Acquired von Willebrand Syndrome with Inhibitors Both to Factor VIII Clotting Activity and Ristocetin-Induced Platelet Aggregation

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.     

Binding of Factor VIII to Platelets in the Presence of Ristocetin

Ristocetin agglutinates platelets in platelet-rich plasma (PRP) containing factor VIII-related von Willebrand factor (WF). When the clumps were separated from the supernatant and resuspended in buffer, the platelets dispersed if the buffer was free of ristocetin and the PRP had been treated with aspirin or EDTA to prevent the platelet release reaction. Binding to platelets was determined by measuring WF (ristocetin cofactor assay), VIIIR-antigen (AG, by radioimmunoassay), and coagulant activity (C) in the initial plasma, the supernatant remaining after removing the clumps formed by ristocetin, and the eluate produced after resuspending the clumped platelets in buffer. No activity was bound to platelets in the absence of ristocetin. After agglutinating platelets in PRP with ristocetin, WF, AG and C activities in the supernatants were about 65% of the initial plasma values, and in the eluates about 9%. Formalin-fixed washed platelets resuspended in 1 in 3 plasma, and platelets in PRP diluted 1 in 3 also agglutinated when shaken with ristocetin. There was no difference between results with fixed and unfixed platelets. In one series of experiments, the supernatants had about 60% of the initial WF and AG activities; since these activities were the same, they may measure the same substance. In a later series, WF and C were compared; in shaken samples the supernatants had 39% of the initial WF activity and 56% of the initial C activity, and in unshaken samples the supernatants had 3% WF and 29% C. The significantly lower values for C binding suggest that this activity is located on a different molecule from WF-AG and that the molecules are more readily separated in diluted than in undiluted plasma. Activity in the eluates was inversely proportional to that in the supernatants although the initial activity was never completely recovered. WF and C binding increased with platelet number and ristocetin concentration. Platelets exposed to ADP before fixation showed less agglutination than control fixed platelets and bound less WF. Fixed platelets incubated for 15 min with 30 μg trypsin/ml and washed, or platelets from a patient with the Bernard-Soulier syndrome failed to agglutinate and bound virtually no WF, AG or C. We conclude that WF, AG and C reversibly bind to fixed or unfixed platelets in the presence of ristocetin, that binding is similar in undiluted plasma but greater for WF and AG than C in plasma diluted 1 in 3, and that ADP and trypsin alter binding, probably by affecting a platelet receptor.     

Characteristics of the Factor VIII Protein and Factor XIII in Various Factor VIII Concentrates

The in vitro properties of 5 factor VIII preparations (AHF-Kabi, Hemofil Hyland, AHF-Profilate Abbott, Kryobulin Immuno and Factorate High Purity Armour) and an ordinary cryoprecipitate were studied with reference to factor VIII clotting activity (VIII:C), factor VIII clotting antigen (VIILCAg), factor VIII related antigen (VIIIR:Ag) (EI, IRMA, CIE), ristocetin cofactor activity (VIIIR:RCF), fibrinogen and factor XIII. All the preparations with the exception of Factorate had higher levels of VIII:CAg than VIII:C indicating inactivation of the biological activity of VIII:C during the procedure. AHF-Kabi (fraction I-0) and the cryoprecipitate, the only preparations capable of normalising the defect in patients with von Willebrand's disease, showed the same level of VIIIR:Ag determined by EI and by IRMA, while all the other preparations (i.e. cryoprecipitates purified further in different ways) had considerably lower levels of VIIIR:Ag determined by IRMA than by EL Based on these in vitro techniques it seems to be possible to predict which preparations can be used successfully in patients with von Willebrand's disease, while no such conclusions can be made from VIIIR:RCF determinations. EI yielded similar concentrations of factor XIII a subunit in all the preparations tested. 3 functional assays showed high factor XIII activities in AHF-Kabi but low or no activities in the others. Thus, considerable differences were found of the in vitro properties of the proteins in 5 factor VIII concentrates and a cryoprecipitate. The action of proteases and the techniques used in the purification procedure are probably of crucial importance for the properties of the various factors.     

