Coagulation Disorders UnitDepartment of Hematology

Comprehensive Cancer CenterFaculty of Medicine

Doctoral Programme in Clinical ResearchUniversity of Helsinki and

Helsinki University Hospital

INSIGHTS INTO CLINICAL ANDLABORATORY PHENOTYPES OF VON

WILLEBRAND DISEASE

Vuokko Nummi

Doctoral DissertationTo be presented for public discussion with the permission of the Faculty of

Medicine of the University of Helsinki, in Biomedicum lecture hall 2on the 27th of September, 2019 at 12 o’clock

SUPERVISORSProfessor Riitta LassilaCoagulation Disorders UnitDepartment of HematologyComprehensive Cancer CenterHelsinki University HospitalHelsinki, Finland

Timea SzantoCoagulation Disorders UnitDepartment of Clinical ChemistryHUSLAB Laboratory ServicesHelsinki University HospitalHelsinki, Finland

REVIEWERSProfessor Risto KaajaInternal MedicineDepartment of Clinical MedicineTurku University HospitalTurku, Finland

Docent Sakari KakkoDepartment of Internal MedicineUniversity of Oulu and Clinical Research CenterOulu University HospitalOulu, Finland

TO BE DISCUSSED WITHProfessor Jan VoorbergMolecular and Cellular HemostasisUniversity of AmsterdamDepartment of Plasma Proteins, Sanquin ResearchAmsterdam, the Netherlands

Cover art: Mouth Full of Rain XI 2008, Kristiina Uusitalo

ISBN 978-951-51-5415-6 (paperback)ISBN 978-951-51-5416-3 (online)ISSN 2342-3161 (paperback)ISSN 2342-317X (online)UnigrafiaHelsinki 2019

4

1 ABSTRACT

Von Willebrand disease (VWD) is an inherited bleeding disorder due toquantitative or qualitative defect of von Willebrand factor (VWF). The bleedingsymptoms of affected patients range from mild spontaneous mucocutaneuousbleeds to severe gastrointestinal, intra-articular, trauma- and surgery-associated bleeds. The reasons behind this intriguing variance of bleedingtendency are not well understood. This thesis aims at characterizing VWF in acohort of VWD patients by means of phenotypic and genotypic studies andglobal coagulation tests, with specific focus on platelets and their VWF-dependent functions. These features were characterized in two patient groups:type 3, the most severe form of VWD with absent VWF (studies I to III), and ina group of patients who carry a historical diagnosis of any type of VWD (studyIV).

In study I, we reported the clinical features and focused on the molecularanalysis of the VWF gene in 10 Finnish patients with type 3 VWD. In study II,type 3 was further characterized in 9 patients in the laboratory by evaluatingthe clotting capacity by thrombin generation assay in the presence and absenceof platelets before and after VWF therapy. Thrombophilia screen and naturalanticoagulants were also measured. Finally, platelet expression of P-selectinand phosphatidylserine were investigated by means of flow cytometry. To coverthe spectrums of treatment challenges encountered in VWD, a case report of apatient with anti-VWF alloantibodies and its management in an acute setting ispresented (study III). Study IV included 83 subjects who carried a historicaldiagnosis of VWD, diagnosed median 20 years ago. All patients were evaluatedaccording to present guideline recommendations and additionally, we assessedwhole blood platelet aggregation using Multiplate® and function using PFA-100®.

In study I, we discovered that two main mutations of VWF, c.2435delC andc.4975C>T, account for 85% of the genetic defects of our Finnish type 3 VWD.Even though the genetic background was homogenous, the clinical expressionvaried. Among our patients, some had severe bleeding tendency and receivedregular VWF prophylaxis while others had only low spontaneous bleedingrates. Out of patients included, 30% (study I) and 44% (study II) had relativelymild bleeding phenotype and received only occasional on-demand therapy. Instudy II, we showed that these type 3 patients expressed less than 50% ofnormal thrombin generation (TG) capacity in plasma, and TG was dependenton FVIII levels. However, in platelet-rich plasma, TG was close to normal,suggesting a compensatory role for platelets. The specific mechanisms behindthis remain unraveled, with enhanced platelet P-selectin andphosphatidylserine expression demonstrated in individual patients. No

5

thrombophilic traits or other clotting factor abnormalities compensating forthe VWF-induced coagulation defect were discovered.

Study III describes the management of a 49-year old woman with type 3VWD and a history of anti-VWF alloantibodies. Following an electivecolonoscopy with biopsy, she suffered from extended life-threateninggastrointestinal bleeding. The administration of intravenous immunoglobulininfusions appeared to eliminate the alloantibody, and the patient wassuccessfully treated with VWF therapy.

In study IV, the historical VWD diagnosis was confirmed 93% of type 2 and3 cases, but only in 26% of type 1. Current guidelines have lowered the cut-offvalues of VWF for VWD and introduced the term ‘low VWF’ for patients withborderline VWF levels, and indeed laboratory results indicated low VWF for26% of our historical type 1 patients. However, 47% of historical type 1 patientshad currently normal VWF and FVIII. Other coagulation defects, such as mildplatelet function disorders were discovered in individual subjects. Out ofhistorical type 1 patients, 21% had a normal bleeding score, indicating clinicallyinsignificant bleeding phenotype. As a conclusion, the bleeding symptoms ofpatients who carry historical type 1 VWD diagnosis are not always explained byVWF deficiency. VWD was associated to abnormal whole blood plateletaggregation and function, depending on the type of VWD and associating alsoto the severity of bleeding symptoms. With these rapid methods, confirmationof VWD diagnosis can be performed promptly in the laboratory.

6

2 CONTENTS

1 Abstract ……………………………………………………………………………4

2 Contents …………………………………………………………………………..6

3 List of original publications…………………………..……………….8

4 Abbreviations………………………………………………………………..…9

5 Introduction…………………………………………………….……………….11

6 Review of the literature…………………………………………………..13

6.1 VWF and hemostasis………………………………………………13

6.1.1 VWF: synthesis, storage, secretion and clearance…………13

6.1.1 VWF and primary hemostasis…………………………………….16

6.1.1 Coagulation system…………………………………………………..17

6.2 VWD: clinical aspects……………………………………………..20

6.2.1 Epidemiology and classification………………………………...20

6.2.2 Clinical features and evaluation of bleeding symptom….22

6.2.3 Treatment…………………………………………………………..……23

6.2.4 Laboratory evaluation of VWD………………………………..…25

6.2.5 Genetic analysis……………………………………………………..…30

6.2.5 Global hemostasis assays……………………………………….....33

7 Aims……………………………………………………………………………….…38

8 Patients and study design……………………….………………….…..39

9 Methods…………………………………………………………………………….41

7

9.1 Clinical evaluation of bleeding symptoms……………….41

9.2 Laboratory analysis……………………………………………...……41

9.2.1 Blood sampling……………………………………………………………41

9.2.2 Coagulation assays………………………………………………………41

9.2.3 Thrombophilia screen………………………………………………….42

9.2.4 Multimer analysis………………………………………………………..43

9.2.5 Platelet function analyser (PFA)…………………………………...43

9.2.6 Genetic analysis ………………………………………………………….43

9.2.7 Thrombin generation ………………………………………………….44

9.2.8 Rotational thromboelastometry (ROTEM)…………………….45

9.2.9 Platelet analysis under flow cytometry…………………………..45

9.2.10 Platelet aggregation……………………………………………………..45

9.3 Statistical analysis……………………………………………………..46

10 Results……………………………………………………………………..……….47

10.1 Study I…………………………………………………………………………..47

10.2 Study II…………………………………………………………………………49

10.3 Study III………………………………………………………………………..57

10.4 Study IV………………………………………………………………………..59

11 Discussion……….….……………………………………………………………..63

12 Summary and conclusions…………………………………………………70

13 Acknowledgements………………………………………………………..…..71

14 References…………………………………………………………………………..73

15 Original publications………………………………………………………….89

8

3 LIST OF ORIGINAL PUBLICATIONS

This thesis is based on the following original publications, referred to in thetext by their Roman numerals (I-IV)

I. Jokela V, Lassila R, Szanto T, Joutsi-Korhonen L, Armstrong E, Oyen F,Schneppenheim S, Schneppenheim R. Phenotypic and genotypiccharacterization of 10 Finnish patients with von Willebrand disease type 3:discovery of two main mutations. Haemophilia 2013; 19: e344-e348

II. Szanto T*, Nummi V*, Jouppila A, Brinkman HJ, Lassila R. Plateletscompensate for poor thrombin generation in type 3 von Willebranddisease. In Press, Platelets

III. Nummi V, Lehtinen E, Mäkipernaa A, Szanto T, Lassila R. Intravenousimmunoglobulin treatment in a type 3 von Willebrand disease patient withalloantibodies and a life-threatening gastrointestinalbleed. Haemophilia 2019; 25: e291-e293.

IV. Nummi V, Lassila R, Joutsi-Korhonen L, Armstrong E, Szanto T.Comprehensive re-evaluation of historical von Willebrand diseasediagnosis in association with whole blood platelet aggregation andfunction. International Journal of Laboratory Hematology 2018; 40:304-311

These articles are reprinted here with the kind permission of their publishers.

*These authors shared first authorship.

9

4 ABBREVIATIONS

ADP adenosine 5-diphosphate

APTT activated partial thromboplastin time

AUC area under the curve

AVWS acquired von Willebrand syndrome

BAT bleeding assessment tool

BS bleeding score

CAT calibrated automated thrombogram

CFT clot formation time

CT closure time

EPI epinephrine

ETP endogenous thrombin potential

F coagulation factor

FVIII:C factor VIII coagulant activity

GI gastrointestinal

GPIbα platelet glycoprotein Ibα

GP IIb/IIIa platelet integrin αIIbβ3

HMWM high-molecular-weight multimers

IVIG intravenous immunoglobulin

MCF maximal clot firmness

HGVS Human Genome Variation Society

PC protein C

pd plasma-derive

PPP platelet-poor plasma

PS protein S

PRP platelet-rich plasma

PFA platelet function analyser

ppVWF von Willebrand factor propeptide

10

ROTEM rotational thromboelastometry

RIPA ristocetin-induced platelet aggregation

TFPI tissue factor pathway inhibitor

TF tissue factor

TEG thromboelastography

TG thrombin generation

TXA tranexamic acid

VWD von Willebrand disease

VWF von Willebrand factor

VWF:Ag von Willebrand factor antigen

VWF:RCo von Willebrand factor ristocetin induced activity

VWF:CB von Willebrand factor collagen binding

VWF:GPIbM von Willebrand factor activity by glycoprotein Ib

binding assay

11

5 INTRODUCTION

Von Willebrand disease (VWD) is an inherited bleeding disorder, firstdescribed by Finnish internist Erik von Willebrand in 1926 (von Willebrand,1926). He reported a novel bleeding disorder in a large family from Föglö onthe islands of Åland. The index case, a 5-year old girl, Hjördis, had severemucosal bleeding symptoms. Twenty-three out of Hjördis’ “family S” of 66were also affected and four of Hjördis’ sisters had died between the ages of 2 to4 due to uncontrolled bleeding. Unlike the x-chromosomal inheritance seen inhemophilia, in this new bleeding disorder both sexes were affected andmucosal bleeding was the dominant symptom. In a landmark publication, Erikvon Willebrand concluded in 1926 that the disorder occurred due to combinedplatelet dysfunction and a defect of a vessel wall and named it “hereditarypseudohemophilia” (Lassila et al., 2013). In 1934, the index case Hjördis bledto death following her fourth menstrual period at the age of 14.

The cause of VWD was discovered later. In the 1950s it was demonstratedthat these patients display reduced FVIII activity. However, it was only in the1970s that it became clear that the cause of this bleeding disorder wasdeficiency or dysfunction of FVIII-associated plasma protein named as vonWillebrand factor (VWF) (Stites et al., 1971, Zimmerman et al., 1971). VWF hascentral role both in primary hemostasis mediating platelet adhesion andaggregation, and also in secondary hemostasis, as VWF is the carrier proteinfor FVIII (Sadler 1998). Quantitative deficiencies (type 1 and 3) or dysfunction(type 2) of VWF lead to VWD.

Two Swedish scientists and pioneers, Inga-Marie Nilsson and MargaretaBlombäck, were first to treat VWD by human plasma fraction (Nilsson et al.,1957). They also visited the Åland Islands in 1957, 1977 and 1990s andexamined the family members of the original “family S”. They confirmed thatthe family carried a deletion in exon 18 of VWF (c.1668delC), leading tolowered VWF levels (Nilsson 1999, Blombäck 1999, Zhang et al., 1993). Hjördishad inherited this defect from both of her parents, and indeed had the mostsevere form of the disease (type 3).

Clinical features of VWD are heterogeneous. Mucocutaneous bleeds,epistaxis, post-surgical bleeds, menorrhagia and delivery related bleeds arecommon among affected adults (de Wee et al., 2012). Typical for VWD is thegeneralized bleeding tendency presented from multiple sites. However,bleeding symptoms can be present also in normal population and thedistinction between type 1 VWD and a healthy individual with mild bleedingsymptoms is not always straightforward. VWF levels show wide variationamong both healthy individuals and also affected patients, and VWF laboratorytests are not infallible (Favaloro et al., 2018). Symptomatic patient with lowVWF not fulfilling the actual VWD criteria may or may not be diagnosed and

12

treated as having VWD, depending on the physician (Flood et al., 2016). Duringthe past years, hematologists have been advised not to overdiagnose type 1VWD, and Sadler provocatively called VWD type 1 `a diagnosis in search of adisease` (Sadler 2003).

The bleeding symptoms are usually more prominent in severe VWD (type 2and type 3) and the initial diagnosis by laboratory analysis is morestraightforward. However, optimal treatment with VWF replacement therapyand management of these patients is often challenging under severalcircumstances. As a treatment complication, 5-10% of patients with type 3develop anti-VWF alloantibodies (James et al., 2013). Hemarthrosis, a typicalsymptom for hemophilia patients, occurs also in severe VWD when FVIII is lowin a frequency of 35-45 % and may lead to severe arthropathy (van Galen et al.,2015). Gastrointestinal bleeds can be life-threatening and respondunpredictably to treatment (Makris et al., 2015). Currently, there is no clearconsensus on the initiation of prophylactic factor concentrate therapy in VWD(Abshire et al., 2015).

In the present series, VWD is investigated both from clinical and laboratorypoint of view. We have first characterized 10 patients with severe type 3 VWDand report the clinical heterogeneity and genetic background of this group. Tofurther elucidate the hemostatic defect among these patients, global hemostaticcapacity (thrombin generation [TG] and thromboelastometry [ROTEM]) andflow cytometric analysis of platelets were performed. A case report focusing onthe treatment of a patient with anti-VWF alloantibodies is reported. Finally, wehave assessed the accuracy of historical VWD diagnoses (given median 20years ago) in 83 patients focusing on VWF-dependent platelet functions inwhole blood.

13

6 REVIEW OF THE LITERATURE

6.1 VWF AND HEMOSTASIS

Hemostasis is the physiological reaction to vascular injury and tissue damage.Normal hemostasis leads restoring blood vessel integrity at vascular injury sitefirst by platelet plug formation (platelet adhesion and aggregation) and thenfibrin formation (coagulation system) and subsequently dissolving the plug. Itis a highly regulated and interactive process, involving vasculature, blood flow,platelets, coagulation factors, coagulation inhibitors and fibrinolytic system.

VWF is the largest human plasma glycoprotein which has a major role bothin primary hemostasis by mediating platelet adhesion and aggregation and alsoin secondary hemostasis, as VWF is the carrier protein for coagulation factorVIII (FVIII). For the purpose of this study, this review will focus on VWF inprimary and secondary hemostasis.

6.1.1 VWF: SYNTHESIS, STORAGE, SECRETION AND CLEARANCE

Unlike most plasma coagulation proteins synthesized by hepatocytes, VWF issynthesized by endothelial cells and platelet precursor cells, megacaryocytes(Nachman et al., 1977, Jaffe et al., 1973). VWF gene is located on the short armof chromosome 12 consisting of 52 exons in the span of approximately 178-180kB (Ginsburg et al., 1985, Sadler et al., 1985, Lynch et al., 1985). Secondpseudogene with 97% homology on chromosome 22 has been identified, withseveral stop codons implying it is not expressed in humans (Mancuso et al.,1991). However, by gene conversions it affects the mutation spectrum detectedin VWD.

VWF is encoded as a large precursor molecule consisting of 22-amino acidsignal pre-propeptide, 741-amino acid propeptide (ppVWF) and mature 2050-amino acid VWF. VWF has highly complex mosaic structure with repeatedhomologous domains and is rich in cysteine: cysteines constitute 8.2% of itsamino acids (4-fold of normal). Historically, VWF has been considered toconsist of four repeated homologous domains (A to D, Figure 1) (Pannekoek etal., 1989). Recently, the structure has been revised after structural andbioinformatical studies (Figure 1) (Zhou et al., 2012). Accounting for thehemostatic functions of VWF, binding sites on VWF have been located forFVIII, platelet glycoprotein Ibα (GPIbα), platelet integrin αIIbβ3 (GP IIb/IIIa),collagen (type I, type III and type VI) and ADAMTS13 (Figure 1). Additionalbinding sites with number of proteins such as TSP-1 (adhesive glycoprotein

14

Figure 1. Structure of VWF as historically recognized in panel A: D1-D2-D’-D3-A1-A2-A3-D4-B1-B2-B3-C1-C2-CK (including presentation of binding sites for various proteins, adapted fromSadler et al., 2006) and the revised structure in panel B (adapted from Zhou et al., 2012). In therevised structure, the D-domains comprise VW domain, C8 fold, a trypsin inhibitor-likestructure and an E-module. D’ domain does not contain VW or C8, and D4 has a unique D4Ndomain instead of the E-module. Also, in the C-terminal of VWF, 6 repeated VW C-domains in aknot-like formation replace the BC-domain. VWD = von Willebrand domain, C8 = 8-cysteine,TIL = trypsin inhibitor- like, CK = cysteine knot.

mediating cell adhesion found in platelet α-granules), OPG (a tumor necrosisfactor receptor) and heparin (inhibiting VWF-GPIbα intercation) have beenlocated, but the biological functions of these interactions are less well known(Pimanda et al., 2002, Zannettino et al., 2005., Sobel et al., 1991).

VWF undergoes complex, highly regulated post-translational modificationsin the endoplasmic reticulum and the Golgi apparatus. First, in theendoplasmic reticulum, the signal peptide is removed, and N-glycosylation isinitiated by addition of 17 N-linked carbohydrate structures. This is also thesite where VWF monomers form dimers in a “tail-to-tail” fashion, linked byinterchain disulfide bonds in the cysteine knot of the C-terminal (Katsumi etal., 2000). During transit to the trans-Golgi network, the pH drops to 6.2 andVWF dimers obtain the shape of a bouquet (Zhou et al., 2011).

Next, in the Golgi VWF dimers are covalently linked in a “head-to-head”fashion by inter-chain disulfide bonds that involve the paired cysteine residuesin the D3-domain of the N-terminal (Sadler 1998). This is a process calledmultimerization. ppVWF facilitates the multimerization as an endogenouschaperone, possibly catalyzing the disulfide bond formation via its proteinisomerase activity (Wise et al., 1998, Mayadas et al., 1992). The multimersformed are heterogeneous in size, containing 2 to >60 subunits (Sadler 1998).Multimer size (classified as low, intermediate, high and ultra large molecularweight multimers) determines the adhesiveness of VWF, ultra large multimersbeing the most potent. The multimers organize into helical structures, whereppVWF and D´-D3 domains form the interior wall of a hollow tubule. The

D1 CKA1 A3A2D2 D´ D3 D4 B1 B2 B3 C1 C2

propeptide mature VWFsignalpeptide

ADAMTS13cleavage site

FVIII collagenTSP-1

GPIbcollagen

OPGANG-2Heparin

IIbdimerizationmultimerization

furin

A

VWD1 C8 -1

E 1

D1 assembly

VWD2 C8 -2

TIL -3

E 2

D2 assembly

TIL` E` VWD3 C8 -3

TIL -2

E 3

D`D3 assembly

TIL -1

VWD4D4N

C8 -4

TIL -4

A1 A3A2

D4 assembly

C1 C2 C3 C4 C5 C6 CTCK

B

15

remainder VWF molecule protrudes outward from the tubule (Huang et al.,2008). Also in the Golgi, maturation of the N-linked glycans proceeds and 10O-linked carbohydrate structures are added to VWF (Canis et al., 2010). Furincleavages ppVWF from VWF, though it remains covalently associated to VWF(Rehemtulla et al., 1992).

Both megakaryocytes and endothelial cells synthesize VWF. In themegakaryocyte cells VWF is packed into α-granules and released upon plateletactivation. In contrast, in the endothelial cells, majority of the newlysynthetized VWF is constitutively released to plasma. Smaller amounts of VWFproduced in the endothelial cells are stored in Weibel Palade bodies, to bereleased during endothelial activation (Valentijn et al., 2013, Nightingale et al.,2013). Ultra-large multimers of VWF are abundant in Weibel-Palade bodiesand α-granules, whereas the constitutively secreted VWF is rich in highmolecular weight multimers (HMWM) (Stockschlaeder et al., 2014). VWFsynthesized in the endothelial cells appears to be sufficient to support normalhemostasis. However, it is possible that in the absence of endothelial VWF, theVWF produced in the megakaryocytes can compensate for the defect (Kanaji etal., 2012).

Stress stimulation, in vivo TG, fibrin, histamine, vascular injury anddesmopressin administration all lead to release of VWF from the Weibel-Palade bodies of the endothelial cells. Different types of exocytosis have beenobserved, depending on the physiological conditions present (Valentijn et al.,2011, Nightingale et al., 2013). Constitutively released VWF originates fromexocytosis single Weibel Palade bodies, whereas under stimulation, Weibel-Palade bodies fuse into secretory pods, and bundles of VWF are released(Giplin et al., 2008).

