized by endothelial cells appeared as two discrete protein bands of different molecular weight in sodium dodecyl sulfate polyacrylamide gels. In contrast the factor VIII related antigen isolated from plasma moved as a single protein band in an identical gel system. The factor VIII related antigen from endothelial cell postculture medium was immunoisolated using monospecific rabbit antibody to normal human plasma factor VIII related antigen, electrophoresed on sodium dodecyl sulfate polyacrylamide gels, radiolabeled with 125I, trypsinized and subjected to peptide mapping using two-dimensional high voltage electrophoresis and thin-layer chromatography. The two forms of endothelial cell factor VIII antigen had virtually identical peptide maps. These studies suggest that the molecular heterogeneity of the factor VIII related antigen system reflects polymeric associations of identical subunits. Circulation in vivo may alter the ratio of polymer subsets.

Immunological and immunofluorescent studies carried out on plasma and platelets of three cases of congenital factor XIII deficiency are reported. Two of these patients were originally thought to have normal factor XIII subunit S and no subunit A. However, repeated assays carried out using different lots of antiserum showed that in reality the patients lacked both subunit S and subunit A. The false positive finding was due to the presence of a anti-factor VIII contaminant in the antiserum originally used. The third patient had a normal subunit S and no subunit A. No factor XIII antigen was found by the indirect immunofluorescent technique in normal, factor XIII deficiency and von Willebrand's disease platelets. On the contrary, by using the non-monospecific antiserum a fluorescent pattern similar to that observed by using an anti-factor VIII antiserum, had been noted. On the basis of the data presented in this paper a tentative classification of factor XIII deficiency in two groups is proposed: Type I is characterized by the lack of both factor XIII subunit S and A. Type II is characterized by a normal subunit S and no subunit A. The need for a re-evaluation of published case of factor XIII deficiency by means of monospecific antisera is indicated.

A plasma protein required for the support of ristocetin-induced platelet aggregation was isolated from antihemophilic factor concentrate and radiolabeled with 125I. A double-antibody radioimmunoassay was developed, with use of specific rabbit anti-VIII related antigen serum and goat anti-rabbit globulin. The assay is sensitive, reproducible, and technically simple to perform. Values obtained in normal subjects ranged from 0.65 to 1.53 units, similar to our normal range for VIII coagulant activity (0.67-1.43 units). However, normal or increased values of VIII-related antigen were observed in VIII coagulant-deficient hemophiliacs. Also, concentrations of VII-related antigen significantly exceeded coagulant concentrations in several patients with liver disease or disseminated intravascular coagulation, or both. Of a broad selection of congenital coagulation disorders examined, only patients with von Willebrand's disease had decreased VIII-related antigen concentrations, and these corresponded to the lowered concentration of ristocetin cofactor in the patients. In three transfused patients, VII-related antigen values correlated with the concentration of the cofactor. Our results suggest that the radioimmunoassay of VIII-related antigen is a simple and valuable adjunct in the study of patients with clotting abnormalities.

Classification of the subtypes of von Willebrand's disease (vWd) has been based on a quantitative deficiency or an abnormal multimeric composition of von Willebrand factor (vWf). Although the co-deficiency of a second protein, von Willebrand's antigen II (vW AgII), had been previously recognized, its concentration in a relatively large number of normal individuals or patients with well-defined vWd variants had not been studied. The plasma from patients with type I, IIA, IIB, IIC, and III (severe) vWd was evaluated, and the concentrations of vW AgII and vWf were determined. Although patients with type I and III vWd had reduced levels of both proteins, the plasma vW AgII concentration was normal in patients with type II vWd. Analysis of the results indicates that type I and type II variants can be discriminated with greater than 80% accuracy by comparison of results of these two antigenic assays. The normal levels of vW AgII in type II variants suggest a possible difference in the pathophysiology of type I and type II vWd.