Binding of Low-Molecular-Weight Canine Factor VIII Coagulant from von Willebrand Plasma to Canine Factor VIII-Related Antigen

S ummary . Plasmas having no detectable factor VIII-related antigen but moderate factor VIII coagulant were obtained from two unrelated dogs homozygous for von Willebrand's disease and with the severe clinical expression of the disease. When these plasmas were gel-filtered in a buffer at physiologic ionic strength, the factor VIII coagulant eluted in the bed volume as a single well-defined peak. Addition of protease inhibitors, including diisopropylfluorophosphate, did not change the elution pattern. Each plasma was then combined individually with plasmas from six different mutants of canine haemophilia, all of which had normal factor VIII-related antigen but no detectable factor VIII coagulant. The factor VIII coagulant elution profile of these combined plasma resembled that of normal canine plasma. Slightly over half of the recovered factor VIII coagulant coeluted in the void volume with the factor VIII-related antigen; the rest eluted as a second, distinct peak of lower molecular weight. These results demonstrated that part of the factor VIII coagulant of the von Willebrand plasmas had bound to the factor VIII-related antigen of the haemophilic plasmas. This finding supports the theory that factor VIII exists as a macromolecular complex of nonidentical components in normal citrated plasma.     

Studies on the Mechanism of Ristocetin-induced Platelet Aggregation: Binding of Factor VIII to Platelets

The effect of ristocetin on the binding of [125I]factor VIII to platelets was studied. High and low affinity F.VIII binding sites exist on platelets. The high affinity sites bind 13 times more F.VIII than the low affinity sites. Ristocetin increased the binding of F.VIII to both types of binding sites by increasing the affinity of F.VIII for the platelet and increasing the total number of platelet binding sites. Chymotrypsin-treated platelets were not aggregated by ristocetin and F.VIII: these platelets have less of the major platelet membrane glycoproteins and bind much less [125I]F.VIII than do buffer-treated platelets with and without ristocetin.     

Variables Involved in Cryoprecipitate Production and their Effect on Factor VIII Activity

S ummary . The Regional Transfusion Centres in England and Wales have studied the following variables in the preparation of cryoprecipitate: blood group of the donor, choice of anticoagulant (ACD or CPD), conditions of freezing, thawing and centrifugation of the plasma and age of the starting plasma.
Mean factor VIII activity in cryoprecipitates derived from group A plasma was significantly higher than from group O plasma. CPD may lead to increased potency of factor VIII when freshly collected plasma is used to prepare cryoprecipitates; rapidity of thawing may also be advantageous.
None of the other variables studied produced significant effects on the factor VIII activity of the cryoprecipitates and particular attention has been drawn to the results of experiments concerned with the age of starting plasma.     

Factor VIII:C and VIII:CAg Response in Patients with Haemophilia A and von Willebrand's Disease after Administration of Different Factor VIII Concentrates or Plasma

S ummary . Factor VIII procoagulant activity (VIII:C) and factor VIII procoagulant antigen (VIII:CAg) were studied in seven patients with haemophilia A after administration of three different factor VIII concentrates or plasma. The in vivo recovery of VIII:CAg was less than that of VIII:C and the disappearance rate of VIII:CAg was much higher either when concentrates or plasma were given. The half-life of VIII:C was thus about 12 h but of VIII:CAg only about 3 h or less. Six patients with von Willebrand's disease were studied after administration of AHF- Kabi. In contrast to haemophilia A the discrepancy between VIII:C and VIII:CAg disappearance rates was not present in von Willebrand's disease, since both VIII:C and VIII:CAg showed a typical progressive increase. We conclude that factor VIII:C given to haemophilia patients does not behave like native VIII:C, not even when fresh plasma is used. Patients with von Willebrand's disease are capable of forming a normal VIII:C when appropriately stimulated.     

Factor VIII Procoagulant Activity, Factor VIII Related Antigen and von Willebrand Factor in Newborn Cord Blood

The mean level of factor VIII procoagulant acitivity (VIII:C) and factor VIII related antigen (VIIIR:AG) was normal in 100 newborn cord plasmas, whereas that of von Willebrand factor (VIIIR:WF) activity was slightly lower than normal. On crossed immunoelectrophoresis, 20 of 50 newborn infants had an increased anodal mobility of VIIIR:AG. When the cord plasma showing an abnormal electrophoretic pattern was mixed with normal plasma, two precipitation peaks with a broad base were found. Similar mixing experiments with the abnormal cord plasma and plasma from a patient with atypical von Willebrand's disease did not normalize the electrophoretic mobility of VIIIR:AG. Gel filtration of the cord plasma with an abnormal electrophoretic pattern of VIII:AG, showed that the three activities were all detected at the position corresponding to a molecular weight of about 800 000. The results suggest the presence of qualitative abnormalities of the factor VIII molecule in half of full-term newborn cord plasma.     