When released, ppVWF and VWF dissociate at the neutral pH of blood. Thehalf-life of VWF shows variation: endogenous VWF has a half-life of 4 to 26hours in plasma following desmopressin administration, whereas plasma-derived VWF concentrates have a fairly standard half-life of approximately 16hours (Millar et al., 2008, Favaloro et al., 2007). The half-life of VWF dependson the clearance rates of VWF, which are dependent on VWF glycosylationpatterns. The effect of ABO blood group to the half-life of VWF is the mostevident. ABO blood group determinants are present on 13% of the N-linkedglycans and 1% of the O-linked glycans of VWF, and individuals with bloodgroup O have 25% lower VWF levels due to increased clearance (Gill et al.,1987).

For a long time, it was not clear how VWF is removed from plasma.Immunohistochemical studies have revealed a cellular clearance of VWF andVWF bound FVIII by macrophage pathway both in the liver and spleen (vanSchooten et al., 2008). This is a somewhat unexpected clearance pathway, ascoagulation proteins are usually removed via scavenger receptors localized onhepatocytes, renal cells or endothelial cells.

16

6.1.2 VWF AND PRIMARY HEMOSTASIS

The main function of VWF in hemostasis is tethering platelets at the site ofvascular injury, enabling the formation of a platelet-plug. In the absence ofinjury, VWF circulates in globular form in the bloodstream and does notappear to interact with platelets.

Platelets are small 2 μm diameter anucleated disc-shaped cells that stemfrom megakaryocytes, with average lifespan of 8-9 days (Harker et al., 2000).Platelets express several membrane-based receptor complexes and containstorage granules, α granules and dense granules. α granules contain VWF andadditionally various other proteins such as P-selectin, fibrinogen, fibronectin,FV, FVIII and fibrinogen. Dense granules contain adenosine diphosphate(ADP), adenosine triphosphate, calcium, magnesium, serotonin and histamine(Rendu et al., 2001). Due to their shape and small size, platelets are located atthe edge of the flowing blood in the circulation, normally without interactionwith the endothelium (Brass et al., 2016). Healthy endothelium releases nitricoxide, prostacyclin and endothelial ecto-nucleotidase that prevent plateletactivation (Marcus et al., 2005).

Upon vascular injury, the subendothelial layer of the vessel, collagen andtissue factor (TF)-expressing cells become exposed. VWF is released from theendothelial cells. Upon high shear stress conditions in the arterial circulation,VWF is required for platelet adhesion, whereas in low shear stress conditionsin the venous blood, exposed matrix proteins (collagens, fibronectin andfibrinogen/fibrin) of the damaged endothelium lead to platelet adhesion(Ruggeri 2002, Farndale et al., 2004). VWF binds to the subendothelial matrixprimarily via immobilized collagen at sites of high shear stress. The larger thesize of the VWF multimer, the higher the affinity to bind collagen is. VWFbinds to fibril forming collagens type I and III via A3 domain, and fibrillarcollagen VI via A1 domain.

The anchored VWF multimers unfold into strings under the shear force ofblood flow, exposing binding sites for platelet GPIbα on the VWF A1 domain(Michaux et al., 2006, Valentijn et al., 2010). Platelets bind to VWF via GPIbαof the GPIb-IX-V receptor complex, but this binding is not stable and has faston/off rates. However, the circulating platelets slow down and roll on the VWFcoated endothelium. This surface enables stable platelet adhesion, whichoccurs via platelet binding to collagen with GPIa-IIa and GPIIb-IIIa receptors.Before binding, these receptors are activated by platelet GPVI receptors (Chenet al., 2002).

VWF itself undergoes conformational change after binding to plateletGPIbα, and the A2 domain of VWF is exposed to ADAMTS13 (a disintegrin andmetalloproteinase with a thrombospondin type 1 motif, member 13) (Budde etal., 2014). ADAMTS13 rapidly cleaves VWF strings and inhibits the formationof elongated VWF multimers. This enzymatic function is critical formaintaining normal hemostasis as it prevents thrombi formation in themicrovasculature. Antibodies for ADAMTS13 lead to thrombotic

17

thrombosytopenic purpura, TTP. Platelet derived VWF shows resistance toADAMTS13 proteolysis, which can have hemostatic effect at sites of vascularinjury and platelet activation (McGrath et al., 2013)

Following platelet adhesion, the exposed collagen triggers the initial plateletactivation. The activated platelets release their secretory granules and therelease of ADP and exposure to thrombin lead to the full platelet activation.Platelets undergo a morphological change of shape, from disc shape to a spherewith dendritic extensions increasing their surface area and engagement tofibrin and clot retraction. More platelets from the bloodstream are recruitedand the initial platelet plug forms via platelet aggregation (Welsh et al., 2017).Platelet-platelet binding occurs via activated GPIIb/IIIa, and this is mediatedby fibrinogen (Beguin et al., 1999). However, in cases of low fibrinogen or if thelocal concentration of VWF is high, also VWF facilitates platelet aggregation bybinding GPIIb/IIIa (Schullek et al., 1984). Finally, phophatidylserine andethanolamine on the activated platelet flip to the outer surface. Thisprocoagulant surface is where assembly of coagulation factors occurs togenerate thrombin and fibrin.

6.1.3 COAGULATION SYSTEM

VWF contributes to secondary hemostasis as a carrier and stabilizer of FVIII.FVIII has a major role in the coagulation system, and deficiency of FVIII leadsto a severe bleeding disorder, hemophilia A. Vice versa, high levels of FVIII arethrombogenic (Jenkins et al., 2012). VWF is present in plasma at anapproximate 50:1 ratio to FVIII. Each VWF multimer contains a high-affinitybinding site to FVIII, and all sizes of VWF multimers are capable ofchaperoning FVIII (Lenting et al., 1998). When bound to VWF, FVIII isprevented from taking part to the coagulation cascade. FVIII is also protectedfrom proteolytic degradation by activated protein C and cellular clearance.VWF has a longer half-life than FVIII, and higher VWF levels increase the half-life of infused FVIII in hemophilia A (Fijnvandraat et al., 1995). In the absenceof VWF (type 3 VWD), FVIII levels are low, as the half-life of FVIII is onlyapproximately 2 hours.

Overview of the coagulation system involving FVIII is presented in briefhere according to the original articles and textbooks (Blombäck et al., 2010,Hoffman et al., 2007, Mann et al., 2005, Roberts et al., 2004).

Two groups proposed the historical “cascade” or “waterfall” model ofcoagulation system in 1964 (Davie et al., 1964, MacFarlane 1964). In thismodel, coagulation consists of two separate pathways, the extrinsic andintrinsic pathways, both capable of independently initiating clot formation byactivating FX in the common pathway. The extrinsic pathway is TF-triggered,and the intrinsic proceeds from FXII. This model is not physiological.Deficiencies of FVIII and FIX (hemophilia A and B) in the intrinsic pathwaylead to severe bleeding tendency and are not compensated by the extrinsic

18

pathway. On the contrary, FXII deficiency does not cause clinical bleeding eventhough FXII has role in thrombosis (Maas et al., 2018). Furthermore, patientswith FXI deficiency have a variable bleeding tendency, not predicted by theirFXI level (Pike et al., 2011). Although not physiological, the standard screeningcoagulation tests rely on the cascade model, as prothrombin time (PT)measures the extrinsic pathway and activated partial thromboplastin time(APTT) the intrinsic pathway. APTT as a screening test detects hemophilia A,and also VWD in cases of low FVIII.

The current view of coagulation system is the cell-based model ofhemostasis as described by Hoffmann and Monroe (Hoffmann et al., 2001). Inthis model, coagulation involves a series of coordinated calcium-dependentproenzyme to serine conversions with overlapping phases, taking place on cellsurfaces: TF-expressing cell membrane and the phospholipid surface of theactivated platelets. Coagulation occurs in three steps: initiation, amplificationand propagation (Figure 2).

Figure 2. The cell-model of coagulation system consists of 3 overlapping phases: initiationphase occurring on TF-expressing cell surface and amplification and propagation phases takingplace on platelet surface. The cascade leads to large-scale thrombin formation and theconversion of fibrinogen into insoluble fibrin mesh. Adapted from De Caterina et al., 2013.

In the initiation phase, TF-expressing cells become exposed to bloodcomponents at vascular injury site. TF activates FVII and forms a complex,TF/FVIIa on the TF-expressing cell surface, which in turn activates FX to FXa.FXa is rapidly inhibited by antithrombin and tissue factor pathway inhibitor(TFPI) if it leaves the cell surface (Mast 2016, Quinsey et al., 2004). The FXa

19

that remains on the cellular surface activates FV, and small amounts ofthrombin are produced. This amount of thrombin is insufficient to promotefibrin generation, but it acts in a positive feedback loop to promote coagulationby activation of platelets, FVIII, FV and FXI in the amplification phase.Platelets become activated and coagulation occurs now on the surface ofactivated platelets, whereas initiation phase occurred on TF-expressing cellsurface. FV is activated by thrombin or FXa. Thrombin activates FVIII andcleaves it from VWF. FXIa, activated by thrombin, in its turn activates FIX, alsoactivated by TF/FVIIa.

As platelets are fully activated and have FVIIIa and FVa bound on theirsurface, propagation phase can start. Tenase (FVIIIa/FIXa) andprothrombinase (FXa/FVa) complexes are assembled on the surface ofplatelets. Tenase complex is formed when FIXa reaches FVIIIa on the plateletsurface. Tenase complex in its turn activates FXa, inducing prothrombinasecomplex with FVa. Prothrombinase complex converts prothrombin tothrombin, leading to large-scale thrombin generation. Thrombin rapidlycleaves soluble fibrinogen into fibrin, leading to insoluble fibrin mesh. Thisoccurs in the presence of FXIII, which thrombin converts to FXIIIa, to befurther stabilized by the aggregated platelets.

In conclusion, in distinction to the cascade model, in the cell based model,the extrinsic pathway occurs on the TF-expressing cell membrane and theintrinsic pathway on the activated platelet surface. In FVIII deficiency(hemophilia A and also severe VWD), there is a failure of platelet surface FXactivation, leading to decreased TG on platelet surface and ineffective clotformation (Monroe 2003).

The function of several coagulation inhibitors beyond antithrombin andtissue factor pathway inhibitor are required to halt uncontrolled thrombosisand maintain normal hemostasis. Protein C and protein S are vitamin Kdependent natural anticoagulants. Protein C is a zymogen that is activated to aserine protease after binding to thrombin in the presence of thrombomodulinin a negative feedback loop driven by thrombin (Bouwens et al., 2013). ProteinS acts as a cofactor for protein C, and this complex proteolytically inhibits FVaand FVIIIa (van de Meer 2014). Protein S circulates both in free form and in acomplex bound to complement C4b protein. In addition to being a cofactor forprotein C, free protein S binds to FVa and FXa and serves as a cofactor fortissue factor pathway inhibitor as well (Hackeng et al., 2006).

20

6.2 VWD: CLINICAL ASPECTS

6.2.1 EPIDEMIOLOGY AND CLASSIFICATION

VWD is frequently described as the most common inherited bleeding disorder.Population study among school children in Northern Italy measuring VWFlevels suggested the prevalence of type 1 VWD to be as high as 1% (Rodeghieroet al., 1987). An even higher prevalence of 1.3% was discovered among multi-ethnic children from the United States in another cross-sectional study(Werner et al., 1993). However, a follow-up study of 13 years of the first Italianstudy reported that only 1 out 10 of the subjects diagnosed with VWD hadpresented with bleeding symptoms (Castaman et al., 1999). Symptomatic VWDis thus less common, and perhaps a more realistic prevalence of 1 per 1000 wasreported based on a cross sectional study of symptomatic VWD in primaryhealth care in Canada (Bowman et al., 2010). In the Nordic region, anestimated prevalence of 8 per 10 000 was reported in 2004 (Berntorp et al.,2005). An even lower prevalence of 1 per 10 000 has been estimated for VWDrequiring transfusion therapy (Sadler et al., 2000).

The prevalence of VWD is dependent on the diagnostic cut-off used forVWF, which in the current guidelines ranges from <30-40 IU/dL (Laffan et al.,2014, Nichols et al., 2008, Lassila et al., 2011, Castaman et al., 2013). Clinicianscan also diagnose patients to have VWD with higher levels as there is no realconsensus. No cut-off is given by the International Society on Thrombosis andHaemostasis (ISTH) guidelines (Sadler et al., 2006). VWD is inherited eitherautosomal dominantly or recessively, suggesting that both males and femalesshould be equally affected. VWD is, however, more commonly diagnosed infemales than males in a ratio of approximately 2:1, due to the hemostaticchallenges women in reproductive age encounter (menstruation and delivery)(Lillicrap et al., 2013).

VWD is presently classified into following types according to the ISTHrecommendations: type 1 with partial deficiency of functionally normal VWF,type 2 with qualitative VWF dysfunction and type 3 with virtually completedeficiency of VWF (Sadler et al., 2006) (Table 1). Type 1 is sometimes furtherclassified into 1C with increased VWF clearance, although 1C is not an officialsubtype of VWD. Type 2 is categorized into following subtypes: 2A with thelack of high molecular weight multimers of VWF, 2B with increased VWFbinding to platelet GP1bα, 2M with defective VWF associated platelet functionin the presence of normal multimers and 2N with binding defect of VWF toFVIII. Platelet type VWD, caused by a defect in the platelet receptor GP1bα andresulting in a phenotype similar to that of subtype 2B VWD, is sometimes alsoclassified as a subtype of VWD.

21

Definition Inheritance

Type 1 Partial quantitative deficiency of VWF Autosomal dominant (variablepenetrance)

Type 2A Deficiency of high molecular weight multimers anddecreased VWF-dependent platelet adhesion

Autosomal dominant or recessive

Type 2B Increased affinity to platelet glycoprotein Ibα Autosomal dominant

Type 2M Decreased VWF-dependent platelet adhesion notassociated to lack of high molecular weightmultimers

Autosomal dominant or recessive

Type 2N Deficiency of VWF:FVIII binding Autosomal recessive

Type 3 Complete deficiency of VWF Autosomal recessive

Table 1. The classification of VWD according to the ISTH guidelines (adapted from Sadler et al.,2006).

The current type 1 VWD diagnostic guidelines are mainly based on threelarge multicenter type 1 VWD studies conducted in Europe, Canada and theUnited Kingdom (Goodeve et al., 2007, James 2007 et al., Cumming et al.,2006). In these studies, VWF mutations were detected only in approximately65% of type 1 patients. However, the percentage increased in cases where VWFlevels were lower than 30 IU/dL, and 88% of these patients had a candidatemutation in the European study. Based on these results, the current guidelinesintroduced lower diagnostic cut-offs for type 1 VWD of <30-40 IU/dL, eventhough the diagnosis of VWD was not restricted to patients who have a VWFmutation (Laffan et al., 2014, Nichols et al., 2008, Lassila et al., 2011,Castaman et al., 2013).

The term `low VWF` was given to patients who have VWF at the level of30-50 IU/dL and present with bleeding symptoms. Since then, in theZimmerman study in the United States (US), it was demonstrated that asignificant portion of patients who carry a historical type 1 diagnosis currentlyhave either ´low VWF` or normal VWF levels (Flood et al., 2016). However,despite the normalized VWF levels, these patients had significant bleedingsymptoms. Recently, results from Ireland `low VWF` study including 126patients with VWF 30-50 IU/dL, indicated that low VWF patients havesignificant bleeding symptoms primarily due to decreased synthesis orsecretion of VWF (Lavin et al., 2017). Also, it was demonstrated these patientshave a good response to desmopressin in terms of VWF rise and hemostaticoutcome. Previous data have indicated that other factors besides low VWF maycontribute to the clinical phenotypes of these patients, such as platelet functiondefects (Daly et al., 2009).

22

6.2.2 CLINICAL FEATURES AND EVALUATION OF BLEEDINGSYMPTOMS

The clinical phenotype of VWD is usually severe in type 3, moderate to severein type 2 and milder in type 1. In children, the most common symptoms areepistaxis and bruising (Sanders et al., 2015). In contrast, in adults the mostfrequently reported symptoms include hematomas, menorrhagia, bleedingfrom minor wounds and bleeding after surgery or dental extraction, as reportedin nationwide studies from Holland (WIN-study) (de Wee et al., 2012) andFrance (Veyradier et al., 2016). Menorrhagia is common, and reported in morethan 80% of women with VWD in the WIN-study, and conversely 5-20% ofwomen with menorrhagia have VWD (de Wee et al., 2012). Joint bleeds, themost common manifestation of hemophilia, are rare in VWD, but occurespecially in subtype 2N and type 3 where FVIII is markedly decreased (vanGalen et al., 2012). Recurrent gastrointestinal bleeds and repeated irondeficiency anemia, often associated to angiodysplasia, are among the mostchallenging clinical complications of VWD and cause significant morbidity(Makris et al., 2015). They have been reported in all types of VWD, but areespecially linked to severe types.

Postpartum bleeding may occur in VWD, but bleeding symptoms can alsoimprove during pregnancy in VWD with increasing VWF levels. VWF levelsincrease by 3-5-fold in healthy women during pregnancy. Majority of womenwith type 1 VWD correct their VWF levels during pregnancy, especially in casesof mild deficiency, whereas patients with VWF <20 IU/dL usually have poorincrease of VWF (Castaman 2013). Type 2 and 3 patients do not typicallycorrect their levels.

It is a well-known challenge to reliably and accurately evaluate bleedingtendency. Bleeding symptoms are subjective, and the interpretation andreporting of the symptoms depend on the individual. In pediatric patients, thevalue of family history is crucial, as bruising and epistaxis are common also inhealthy children. Several bleeding assessment tools (BATs), which enable thecalculation of a bleeding score (BS), have been developed for standardizedevaluation of bleeding symptoms. The suggested advantages of BATs includejustifying extensive laboratory studies for symptomatic patients, avoidingunwarranted laboratory testing in asymptomatic patients and the potential topredict individual bleeding risk in the future. They also enable rapidcommunication between physicians.

The first standardized BAT published in 1982 was aimed at screening forVWD. It consisted of 10 questions on procedure- and surgery-related bleedsand post-delivery and spontaneous bleeds (Wahlberg 1982). Later a group ofinvestigators from Vicenza, Italy developed and validated a BS for type 1 VWD,with condensed questionnaire published later as part of European Molecularand Clinical Markers for the Diagnosis and Management of type 1 VWD(MCMDM-1 VWD) study (Tosetto et al., 2006, Rodeghiero et al., 2005). Eachbleeding symptoms was scored from -1 to 4 in the condensed version, andhigher score was associated to type 1 VWD and likelihood of future bleeds. The

23

bleeding score inversely correlated with VWF levels. Also, the usefulness of theBS was demonstrated in a prospective Italian study following 796 VWDpatients for 1 year and reporting all bleeds that required desmopressin or factorconcentrate (Federici et al., 2014). Only Vicenza-based BS higher than 10 wasable to predict future bleeds in the cohort. However, this BS was criticized asthe score was based on the worst bleeding episode and did not take intoaccount the frequency of the symptoms.

In 2010, a working group of ISTH/SSC established the ISTH-BAT, with theaim to have one standardized BAT with improved accuracy for mild bleedingdisorders and improved grading for patients with known bleeding disorders(Rodeghiero et al., 2010).

Most studies have investigated BATs in the context of screening for mildbleeding disorders, especially VWD type 1 (Azzam et al., 2012, Tosetto et al.,2011, Deforest et al., 2015, Phillip et al., 2011). It has also been demonstratedtype 2 and 3 VWD are associated with higher BAT compered to type 1, thoughthere is an overlap between the patient groups (Bowman et al., 2008). A recentsystematic review comprised of 9 studies (4 investigating Vicenza based BS and5 ISTH-BAT) concluded that BATs have only low to moderate diagnostic oddsratio for bleeding disorders (Moenen et al., 2018). Studies with goodmethodological quality had generally lower diagnostic odds ratio. Nevertheless,BATs are useful tools for clinicians enabling a structured interview.

6.2.3 TREATMENT

The current available treatment options for VWD include tranexamic acid,desmopressin and factor concentrates that contain either high purity VWF orintermediate purity VWF/FVIII. For menorrhagia, hormonal treatment shouldbe also considered. Treatment can be administered either on-demand,following a spontaneous or post-traumatic bleed, or prophylactically, beforeinvasive procedures or surgeries.

TRANEXAMIC ACIDTranexamic acid is an antifibrinolytic that can be administered topically, orallyor intravenously. It can be used as an adjuvant therapy for desmopressin orVWF concentrate, or on its own for example in small procedures (Pasi et al.,2004). It should be noted tranexamic acid may accumulate in renalinsufficiency (Wellington et al., 2003)

DESMOPRESSINDesmopressin is a synthetic analog of vasopressin that leads to release ofendogenous VWF and increase of FVIII and tissue plasminogen activatorplasma levels. Desmopressin also enhances platelet procoagulant activity(Colucci et al., 2014). Desmopressin is the first line treatment for mild VWD

24

and can be administered intranasally, subcutaneously or intravenously.Interindividual response to desmopressin is variable, and trial withdesmopressin treatment is recommended in type 1, 2A, 2M and 2N VWD. Thebest responses are usually measured in type 1 (Mannucci et al., 2004).Response should be measured at 30-60 minutes and at 4-6 hours afteradministration to recognize patients who have enhanced clearance.Desmopressin is not useful in type 3, and it is not recommended in 2B either,as the response is generally poor and it may cause thrombocytopenia (Federiciet al., 2008). Disadvantages of desmopressin include flushing, hypotension,risk for fluid retention and risk for hyponatremia. Desmopressin iscontraindicated in patients who have atherosclerosis, high risk for thrombosisor history of seizures.