Unresponsiveness to skin testing with PPD and tetanus toxoid was commonly seen in patients with haemophilia A but not infected with human immunodeficiency virus but was uncommon in controls. Vaccination history indicated that the unresponsive patients had not been immunised in childhood. Other tests of immune competence (skin tests with other antigens, lymphocyte stimulation with mitogens and antigens, and viral serology) showed that the haemophilia A patients had an adequate response to pathogens to which they had been exposed. Five of 12 such patients had a mild T4 lymphopenia, and this may have been related to parenteral administration of large quantities of protein.

Evidence for the presence of circulating immune complexes was found in thirty-four out of fifty-five samples from forty-seven patients with haemophilia A. In eleven patients the complexes, precipitated from the blood with polyethylene glycol, were digested with pepsin. The F(ab')2 antibody was tested, and found to have neutralizing activity against coagulant Factor VIII in two patients. In one of these no free antibody had ever been found in the plasma, while in the other the antibody was concentrated tenfold in the complex. In two other samples free without complexed antibody was found. In comparison, IgG-containing complexes were found in nine out of nineteen patients with von Willebrand's disease and no complexes were found in the sera from twelve multi-transfused thalassaemics. PEG precipitation is a useful technique for the preparation of concentrated immune complexes for further study such as antigen identification.

A form of von Willebrand's disease has been described with enhanced ristocetin-induced platelet aggregation and anodal migration of the factor VIII/von Willebrand factor protein (type IIb). We studied two families with this form of von Willebrand's disease and macrothrombocytopenia. We have found that these platelets bind more of the normal and intermediate-sized multimers of the factor VIII/von Willebrand factor than normal platelets. Analysis of the binding data show an increased affinity of these vWd platelets for the factor VIII/von Willebrand factor. These findings are consistent with an increased number of platelet receptors, which, either by their native topography or migration on the platelet surface, bind factor VIII/von Willebrand factor protein with greater affinity than normal platelets, platelets of other vWd patients, and large platelets of other etiologies.

A rapid method for measuring factor VIII antigen in plasma is presented, based upon optical measurement of the rate of agglutination of a stirred suspension of latex particles coated with anti-factor VIII immunoglobulin, when mixed with plasma containing factor VIII antigen. A log-log relationship exists between factor VIII concentration and the reciprocal of the time required for a fixed absorbance change. The assay is rapid, requiring less than 3 min for normal plasma, involves stable reagents, and can be performed with conventional instrumentation. Antigen values in normal plasma a well as plasma from patients with von Willebrand's disease (angiohemophilia) correlate well with values measured by electroimmunoassay or by radioimmunoassay.

A case of acquired von Willebrand's syndrome (vWs) is described which appeared to be due to antibodies directed against factor VIII clotting activity (FVIIIC), factor VIII-related antigen (FVIIIRAg) and von Willebrand factor. The antibodies directed against FVIIIRAg was demonstrated by the inhibitory effect of a platelet eluate on Ristocetin-induced aggregation of normal platelets. This effect was not shown by the patient's platelet-poor plasma alone, nor could it be demonstrated in platelet eluates from 13 other patients who had antibodies to FVIIIC but in whom there was no evidence of an acquired vWs.

This report describes a patient without evident underlying disease, in whom an acquired von Willebrand's syndrome was discovered before surgery. Coagulation abnormalities included a borderline bleeding time, a low retention of platelets on glass beads, decreased levels of factor VIII procoagulant activity (VIIIAHF), factor VIII-related antigen (VIIIAg), and ristocetin-induced agglutination cofactor (VIIIVWF). After cryoprecipitate infusion the patient did not have the expected rise and there was no secondary increment in VIIIAHF. The patient was treated with prednisone for three weeks without significant improvement in the laboratory findings. Spontaneous resolution was observed long after this therapy. The haemostatic abnormalities were attributable to the presence of an inhibitor directed against VIIIVWF. The inhibitor was found in the IgM fraction. Its autoimmune nature is probable although we failed to demonstrate any inhibitory effect of Fab obtained from the patient's purified IgM. Despite the lack of inhibitory effect against VIIIAHF and VIIIAg, the low levels of all three activities of the factor VIII complex could be explained by the short half-life of immune complexes between factor VIII and the inhibitor.