Ultracentrifugal Analysis of Factor VIII and von Willebrand Factor in Therapeutic Preparations

Plasma and therapeutic preparations of factor VIII (1 recombinant factor VIII and two monoclonally purified plasma-derived factor VIII preparations, Kogenate, and AHF-M and Monoclate, respectively) were centrifuged in a sucrose density gradient, and the fractions were analyzed for factor VIII and von Willebrand factor (vWF). The residual vWF in the monoclonally purified factor VIII preparations sediments more slowly than the vWF of plasma. In the absence of added vWF, the factor VIII in all preparations sediments more slowly than plasma factor VIII. These same preparations of factor VIII added to hemophilic plasma as a source of vWF sediment differently. The addition of either recombinant factor VIII or AHF-M results in sedimentation of the factor VIII with the plasma vFW and in a position indistinguishable from factor VIII in plasma. In contrast, when Monoclate is added to hemophilic plasma in vitro, the factor VIII sediments more slowly than the vWF of the hemophilic plasma. However, 5 min after the infusion of Monoclate into a patient with hemophilia A, the factor VIII sediments with the plasma vWF. These results indicate that the addition of recombinant factor VIII and AHF-M results in random binding to all vWF multimers of plasma, while there is little exchange between the added factor VIII in Monoclate and the plasma vWF in vitro. In contrast, when the Monoclate is infused, there is rapid binding of factor VIII to the plasma vWF. This is presumably due to tight binding of the factor VIII to the vWF in Monoclate, whereas the factor VIII in Kogenate and AHF-M, containing no or very low amounts of vWF, are free to bind to plasma vWF. In vivo there are other conditions, as yet uninvestigated, which allow for the binding of factor VIII to vWF.     

Clinical Investigation of Intermediate- and High-Purity Antihaemophilic Factor (Factor VIII) Concentrates

S ummary . Intermediate- and high-purity antibaemophilic factor (factor-VIII) concentrates developed by the American National Red Cross (ANRC) have undergone extensive clinical investigation. When administered to severely affected haemophiliacs, the recovery in vivo and the metabolic half-disappearance time were similar to those achieved with plasma.     

Relationship between the Binding Sites for von Willebrand Factor,Phospholipid, and Human Factor VIII C2 Inhibitor Alloantibodies within the Factor VIII C2 Domain

Some factor VIII (FVIII) inhibitor alloantibodies block FVIII binding to von Willebrand factor (VWF) and phospholipid (PL) and recognize a C2 domain epitope that overlaps both binding sites. We previously showed that FVIII peptide 2315-2330 neutralized FVIII inhibitors and that Cys2326 and Glu2327 contributed to the maximum neutralizing effect. In the present study, we investigated the relationship between the essential binding sites for VWF, PL, and anti-C2 inhibitors by means of competitive-inhibition assays with overlapping synthetic peptides that span the C terminus of the C2 domain (residues 2288-2332). We identified 2 peptides (residues 2303-2317 and 2315-2330) that specifically blocked FVIII binding to VWF or PL by approximately 80% (50%-inhibitory concentration [IC50], 9.0 microM) and 95% (IC50, 0.12 microM), respectively. To examine in detail the residues responsible for PL binding, we prepared mutants of peptide 2315-2330 in which we sequentially substituted each residue with Gly. Two residues, Ile2317 and Met2321, were shown to be essential for PL binding. Their substitution with Gly reduced the inhibitory effect by >90%. The data suggest that the binding sites for VWF, PL, and anti-C2 inhibitors in the C2 domain are in very close proximity but are not identical.     

Influence of the Primary Anticoagulant on the Recovery of Factor VIII in Cryoprecipitate

The influence of anticoagulant on overall factor VIII-yield was measured by drawing blood from one donor simultaneously in three bags containing ACD, CPD and heparin, respectively. After parallel processing factor VIII:C and factor VIII:CAg were measured. It is concluded that, under the circumstances used in this experiment, CPD gives the highest yield of factor VIII.     

Quantitation of Binding of Factor VIII Antigen to Concanavalin A

This study establishes a convenient method for screening plasma samples for abnormalities of the carbohydrate content of the factor VIII (FVIII) molecule. A radioimmuno-electrophoretic technique has been developed to quantitate the percentage binding of FVIII-related antigen (VIII-Ag) to the lectin concanavalin A (Con A). Plasma samples were electrophoresed through a strip of agarose containing Con A into agarose containing a mixture of unlabelled anti-FVIII and 125I-anti-FVIII where precipitant lines formed, the height of which was dependent upon the degree of VIII-Ag binding to Con A in the first gel. Using this system reduced binding of VIII-Ag to Con A was found in the plasma of 12 patients with moderate classical von Willebrand's disease (vWd), while the Con A binding of six haemophilia A patients fell within the normal range. The VIII-Ag in normal cryoprecipitate showed increased % binding to Con A while the VIII-Ag remaining in the cryo-supernate demonstrated reduced Con A % binding.