VWF CONCENTRATESThere are several virally inactivated plasma-derived VWF concentrates(pdVWF) with excellent safety profiles available. Administration of pdVWFresults in a fairly constant increase of VWF concentration of 0.021-0.024(IU/mL) (IU/kg) (Mannucci et al., 1992). However, concentrates differ in theirVWF multimer compositions and the amount of FVIII (Batlle et al., 2009,Budde et al., 2006). VWF:RCo/VWF:Ag ratio reflects the multimericcomposition of the concentrate, and a ratio close to 1 is preferred as indicatesnormal multimeric structure. The appropriate amount of FVIII in relation toVWF is not well defined. A ratio of 1:1 is easy to dose, but all availableconcentrates have shown good clinical results (Castaman et al., 2011). In type 3VWD, after a high purity VWF concentrate, it takes more than 12 hours forFVIII levels to normalize (Borel-Derlon et al., 2007). While this appears to beconvenient when administrating prophylaxis to avoid accumulation of FVIII,upon acute bleeding or emergency surgery, VWF/FVIII or combination of highpurity VWF and FVIII concentrate should be used. Recombinant VWF, vonicogalfa, has been recently registered and approved for treatment of VWD.Intriguingly, as VWF is synthesized in the absence of ADAMTS13, the productcontains also ultra large multimers (Gill et al., 2015).

Regular prophylactic treatment with VWF concentrate in VWD is rarecompared to hemophilia, and usually given only in type 3 (Abshire et al., 2013).In a Swedish cohort, 74% of severe VWD patients were on prophylaxis, whereasamong the multicenter VWD prophylaxis network survey, the respectivepercentage was only 22% (Berntorp et al., 2005, Abshire et al., 2013). Theinitiation of prophylaxis depends on the availability of concentrate, and in anIndian cohort of 102 patients none of the patients received prophylaxis(Elayaperumal et al., 2018). Typical reasons for initiating regular prophylaxisinclude epistaxis/oral bleeding, gastrointestinal bleeding, joint bleeding andmenorrhagia. There is only one prospective study on prophylaxis in VWD,demonstrating good efficacy in mucosal and joint bleed rates (Abshire et al.,2015). From clinical experience, it is known that the response to prophylaxis is

25

less effective to control gastrointestinal bleeds, and these may remain difficultto manage.

TREATMENT OF PATIENTS WITH ANTI-VWF ALLOANTIBODIESAlloantibody formation against VWF is a rare life-threatening complication inVWD occurring in 6-10% of type 3 patients (Iorio et al., 2008, Mannucci et al.,1995). The appearance of alloantibodies was associated to partial or completeVWF gene deletions in the first case descriptions published, but since then alsocases of nonsense and frameshift mutations have been reported (Shelton-Inloes et al., 1987, Ruggeri 1979, James et al., 2013).

Patients who develop alloantibodies become unresponsive to VWF therapy.Moreover, re-exposure to VWF containing products has been described to leadto immune-complex mediated complement activation and life-threateninganaphylactic reactions (Mannucci et al., 1987). To date, treatment of thesepatients is based on case reports and expert opinions. Currently, as it is notknown which patients develop these severe allergic reactions, VWF containingproducts are often considered contraindicated in cases of known VWFalloantibodies (James et al., 2013). Both recombinant FVIII (rFVIII) andrecombinant factor VIIa (rFVIIa) have been effectively used to treat bleeds andmanage surgery and delivery among these patients (Mannucci et al., 1995,Ciavarella et al., 1996).

Use of intravenous immunoglobulin (IVIG) has been reported in acquiredvon Willebrand syndrome (AVWS), unlike in cases of inherited VWD-associated antibodies (Federici et al., 2000). However, immunomodulativetherapy has been used as adjuvant therapy in successful pediatricimmunotolerance induction regimens in a 9-year-old boy, who also receivedIVIG and in 2 brothers receiving mycophenolate mofetil and rituximab(Pergantou et al., 2012, Berntorp et al., 2018).

6.2.4 LABORATORY EVALUATION OF VWD

The diagnosis of VWD is based on the presence of both bleeding symptoms ofthe patient and/or family members and laboratory evidence of abnormal VWFand/or FVIII. There are a number of diagnostic algorithms for laboratoryevaluation of VWD, but the search of a diagnosis should only be initiated ifthere is evidence of abnormal bleeding tendency.

Currently, no rapid screening tests are available for VWD and the availablescreening tests are mainly used to exclude other bleeding disorders. Plateletcount is usually normal but can be lowered in 2B and platelet type VWD. PT isnormal in VWD. APTT is usually also normal, but can be prolonged to avariable degree, depending on the FVIII levels and the sensitivity of the APTTreagent.

26

Classic diagnostic tests for VWD diagnosis comprise VWF antigen(VWF:Ag) measuring VWF concentration, VWF activity and FVIII activity(FVIII), and for the first-line investigation they are sufficient (Figure 3, Table2). For secondary classification of VWD, further analysis is often required,including multimer analysis, ristocetin-induced platelet aggregation (RIPA),and VWF:FVIII binding assay and genetic analysis.

Figure 3. A diagnostic algorithm for laboratory investigation of von Willebrand disease,adapted from Laffan et al., 2014.

ScreeningPT, APTTPFA-100/PFA-200Complete blood countBlood group

First levelVWF:AgVWF:RCoVWF:CBFVIII:C

If only FVIII:C is low,consider 2Nor mild hemophilia A

Undetectable VWF:Ag

Type 3

Reduced VWF:RCoand/or VWF:CB

VWF:RCo/VWF:Agratio >0.5-0.7

Type 1

VWF:RCo/VWF:Agratio <0.5-0.7

Second level

Response tolow concentration RIPA

Type 2Bor platelet-type

Normal/reduced RIPA

Loss of HMWM,VWF:CB/VWF:Ag ratio low

Preservation of HMWM,normal VWF:CB/VWF:Ag ratio

Type 2AType 2M

27

SCREENING OF VWD BY PFAPlatelet function analyzers PFA-100® and PFA-200® (PFA) (Siemens,Marburg, Germany) are primary hemostasis screening devices, measuringplatelet and VWF function under flow. The instrument uses whole bloodsamples and is fast and easy to use. A sample is aspired through a capillarytube spiked with collagen/epinephrine or collagen/adenosine diphosphatecoated membrane with central aperture. The time to close the aperture (closuretime, CT) by platelet plug formation is measured.

Several studies have demonstrated a high sensitivity (96.5-100%) of PFAfor VWD and specificity of 40%, especially in type 2 and 3 VWD (Fressinaud etal., 1998, Ardillon et al., 2015). The sensitivity for type 1 is only around 80%,but possibly higher in cases with more reduced VWF (<25 IU/dL) (Favaloro etal., 2008, Sadler et al., 2006). In contradiction to other reports, one studyreported a sensitivity of only 61.5% for VWD (Quiroga et al., 2004). Also, PFAis less sensitive for specific platelet function defects as compared with lighttransmission aggregometry and flow cytometry analysis (Quiroga et al., 2004,Favaloro et al., 2015).

FIRST-LINE LABORATORY ANALYSISVWF antigen measures the quantity of VWF protein in the plasma byimmunological method, either by ELISA using poly- and/or monoclonalantisera or by immunoturbidometric methods that rely on latex particleagglutination (Cejka et al., 1982, Veyradier et al., 1999).

VWF activity is measured by analysis of VWF binding to platelet GPIb andcollagen. Traditional VWF ristocetin cofactor activity (VWF:RCo) test wasdeveloped in 1970s. The assay uses ristocetin, an antibiotic sulphate thatpromotes interaction between VWF A1 domain and platelet GPIb receptorunder static conditions, leading to platelet agglutination. The use of VWF:RCoassay has disadvantages that include high inter-laboratory coefficient ofvariation of 20-50% and lower limit of detection often as high as 10-20 U/dL,although automation of the assay has led to improvements (Kitchen et al.,2006, Meijer et al., 2006). VWF:RCo is also affected by some benign VWFsequence variations that hinder VWF’s binding to ristocetin, such as p.D1472H(Flood et al., 2010).

The new VWF activity assays, being independent of ristocetin andlyophilized platelets, may overcome some of these disadvantages, and haveemerged during past years to replace VWF:RCo (Lawrie et al., 2013). Theseassays rely either on the gain-of-function GPIb mutants that bind VWF withoutthe site-specific change of the charge caused by ristocetin, or on a monoclonalantibody that recognizes the functional GPIb-binding epitope on VWF(VWF:GPIbM) (Timm et al., 2015).

Measurement of VWF activity by VWF and collagen interaction (VWF:CB)is sensitive to VWD variants that are characterized by lack of high moleculareight multimers. Elisa-based assays that measure the interaction between VWFand immobilized collagen have been available for several years, using

28

preferably type I/III collagen mixtures (Favaloro et al., 2007). MeasuringVWF:CB has been complimentary to VWF:RCo, and used more in Australiaand Europe compared to the US (Favaloro 2018). VWF:CB/VWF:Ag ratio lowerthan 0.6 is suggestive of 2A/2B with reduced high molecular eight multimers,whereas higher ratio indicates 2M (Favaloro et al., 2010). Recently, anautomated chemiluminescence-based assay has become available, withreduced variability and higher sensitivity at low VWF levels (Favaloro et al.,2016, Stufano et al., 2018). The automated VWF:CB assay may, however, havelower sensitivity for discrimination between 2A/B and 2M (Favaloro 2018).

FVIII is measured using an APTT-base one-stage clotting assay or bychromogenic assay (Kitchen et al., 2016). The assays measure the functionalactivity of FVIII.

VWF PROPEPTIDEMeasurement of VWF propeptide can be used to analyse the pathophysiologyof type 1 (Eikenboom et al., 2013). VWF propeptide and VWF remainnoncovalently bound when stored in Weibel Palade bodies of the endotheliumor α-granules of platelets. When released in plasma, these molecules dissociate(Borchiellini et al., 1996). VWF propeptide has significantly shorter half-life of2-3 hours compared to VWF, and because of this measuring VWF propeptideratio to antigen ratio offers insight into the clearance of VWF. A high ratio ofVWF propeptide to antigen indicates increased clearance of VWF. In the WINstudy, it was demonstrated that absence of VWF propeptide discriminates type3 from severe type 1 patients, who have rapid VWF clearance (Sanders et al.,2015).

MULTIMER ANALYSISLow-resolution agarose gel electrophoresis can be used to study plasmamultimeric patterns of VWF and to distinguish subtypes of VWD. For the strictclassification of type 2 VWD, multimer analysis should be performed, thoughthis is not necessarily required from clinical perspective. There are severalvisualization techniques (Krizek et al., 2000, Ott et al., 2010). Low agaroseconcentration of 1% can be used in the gels to detect the presence of highmolecular weight multimers. Type 1, 2M and 2N have all sizes of multimerspresent, whereas in type 2A intermediate and high molecular weight multimersare missing. High molecular weight multimers are usually absent in type 2B,but patients can also have normal multimer pattern (Federici et al., 2009).Abnormal multimer pattern in 2B may predict increased bleeding risk(Casonate et al., 2017).

Higher concentration 1.5-3% agarose gel can be used to detectabnormalities of VWF satellite bands (triplets). By measuring subtle alterationsof the inner structure of multimers, subtype 2A can be further subclassifiedaccording to previous guidelines, no longer endorsed by the current guidelines(Sadler et al., 1994).

29

Assay

PFA-100 / PFA-200 Primary hemostasis screening test

VWF:Ag Quantity of VWF protein in the plasma measured by immunologic assays

VWF:RCo VWF activity via ristocetin induced VWF binding to platelet GPIb understatic conditions

VWF:GPIbM VWF activity via monoclonal antibody directed against GPIb bindingepitope of VWF

VWF:CB VWF binding to collagen, sensitive to the presence of HMWM

FVIII:C Coagulant activity of FVIII

VWF:pp The quantity of VWF propeptide. High ratio of VWF:pp/VWF:Ag indicatesincreased clearance rate of VWF

Multimer analysis Analysis of structural features of VWF by visualization of VWF multimers.

RIPA Ristocetin induced platelet aggregation. Can be measured under variableristocetin concentration in PRP or in whole blood.

VWF:FVIII binding Quantitative analysis of VWF activity by assessment of VWF adhesion toFVIII.

Table 2. Laboratory assays used to diagnose and classify VWD. Adapted from Sadler et al.,2006

New rapid and semi-automated assays, providing quantitativedetermination of VWF multimers have been developed to standardize themethod and overcome technical difficulties (Favaloro et al., 2017, Pikta et al.,2018).

RISTOCETIN INDUCED PLATELET AGGREGATION (RIPA)VWD 2B and platelet type VWD are characterized by pathological enhancement ofVWF and platelet GPIb interaction. Gold standard for the diagnosis of 2B has been todemonstrate increased platelet aggregation at low ristocetin (<0.6 mg/mL) inplatelet-rich plasma in Bonn aggregometry (Born 1962). In contrast, RIPA at higherconcentration is usually lowered in other types of VWD and absent in type 3(Mannucci 1977).

Born aggregometry is technically challenging and time-consuming. While notpractical to perform for all VWD cases, RIPA is recommended in cases with low VWFactivity to antigen or VWF collagen binding to antigen ratios or thrombocytopenia(Federici et al., 2009, Laffan et al., 2014).

Multiplate® is a new generation whole blood aggregometer, a relatively fastmethod and easy to use. In a comparative study, Multiplate results weredemonstrated to correlate with Bonn aggregometry among 30 VWD patients

30

(Valarche et al., 2011). Furthermore, a recent study of 100 VWD patients suggestedthat RIPA by Multiplate may predict the bleeding tendency of VWD patients(Schmidt et al., 2017).

VWF BINDING TO FVIIIVWF binding to FVIII can be measured by in vitro analysing the binding ofVWF to exogenous FVIII or by ELISA-based technique (Mazurier et al., 1990,Zhukov et al., 2009). In type 2N VWD, VWF binding to FVIII is impaired.

VWD type VWF:Ag VWF:RCo FVIII Other features

low VWF 30-50 IU/dL 30-50 IU/dL Normal orreduced

Type 1 reduced reduced normal orreduced

VWF:Ag and VWF:RCo equally reducedFVIII:VWF:Ag 1.5all sizes of multimers present, uniformlydecreased

Type 2A reduced ornormal

reduced normal orreduced

VWF:RCo/VWF:Ag <0.6 lack of HMWM VWF:CB /VWF: RCo <0.6

Type 2B reduced ornormal

reduced(occasionallynormal)

normal orreduced

VWF:RCo/VWF:Ag <0.6 increased RIPA at low ristocetin thrombocytopenia

Type 2M reduced ornormal

reduced normal orreduced

VWF:RCo /VWF:Ag <0.6normal multimers VWF:CB /VWF: RCo >0.6

Type 2N normal orreduced

normal orreduced

reduced FVIII/VWF:Ag ratio of <1 defective VWF:FVIII binding

Type 3 absent absent reduced absent multimers

Table 3. The characteristics of VWD types and subtypes in laboratory analysis. VWF: RCo =ristocetin induced VWF activity, VWF:Ag = VWF antigen, VWF:CB = VWF collagen binding.Adapted from Laffan et al, 2014.

6.2.5 GENETIC ANALYSIS

Phenotypic assays are usually sufficient to subtype VWD patients. Currently,genetic analysis of VWF is recommended routinely only in 2B, 2N and withconsideration in VWD patients having inconsistent clinical/laboratory findings,and in type 3 cases in connection with genetic counseling. Genetic analysis doesprovide further information on the mechanisms of the VWF defect. Byconventional polymerase chain reaction sequencing and multiple ligand-

31

dependent probe amplification, mutation detection remains both timeconsuming and challenging. Next-generation sequencing enables fasterdiagnostics and is rapidly gaining popularity in VWD (Batlle et al., 2016,Fidalgo et al., 2016, Veyradier et al., 2016).

Various kinds of mutations in VWF lead to VWD. Causative mutationsreported include transcriptional, splice site, missense and nonsense mutations,small deletions, small insertions and small duplications, gene conversions andlarge deletions (Goodeve 2010). Novel variants detected may not fall intospecific types or subtypes of VWD, and sometimes a dual classification can beused in cases with compound heterozygous mutations (Sadler et al., 2006).

GENETIC ANALYSIS IN TYPE 1In type 1 VWD, genetic testing is usually not encouraged. Genetic defects arescattered across the VWF gene and analysis of all 52 exons is needed. Type 1 isusually inherited autosomally dominantly, but several mutations showincomplete penetration. Main mutation mechanisms behind type 1 includeincreased clearance, decreased synthesis/secretion and intracellular retention(Goodeve 2010, Veyradier et al., 2016, Wang et al., 2011). Type 3 carriers canpresent with a type 1 phenotype as demonstrated in the Canadian type 3 studyand French VWD study (Bowman et al., 2013, Veyradier et al., 2016).

Genetic defects were detected in approximately 65% of the cases in theEuropean, UK and Canadian type 1 studies (Cumming et al., 2006, Goodeve etal., 2007, James et al., 2007). The percentage increased to up to 88% in caseswhere VWF levels were lower than 30 IU/dL. In up to 15% of the patients, twoor more candidate mutations were detected. More recently, in the French studyincluding type 1 patients with VWF level <30 IU/dL, mutations werediscovered in 96% (161/167) of the subjects by next generation sequencing(Veyradier et al., 2016).

GENETIC ANALYSIS IN 2In type 2 VWD, mutations are located in specific functional domains of VWF.Missense mutations, small deletions, insertions or duplications are usuallydetected. Mutations are identified in exons 11-16, 24-26, 28 and 51 in type 2A,exon 28 in 2B and 2M and exons 18-25 in 2N, and thus the genetic analysis ismore straightforward than in type 1 (Figure 4) (Goodeve 2010, Corrales et al.,2012). Furthermore, in type 2 genetic analysis helps in differential diagnostics,in particular distinguishing subtype 2B from platelet-type VWD and type 2Nfrom hemophilia A. As this affects the treatment options and geneticcounseling of the patients, genetic confirmation for diagnosis of 2B and 2N isencouraged. Especially in cases where multimer study is not available, genetictesting is useful for confirmation of a specific subtype.

32

Figure 4. The locations of the reported VWF mutations among VWD types and subtypes in theVWF protein. Adapted from Goodeve, 2000.

GENETIC ANALYSIS IN TYPE 3Type 3 has historically been described to be inherited via an autosomalrecessive trait with either compound heterozygous or homozygous mutations,the latter discovered especially in nations with consanguineous marriages. TheCanadian type 3 and French VWD study have demonstrated that carriers oftype 3 can have type 1 phenotype (Bowman et al., 2013, Veyradier et al., 2016).Indeed, in the family of Hjördis, the patient described by Erik von Willebrand,several of the carriers of the disease were also symptomatic (Lassila et al.,2013). Genetic diagnostics is useful in this patient group enabling prenataldiagnostics and genetic counseling.

Gene defects in type 3 are scattered throughout the gene and detected in upto 90% of patients with conventional sequencing and mutations scatteredthroughout the 52 exons (Goodeve 2010). In the Canadian type 3 study,mutations in propeptide region of the gene resulted with higher bleeding scoresbut this observation has not been repeated in further cohorts (Bowman et al.,2013).

In most screened nations, no founder mutations have been discovered, andtypically several novel mutations are identified in nationwide cohorts.However, the C.2465delC mutation in exon 18 has been observed in countriessurrounding the Baltic Sea frequently. Also a few mutations have been detectedonly in specific national cohorts, implying local founder effect for these specificmutations. A deletion of 35kB covering exons 1 to 3 has been discovered only inHungary (Mohl et al., 2008). In a large Indian cohort of 102 patients, 5mutations were also noted repeatedly (Elayaperumal et al., 2018).

OTHER GENETIC LOCI MODIFYING VWF LEVELSVWF gene variants fail to entirely explain the heterogeneity detected in VWFlevels in patients and healthy individuals, as even in patients with identicalmutations the VWF levels vary (Goodeve et al., 2007, Robertson et al., 2011).Several studies have researched for additional loci that could contribute tothese variable levels and the incomplete penetrance of mutations.

D1 CKA1 A3A2D2 D´ D3 D4 B1 B2 B3 C1 C2

propeptide mature VWFsignalpeptide

exon: 2type 1type 2Atype 2Btype 2M

type 2N

type 3

3 - 10 11 - 17 18 - 20 2820 - 28 28 28 - 32 33 - 34 35 - 39 42 - 4440 - 42 45 - 48 49 - 52

33

It has been known for long that ABO blood group variant affects VWFlevels, with O blood group resulting in lower VWF antigen levels (Gill et al.,1987). Type O blood group is more prevalent in type 1 and type 2 VWDcompared to normal population. ABO-blood group appears to have a role in thepathophysiology of low VWF as well. In a Canadian type 1 study, 66% ofpatients who had VWF <30 IU/dL had type O blood group compared to 46% inthe normal population (James et al., 2007). In the EU type 1 study, 76% ofpatients without a causative mutation in VWF had type O blood groupcompared with its frequency of 38% in the normal population (Goodeve et al.,2007). Moreover, in the Irish `low VWF` study, 89% of patients had bloodgroup O (55% in normal Irish population) (Lavin et al., 2017).

Single nucleotide polymorphism (SNPs) affecting VWF levels have beenaslo identified. CLEC4M and LRP1 are associated to increased clearance ofVWF, and STXBP5 affects VWF exocytosis (Smith et al., 2010, Rastegarlari etal., 2012, Rydz et al., 2013). There are various other candidate genes discoveredin next generation sequencing (Smith et al., 2010, van Loon et al., 2016).Additional genetic loci unassociated to VWF levels can also modify the clinicalexpression of VWD.