A patient is described with clinical and laboratory evidence of von Willebrand's disease (VWD) in association with lymphosarcoma. He consistently had a bleeding time of over 20 min; factor VIIAHF, VIIAGN, and VIIVWF under 20%; and severe, diffuse hemorrhage, Following transfusion with cryoprecipitate, the patient had the expected rise in VIIAGN and VIIAHF. The patient's plasma contained an inhibitor which prevented aggregation of normal platelets by ristocetin, but which did not interfere with the measurement of VIIAGN or inactivate VIIAHF activity. The inhibitor was present in serum heated at 56 degrees C for 30 min, was present in a purified IgG fraction of serum, and was precipitated by anti-human IgG. The antibody did not interact directly with washed platelets, but appeared to interact with that portion of the factor VII protein that supports ristocetin aggregation (VIIIVWF). The data provide the first explanation for the pathophysiology of the acquired von Willebrand's syndrome.

Factor VIII in preparations from normal plasma is a large glycoprotein of greater than 2 million molecular weight which elutes in the exclusion volume of 4% agarose gels at an ionic strength of 0.15. Recent studies have demonstrated that the factor VIII in canine and bovine plasma is a macromolecular complex composed of a large inert carrier protein and a noncovalently bound small fragment which contains the procoagulant active site. This complex is dissociable in 0.25 M CaCl(2), and conditions for its recombination have been reported. The present study reports the dissociation characteristics of normal human factor VIII preparations in 0.25 M CaCl(2) and the ability to achieve quantitative recombination of the dissociated fragments of normal human and bovine factor VIII after the removal of Ca(2+). The recombination technique was used to characterize further the defect in hemophilia and von Willebrand's disease. Void volume preparations from human hemophilia A(-), canine hemophilia A, and human von Willebrand's plasma, with no factor VIII procoagulant activity, were mixed with the small active fragment prepared from the normal plasma of their respective species. Chromatography of the three mixtures in agarose gel showed that the fractions from the human hemophilic plasmas contained a molecule that bound the small active normal fragment, but neither the fractions from canine hemophilia A plasmas nor the fractions from the human von Willebrand's plasmas demonstrated evidence of such material. These data suggest that there is present in human hemophilia A plasma a normal functional carrier molecule which is absent or nonfunctional in the plasma of hemophilic dogs and humans with von Willebrand's disease.

A 16-yr-old girl with severe von Willebrand disease complicated by the development of precipitating alloantibodies to von Willebrand factor (anti-VWF) had a life-threatening anaphylactoid reaction immediately after the infusion of a commercial plasma concentrate of factor VIII/von Willebrand factor. An early post-infusion activation of the complement system was demonstrated by the appearance of C3 split products and by the drop of serum CH50 activity, occurring in parallel with a post-infusion drop in the anti-VWF antibody levels. Immune complexes remained unchanged in the early post-infusion period and rose to a moderate extent only after 24 h. We conclude that biologically active products of the complement system contributed to the onset of this life-threatening reaction which occurred after concentrate infusion.

Helper (OKT4) and suppressor (OKT8) T lymphocyte populations and functional assays of cellular immunity were studied in 37 patients with haemophilia and related disorders in parallel with age matched control subjects. The study included 26 patients with factor VIII (FVIII:C) deficiency, eight patients with factor IX deficiency and three patients with severe von Willebrand's disease (vWd). In patients with factor VIII deficiency low helper T lymphocyte counts, low T helper: suppressor ratios and a diminished response to the lymphocyte mitogen phytohaemagglutinin with decreased natural killer cell activity were observed. Individuals with factor IX deficiency had low absolute T lymphocyte (OKT3) counts and T helper cell counts. Patients with severe factor VIII deficiency (FVIII:C less than 1 u/dl) had lower T helper suppressor ratios and lower killer cell and natural killer cell activity in comparison to mildly affected individuals and all of the severely affected factor IX deficient cases. Studies of humoral immunity revealed a generalized increase in immunoglobulin levels in patients with coagulation disorders of any type. The total haemolytic complement activity was reduced in a significant proportion of haemophilic subjects. Levels of alpha-1-interferon were elevated in the groups of haemophilic subjects studied. The abnormalities of cellular and humoral immunity observed did not correlate with the amount or type of coagulation factor administered to individual patients in the preceding 2 years. The most marked abnormalities of immune function occurred in the one patient diagnosed as suffering from AIDS on clinical grounds.