6.2.5 GLOBAL HEMOSTASIS ASSAYS

Traditional coagulation tests are adjusted for measuring specific parts of thecoagulation system. They do not capture the complexity of hemostasis and areoften insensitive to the clinical phenotypes of bleeding disorders andprothrombotic states (Young et al., 2013). Global hemostasis assays have beendeveloped with the aim to fulfill this gap: to assess the overall hemostaticsystem and the balance between procoagulants and anticoagulants. However,the important exclusions of hemorheology and vascular wall remain in theseassays as well.

For the purpose of the present studies, this review will focus on the twotypes of global hemostasis assays that have this far been studied in bleedingdisorders but are not in routine clinical use: thrombin generation (TG) testsand rotational thromboelastography/metry.

THROMBIN GENERATION (TG) TESTS

Thrombin plays a central role in the coagulation cascade, with concentration-dependent procoagulant, anticoagulant, fibrinolytic and antifibrinolyticproperties. There are no bypassing mechanisms to TG and the extent of TGcontributes to the final formation of a hemostatic plug. The most commonlyused TG test, calibrated automated thrombogram (CAT), was developed byHemker (Hemker et al., 1993). A trigger is added to recalcified plasma in the

34

presence of phospholipids to initiate TG. Thrombin formation leads to cleavingof a fluorogenic substrate recorded by the device.

Figure 5. Thrombin generation curve by calibrated automated thrombogram (CAT) in aplatelet-poor plasma sample of a normal control and a type 3 VWD patient.

The calibrated automated thrombogram measures thrombin activitycontinuously and records a TG curve (Figure 5). The usual parameters ofinterest include lag time (time to initiate TG), peak thrombin (peak height ofthe curve), time to peak (time to reach peak thrombin) and endogenousthrombin generation (ETP, total amount of thrombin formed). TG assay can beperformed in platelet-poor plasma (PPP) but platelet-rich plasma (PRP) allowsevaluating the effect of platelets and is considered to better reflect physiologicalconditions. In most studies, the platelet number is standardized, usually to150x109/L. TG is traditionally triggered by tissue factor (TF) (Baglin et al.,2011). Also, FXIa has been used as an alternative trigger (Waters et al., 2015).Studies have applied several different TF (0.5-5 pM) and phospholipidconcentrations (van Veen et al., 2008a). Studies involving patients withhemophilia are generally performed under low TF stimulation, which issensitive to defects of intrinsic complex formation (van Veen et al., 2008a).However, under low TF stimulation it is sometimes not possible to obtain fullTG curves due to delayed reaction.

The Subcommittee on Control of Anticoagulation of the SSC of the ISTHhas published a communication towards a recommendation forstandardization of the TG tests (Baglin et al., 2011). Performing TG tests issubject for various confounders and requires careful consideration of pre-analytical conditions (Rodgers et al., 2014). PPP and PRP should be preparedand the assay performed within 30 minutes to 2 hours after sampling. Samplesshould be pre-heated for 10 minutes before assay initiation. Recently, a

Thro

mbi

n(n

M)

35

multicenter study in the Nordic countries for standardization of TG bycalibrated automated thrombogram has produced results which are as robustas standard coagulation assays used in the routine laboratories (Ljungkvist etal., 2018).

TG has been studied extensively in hemophilia and other bleeding disorders(van Veen et al., 2008a). It has been established that the measured TG showsgood correlation to FVIII and FIX levels in hemophilia (Nummi et al., 2017).The association to clinical phenotypes and adjusting treatment by TG testsremains uncertain, with promising results from a few studies. A comprehensivestudy investigating 22 severe hemophilia patients with mild bleedingphenotype demonstrated that higher endogenous thrombin potential inplatelet-rich-plasma identified the patients with mild phenotype from controls(Santagostino et al., 2010). Earlier, similar results were obtained in a study ofpatients with rare coagulation factor (factor II, V, X and XI) defects (Al Dieri etal., 2002). A study including 39 FXI deficient patients also indicated that TGtest distinguished between bleeding and non-bleeding phenotypes in FXIdeficiency (Livnat et al., 2015). These results remain to be confirmed in largemulticenter studies, which remain difficult to perform considering thetechnical and standardization requirements of TG tests. There are also studieswith negative results of using TG tests to predict the clinical phenotypes in FVIIdeficiency and hemophilia A patients with inhibitors (Tran et al., 2014, Salinaset al., 2015). Recent studies have focused on measuring TG in cases wherecoagulation factor levels cannot be measured: inhibitor patients andmonitoring new non-factor therapies.

There are two studies investigating TG in VWD. The first study examinedtwo type 1 patients with mild to moderate VWF deficiency(VWF:RCo/VWF:Ag/FVIII 17/21/50% in patient 1 and 49/52/72% in patient2, respectively) (Keularts et al., 2000). TG was measured by sub-samplingmethod in stirred plasma at low shear rate. TG was initiated by kaolin in PPP,in the presence of added phospholipids and in PRP, but without a trigger. TGwas impaired in the two patients in PRP in contrast to PPP. After scramblingthe platelet membrane with ionomycin, TG normalized in PRP also.Desmopressin normalized TG in PRP, whereas TG in plasma was unaffected.The authors concluded VWF is required for normal coagulation in the presenceof fibrin and platelets, and that TG in PRP appeared to be a suitable test toassess VWF defects.

In the largest study to date, Rugeri et al. measured TG by the calibratedautomated thrombogram in 53 VWD patients, including 28 type 1, 23 type 2and 2 type 3 patients (Rugeri et al., 2007). TG was initiated by tisuue factor at1 pM in the presence of 4 μM phospholipids in PPP and by TF 0.5 pM in PRPsamples. In the study, TG was impaired in VWD compared to normal controlsboth in PPP and PRP, and correlations between VWF activity and FVIII and TGvariables were demonstrated. TG in type 1 and type 2 VWD was at a similarlevel. Patients with a higher historical bleeding score had low TG compared topatients with lower bleeding score. By spiking type 3 PRP and plasma with

36

recombinant FVIII and/or VWF, the study demonstrated FVIII had greatercapacity to correct TG compared to VWF. At normalized FVIII level, addingVWF had no effect on TG. The authors concluded that the decreased TG inVWD in mainly attributed to the decreased FVIII levels. The authors suggestedthat low TG is associated to a more severe bleeding phenotype in VWD.However, evaluation of platelet adhesion and aggregation are required toactually predict bleeding risk in VWD patients. The study included only twotype 3 patients, and TG in severe VWD is largely unexplored.

THROMBOELASTOGRAPHY (TEG) ANDTHROMBOELASTOMETRY (ROTEM)Thromboelastography (TEG) and rotational thromboelastometry (ROTEM) aretwo global hemostatic devices measuring clot formation, strength of the formedclot and subsequent clot lysis in whole blood (Whiting et al., 2014). Theseassays have a rapid turnaround time of 15-20 minutes and are increasinglyused in emergency and perioperative setting.

The principle of TEG was introduced already in 1948 (Hertert 1948). In theTEG system, a whole blood sample is placed into an oscillating cylindrical cup,containing a pin attached via a torsion wire. During analysis, anelectromechanical transducer measures the viscoelastic strength of the clot viathe torsion wire continuously. ROTEM is a modern modification of TEG withautomated pipetting in which the pin oscillates instead of the cup. The assaysproduce a figure/curve as an output demonstrating the clot dynamics (Figure6). Both devices measure essentially the same parameters but have distinctnomenclature. Usual variables of interest include the time to reach a clot with a2 and 20-millimeter size, the angle of the developing curve, maximal peakamplitude of the clot and the timing of the clot lysis.

By using different reagents, both systems can analyze different aspects ofhemostasis. In TEG, the available tests include kaolin vials to measure contactpathway and TF activation to measure common pathway, functional fibrinogentest, native analysis and platelet mapping (measurement of platelet response toADP/arachidonic acid). Respectively, ROTEM can be used to measure thecontribution of the intrinsic pathway (INTEM) and the tissue factor-inducedextrinsic pathway (EXTEM) to clot formation. The degree of fibrinolysis(APTEM), and the contribution of fibrinogen to clot strength independently ofplatelets (FIBTEM) can be assessed. Also, native samples with recalcificationonly can be assessed (NATEM). There are also assays to analyze the effects ofanticoagulants for both devices. Studies indicate that TEG can provide usefuldata for monitoring treatment of hemophilia patients with inhibitors (Young etal., 2006). However, the role of TEG/ROTEM in screening for inheritedbleeding disorders is not well defined. Recently, Schmidt et al investigatedROTEM and TEG comprehensively in 100 VWD patients (27 type 1, 53 type 2and 20 type 3) (Schmidt et al., 2017). The study demonstrated that both, theprolonged reaction time (time to reach a clot size of 2 millimeters) and clottingindex (global assessment of parameters) in TEG, have excellent discriminative

37

value for VWD, partly due to lowered FVIII levels. Association to clinicalbleeding phenotype was not reported. In contrast, ROTEM had only limitedaccuracy. Only five patients had abnormal clotting time (time to reach a clotsize of 2 millimeters) in ROTEM.

Figure 6. A depiction of thromboelastography (TEG) and thromboelastometry (ROTEM) output.The devices have distinct nomenclature for the captured parameters. R = reaction time, K =kinetics of clot formation, MA = maximum amplitude, CL = clot lysis, CT = clotting time, CFT =clot formation time, MCF = maximum clot firmness, LY = lysis, α = alpha angle.

38

7 AIMSThe general aim of the study was to evaluate clinical phenotypes and theirvariability in VWD in association to traditional clotting factor activities, globalhemostasis assays, platelet function analysis, and genetic analysis.

The specific aims of the study were to:

I. Describe the genotypic and phenotypic heterogeneity of Finnishpatients with severe type 3 VWD (I)

II. Assess thrombin generation, rotational thromboelastometry andplatelet activation by flow cytometry in relation to clinicalphenotypes in type 3 VWD (II)

III. Present a case report on the management and outcome of a life-threatening bleed in a patient with type 3 VWD and anti-VWFalloantibodies (III)

IV. Evaluate whole blood platelet aggregation and function in anassociation with bleeding phenotype and VWD type in a cohort ofpatients with historical VWD disease diagnosis (IV)

39

8 PATIENTS AND STUDY DESIGN

8.1 PATIENTS AND STUDY DESIGN (I-III)For studies I to III, patients diagnosed with type 3 VWD and followed up at theCoagulation Disorder Unit in the Helsinki University Hospital were enrolled(Table 4).

No. Age(years)and sex

VWF:RCo

(IU/dL)

VWF:GPIbM(IU/dL)

VWF:Ag

(IU/dL)

VWF:CB

(IU/dL)

FVIII(IU/dL)

Study I StudyII

1. 76 / F <6 <4 <5 <3 1 x x

2. 74 / F <6 <4 <5 <3 2 x x

3. 70 / M <6 <4 <5 <3 3 x x

4. 67 / F <6 <4 <5 <3 1 x x

5. 65 / F <6 <4 <5 <3 1 x -

6. 64 / M <6 <4 <5 <3 4 x x

7. 56 / F <6 <4 <5 <3 1 x x

8. 49 / F <6 <4 <5 5 4 x x

9. 44 / F <6 <4 <5 <3 14 x x

10. 36 / M <6 - <5 <3 3 x -

11. 20 / F <6 <4 <5 <3 3 - x

Table 4. Type 3 VWD patients included in studies I-III with baseline laboratory results. VWFlevels are unmeasurable in all. In patient 8, VWF:CB was available only after 5 day washout,indicating measurable activity of 5 IU/dL. F = female, M = male.

In study I, we investigated the clinical features, standard laboratoryphenotypes, multimers and genotypes of 10 unrelated type 3 patients. Thestudy was performed as a collaborative project with prof. ReinhardSchneppenheim’s laboratory in University Medical Center Hamburg-Eppendorf and Medilys Laboratory (Hamburg, Germany).

In study II, we analyzed TG, ROTEM, and platelet activation using flowcytometry in type 3 VWD. Nine patients were recruited. Samples were collectedafter 3-7 days washout from replacement therapy in patients on prophylactic

40

therapy. Response to treatment at 30 minutes post therapy was investigated in6 patients.

In study III, the management of a life-threatening gastrointestinal bleedcomplication of patient 7 with anti-VWF alloantibodies is reported.

8.2 PATIENTS AND STUDY DESIGN (IV)Patients with historical VWD diagnosis referred to and assessed at CoagulationDisorder Unit in the Helsinki University Hospital during years 2010-2015 wereenrolled to study IV: a total of 83 patients, 38 type 1, 32 type 2, and 13 type 3cases. The previous VWD diagnosis were made outside the setting ofcomprehensive bleeding disorder care.

Patients’ VWD diagnoses were re-evaluated according to the NordicHemophilia Council guidelines, with a significant bleeding history associatedwith lowered VWF levels (<35 IU/dL) as the primary diagnostic basis. Type 3was defined as undetectable VWF antigen and type 2 was defined as abnormalVWF:RCo/VWF:Ag ratios of 0.7 (excluding 2N).

A quantitative standardized Vicenza based BAT was calculated tocharacterize the bleeding phenotype covering all historical bleeds. Centralizedmeasurements of VWF and FVIII were carried out and repeated a minimum of3 times, at least 2 months apart. Whole blood platelet functions were assessedwith PFA and Multiplate in all.

8.3 ETHICAL ASPECTSStudies were conducted according to the declaration of Helsinki and approvedby the ethical committee of Helsinki University Hospital. All patients gavewritten consent.

41

9 METHODS

9.1 CLINICAL EVALUATION OF THE BLEEDINGSYMPTOMS (I, II, IV)In study I, Vicenza based bleeding score (BS) (Tosetto et al., 2006) wascalculated for included patients. Additionally, data on bleeding symptoms andtype and mode of treatment (on-demand or prophylaxis) were collected fromthe medical records. In study II, we determined the BS with ISTH BAT(Rodeghiero et al., 2010) and calculated the annual bleed rate (ABR) based onpatient reports.

In study IV, Vicenza based BS (Tosetto et al., 2006) was determined for allpatients carrying a historical VWD diagnosis. The score was consideredabnormal at a score > 5 in females and >3 in males. The calculation of BAT wasbased on an interview with a specialized physician, reporting all bleeds prior tothe evaluation.

9.2 LABORATORY ANALYSIS

9.2.1 BLOOD SAMPLING (I-IV)

Analyses were performed in Coagulation Disorder Unit, Department of ClinicalChemistry, HUSLAB Laboratory Service, Helsinki University Central Hospital,unless otherwise stated. Under fasting conditions, peripheral venous bloodsamples were collected without any known acute traumas, bleeds or surgeries.The samples were obtained into vacuum collection tubes containing:

(1) EDTA for blood cell counts and genetic analysis;

(2) sodium citrate (109 mM sodium citrate, 3.2%) for coagulation factorassays, PFA-100, TG assay, ROTEM, flow cytometric analyses and lighttransmission aggregometry study in PRP;

(3) recombinant hirudin (lepirudin 25 μg/mL, 400 ATU/mL; DynabyteMedical, Munich, Germany) for whole-blood aggregometry studies byMultiplate.

42

9.2.2 COAGULATION ASSAYS (I-IV)

Citrated samples were centrifuged (2000 g, 10 min) and frozen in aliquots at−70 °C, if not tested immediately. VWF activity was measured either byVWF:RCo assay using BS VWF reagent (I, III-IV) or by VWF:GPIbM assay (II).VWF:Ag was quantified with VWF Ag Latex reagent and FVIII activity by one-stage clotting assay (Pathromtin SL and Coagulation FVIII Deficient Plasma).Reagents for VWF:RCo, VWF:GPIbM, VWF:Ag and FVIII:C were obtainedfrom Siemens Heathcare Diagnostics and the analyses were performed on aBCS-XP Analyzer (Siemens Healthcare Diagnostics, Marburg, Germany).VWF:CB was determined by Technozym® VWF:CB enzyme-linkedimmunosorbent assay (Technoclone GmbH, Vienna, Austria) using an EVOLISAnalyzer (Bio-Rad Laboratories, Berkeley, CA, USA). The reference rangeswere as follows (IU/dL): VWF:RCo 44-183, VWF:GPIbM 50-190, VWF:Ag 51-169, and FVIII 52- 148 (120 controls) and VWF:CB 50-170 (20 controls).

PT (Nycotest PT®, Oslo, Norway), APTT (Actin FSL®), factors V, VII,IX, X, XII, XIII (reagents from Siemens, Pathromtin SL and factor deficientplasma) and fibrinogen (Clauss method, Multifibren® U) were measuredon a BCS XP Analyzer (II, IV).

Prothrombin fragments F1+2 indicating in vivo TG were measured byenzyme immunoassay Enzygnost® (Siemens) with reference range of 69-229 pM (II). ADAMTS13 activity was determined by Technozym® ELISA,with reference range of 40-130 % (II). Total and free TFPI antigens weremeasured by Asserachrom ELISA (Stago, Asnières sur Seine, France) at theDepartment of Hematology, Oslo University Hospital (kind courtesy of Dr.Per Morten Sandset), and compared to in-house controls (13.4 ng/mL forfree 89.7 ng/mL for total TFPI) (II).

VWF binding to FVIII was evaluated by comparing the binding tonormal control, 2N homozygote, and 2N heterozygote (Finnish Red CrossBlood Service laboratory, Helsinki, Finland) (IV).

9.2.3 THROMBOPHILIA SCREEN (II)

Thrombophilia screen was performed for all patients in study II. Freeprotein S (PS) antigen was measured by latex-based method (IL Test FreeProtein S; Instrumentation Laboratory) and protein C and antithrombinactivities by chromogenic assays (Berichrom Antithrombin III andBerichrom Protein C; Dade Behring). The lupus anticoagulant was detectedusing two screening tests based on APTT (IL Test APTT-SP,Instrumentation Laboratory, Italy) and Russell Viper Venom activatedclotting time (DVVtest 10, American Diagnostica Inc). Anti-cardiolipin andanti-beta2 glycoprotein I -specific phospholid antibodies of IgG class weremeasured with immunological assays (Varelisa Cardiolipin IgG Antibodies

43

and Beta-2-Glycoprotein I IgG Antibodies, Phadia GmbH, Freiburg,Germany, respectively) with reference values of < 15 U/mL. The variants FVLeiden (R506Q) and prothrombin G20210A were detected using automatedPCR-based mini-sequencing.

9.2.4 MULTIMER ANALYSIS (I, IV)

In study I, VWF multimer analyses for type 3 patients were carried out by SDSagarose electrophoresis of patients' plasma, Western Blotting and luminescentvisualization recorded by photo-imaging in Medilys Laboratory (Hamburg,Germany), as described in detail previously (Schenneppenheim et al., 1988,Budde et al., 2008).

Multimer analyses for type 2 patients (VWF:RCo:VWF:Ag <0.7) in studyIV with similar technique were performed at Lund University, Malmö aspart of traditional clinical collaboration.

9.2.5 PLATELET FUNCTION ANALYSER (PFA) (I, IV)

In the PFA-100® (Siemens Healthcare Diagnostics, Marburg, Germany),closure times (CTs) on collagen cartridges spiked with either epinephrine(CT/EPI) or adenosine diphosphate (CT/ADP) were measured, using referenceranges CT/EPI 82-150 s and CT/ADP 62-100 s which were provided by themanufacturer. We had additional 27 healthy volunteer samples as controlgroup.

9.2.6 GENETIC ANALYSIS (I, IV)

In study I, genomic DNA was isolated from frozen buffy coats with standardprocedures for all type 3 patients. Molecular studies were carried out asdescribed previously in University Medical Center Hamburg-Eppendorf(Hamburg, Germany) (Schneppenheim et al., 2010). PCR products weresequenced with using the ABI Prism Big Dye Terminator Cycle sequencingready reaction kit. Sense and antisense PCR primer were used as sequencingprimers on an ABI 377 sequencer (Applied Biosystems, Weiterstadt, Germany).All patients were first screened for c.2435delC mutation in exon 18 of VWF. Incases of heterozygosity for c.2435delC or negative results, all the remainingexons of VWF were sequenced. Gene dosage analysis by Multiplex Ligation-Dependent Probe Amplification (MLPA, MRC-Holland, Amsterdam, TheNetherlands) was performed for all to exclude heterozygous deletions.Mutations were named according to the Human Genome Variation Society(HGVS) system. The ISTH/SSC VWF mutation and polymorphism database

44

and published literature were screened for previous descriptions of thediscovered mutations (http://www.vwf.group.shef.ac.uk/). The validity ofsplice site mutations was tested by two different splice site prediction programs(http://www.fruitfly.org/seq_tools/splice.html;http://cbs.dtu.dk/services/NetGene2/).

In study IV, for the purpose of the study, genetic testing was limited tospecific patients in accordance with the VWD guidelines (Laffan et al., 2014):(i) patients in whom phenotypic analysis did not clarify VWD type or subtype,(ii) patients who were previously diagnosed with or suspected of 2B or 2N todifferentiate from platelet-type VWD or hemophilia A, and (iii) patients withtype 3 to provide genetic counseling and to identify patients at higher risk ofdeveloping neutralizing antibodies. The analyses were performed by PCR basedsequencing in one of 4 laboratories, depending on the type of VWD and time ofthe analysis: Lund University Hospital (Malmö, Sweden) as part of traditionalclinical collaboration, Addenbrooke's Hospital Genetics Laboratories(Cambridge, UK), Sheffield Diagnostic Genetics Service (Sheffield, UK), andHamburg-Eppendorf University Clinic (Hamburg, Germany) for the type 3patients.

9.2.7 THROMBIN GENERATION BY CALIBRATED AUTOMATEDTHROMBOGRAM (CAT) (II)

TG was measured in platelet-poor (PPP) and –rich plasma (PRP) by CalibratedAutomated Thrombography (CAT, Thrombinoscope, Maastricht, theNetherlands). PPP was prepared by double centrifugation of 2000 g for 10 minand 10 000 g for 10 min and immediately stored at -80 °C. PRP was preparedby single (180 g, 10 min) centrifugation and platelet count was adjusted to 150x 109/L by autologous plasma. Measurements were performed as described byHemker et al (Hemker et al., 2000). Briefly, 80 μL of PPP was supplementedwith 20 μL of 1 pM TF and 4 μM phospholipids (PPP Reagent Low) andrespectively 80 μL of PRP with 20 μL of 1 pM TF without exogenousphospholipids (PRP reagent) or with inner method standard ThrombinCalibrator (all reagents from Thrombinoscope, Stago). Analyses were carriedout in triplicates for at least 60 min. We followed endogenous thrombinpotential (ETP), peak thrombin, and lag time. We had 14 normal controls.

The contribution of FVIII and protein S on TG in PPP were further analysedunder the same low 1pM TF activation (PPP reagent Low) in the Department ofMolecular and Cellular Hemostasis, Sanquin Research Laboratory, by Herm-Jan Brinkman (Sanquin, Amsterdam, the Netherlands). The response toexogenous FVIII was evaluated by spiking PPP with 1 IU/mL plasma-derivedhuman FVIII, as described previously (Bloem et al., 2012). The role ofendogenous FVIII was measured in the presence of VK34 (5 μg/mL), a FVIII

45

inhibitory antibody directed to the A2 domain of FVIII (van den Brink et al.,2000). The impact of the APC-independent activity of protein S was exploredby adding a protein S inhibitory antibody CLB-PS13 directed to the Gla domain(100 μg/mL), as described previously (Pitkänen et al., 2015). The results werecompared normal to pooled plasma and commercially available FVIII-depletedplasma (Siemens Healthcare GmbH, Erlangen, Germany).

9.2.8 ROTATIONAL THROMBOELASTOMETRY (ROTEM) (II)

Thromboelastometry in whole blood was measured within 2 hours of bloodcollection with automated pipetting program by ROTEM (InstrumentationLaboratory, Bedford, MA, USA), as recommended by the ISTH (Chitlur et al.,2014). We assessed the INTEM, EXTEM, and FIBTEM. The following variableswere recorded: clotting time (CT, s), clot formation time (CFT, s), alpha angle(α), maximum clot firmness (MCF, mm), and the area under the curve (AUC).The reference ranges were provided by the manufacturer: for INTEM, CT 100-240 s, CFT 30-110 s, α angle 70-83º, MCF 50-72 mm; and for EXTEM, CT 38-79 s, CFT 34-159 s, α angle 63-83º, MCF 50-72 mm and for FIBTEM, MCF 9-25 mm.

9.2.9 PLATELET ANALYSIS UNDER FLOW CYTOMETRY (II)

Platelet degranulation via adenosine-5-diphosphate (ADP) induced P-selectinexpression (PE-labelled anti-CD62P, Becton Dickinson) and the expression ofprocoagulant phosphatidylserine via GPVI-induced annexin V binding (FITC-labelled; Becton Dickinson, activation by cross-linked collagen-related peptide[CRP-XL], department of Biochemistry, University of Cambridge, UK) weremeasured by flow cytometry. A concentration series of ADP (1, 2, and 5 μM;Sigma) or CRP-XL (100 and 500 nM, 1.0, 7.5, and 15 μM) were incubated with5 μL of fresh, citrated whole blood for 20 minutes. Samples were fixed with200 μL of 0.2% formaldehyde, excluding the annexin V-stained samples.HBS/Ca2+ buffer was used in all experiments. The samples were analysed withFACScan (BD Accuri C6 Flow Cytometer). Data were expressed as percentageof fluorescence-positive platelets and as median fluorescence intensity(arbitrary units, AU). Patient samples were compared to 6 healthy controls.

9.2.10 PLATELET AGGREGATION (IV)

In study IV, whole blood platelet aggregation with the Multiplate® analyzer(Dynabyte Medical, Munich, Germany) was studied within 180 minutes ofvenipuncture. The following tests were carried out: standard 0.8 mg/mL ofristocetin (RistoHigh®) and 0.2 mg/mL of ristocetin (RistoLow®), and in-house modification of ristocetin at 0.6 mg/mL (Helena Biosciences, Gateshead,

46

UK) for analysis of VWF dependent platelet aggregation. ASPItest® (0.5 mmol/Lof arachidonic acid); ADPtest® (6.4 μmol /L of adenosine diphosphate);TRAPtest® (32 μmol/L of thrombin activating peptide); and COLLtest® (3.2μg/mL of collagen) were also performed. The results were expressed as the areaunder the curve (AUC, arbitrary units). Local reference ranges were establishedwith 20 controls (AUC): RistoHigh®: 70-210, RistoLow®: 0-9, ristocetin 0.6mg/mL: 60-110, and ADP: 40-110, ASPI: 70-125, TRAP: 70- 130, and COLL:40-110.

For clarification of subtype 2B in study IV, a platelet aggregation study inPRP (adjusted platelet count of 300 × 109 /L) with LTA (turbidimetricPPACKS-4 aggregometer, Helena Laboratories, Beaumont, TX, USA) wascarried out in 2 patients. Maximal aggregation (% of change in lighttransmission) at 5 minutes was recorded using several concentrations (0.4, 0.6,0.8, 1.0, and 1.2 mg/mL) of ristocetin (Helena Bioscience).

9.3 STATISTICAL ANALYSIS (I-IV)

For the recorded data in all studies, median, interquartile range (IQR) andrange were calculated. The significance of difference between two independentdata groups was examined with Mann-Whitney test and the significance ofdifference between two dependent data groups with Wilcoxon signed-rank test.The correlation between two variables was analyzed using Spearman’s test. Ap-value of <0.05 was considered significant in all studies

In study II, for samples analysed under variable concentrations ofactivation in flow cytometry, the significance of difference between groups wasexamined with Kruskar-Wallis test.

All calculations were performed with Prism 7.0 (GraphPad Software, LaJolla, CA, USA).

47

10 RESULTS

10.1 STUDY I

10.1.1 CLINICAL PHENOTYPES

At the time of the study, 7 of the 10 type 3 patients included were on regularprophylaxis with plasma derived (pd) VWF/FVIII concentrate Haemate® (CSLBehring), administered 1 to 3 times per week (Table 5). Three patients hadmilder clinical phenotype and received on-demand therapy in case of bleedsand ahead of procedures/surgery.

Patient 5 had history of an anti-VWF alloantibody, diagnosed after lack ofresponse to pdVWF/FVIII concentrate (no history of allergic reactions), andwas treated with dual therapy: Haemate® and pdFVIII Amofil® (Sanquin).Patient 7 had a history of an appearance of an anti-VWF alloantibody following2 deliveries and post surgery, and had had once allergic reaction topdVWF/FVIII as described in detail in study III. She had only low spontaneousbleeding tendency and was treated on on-demand.

Vizenca based BS exceeded 10 and indicated severe bleeding phenotype inall (Table 5). All patients had presented with severe epistaxis and all 7 femalesreported menorrhagia. Eight patients had history joint bleeds and consequentlyhad developed hemophilia-like arthropathy. None had a history of CNS bleeds.

Out of the 7 women included to the study, 3 had had pregnancies withdeliveries. All 3 women had previously 2 vaginal deliveries. The deliveries weremanaged with either cryoprecipitate or pdVWF, with therapy administered 7 to10 days postpartum in each case. No abnormal bleeding was associated to thesedeliveries.

10.1.2 MULTIMER ANALYSIS

Multimers were absent in plasma of all 3 patients treated on-demand, whereasmurky multimer pattern elucidated traces of factor concentrate in plasma ofthe 7 patients on prophylaxis (Figure 5).

10.1.3 GENETIC ANALYSIS

Of the 10 patients investigated, 7 were identified with homozygous and 3 withcompound heterozygous VWF mutations, all expected to lead to null alleles(Table 5). No large deletions were observed in any of the patients by MLPA.

The mutation originally described in Föglö of the Åland Islands, c.2435delC(p.Pro812ArgfsX31) in exon 18 was detected in 45% of the alleles (9/20). Themutation was homozygous in 4 patients, and patient 2 was compound

48

heterozygous for c.2435delC and a novel frameshift mutation c.1668delC inexon 13.

The nonsense mutation c.4975C>T (p.R1659X) was discovered in 40% ofthe alleles (8/20) and was homozygous in 3 patients. The families of thesehomozygous patients were all traced to Southern Ostrobothnia region ofFinland. Two patients had compound heterozygous mutations of c.4975C>Tand another mutation, in patient 1 a novel frameshift mutationc.2072delCCinsG in exon 14 and 7 in patient 8 a novel splice site mutationc.874+2T>C in intron. The c.874+2C>T variant destroys the canonical GT ofthe donor splice site and analysis by two different splice site predictionprograms predicted the complete loss of the intron 7 donor splice site.

Table 5. VWF variants detected in genetic analysis in the patient group. * Denotes novelpreviously not reported VWF variant. Variants are marked according to NCBI referencesequence NM_000552.3 of VWF cDNA, HGVS nomenclature.

No. Nucleotidechange

Aminoacid change Exon Mutation type TosettoBAT

1. c.4975C>T/c.2072delCCinsG*

p.R1659X/p.Pro691GInfsX50

28/14 Nonsense /Small deletion

22

2. c.2435delC/c.1668delC*

p.Pro812ArgfsX3/p.His556GInfsX21

18/13 Small deletion/Small deletion

18

3. c.2435delC/c.2435delC

p.Pro812ArgfsX3/p.Pro812ArgfsX31

18/18 Small deletion/Small deletion

20

4. c.4975C>T/c.4975C>T

p.R1659X/p.R1659X

28/28 Nonsense /Nonsense

19

5. c.2435delC/c.2435delC

p.Pro812ArgfsX3/p.Pro812ArgfsX31

18/18 Small deletion/Small deletion

25

6. c.4975C>T/c.4975C>T

p.R1659X/p.R1659X

28/28 Nonsense /Nonsense

13

7. c.4975C>T/c.4975C>T

p.R1659X/p.R1659X

28/28 Nonsense /Nonsense

23

8. c.4975C>T/c.874+2T>C*

p.R1659X/-

28/intron 7

Nonsense/Splice site

27

9. c.2435delC/c.2435delC

p.Pro812ArgfsX3/p.Pro812ArgfsX31

18/18 Small deletion/Small deletion

13

10. c.2435delC/c.2435delC

p.Pro812ArgfsX3/p.Pro812ArgfsX31

18/18 Small deletion/Small deletion

18

49

Figure 5. Multimer analysis of patient and normal plasma (NP) demonstrating absence of VWFin patients on on-demand (2, 4 and 7) and traces of factor concentrate in patients onprophylaxis (1, 3, 5, 6, 8, 9, and 10).

10.2 STUDY II

10.2.1 PATIENT CHARACTERISTICS

Of the 9 type 3 patients enrolled, at the time of the study 4 were treated on on-demand basis and 5 were on regular prophylaxis. Bleeding scores (BS, ISTHBAT) were high (≥15) in all indicating severe bleeding phenotype. Thus, alsoon-demand treated patients had high BS results due to historical bleeds relatedto surgeries, tooth extractions, deliveries, and previous episodes of epistaxisand menorrhagia (Table 6). However, currently all four on-demand treatedpatients reported milder bleeding tendencies and rarely used factorconcentrate (annual bleed rate ABR requiring factor concentrate 0 – 1).

Standard laboratory data at the time of the study sample withdrawal arepresented in Table 6. On-demand treated patients had undetectable VWF andlow FVIII levels. Patients on prophylactic treatment had their last dose infused3-7 days previously and trace amounts of factor concentrate were present.Median VWF levels (IU/dL) were low: VWF:Ag was <5 (<5 – 17), VWF:GPIbM4 (<4 – 14), and VWF:CB <3 (<3 – 8). FVIII varied between 4 – 53 IU/dL, andthe levels were surprisingly well-conserved 50 and 53 IU/dL in 2 patients.

50

The VWD genotypes of patients 1 to 8 were described in study I. Patient 9,who was not included into study I, was compound heterozygous for two nullmutations, c.2435delC and c.5312-2delA, a novel small deletion in intron 30.

No factor V Leiden, prothrombin nucleotide 20210 variant or otherthrombophilic traits compensating for the VWF defect were discovered.Antithrombin, free protein S and protein C levels were normal in all (Table 7).ADAMTS13 activity was within normal range and replacement therapy did notaffect the levels. Free and total tissue factor pathway inhibitor (TFPI) varied 5-fold (medians 11 and 104 ng/mL, compared to in-house controls of 13.4 and89.7 ng/mL, respectively), with the lowest levels recorded in patient 9. F1+2fragment levels ranged between 135 and 360 pM (reference 69 – 229 pM),indicating normal to moderately increased in vivo thrombin generation. HigherF1+2 levels were associated with older age (r=0.70, p=0.006). Factorconcentrate infusion did not modify the F1+2 levels.

Table 6. Mode of treatment, washout from factor concentrate, dose of factor concentrategiven in the study, VWF and FVIII levels and ISTH bleedings assessment tool (BAT) results. OD =on-demand, PRO = prophylaxis. For the patients whose treatment response to factorconcentrate was studied (n=6), measurements before and 30 minutes after the replacementtherapy are given.

No.Treat-ment

Washout Studytreatmentdose

VWF:AgIU/dL

VWF:GPIbMIU/dL

VWF:CBIU/dL

FVIIIIU/dL

ISTHBAT

1. OD over 14days

1000 IU VWF <5 –37

<4 –21

<3 –26

2 –12

29

2. OD over 14days

- <5 <4 <3 2 18

3. PRO 7 days 3000/7200 IUFVIII/VWF

17 –249

<4 –117

8 –202

50 –136

23

4. OD over 14days

- <5 <4 <3 2 22

5. PRO 3 days 2000/4800 IUFVIII/VWF

<5 –179

<4 –73

<3 –131

23 –99

15

6. OD over 14days

- <5 <4 <3 2 21

7. PRO 3 days 2000 IU VWF 13 –69

14 –35

5 –49

53 –57

30

8. PRO 4 days 1000/2400 IUFVIII/VWF

<5 –174

9 –66

<3 –109

9 –73

13

9. PRO 4 days 1000/2400 IUFVIII/VWF

<5 –72

4 –43

<3 –52

4 –43

19

51

Table 7. Further coagulation assessments including prothrombin fragment (F1+2), ADAMTS13,protein C (PC), protein S (PS) and total and free tissue factor pathway inhibitor (TFPI) levels.Results before and after therapy indicated in patients whose response to factor therapy wasstudied.

10.2.2 THROMBIN GENERATION (TG) IN PLATELET-POOR (PPP) AND -RICH PLASMA (PRP)

In the plasma of type 3 VWD patients, both peak thrombin and ETP werereduced when compared with healthy controls (Figure 6 and 7). Median peakthrombin was 16% (6 – 93%) and ETP 33% (17 – 130%) of normal. Time toinitiate TG (lag time) was also prolonged. TG was especially poor amongpatients on on-demand therapy, who had 69% lower peak thrombin than theones on prophylaxis. One outlier patient on prophylaxis (patient 9) had anormal TG curve, despite a washout of 4 days and undetectable VWF and lowFVIII:C of 4 IU/dL. This was a patient with the lowest measured TFPI levels,but no other abnormalities in coagulation parameters.

In plasma, peak thrombin showed significant correlation with the measuredFVIII level (r=0.77, p=0.02) and VWF:CB (r=0.81, p=0.016), but not withVWF:Ag or VWF:GPIbM. ETP and lag time did not show association to any TG

No.F1+2pM/L(69-229)

ADAMTS13 %(40-130)

PC %(74-141)

PS %(50-150)

AT3 %(85-125)

TotalTFPIng/mL

FreeTFPIng/mL

1. 360 –343

66.3 –56.9

150 86 117 123 25.5

2. 324 99.5 158 59 107 113.9 10.6

3. 311 –292

145.1131.5

128 90 112 110.6 26.3

4. 239 89.3 94 68 111 86.0 7.2

5. 178 –176

95.0 –97.9

91 74 111 85.1 10.9

6. 146 99.5 131 69 110 104.4 16.2

7. 230 –245

106.4 –114.4

208 80 122 115.9 19.2

8. 225 –226

115.9 –128.5

150 60 110 96 10.8

9. 135 –132

98.1 –106.2

115 112 94 55.3 4.3

52

parameter. None of the TG variables correlated with the ADAMTS13, TFPI,protein C, protein S or antithrombin levels.

In contrast with the measurements in plasma, in PRP of type 3 VWDpatients TG was not significantly lower in comparison with normal controls(Figure 6 and 7). Median peak thrombin was 77% (20 – 140%) and ETP 106%(50 – 150%) of normal. Lag time was normal. Similarly as also demonstrated inplasma, peak thrombin was 30% lower in on-demand compared withprophylaxis treated patients. The measured TG variables in PRP failed tocorrelate with VWF, FVIII or any of the other coagulation parameters.

Thro

mbi

n(n

M)

Thro

mbi

n(n

M)

Figure 6. Thrombin generation (TG) curves of type 3 patients and normal controls in platelet-poor plasma (PPP)(A) and platelet-rich plasma (PRP)(B).

10.2.3 THROMBIN GENERATION AFTER ADMINISTRATION OFREPLACEMENT THERAPY

We measured the impact of replacement therapy at 30 minutes after the factorconcentrate infusion in 5 patients on prophylaxis and in 1 patient on on-demand therapy. TG was measured after a typical prophylactic or on-demanddose (VWF 16-95 IU/kg, Table 6). Replacement therapy restored VWF levels tonormal (Table 6 and Figure 8). FVIII levels were normalized after VWF/FVIIIconcentrate, and showed minor increase after pure VWF infusion, as expected.

In PPP, the administration of VWF/FVIII concentrate in 4 patients restoredpeak thrombin and ETP (Figure 8). Administration of pure VWF in 2 patientsincreased TG in PPP less.

The response to VWF/FVIII concentrate measured by TG in PRP was lowerand inconsistent. Also in the 2 patients receiving pure VWF, the responsemeasured in PRP was variable: TG increased in one and was unaffected in theother.

53

Figure 7. Thrombin generation (TG) in type 3 VWD (VWD3). In PPP, all TG variables wereabnormal. The median peak thrombin (vs controls) was 19 nM (range 7-109 nM) vs 120 nM(50-190 nM), ETP 330 nM×min (180-1300 nM×min) vs 1000 nM×min (560-1600 nM×min) andlag time 8.0 min (3.0-16 min) vs 5.2 min (4.3-9.9 min). In contrast, PRP there was no significantdifference between VWD3 and controls. Peak thrombin (vs. controls) was 64 nM (15-120 nM)vs. 83 nM (43-130 nM), ETP 1400 nM×min (700-2000 nM×min) vs. 1300 nM×min (840-2000nM×min) and lag time 9.1 min (5.3-24 min) vs. 10 min (5.6-13 min). Median and interquartilerange are indicated by the lines.

54

Figure 8. Impact of replacement therapy on VWF and FVIII (a) and TG (b and c) at 30 min afterinfusion in 6 patients. Four were on FVIII/VWF therapy (black lines) and 2 on pure VWFconcentrate (grey lines). In PPP, the administration of VWF/FVIII normalized peak thrombinand ETP, whereas the administration of VWF increased TG only modestly (recovery less than50% of normal). In PRP TG increased modestly or was unaffected by VWF/FVIII or VWFinfusion.

10.2.4 IMPACT OF FVIII AND PROTEIN S ON TG IN PPP

To investigate the contribution of FVIII to the TG in PPP, we assessed theimpact of addition and inhibition of FVIII by in vitro spiking studies. SpikingPPP with FVIII (1 IU/mL) increased peak thrombin in all patient samplesregardless of the baseline FVIII levels (Figure 9). However, the TG measuredafter the addition FVIII was lower in type 3 VWD patients compared to thecontrols (normal plasma or FVIII–depleted plasma).

Conversely, inhibiting FVIII with the VK 34 antibody reduced the peakthrombin to <10 nM (with the exception of patient 9, where peak thrombindecreased by 3-fold to 52 nM). As a negative control, VK 34 had no impact onthe peak thrombin in the FVIII-depleted plasma.

The contribution of protein S on TG in PPP was investigated by addition ofPS 13, a protein S inhibitor. This enhanced peak thrombin in all type 3 VWDsamples, especially in on-demand treated patients (Figure 9). The increase ofTG capacity after blocking protein S was greater type 3 VWD than in normalplasma, but less than in FVIII-depleted plasma.

55

Figure 9. Peak thrombin after addition and inhibition of FVIII (A) and after inhibition of proteinS (B) in patients 1 to 9. After FVIII addition, peak thrombin increased from a median 12 to 101nM. Inhibition of FVIII (VK 34) resulted in a decrease of TG from median 12 to 5 nM. Inhibitionof protein S (PS13) increased TG from median 12 to 39 nM.

10.2.5 ROTEM

ROTEM analysis was carried out for all. All ROTEM results by INTEM,EXTEM and FIBTEM activation were normal and unaffected by factorconcentrate infusion at 30 minutes (Table 8).

Table 5. Thromboelastometry (ROTEM) results in all (n=9) and before and after treatment(n=4). No significant change was detected between pre and post samples. Reference rangesprovided by the manufacturer given in the top column.