The majority of patients with congenital clotting disorders who use clotting factor concentrate exhibit lymphocyte subpopulation abnormalities. A subset of these patients develop lymph node enlargement (LNE), part of the spectrum of clinical disease associated with the acquired immune deficiency syndrome (AIDS). It is therefore important to determine if these patients with LNE exhibit specific immune alterations suggestive of early infection with the AIDS agent. We used one- and two-color immunofluorescence to distinguish the lymphocyte subpopulation alterations associated with concentrate use from those associated with LNE. Patients who use concentrate had elevated levels of Leu-2+ (T suppressor phenotype) cells and Leu-7+ (phenotype of some natural killer) cells. These increased levels were largely caused by a dramatic (2.6-fold) increase in the number of lymphocytes co-expressing Leu-2 and Leu-7 (2+7+). A dose-response effect between amount of concentrate infused during the preceding year and level of 2+7+ cells was observed. Concentrate recipients, as a group, also showed increased levels of T cells expressing Dr antigen (T+Dr+ phenotype, characteristic of activated or immature T cells) and cells expressing the T10 antigen (phenotype of some null cells and activated/immature T cells). Patients with LNE showed a further increase in T10+ cells as well as a distinctive decrease in Leu-3+ (T helper phenotype) lymphocytes. All LNE patients exhibited either low Leu-3+ levels, high T10+ levels, or both. Thus, concentrate use was associated with increased levels of Leu-2+ (particularly 2+7+) cells and T+Dr+ cells, whereas LNE was associated with decreased levels of Leu-3+ cells and high levels of T10+ cells.

Using one-stage quantitative factor VIII assay and the quantitative immunoelectrophoretic method of LAURELL a good correlation has been found between factor VII activity and protein concentration. In 80% of the known hemophilia A carriers the protein concentration was about twice as high as in normals as compared to the factor VIII activity, whereas about 20% of the known carriers have the same ratio between factor VIII protein and factor VIII activity as normals. The reliability of the test and its usefulness for the detection of hemophilia A carriers are discussed.

An enzyme immunoassay system basing on a competitive method has been developed to measure factor VIII related antigen (F. VIII R:Ag). A sufficient discrimination at low F. VIII R:Ag concentrations was gained. This method appears to be sensitive to 7,8 X 10(-3) U/ml F. VIII R:Ag showing an intraassay coefficient of variation (CV) of 0,11. In comparison to the commonly used Laurell electroimmunodiffusion assay for factor VIII significant less antisera per sample for the enzyme immunoassay technique is necessary.

Using an immune system isolated from a defined NANB epidemy we could find such markers in 29 of 71 patients (41%). The greater number of the antigen speaks in favour of a large pool of healthy carriers. The small number of patients with antibodies may be explained by the low sensibility of our test which detects only antibodies with high titers. According to this analysis of NANB associated immune markers the clinical importance of this infection in haemophiliacs seems to be small. Statements concerning the chronicity of the disease can not yet be made because of the preliminary character of this random test made in a holiday camp of haemophiliac children.

An immunoradiometric assay (I.R.M.A.) for antihaemophilic factor (factor VIII) has been developed which is based on a human inhibitor to procoagulant factor VIII (FVIIIC) which arose in a polytransfused haemophiliac. The assay detected and measured the antigen related to FVIIIC (FVIIICAg) in normal plasma and serum but failed to detect any antigen in samples from five patients with severe haemophilia. In a sixth patient with severe haemophilia a non-parallel dose-response curve was obtained. FVIIICAg was also present in plasma and serum from six samples of cord blood and in two samples of fetal serum obtained after termination of pregnancy. The assay may, therefore, be used to measure FVIIICAg in small serum samples and provides a potential practical method for prenatal diagnosis of haemophilia.