EXTEMCT38-79 s

EXTEMCFT 34-159 s

EXTEMα 63-83°

EXTEMMCF50-72mm

INTEMCT 100-240 s

INTEMCFT 30-110 s

INTEMα70-83°

INTEMMCF50-72mm

FIBTEMMCF 9-25 mm

Alln=9

4944-52

6951-106

7971-85

6862-72

186162-207

6541-109

7770-81

7061-73

2112-27

PREn=6

4947-52

6951-106

7871-82

6962-72

184172-207

6841-109

7670-81

6961-73

2012-27

POSTn=6

4843-62

8451-136

7864-83

6862-72

190160-235

7765-81

7570-80

6763-73

1913-26

56

10.2.6 FLOW CYTOMETRIC ANALYSIS OF PLATELETS

In the flow cytometric analysis of platelets, individual type 3 VWD patientsdemonstrated higher platelet activation status when compared with normalcontrols, but this finding was not apparent in all. Increased platelet activationby P-selectin expression was detected in 67% of the type 3 VWD patients (4 on-demand and 2 prophylaxis treated) and also 56% (3 on-demand and 2prophylaxis treated) of the patients demonstrated stronger annexin V bindingthan normal conrols (Figure 10). This was demonstrated both bymeasurements of percentage of positive platelets expressing P-selectin/annexinV binding and as increased median fluorescence intensity. However, the overalldifference between type 3 VWD patients and normal controls did not reachstatistical significance. Platelet activation status did not associate with the TGresults, or other coagulation factors.

Figure 10. Platelet reactivity by P-selectin expression under ADP stimulation as expression ofpositive platelets and median fluorescence intensity in panels A and B, and correspondinglyannexin V-binding upon stimulation with CRP-XL in panels C and D.

57

10.3 STUDY III

10.3.1 BACKGROUND

This case study reports the management of an acute gastrointestinal bleeding(GI) episode in a 49-year old patient with anti-VWF alloantibodies (Patient 7,Table 5).

Previously, the patient had been treated for bleeds and surgeries withcryoprecipitate and plasma derived factor concentrates, both VWF/FVIII(Haemate®) and pure FVIII (Amofil®, manufactured with the method M,Sanquin). The antibodies were first demonstrated following a delivery at theage of 35, when response to cryoprecipitate therapy disappeared 8 days aftergiving birth, suggesting a neutralizing antibody. During this episode, thepatient also had a possible allergic reaction (rash and fever) followingreplacement therapy. However, there were no hemostatic problems. A seconddelivery two years later was handled with Haemate®, and again the responseto the replacement therapy was lost at 6 days post delivery.

At the age of 46, the patient had a prolonged GI bleeding episode followingpolypectomy in a local hospital. An adenoma with mild dysplasia was removed.The patient was hospitalized for 3 weeks due to bleeding, receiving extendedtherapy with VWF/FVIII (Haemate®) and pdFVIII (Amofil®), and several redblood cell (RBC) transfusions. Subsequently, the patient was referred to ourcomprehensive care center.

Initial laboratory results at our center revealed absent VWF andconsequently low FVIII (2 IU/dL). The patient had low spontaneous bleedingrate, and for the next 3 years, had only minor bleeding problems managed withtranexamic acid. DNA analysis showed homozygous c.4975C>T (p.R1659X)mutation in exon 28 of VWF.

10.3.2 THE CURRENT BLEEDING EPISODE

At the age of 49, a control colonoscopy was planned in our hospital. Theprocedure was decided to be managed with a different VWF/FVIII concentrateWilate® (Octapharma) which contained a higher ratio of FVIII (VWF/FVIII1:1) compared to Haemate®.The patient received a dose of 40 IU/kg of Wilate® prior to the colonoscopy(day 1, Figure 11). The procedure was initially uneventful with good localhemostasis, and 6 minor polyps (later confirmed benign hyperplastic), wereresected. Wilate® (20 IU/kg) replacement therapy was continued twicedailyfor 2 post-operative days with good response in the laboratory.

58

Figure 11. Timelines demonstrating the treatment of our patient with anti-VWF alloantibodies.The patient underwent an elective colonoscopy with biopsy on day 1, leading to severegastrointestinal (GI) bleeding complication from day 3 to 11. The bleed resolved and theresponse to VWF treatment was achieved following intravenous immunoglobulin (IVIG)treatment (arrows). pdVWF/FVIII1 = Wilate®, pdVWF/FVIII2 = Haemate®.

On day 3 the patient was discharged from the hospital without signs ofbleeding, alloantibodies or other complications. However, on her way home, anacute lower GI bleed started, and she returned to the hospital. The Wilate®therapy was restarted with a bolus of 40 IU/kg, followed by 20 IU/kg every 8hours. During hospital days 3 to 6, VWF and FVIII levels were maintained withthe therapy, but the GI bleeding persisted. The patient received also daily RBCand platelet transfusions.

On day 7, the response to replacement therapy suddenly disappeared in thelaboratory (Figure 11). At the same time, the GI bleeding got worse, with severemelena 10 times and 1700 mL daily. This was considered to represent thereappearance of a VWF inhibitor, and the patient was transferred to theintensive care unit. Therapy with a bypassing agent rFVIIa (NovoSeven®,NovoNordisk) at 9.6 mg (109 μg/kg) every 2 hours was initiated. A controlcolonoscopy on day 7 discovered bleeding from multiple biopsy sites, and in adifficult procedure with poor local hemostasis and visibility, these were treatedwith 7 clips. With the uncontrolled bleed continuing, Haemate® and pdFVIII(Amofil®) were added to the therapy on day 8, administered 4-6 times dailyalternating with the rFVIIa therapy. Despite transient VWF and FVII response

0 1 2 3 4 5 6 7 8 9 1011121314151617181920210

255075

100125150175200225250275300

IU/d

LVWF:RCoFVIII

VWF:Ag

colonoscopy: x x x

GI bleed:

epistaxis:pdVWF/FVIII1

pdVWF/FVIII2

pdFVIII

rFVIIa

IVIGinfusion

DAY:

59

on day 9, severe GI bleeding persisted from day 7 to 10. RBC units had to beadministered at 2-6 units per day to maintain hemoglobin levels combinedwith platelet transfusions and fresh frozen plasma (Octaplas®, Octapharma). Anew control colonoscopy on day 11 found active bleeding at a single biopsy site,again treated with clips.

Without detectable treatment response, to support the local hemostasisachieved in the colonoscopy on day 11, intravenous immunoglobulin (IVIG,Nanogam®, Sanquin) was administered at 1 g/kg. Now, VWF and FVIII levelsincreased immediately a few hours after the infusion, and rFVIIa therapy wasstopped. Subsequently, the bleeding finally ceased on day 11, and the patientwas transferred back to the regular ward. IVIG was continued every 24 hoursfor a total of four doses, together with Haemate® and pdFVIII therapy every 8-12 hours. The effect of IVIG infusions on VWF and FVIII levels increasedduring the infusions (VWF:Ag reaching up to 300 and VWF:RCo 200 IU/dL),and lasted for 2 days after the fourth infusion.

The GI bleed had resolved, but the hospitalization was prolonged due toepistaxis starting on day 13. This was treated by posterior tamponation. Someepistaxis recurred after the removal of the tamponade on day 17 and 2 moreIVIG-infusions were administered on day 17 and 18. The impact on VWF andFVIII levels was again immediate and enough to control the bleed. The patientwas finally discharged from the hospital after 20 days of treatment, continuingonly tranexamic acid at home.

Unfortunately, at the time of the bleeding episode, no antibodymeasurement was available, but laboratory results clearly indicated thepresence of a neutralizing anti-VWF alloantibody. Since then, an anti-VWF:CBtiter of 5 BU/mL but no anti-VWF IgG or IgM antibodies were measured aspart of the International Registries and Inhibitor Prospective (IPS3-WINTERS-IPS3) study, demonstrating an inhibitor.

Since the bleeding episode, the patient has experienced repeated epistaxis,usually managed with tranexamic acid. Once due to epistaxis, the patientreceived Haemate® in a local hospital and developed shortness of breath,dizziness and itchiness in the throat post infusion, resolving afterdiscontinuation of the treatment. She has not experienced further GI bleeds.

10.4 STUDY IV

10.4.1 RE-EVALUATION OF PAST VWD DIAGNOSES

Historical VWD diagnoses given a median of 20 years (range 1-54) ago were re-evaluated by a comprehensive clinical and repeated laboratory assessment inthe 83 subjects included to the study. The median age of the patients enteringthe study was 44 years (16-76). The original diagnostic VWF:RCo and FVIIIresults were available in all type 1 patients, but only in 16/32 of type 2 and 3/13

60

of type 3 patients. The initial VWF:Ag levels were unknown in the majority ofthe cases.

According to current guidelines and laboratory data, we could confirm adiagnosis of VWD in 52/83 (63%) of the patients. Specific VWD types wereconfirmed in all 13/13 (100%) type 3 patients, in 29/32 (90%) type 2 patients,but only in 7/38 (18%) type 1 patients. In addition, in 3 patients with historicaltype 1 VWD, the diagnosis was changed to type 2 VWD. Ten patients withprevious type 1 diagnosis currently fulfilled criteria for low VWF instead of type1.

Including all patients with historical VWD, the Vicenza-based BAT score atre-evaluation showed a negative correlation with VWF:RCo (r = −.43), VWF:Ag(r = −.51), VWF:CB (r = −.54), FVIII (r = −.44), RIPA 0.6 mg/mL (r = −.34),and RIPA 0.8 mg/mL (r = −.54) and a positive correlation with PFA CT/EPI(r = .45) and CT/ADP (r = .46) (in all p≤.001).

10.4.2 PFA-100

In the PFA, CTs with both ADP and EPI cartridges were prolonged in allconfirmed VWD cases, confirming a primary hemostatic defect (Figure 12). Thedegree of reduction was dependent on the VWD type, with the type 3 and type2 patients showing markedly prolonged closure times. All type 1 patients alsohad abnormal PFA-100, whereas in the low VWF group, both CT/EPI andCT/ADP CTs were prolonged in 6/10 cases, CT/ADP marginally in 2/10, andboth normal in 2/10.

Figure 12. PFA-100 results in different patient groups, illustrating abnormal results in all confirmed VWD patients, irrespective of the VWD type by both epinephrine (CT/EPI) and adenosine diphosphate (CT/ADP) cartridges compared to normal controls.

norm

alco

ntrol

lowVW

F

VWD

1

VWD

2

VWD

350

100

150

200

250

300

CT

C/E

PI(s

)

A

norm

alco

ntrol

lowVW

F

VWD

1

VWD

2

VWD

350

100

150

200

250

300

CT

C/A

DP

(s)

B

61

10.4.3 RISTOCETIN INDUCED PLATELET AGGREGATION (RIPA)

Whole blood RIPA at 0.6 mg/mL ristocetin was reduced in all VWD cases inMultiplate® with sensitivity of 100%, with the exception of subtype 2B (Figure13). The sensitivity for VWD (excluding 2B) was 89% at the higher, standardristocetin concentration of 0.8 mg/m, where normal results were obtained in 6VWD patients (4 type 1 and 2 type 2A). RIPA was variable in the 10 patientswith low VWF and overlapped with that of healthy controls. Four patients haddiminished RIPA at ristocetin 0.6 mg/mL, but normal RIPA at 0.8 mg/mLristocetin concentrations.

Increased RIPA by Multiplate at low ristocetin (0.2 mg/mL) identifiedsubtype 2B in 7 of the 8 patients, distinct from other type 2 subtypes (2A, 2M,2N) and normal controls. RIPA by Multiplate at low ristocetin was normal inone 2B patient with thrombocytopenia, who also needed a higher minimumaggregating dose of ristocetin in LTA (0.8 mg/mL compared to standard0.4 mg/mL). 2B subtype was confirmed in all patients by genetic analysis.

Figure 13. Ristocetin-induced platelet aggregation (RIPA) in whole blood by Multiplate in lowVWF, different VWD types and subtypes, and normal controls at 0.2, 0.6 and 0.8 mg/mLristocetin. The low concentration identified subtype 2B. The higher concentrations demonstrateprogressively lowered RIPA according to the type of disease, with absent RIPA in type 3 VWD.

normal

contr

ol

lowVW

F

VWD1

VWD

2A

VWD

2M

VWD

2B 2N

VWD3

0

50

100

150

200

risto

cetin

0.8

mg/

mL,

AUC

C

norm

alco

ntrol

lowVW

F

VWD1

VWD

2A

VWD

2M

VWD

2B

VWD

2N

VWD3

0

50

100

150

200

risto

cetin

0.6

mg/

mL,

AUC

B

norm

alco

ntrol

VWD2A

VWD2M

VWD2B

0

10

20

30

40

50

risto

cetin

0.2

mg/

mL,

AUC

A

62

10.4.4 PATIENTS WITH HISTORICAL TYPE 1 VWD

Type 1 was confirmed in 7 of the 38 patients with previous type 1 diagnosis(Figure 14). Three out of these 38 cases had type 2 VWD instead. All confirmedVWD cases had original diagnostic levels of VWF:RCo <35 IU/dL. Alldemonstrated only modest increase of VWF:RCo and FVIII (median increase of5 IU/dL and 12 IU/dL) from the diagnostics levels, in a median 17-years (1-34)follow-up since diagnosis.

Prior type 1 VWD was reclassified as low VWF in 10 of the 38 cases basedon VWF of 35-50 IU/dL, and a bleeding tendency. Nine out of these 10 patientshad abnormal bleedinf tendynce by BAT, but the median was lower than intype 1 (6 vs 11, P = 0.003). Six patients had VWF:RCo levels of 35-50 IU/dLalso at the original diagnosis, whereas in 4 cases, the levels had increased fromthe original <35 IU/dL.

We were unable to confirm VWF defect in 18 of the 38 patients previouslydiagnosed with type 1, as the VWF levels were repeatedly normal. At originaldiagnosis, 10/18 cases had VWF:RCo levels higher than the current diagnosticcutoff of 35 IU/dL, and for 8/18, VWF:RCo levels were originally <35 IU/dL. Atthe re-evaluation, 8/18 subjects were asymptomatic (normal BS) and hadnormal laboratory findings, and they were determined not to have a bleedingdisorder. Five of 18 had abnormal BS and altered platelet function asdetermined by Multiplate analysis (3 ADP and 2 ASPI) suggesting a plateletfunction defect. Four of them had also abnormal PFA analysis. Five patientswith abnormal BS without laboratory abnormalities were concluded to have ableeding disorder of unknown origin. No other clotting factor deficiencies werediscovered.

Figure 14. A flow chart illustrating the re-evaluation of the previous type 1 VWD diagnoses.

previous VWD type 1n = 38

normal VWFn = 18 (48%)

low VWFn = 10 (26%)

VWDn = 10 (26%)

abnormal BS,no laboratory

defectsn = 5

defec veplateletfunc on

n = 5

no bleedingdisorder

n = 8

type 2An = 1

type 1n = 7

type 2Mn = 1

type 2Nn = 1

63

11 DISCUSSION

11.1 THE INHERITANCE OF TYPE 3 VWD IN FINLAND

The genetic background of 10 Finnish type 3 VWD patients is reported in studyI. Type 3 VWD is a rare disease, and the reported prevalence has ranged from0.1 to 5.3 cases per million of the population in different nations. The lowestprevalence has been reported in Southern Europe and the highest in Arabiccountries and Scandinavia (Eikenboom et al., 2001). In Finland, the prevalenceof type 3 was reported to be relatively high 4 per million in 1999, with 18diagnoses (Kekomäki et al., 1999). Since then, according to our knowledgealtogether 27 patients have been diagnosed with type 3 VWD.

In our study, although our 10 patients were from different families, therewas only little genetic heterogeneity. Two main mutations, c.2435delC andc.4975C>T were detected in 85% of the analyzed alleles. In most studies, wheregenetic analysis has been performed nationwide or in large cohorts, nocommon founder mutations have been discovered. However, c.2435delC hasbeen commonly reported at high frequency in countries surrounding the BalticSea and only rarely in other populations. The mutation constitutes 50% of themutations detected in type 3 in Sweden, 20% in Germany, 75% in Poland andalso 12% in Hungary (Schneppenheim et al., 1994, Zhang et al., 1994, Gazda etal., 1997, Mohl et al., 2008).

The other main mutation discovered in 40% of the alleles in our patients,c.4975C>T, has been reported in individual patients from Japan, Iran, Indiaand Turkey, suggesting that the mutation is located on a ´hotspot´, where therisk for mutations is elevated (Hagiwara et al., 1996, Baronciani et al., 2000,Berber et al., 2013).

Since our publication, according to our knowledge 6 patients with type 3from Finland have undergone genetic analysis. Among these 6 patients,c.2435delC was detected in 33% (4/12) and c.4975C>T in 50% (6/12) of thealleles, in agreement with our results.

Our findings are interesting also considering the historical aspects of VWD.Erik von Willebrand, a Finnish internist working in the Helsinki DiaconiaHospital, was the first to describe VWD. The original family S from Föglö in theislands of Åland remains one of the largest families with VWD. In the currentstudy, we have confirmed that the Åland Island mutation c.2435delC is alsocommon among type 3 patients in mainland Finland. This deletion leads tosevere bleeding tendency, and as described by Erik von Willebrand, alsocarriers of the mutation can be symptomatic. With this background, the geneticanalysis of type 1 VWD in Finland would be of interest considering therelatively high incidence of type 3 in Finland.

64

11.2 CLINICAL HETEROGENEITY IN TYPE 3 VWD

There is a great heterogeneity in the clinical expression of VWD, including themost severe form of the disease, type 3. In contrast to hemophilia, whereprophylaxis with factor treatment is the gold standard of treatment, in VWDthere is no clear evidence on which patients should be treated prophylactically(Abshire et al., 2015). While some patients present with severe bleedingtendency and suffer from joint bleeds leading to hemophilia-like arthropathy,other patients do not bleed spontaneously and do not need prophylaxis.

In the Canadian type 3 VWD study, it was reported that patients whocarried mutations in the propeptide region of VWF had higher BAT score(Bowman et al., 2013). However, this result remains to be confirmed in furthercohorts. FVIII level has been considered to be a major determinant of bleedingphenotype in type 3, but this was not confirmed among our patients, who allhad low native FVIII levels of less than 5 IU/dL.

The likelihood of joint bleeds may depend factors such as the physicalactivity of patients and occurrence of traumas as well as the hemostatic factors.An interesting comparison would be moderate hemophilia A, in which FVIIIlevels are similar.

In our study, we have examined several features that could contribute to thebleeding risk in type 3 VWD. In addition to VWF and FVIII levels, we assessedthrombophilia screen, TG, ROTEM and platelet activation status by flowcytometry. In patients with low spontaneous bleeding rate and low ABR, we didnot discover thrombophilic traits or other coagulation abnormalities that couldcompensate for the VWF/FVIII defect. Our patient number is too small tomake any definite conclusions on whether these factors may have role in theclinical heterogeneity of severe VWD. However, the lack of these knownthrombophilic traits among our patients with variable bleeding phenotypessuggests that there are possibly other, unexplored mechanisms involved. Thesepatients with milder clinical phenotype had reduced TG results in both plasmaand less evidently in the presence of platelets. However, procoagulantproperties of platelets could compensate VWF-associated defect, both byinherited and acquired mechanisms. The contribution of platelets on theclinical phenotypes in VWD calls upon further studies.

It seems unlikely that a single laboratory examination would be able toexplain the diversity of bleeding symptoms in type 3, and larger multicentermultinational prospective studies are required to increase understanding of theclinical course of type 3 VWD. Such study is currently ongoing under thesupervision of Dr. Augusto B. Federici, with especial focus on patients withinhibitors. The study is entitled 3WINTERS-IPS (Type 3 von WillebrandInternational Registries Inhibitor Prospective Study) and the aim is to assessthe frequency, types and risk factors for bleeding episodes and the efficacy andsafety of VWF concentrates. Also patients from our clinic are included to this

65

study. Hopefully, this study will provide guidelines for better follow-up andtreatment of these patients.

10.3 GLOBAL HEMOSTATIC ASSAYS: TG AND ROTEM

Previous study on TG in VWD indicated that VWD is associated to decreasedTG capacity both in plasma and in the presence of platelets, and that the defectin TG is mainly attributed to the low FVIII levels (Rugeri et al., 2007). Theprevious study, however, only included 2 type 3 patients. Our study included 9type 3 patients with variable bleeding phenotypes. In line with the previousstudy, we detected reduced TG in plasma. Spiking studies with adding andinhibiting FVIII also supported the conclusion that the reduced TG in plasma ismainly dependent on the low FVIII level.

However, other regulatory mechanisms may contribute to TG in type 3VWD beyond the control of FVIII. Interestingly, we observed that protein Scauses further dampening of TG, especially in patients with the lowest TG. LowTFPI may also associate with higher TG, as demonstrated in one outlier patientwith normal TG despite low FVIII. It is remarkable that both these regulatorsreside in platelets. The microenvironment of thrombin generation on activatedplatelets may impact the stability of the forming clot.

Our studies of TG in the presence of platelets in PRP contrast the resultsfrom the earlier study. We found TG is not significantly reduced in PRP of type3, with only one patient presenting with less than 50% of normal TG. Thissuggests that platelets assist the poor TG in plasma, although the specificmechanisms behind this remains unclear. Even though we measured higherthan normal platelet activation status in individual patients, this was notdirectly associated to TG results in this small series of patients.

TEG and ROTEM, measuring clot formation and lysis in whole blood, areincreasingly popular in emergency and perioperative setting in the clinics. Insupport of an earlier study by Schmidt et al, we have demonstrated thatstandard ROTEM is insensitive to severe VWD (Schmidt et al., 2017). Thecurrent standard ROTEM reagents (EXTEM, INTEM, FIBTEM) fail to detectVWD, and modification of reagents would be necessary to identify VWD byROTEM. However, screening for VWD by TEG appeared promising in thestudy by Schmidt et al. Whether the TEG results can predict the clinicalbleeding phenotype remains unknown.

66

11.4 WHOLE BLOOD PLATELET FUNCTION IN VWD

Currently, analysis of platelet function is not recommended routinely in VWDby the guidelines. This far, PFA and platelet aggregation have been studiedmost extensively in VWD.

PFA-100, a screening tool for primary hemostasis, has been reported tohave a sensitivity of approximately 90% for VWD in a 700 hundred patient caseseries (Favaloro et al., 2008). The sensitivity is higher for the more severeforms of the disease, with 100% sensitivity for type 3 and 2A. Our resultssupport these data, with 100% sensitivity for all cases of confirmed VWD. Theassay is easy-to-use and fast, requiring only 5 minutes per test, enabling rapidscreening for VWD. However, our results also suggest that the assay is lesssensitive to identify patients with low VWF, and the clinical consequences ofthis are not straightforward.

For platelet aggregation studies, the main application in VWD has been toidentify patients with subtype 2B, who show increased RIPA at low ristocetinconcentration in PRP by Born aggregometry. Reduced RIPA in patients withVWD (subtypes other than 2B) was already reported by Mannucci in 1977, butthe method failed to identify all cases (Mannucci 1977). Born aggregometryremains technically a challenging and time-consuming. Multiplate, a wholeblood aggregometer, is a faster, modern technique. Currently, there are onlylimited numbers of studies on VWD using RIPA by Multiplate. In the firststudy published, Multiplate was found to agree with the results obtained fromBorn aggregometry in 30 VWD patients, and increased RIPA wasdemonstrated in subtype 2B of 3 subjects (Valarche et al., 2011). In the largeststudy to date, including 100 VWD patients, an association between RIPA andBAT score was detected in type 1 VWD (Schmidt et al., 2017). As Multiplate is afast method and easy to perform and does not require centrifugation steps incontrast to Born aggregometry, the authors suggested that Multiplate could beused as a test to exclude VWD.

In our study, we have demonstrated that RIPA by Multiplate isprogressively reduced in VWD according to the subtype and VWD severity. Wealso found an overall correlation of BAT (Vizenca based) with RIPA in ourcohort of patients carrying a historical VWD diagnosis. Our in-housemodification applying 0.6 mg/ mL ristocetin as the final concentration was100% sensitive to detect type 1 VWD, whereas standard concentration 0.8mg/mL failed to identify all type 1 and 2 cases. Additional studies are requiredto determine the optimal ristocetin concentration in Multiplate. Subtype 2Brequires a low concentration of ristocetin for detection, as we have illustratedhere also in whole blood. Further studies are needed to assess correlationsbetween clinical features and findings in Multiplate and to confirm whether itadds value to the traditional laboratory analysis performed in VWD.

67

11.5 PLATELET PROCOAGULANT ACTIVITY IN VWD

Beyond platelet function and aggregation studies in VWD, there is also someadditional data available on platelet contribution to the clinical phenotypes ofVWD. Genetic differences in the expression of the integrin subunit alpha2responsible for platelet adhesion to collagen have been demonstrated to modifythe bleeding phenotype of type 1 and type 2 (Kunicki et al., 2004, Kunicki et al.,2006). One study demonstrated that platelet enzymes that hydrolyze adeninenucleotides are lowered in VWD (Santos 2010). Also, mutations in platelet G-protein coupled receptor genes, such as P2RY12, P2RY1, F2R, F2RL3, TBXA2Rand PTGIR have been shown to affect bleeding symptoms in type 1 VWDpatients (Daly et al., 2009, Stockley et al., 2016).

We analyzed platelet activation status in flow cytometry by P-selectinexpression under ADP stimulation and by annexin V binding via theprocoagulant GPVI agonist under CRP activation in type 3 VWD. Compared tonormal controls, six patients (67%) were high responders in their platelet P-selectin, and five (56%) in phosphatidylserine expression. We could notdetermine a direct association between the bleeding phenotype and plateletactivation status, as both on-demand and prophylactically treated patients hadhigher than normal results. Also, we could not detect a direct associationbetween the TG test results in plasma or in the presence of platelets. Based onthese results, it is not possible to conclude whether enhanced platelet activationstatus can modify the bleeding phenotype in type 3 VWD and compensate tothe VWF defect. This requires further studies in larger patient cohorts.

11.6 TREATMENT OF PATIENTS WITH ANTI-VWFALLOANTIBODIES

Currently, recommendations on treatment of VWD patients who develop anti-VWF antibodies are based on case reports and expert opinions. It is not knownwhich of the patients with alloantibodies develop severe life-threateningreactions, and a review article recommended against using VWF containingproducts among these patients (James et al., 2013). There are reports of usingrecombinant FVIII and bypassing agent recombinant FVIIa to treat bleeds andmanage surgery among these patients. However, alternative treatmentstrategies also have to be considered, and VWF-containing products are alsoused (Jenkins et al., 2016).

There are reports of immunnotolerance induction therapy in pediatricpatients with type 3 VWD and alloantibodies (Pertangou et al., 2012, Berntorpet al., 2018). IVIG, rituximab and mycophenolate mofetil have been used asadjuvant therapies in these cases. However, according to our knowledge, ourcase report describes for the first time, how IVIG therapy preceding

68

VWF/FVIII infusion helped to resolve an acute bleed in a type 3 VWD patientwith anti-VWF antibodies. IVIG therapy may enable administration of VWFcontaining products, at least in some patients. Similar expriences have beenreported in AVWS (Federici, 2005). Possible mechanisms of action for IVIGinclude elimination of circulating immune complexes, inhibitory effect on theantibody producing cells or preventing the clearance of inhibitor-VWF complexby blocking Fc receptors, but other mechanism may also be involved, as inother indications in which IVIG is used.

The use of recombinant FVIII alone at high doses of continuous infusioncorrects FVIII levels in type 3 VWD, though with a short half-life. As expertopinions suggest that the haemostatic control in VWD is more dependent onFVIII than VWF activity levels, treatment with recombinant FVIII appears tobe a possible future strategy for treating bleeds in our patient as well. Thehemostatic effect in the context of gastrointestinal bleeds needs to beconfirmed.

11.7 TYPE 1 VWD AND ´LOW VWF´

In our study IV including 38 patients with historical type 1 diagnosis, only 7have confirmed to have type 1 and 3 type 2 VWD. 10 patients had borderlineVWF levels indicating a diagnosis of low VWF instead of VWD. In 18 patients,we measured repeatedly normal VWF levels. While in 5 patients we foundabnormal platelet function, no laboratory abnormalities were detected in 13subjects, out of whom 5 were symptomatic bleeders. Our study demonstratespatients who carry a historical VWD diagnosis are a heterogeneous group, andtheir bleeding symptoms are not always associated to reduced VWF levels.Other defects of primary hemostasis should also be considered. This has beendemonstrated in other cohorts as well (Daly et al., 2009).

One key issue in evaluating historical type 1 VWD diagnoses is the difficultyto reliably assess and report mild bleeding symptoms. This becomesincreasingly challenging in patients who carry the diagnosis of a bleedingdisorder, as the reporting of bleeding symptoms is subjective. In a recent meta-analysis, the diagnostic odds ratio of a BAT for a bleeding disorder appeared tobe low to moderate (Moenen et al., 2018).

As our study was underway, the understanding of pathophysiology of lowVWF, detected in 10 of our patients, has increased. The low VWF Ireland studyincluded 126 patients with VWF levels of 30-50 IU/dL. Majority of thesepatients were symptomatic bleeders, not explained by other coagulationdefects. Increased VWF clearance was discovered only in 6% of the patients,and majority appeared to have decreased synthesis and/or constitutivesecretion of VWF instead. Desmopressin administration corrected the VWFlevels. Further studies are needed to optimize the follow-up and treatment ofthis patient group.

69

Another unresolved issue involves VWD patients who correct their VWFlevels with aging, and their treatment when facing hemostatic challenges (Rydzet al., 2015). Our study also included patients who by previous laboratoryanalysis appeared to have type 1 VWD, but have since corrected their levels. Itis under debate whether the normalization of VWF levels leads to reduction ofbleeding symptoms. In a 1-year follow-up study, elderly patients with higherVWF levels still reported similar bleeding rates as younger patients with lowerVWF levels (Sanders et al., 2014). In contrast, a retrospective study of 39 type 1patients recently demonstrated that aging leads to reduction of bleedingsymptoms (Seaman et al., 2018). Treatment of elderly patients withdesmopressin and/or VWF concentrate is not straightforward, as it can lead tothrombotic or other vascular complications.

11.8 OVERALL STUDY STRENGTHS AND WEAKNESSES

In studies I and II, we have investigated the features of severe type 3 VWD inFinland including analysis of clinical phenotype, genotype, VWF, FVIII andother coagulation factors, thrombophilia screen, and also experimental studiesincluding TG test, ROTEM and platelet activation in flow cytometry. Thestrength of our study includes the comprehensive evaluation and assessment oflaboratory assays not previously investigated in this patient group. Thesestudies were carried out in patient cohorts of 10 and 9 patients, respectively,and unfortunately, not all type 3 patients from Finland could be recruited.Further studies among type 3 patients in Finland have since confirmed ourresults on the genetic background of type 3 from study I. For study II,considering the technical requirements for performing TG test, especially inPRP, larger patient cohorts will be challenging to reach considering the lowprevalence of type 3 VWD. As patients on prophylaxis had higher TG than theones treated on-demand, likely reflecting treatment remnants, a longerwashout from treatment would have been desirable to evaluate the native TGcapacity in this patient group. However, this is difficult to achieve, as longerwashouts would increase the likelihood of bleeds.

In study III, during the acute bleeding episode of the patient, an anti-VWFantibody measurement was not available to us. A detailed information aboutthe alloantibody could possibly help to predict which patients have goodresponse to IVIG treatment. Later, a study of the alloantibody implied toinhibition of VWF’s interaction with collagen.

The strength of the study IV is the comprehensive analysis of both clinicaland repeated laboratory measurements of all 83 patients. The study design is acase-series including patients referred to and evaluated at our center, and theconclusions are based on our single center experience. All included studysubjects were referred to a tertiary care unit, likely representing symptomaticbleeders. Not all patients could be given a diagnosis with the available assays,and the abnormal bleeding tendency remained of unknown origin.

70

12 SUMMARY AND CONCLUSIONS

The results from studies I-IV can be summarized with the followingconclusions:

I. Type 3 VWD is associated to absent VWF and low FVIII, but despitethe dual hemostatic defect, the patients have variable bleedingphenotypes. While some develop hemophilia-like arthropathy andrequire prophylactic treatment, others have relatively lowspontaneous bleeding tendency and require treatment with VWFonly rarely. In Finland, with a relatively high prevalence of type 3VWD, the disease is mainly caused by two mutations of VWF:c.2435delC and c.4975C>T, comprising 85% of the genetic defects.

II. TG in type 3 VWD is markedly reduced in plasma and close tonormal in the presence of platelets. It is unclear how plateletscompensate for the poor TG. FVIII is the key determinant of TG inplasma, but also natural anticoagulants (protein S and TFPI) canhave role in the overall TG capacity of these patients.

III. Alloantibody formation in type 3 VWD is a life-threateningtreatment complication. Administration of IVIG helped to resolvean acute severe GI bleeding in our case report.

IV. a) VWD is associated to reduced platelet function by PFA-100 andRIPA by Multiplate® in whole blood platelet aggregation studies.VWF-dependent platelet functions are dependent on the type ofVWD, with the most severe defect demonstrated in type 3 VWD.These tools may provide additional help in the challenginglaboratory diagnostics of VWD.

b) Patients who carry a historical type 1 VWD diagnosis are aheterogeneous group of patients with variable bleeding symptoms,and current VWF levels range from reduced to normal. Carefulanalysis of bleeding phenotype is recommended. The bleedingsymptoms of this group are not always explained by VWFdeficiency, and some have platelet function disorders or bleedingdisorders of unknown origin.

71

13 ACKNOWLEDGEMENTSThis work was carried out at the Coagulation Disorder Unit, Department ofHematology and Comprehensive Cancer Center, Helsinki University Hospitalin 2012-2018.

Financial support received from Helsinki University Governmental Grant,Blood Disease Research Foundation and Finnish Medical Society is sincerelyacknowledged.

My deepest gratitude goes to my thesis supervisor, Professor Riitta Lassila.Thank you for mentoring me and having faith in me ever since I was a secondyear medical student – I feel lucky for this. Your enthusiasm, extensiveresearch experience, and profound knowledge in the field of hemostasis are allgenuinely appreciated.

My gratitude is extended to my second thesis supervisor Timea Szanto.Your broad knowledge in von Willebrand disease was essential to this thesis.You always took the time to go through things without any sign of hurry ordistress, always maintaining good humor.

Professor Risto Kaaja and Docent Sakari Kakko, the official supervisors ofthis dissertation are warmly thanked for agreeing to take the time to review thisthesis. Comments from both of you led to significant improvements in thiswork.

I would like to thank all my coauthors. Annukka Jouppila, without yourcontribution this work would not have been possible. Not only was yourexpertise required in the laboratory work, but also your sense of humor andfriendship are most sincerely appreciated. Elina Lehtinen, thank you for yourcontribution, for always taking the time to carefully review and comment eachmanuscript multiple times and for the kind guidance in the clinic. Lotta Joutsi-Korhonen and Anne Mäkipernaa are also gratefully acknowledged.

Working with rare diseases, international collaboration is essential. I amgrateful to Professor Reinhard Schneppenheim and Florien Oyen in HamburgMedical Center for introducing me to the genetics of von Willebrand disease.Herm-Jan Brinkman from Sanquin, the Netherlands, is acknowledged for hismajor contribution to the second work in this thesis.

I am sincerely grateful to the personnel at Coagulation Disorder Unit andCoagulation Laboratory at HUSLAB Laboratory Service for their contributionto this work. In particular, I would like to thank Susanne Johansson, MarjaLemponen, Heidi Asmundela, Aino Lepäntalo, Kirsi Laasila, Pia Ettala, KlausÖsterholm, Tuukka Helin, Maria Patronen, Salla Valtonen and Sini Tuomisto.Nora Mattila and Hanna Pitkänen, thank you for the peer support over theyears.

I am grateful to both Central Finland Central Hospital and HelsinkiUniversity Hospital Internal medicine clinics for allowing me to combineclinical work with research.

72

All patients and healthy controls are gratefully acknowledged, in particularthe type 3 von Willebrand patients who generously agreed to all the bloodsamples and interviews.

I would like to thank all my friends. Especially I would like to mentionAnni, Maija and Tiisa for sharing all the laughs, cries and Friday nights duringmedical school and after. Annaleena, thank you for being such a good friendand always coming up with the positives. Jarno is acknowledged for thescientific coffee breaks in Biomedicum. Magnus, thank you for your friendshipand words of encouragement. Markus and Riikka, I am happy to have workedwith both of you in the clinic, you were kind neighbors and friends, especiallyappreciated during the time I was writing this thesis in Jyväskylä. Satu, youshowed a great example for me with your research work and were always happyto share your practical advises with me.

Finally, I would like to thank my family, my sisters Viivi and Varpu, andKristiina, and the family schnautzers, Suti, Tiuku and Fia. In particular, Iwould like to acknowledge my dad, who took care of our mom during the yearsof her long-term sickness and raised me up by himself, always supporting mein my studies. Kalle, my husband, thank you for making the coffee in themorning and being the best thing that ever happened to me.

Helsinki, August 2019Vuokko Nummi

73

14 REFERENCES

Abshire T, Cox-Gill J, Kempton CL, Leebeek FW, Carcao M, Kouides P, Donfield S, Berntorp E.(2015) Prophylaxis escalation in severe von Willebrand disease: a prospective study fromthe von Willebrand Disease Prophylaxis Network. J Thromb Haemost. 13, 1585-9.

Abshire TC, Federici AB, Alvárez MT, Bowen J, Carcao MD, Cox Gill J, Key NS, KouidesPA, Kurnik K, Lail AE, Leebeek FW, Makris M, Mannucci PM, Winikoff R, Berntorp E; VWD PN.(2013) Prophylaxis in severe forms of von Willebrand's disease: results from the vonWillebrand Disease Prophylaxis Network (VWD PN). Haemophilia 19, 76-81.

Al Dieri R, Peyvandi F, Santagostino E, Giansily M, Mannucci PM, Schved JF, Beguin S, HemkerHC. (2002) The thrombogram in rare inherited coagulation disorders: its relation to clinicalbleeding. Thromb Haemost 88, 576–582.

Ardillon L, Ternisien C, Fouassier M, Sigaud M, Lefrançois A, Pacault M, Ribeyrol O, FressinaudE, Boisseau P, Trossaërt M. (2015) Platelet function analyser (PFA-100) results and vonWillebrand factor deficiency: a 16-year 'real-world' experience. Haemophilia 21, 646-52.

Azzam HA, Goneim HR, El-Saddik AM, Azmy E, Hassan M, El-Sharawy S. (2012) The condensedMCMDM-1 VWD bleeding questionnaire as a predictor of bleeding disorders in women withunexplained menorrhagia. Blood Coagul Fibrinolysis. 23, 311-315.

Baglin T, Besser M, Cattaneo M, Dargaud Y, Gray E, Key N, Lecompte T, Luddington R, PetrosS, Siegemund T, Wolberg A. (2011) Subcommittee on Control of Anticoagulation of the SSC ofthe ISTH. Towards a recommendation for the standardization of the measurement of platelet-dependent thrombin generation. J Thromb Haemost. 9, 1859-61.

Batlle J, Pérez-Rodríguez A, Corrales I, López-Fernández MF, Rodríguez-Trillo Á, Lourés E, CidAR, Bonanad S, Cabrera N, Moret A, Parra R, Mingot-Castellano ME, Balda I, Altisent C, Pérez-Montes R, Fisac RM, Iruín G, Herrero S, Soto I, de Rueda B, Jiménez-Yuste V, Alonso N, VilariñoD, Arija O, Campos R, Paloma MJ, Bermejo N, Toll T, Mateo J, Arribalzaga K, Marco P, PalomoÁ, Sarmiento L, Iñigo B, Nieto Mdel M, Vidal R, Martínez MP, Aguinaco R, César JM, Ferreiro M,García-Frade J, Rodríguez-Huerta AM, Cuesta J, Rodríguez-González R, García-Candel F,Cornudella R, Aguilar C, Borràs N, Vidal F. (2016) Molecular and clinical profile of vonWillebrand disease in Spain (PCM-EVW-ES): Proposal for a new diagnostic paradigm. ThrombHaemost. 115, 40-50.

Batlle J, Lopez-Fernandez MF, Fraga EL, Trillo AR, Perez-Rodriguez MA. (2009) Von Willebrandfactor/factor VIII concentrates in the treatment of von Willebrand disease. Blood CoagulFibrinolysis 20, 89–100.

Beguin S, Kumar R, Keularts I, Seligsohn U, Coller BS, Hemker HC. (1999) Fibrin-dependentplatelet procoagulant activity requires GPIb receptors and von Willebrand factor. Blood 93,564–70

74

Bergamaschini L, Mannucci PM, Federici AB, Coppola R, Guzzoni S, Agostoni A. (1995)Posttransfusion anaphylactic reactions in a patient with severe von Willebrand disease: role ofcomplement and alloantibodies to von Willebrand factor. J Lab Clin Med. 125, 348-55.

Berntorp E, Onundarson P. (2005) Prevalence of von Willebrand disease in the Nordic Region.Hematology Meeting Reports (formerly Haematologica Reports). 1, 4-6.

Berntorp E, Petrini P. (2005) Long-term prophylaxis in von Willebrand disease. Blood CoagulFibrinolysis 16, S23-6.

Berntorp E, Ågren A, Aledort L, Blombäck M, Cnossen MH, Croteau SE, von Depka M, FedericiAB, Goodeve A, Goudemand J, Mannucci PM, Mourik M, Önundarson PT, Rodeghiero F, SzántóT, Windyga J. (2018) Fifth Åland Island conference on von Willebrand disease. Haemophilia 4,5-19.

Blombäck M, Antovic JP. (2010) Essential guide to blood coagulation: Blackwell Publishing Ltd.

Blombäck M. (1999) Scientific visits to the Aland Islands. Haemophilia 5, 12-18.

Borchiellini A, Fijnvandraat K, ten Cate JW, Pajkrt D, van Deventer SJ, Pasterkamp G, Meijer-Huizinga F, Zwart-Huinink L, Voorberg J, van Mourik JA. (1996) Quantitative analysis of vonWillebrand factor propeptide release in vivo: effect of experimental endotoxemia andadministration of 1-deamino-8-D-arginine vasopressin in humans. Blood 88, 2951-8.

Borel-Derlon A, Federici AB, Roussel-Robert V, Goudemand J, Lee CA, Scharrer I, Rothschild C,Berntorp E, Henriet C, Tellier Z, Bridey F, Mannucci PM. (2007) Treatment of severe vonWillebrand disease with a high-purity von Willebrand factor concentrate (Wilfactin®): aprospective study of 50 patients. J Thromb Haemost. 5, 1115–1124.

Born GVR. (1962) Aggregation of blood platelets by adenosine diphosphate and its reversal.Nature 194, 927-9.

Bouwens EA, Stavenuiter F, Mosnier LO. (2013) Mechanisms of anticoagulant andcytoprotective actions of the protein C pathway. J Thromb Haemost. 11, 242-53.

Bowman M Tuttle A, Notley C, Brown C, Tinlin S, Deforest M, Leggo J, Blanchette VS, LillicrapD, James P; Association of Hemophilia Clinic Directors of Canada. (2013) The genetics ofCanadian type 3 von Willebrand disease: further evidence for co-dominant inheritance ofmutant alleles. J Thromb Haemost. 11, 512-20.

Bowman M, Hopman WM, Rapson D, Lillicrap D, James P. (2010) The prevalence ofsymptomatic von Willebrand disease in primary care practice. J Thromb Haemost. 8, 213-6.

Brass LF, Diamond SL. (2016) Transport physics and biorheology in the setting of hemostasisand thrombosis. J Thromb Haemost. 14, 906-17.

Budde U, Metzner HJ, Muller HG. (2006) Comparative analysis and classification of vonWillebrand factor/factor VIII concentrates: impact on treatment of patients with vonWillebrand disease. Seminars in Thrombosis & Hemostasis 32, 626–635

Budde U, Schneppenheim R, Eikenboom J, Goodeve A, Will K, Drewke E, Castaman G,Rodeghiero F, Federici AB, Batlle J, Pérez A, Meyer D, Mazurier C, Goudemand J, Ingerslev J,Habart D, Vorlova Z, Holmberg L, Lethagen S, Pasi J, Hill F, Peake I. (2006) A quantitative

75

analysis of bleeding symptoms in type 1 von Willebrand disease: results from a multicenterEuropean study (MCMDM-1VWD). J Thromb Haemost. 4, 766-773.

Budde U, Schneppenheim R. (2014) Interactions of von Willebrand factorand ADAMTS13 in von Willebrand disease and thrombotic thrombocytopenic purpura.Hamostaseologie 34, 215-25

Canis K, McKinnon TA, Nowak A, Panico M, Morris HR, Laffan M, Dell A. (2010) The plasma vonWillebrand factor O-glycome comprises a surprising variety of structures including ABHantigens and disialosyl motifs. J Thromb Hameost. 8, 137-145.

Casonato A, Daidone V, Galletta E, Bertomoro A. (2017) Type 2B von Willebrand disease withor without large multimers: A distinction of the two sides of the disorder is long overdue. PLoSOne 12, e0179566.

Castaman G, Goodeve A, Eikenboom J. (2013) European Group on von Willebrand Disease.Principles of care for the diagnosis and treatment of von Willebrand disease.Haematologica 98, 667-74.

Castaman G. (2011) Treatment of von Willebrand disease with FVIII/VWF concentrates. BloodTransfusion 9, s9-13.

Castaman G, Eikenboom JC, Bertina RM, Rodeghiero F. (1999) Inconsistency of associationbetween type 1 von Willebrand disease phenotype and genotype in families identified in anepidemiological investigation. Thromb Haemost. 82, 1065-70.

Cejka J. (1982) Enzyme immunoassay for factor VIII-related antigen. Clin Chem 28, 1356-8.

Chen H, Locke D, Liu Y, Liu C, Kahn ML. (2001) The platelet receptor GPVI mediates bothadhesion and signaling responses to collagen in a receptor density-dependent fashion. J BiolChem. 277, 3011-9

Ciavarella N, Schiavoni M, Valenzano E, Mangini F, Inchingolo F. (1996) Use of recombinantfactor VIIa (NovoSeven) in the treatment of two patients with type III von Willebrand’s diseaseand an inhibitor against von Willebrand factor. Haemostasis 26, 150-4.

Colucci G, Stutz M, Rochat S, Conte T, Pavicic M, Reusser M, Giabbani E, Huynh A, ThürlemannC, Keller P, Alberio L. (2014) The effect of desmopressin on platelet function:a selective enhancement of procoagulantCOAT platelets in patients with primary platelet function defects. Blood 123, 1905-16.

Corrales I, Catarino S, Ayats J, Arteta D, Altisent C, Parra R, Vidal F. (2012) High-throughputmolecular diagnosis of von Willebrand disease by next generation sequencing methods.Haematologica 97, 1003–1007.

Cumming A, Grundy P, Keeney S, Lester W, Enayat S, Guilliatt A, Bowen D, Pasi J, Keeling D, HillF, Bolton-Maggs PH, Hay C, Collins P; UK Haemophilia Centre Doctors' Organisation. (2006) Aninvestigation of the von Willebrand factor genotype in UK patients diagnosed to have type 1von Willebrand disease. Thromb Haemost. 96, 630–41.

Daly ME, Dawood BB, Lester WA, Peake IR, Rodeghiero F, Goodeve AC, Makris M, Wilde JT,Mumford AD, Watson SP, Mundell SJ. (2009) Identification and characterization of a novel P2Y

76

12 variant in a patient diagnosed with type 1 von Willebrand disease in the EuropeanMCMDM-1VWD study. Blood 113, 4110-3.

Davie EW, Ratnoff OD. (1964 )Waterfall sequence for intrinsic blood clotting. Science 145,1310-2.

De Caterina R, Husted S, Wallentin L, Andreotti F, Arnesen H, Bachmann F, Baigent C, Huber K,Jespersen J, Kristensen SD, Lip GY, Morais J, Rasmussen LH, Siegbahn A, Verheugt FW, Weitz JI;European Society of Cardiology Working Group on Thrombosis Task Force on Anticoagulants inHeart Disease. (2013) General mechanisms of coagulation and targets of anticoagulants(Section I). Position Paper of the ESC Working Group on Thrombosis – Task Force onAnticoagulants in Heart DiseaseThromb Haemost. 109, 569-79.

de Wee EM, Sanders YV, Mauser-Bunschoten EP, van der Bom JG, Degenaar-Dujardin ME,Eikenboom J, de Goede-Bolder A, Laros-van Gorkom BA, Meijer K, Hamulyák K, Nijziel MR,Fijnvandraat K, Leebeek FW; WiN study group. (2012) Determinants of bleeding phenotype inadult patients with moderate or severe von Willebrand disease. Thromb Haemost. 108, 683-92.

Deforest M, Grabell J, Albert S, Young J, Tuttle A, Hopman WM, James PD. (2015) Generationand optimization of the self-administered bleeding assessment tool and its validation as ascreening test for von Willebrand disease. Haemophilia. 21, e384-e388.

Eikenboom J, Federici AB, Dirven RJ, Castaman G, Rodeghiero F, Budde U, Schneppenheim R,Batlle J, Canciani MT, Goudemand J, Peake I, Goodeve A and the MCMDM-1VWD Study Group.(2013) VWF propeptide and ratios between VWF, VWF propeptide, and FVIII in the characterization oftype 1 von Willebrand disease. Blood 121, 2336-9.

Elayaperumal S, Fouzia NA, Biswas A, Nair SC, Viswabandya A, George B, Abraham A,Oldenburg J, Edison ES, Srivastava A. (2018) Type-3 von Willebrand disease in India-Clinicalspectrum and molecular profile Haemophilia 24, 930-940

Farndale RW, Sixma JJ, Barnes MJ, De Groot PG. (2004) The role of collagen in thrombosis andhemostasis. J Throm Haemost. 2, 561-573

Favaloro EJ. (2008) Clinical utility of the PFA-100. Semin Thromb Hemost. 34, 709-33.

Favaloro EJ, Lloyd J, Rowell J, Baker R, Rickard K, Kershaw G, Street A, Scarff K, Barrese G,Maher D, McLachlan AJ. (2007) Comparison of the pharmacokinetics of two von Willebrandfactor concentrates (Biostate and AHF (High Purity)) in people with von Willebrand disorder. Arandomized, cross-over multicenter study. Thromb Haemost. 97, 922-930.

Favaloro EJ, Mohammed S. (2016) Evaluation of a von Willebrand factor three test panel andchemiluminescent- based assay system for identification of, and therapy monitoring in, vonWillebrand disease. Thromb Res. 141:, 202-211.

Favaloro EJ, Oliver S. (2017) Evaluation of a new commercial von Willebrand factor multimerassay. Haemophilia 23, e373-e7.

Favaloro EJ. (2007) An update on the von Willebrand factor collagen binding (VWF: CB) assay:21 years of age and beyond adolescence, but not yet a mature adult. Semin Thromb Hemost.33, 727-744.

77

Favaloro EJ. (2015) The Platelet Function Analyser (PFA)-100 and von Willebrand disease: astory well over 16 years in the making. Haemophilia 21, 642-5. d

Favaloro EJ, Lippi G. (2018) Preanalytical issues that may cause misdiagnosis in haemophiliaand von Willebrand disease. Haemophilia 24, 198-210.

Federici AB, Mannucci PM, Castaman G, Baronciani L, Bucciarelli P, Canciani MT, Pecci A.,Lenting PJ, De Groot PG. (2009) Clinical and molecular predictors of thrombocytopenia and riskof bleeding in patients with von Willebrand disease type 2B: a cohort study of 67 patients.Blood 113, 526–534.

Federici AB, Rand JH, Bucciarelli P, Budde U, van Genderen PJ, Mohri H, Meyer D, RodeghieroF, Sadler JE; Subcommittee on von Willebrand Factor. (2000) Acquired von Willebrandsyndrome: data from an international registry. Thromb Haemost. 84, 345-9.

Federici AB, Bucciarelli P, Castaman G, Mazzucconi MG, Morfini M, Rocino A, SchiavoniM, Peyvandi F, Rodeghiero F, Mannucci PM. (2014) The bleeding score predicts clinicaloutcomes and replacement therapy in adults with von Willebrand disease. Blood. 123, 4037-44.Federici AB. Use of intravenous immunoglobulin in patients with acquired von Willebrandsyndrome. Hum Immunol. 2005;66:422-430.

Federici AB. (2008) The use of desmopressin in von Willebrand disease: the experience of thefirst 30 years (1977–2007). Haemophilia, 14(Suppl. 1), 5–14.

Federici AB, Mannucci PM, Castaman G, Baronciani L, Bucciarelli P, Canciani MT, Pecci A,Lenting PJ, De Groot PG. (2009) Clinical and molecular predictors of thrombocytopenia and riskof bleeding in patients with von Willebrand disease type 2B: a cohort study of 67 patients.Blood 113, 526–534.

Fidalgo T, Salvado R, Corrales I, Pinto SC, Borràs N, Oliveira A, Martinho P, Ferreira G, AlmeidaH, Oliveira C, Marques D, Gonçalves E, Diniz M, Antunes M, Tavares A, Caetano G, KjöllerströmP, Maia R, Sevivas TS, Vidal F, Ribeiro L. (2009) Genotype-phenotype correlation in a cohort ofPortuguese patients comprising the entire spectrum of VWD types: impact of NGS. ThrombHaemost. 116, 17-31.

Fijnvandraat K, Peters M, ten Cate JW. (1995) Inter-individual variation in half-life of infusedrecombinant factor VIII is related to pre-infusion von Willebrand factor antigen levels. Br JHaematol. 91, 474-6.

Flood VH, Christopherson PA, Gill JC, Friedman KD, Haberichter SL, Bellissimo DB, Udani RA,Dasgupta M, Hoffmann RG, Ragni MV, Shapiro AD, Lusher JM, Lentz SR, Abshire TC, LeissingerC, Hoots WK, Manco-Johnson MJ, Gruppo RA, Boggio LN, Montgomery KT, Goodeve AC, JamesPD, Lillicrap D, Peake IR, Montgomery RR. (2016) Clinical and laboratory variability in a cohortof patients diagnosed with type 1 VWD in the United States. Blood 127, 2481-8.

Flood VH, Gill JC, Morateck PA, Christopherson PA, Friedman KD, Haberichter SL, BranchfordBR, Hoffmann RG, Abshire TC, Di Paola JA, Hoots WK, Leissinger C, Lusher JM, Ragni MV,Shapiro AD, Montgomery RR. (2010) Common VWF exon 28 polymorphisms in AfricanAmericans affecting the VWF activity assay by ristocetin cofactor. Blood 116, 280-286.

78

Fressinaud E, Veyradier A, Truchaud F, Martin I, Boyer-Neumann C, Trossaert M, MeyerD. (1998) Screening for von Willebrand disease with a new analyzer using high shear stress: astudy of 60 cases. Blood 91, 1325-31.

Gazda H, Budde U, Krey S, Rokicka-Milewska R, Schneppenheim R. (1997) Delta C in exon 18 ofthe von Willebrand factor gene is the most common mutation in patients with severe vonWillebrand disease type 3 in Poland. Blood 90: 566A (abstr).

Giblin JP, Hewlett LJ, Hannah MJ. (2008) Basal secretion of von Willebrand factor from humanendothelial cells. Blood 112. 957-964.

Gill JC, Castaman G, Windyga J, Kouides P, Ragni M, Leebeek FW, Obermann-Slupetzky O,Chapman M, Fritsch S, Pavlova BG, Presch I, Ewenstein B. (2015) Hemostatic efficacy, safety,and pharmacokinetics of a recombinant von Willebrand factor insevere von Willebrand disease. Blood 126, 2038-46.

Gill JC, Endres-Brooks J, Bauer PJ, Marks WJ Jr, Montgomery RR. (1987) The effect of ABOblood group on the diagnosis of von Willebrand disease. Blood 69 1691-5.

Ginsburg D, Handin RI, Bonthron DT, Donlon TA, Bruns GA, Latt SA, Orkin SH. (1985) Humanvon Willebrand factor (vWF): isolation of complementary DNA (cDNA) clones andchromosomal localization. Science. 228: 1401-6

Goodeve A, Eikenboom J, Castaman G, Rodeghiero F, Federici AB, Batlle J, Meyer D, MazurierC, Goudemand J, Schneppenheim R, Budde U, Ingerslev J, Habart D, Vorlova Z, Holmberg L,Lethagen S, Pasi J, Hill F, Hashemi Soteh M, Baronciani L, Hallden C, Guilliatt A, Lester W, PeakeI. (2007) Phenotype and genotype of a cohort of families historically diagnosed with type 1 vonWillebrand disease in the European study, Molecular and Clinical Markers for the Diagnosisand Management of Type 1 von Willebrand Disease (MCMDM1VWD). Blood 109, 112–21.

Goodeve AC. (2010) The genetic basis of von Willebrand disease. Blood Rev. 24, 123–134.

Hackeng TM, Ser’e KM, Tans G, Rosling J. (2006) Protein S stimulates inhibition of the tissuefactor pathway by tissue factor pathway inhibitor. Proc Natl Acad Sci USA 103, 3106-3111

Harker LA, Roskos LK, Marzec UM, Carter RA, Cherry JK, Sundell B, Cheung EN, Terry D,Sheridan W (2000). Effects of megakaryocyte growth and development factor on plateletproduction, platelet life span, and platelet function in healthy human volunteers. Blood 95,2514 LP-2522.

Hartert H. (1948) Blutgerinnungsstudien mit der Thrombelastographie; einem neuenUntersuchungs verfahren. Klin Wochenschr. 26, 577-583.

Hemker HC, Wielders S, Kessels H, Beguin S. (1993) Continuous registration of thrombingeneration in plasma, its use for the determination of the thrombin potential. Thrombosis andHaemostasis. 70, 617-24.

Hoffman M, Monroe DM. (2007) Coagulation 2006: A Modern View of Hemostasis. HematolOncol Clin North Am. 21, 1-11.

Hoffman M. (2003) A cell-based model of coagulation and the role of factor VIIa. BloodReviews 17, S1-S5.

79

Huang J, Roth R, Heuser JE, Sadler JE. (2009) Integrin alpha(v)beta(3) on human endothelialcells binds von Willebrand factor strings under fluid shear stress. Blood 113, 1589-97.

Huang RH, Wang Y, Roth R, Yu X, Purvis AR, Heuser JE, Egelman EH, Sadler JE. (2008) Assemblyof Weibel-Palade body-like tubules from N-terminal domains of von Willebrand factor. ProcNatl Acad Sci U S A. 105, 482-487.

Iorio A, Oliovecchio E, Morfini M, Mannucci PM. (2008) Italian registry of haemophilia andallied disorders. Objectives, methodology and data analysis. Haemophilia 14, 444–53.

Jaffe EA, Hoyer LW, Nachman RL. (1973) Synthesis of antihemophilic factor antigen by culturedhuman endothelial cells. J Clin Invest. 52, 2757-2764

James PD, Notley C, Hegadorn C, Leggo J, Tuttle A, Tinlin S, Brown C, Andrews C, Labelle A,Chirinian Y, O'Brien L, Othman M, Rivard G, Rapson D, Hough C, Lillicrap D. (2007) Themutational spectrum of type 1 von Willebrand disease: Results from a Canadian cohort study.Blood 109, 145–54.

James PD, Notley C, Hegadorn C, Leggo J, Tuttle A, Tinlin S, Brown C, Andrews C, Labelle A,Chirinian Y, O'Brien L, Othman M, Rivard G, Rapson D, Hough C, Lillicrap D. (2007). Themutational spectrum of type 1 von Willebrand disease: Results from a Canadian cohort study.Blood 109, 145–54.

James PD, Lillicrap D, Mannucci PM. (2013) Alloantibodies in von Willebrand disease. Blood122, 636-40.

Jenkins PV, Rawley O, Smith OP, O'Donnell JS. (2012) Elevated factor VIII levels and riskof venous thrombosis. Br J Haematol. 157, 653-63.

Jokela V, Lassila R, Szanto T, Joutsi-Korhonen L, Armstrong E, Oyen F, Schneppenheim S,Schneppenheim R. (2013) Phenotypic and genotypic characterization of 10 Finnish patientswith von Willebrand disease type 3: discovery of two main mutations. Haemophilia 19, 344-8.

Kanaji S, Fahs SA, Shi Q, Haberichter SL, Montgomery RR. (2012) Contribution of platelet vs.endothelial VWF to platelet adhesion and hemostasis. J Thromb Haemost.10, 1646-52.

Katsumi A, Tuley EA, Bod´o I, Sadler JE. (2000) Localization of disulfide bonds in the cystineknot domain of human von Willebrand factor. J Biol Chem. 275, 25585-25594.

Keularts IM, Hamulyak K, Hemker HC, Béguin S. (2000) The effect of DDAVP infusionon thrombin generation in platelet-rich plasma of von Willebrandtype 1 and in mildhaemophilia A patients. Thromb Haemost. 2000 84, 638-42.

Kitchen S, Jennings I, Woods TA, Kitchen DP, Walker ID, Preston FE. Laboratory tests formeasurement of von Willebrand factor show poor agreement among different centers: resultsfrom the United Kingdom National External Quality Assessment Scheme for Blood Coagulation.Semin Thromb Hemost. 2006;32(5): 492-498.

Kitchen S, Kershaw G, Tiefenbacher S. (2016) Recombinant to modified factor VIII and factor IX- chromogenic and one-stage assays issues. Haemophilia 22, 72-7.

80

Krizek DR, Rick ME. (2000) A rapid method to visualize von Willebrand factor multimers byusing agarose gel electrophoresis, immunolocalization and luminographic detection.Thrombosis Res. 97, 457–462.

Kunicki TJ, Federici AB, Salomon DR, Koziol JA, Head SR, Mondala TS, Chismar JD, Baronciani L,Canciani MT, Peake IR. (2004) An association of candidate gene haplotypes and bleedingseverity in von Willebrand disease (VWD) type 1 pedigrees. Blood 104, 2359–2367

Kunicki TJ, Baronciani L, Canciani MT, Gianniello F, Head SR, Mondala TS, Salomon DR, FedericiAB.(2006) An association of candidate gene haplotypes and bleeding severity in vonWillebrand disease type 2A, 2B, and 2M pedigrees. J Thromb Haemost. 4, 137-47.

Laffan MA, Lester W, O'Donnell JS, Will A, Tait RC, Goodeve A, Millar CM, Keeling DM. (2014)The diagnosis and management of von Willebrand disease: a United Kingdom HaemophiliaCentre Doctors Organization guideline approved by the British Committee for Standards inHaematology. Brit J Haematol. 167, 453–465.

Lassila R, Holme PA, Landorph A, Petrini P, Onundarson PT, Hillarp A. (2011) NordicHaemophilia Council's practical guidelines on diagnosis and management of von Willebranddisease. Semin Thromb Hemost. 37, 495-502.

Lassila R, Lindberg O. (2013) Erik von Willebrand. Haemophilia 9, 643-7.

Lavin M, Aguila S, Schneppenheim S, Dalton N, Jones KL, O'Sullivan JM, O'Connell NM, RyanK, White B, Byrne M, Rafferty M, Doyle MM, Nolan M, Preston RJS, Budde U, James P, Di PaolaJ, O'Donnell JS. (2017) Novel insights into the clinical phenotype and pathophysiologyunderlying low VWF levels. Blood 130, 2344-2353.

Lawrie AS, MacKie IJ, Machin SJ, Peyvandi F. (2011) Evaluation of an automated platelet-basedassay of ristocetin cofactor activity. Haemophilia 17, 252–256.

Lawrie AS, Stufano F, Canciani MT, MacKie IJ, Machin SJ, Peyvandi F. (2013) A comparativeevaluation of a new automated assay for von Willebrand factor activity. Haemophilia 19, 338–342.

Lenting PJ, van Mourik JA, Mertens K. (1998) The life cycle of coagulation factor VIII in view ofits structure and function. Blood 92, 3983-3996

Lillicrap D. (2013) von Willebrand disease: advances in pathogenetic understanding, diagnosis,and therapy. Blood 122, 3735-40.

Livnat T, Shenkman B, Martinowitz U, Zivelin A, Dardik R, Tamarin I, Mansharov R, Budnik I,Salomon O. (2015) The impact of thrombin generation and rotation thromboelastometry onassessment of severity of factor XI deficiency.Thromb Res. 2015 136, 465-73.

Ljungkvist M , Strandberg K, Berntorp E , Chaireti R , Holme PA , Larsen OH , Lassila R ,Jouppila A , Szanto T ,Zetterberg E. (2018) Evaluation of a standardized protocol for thrombingeneration using the calibrated automated thrombogram: A Nordic study . Haemophilia, inpress

Lynch DC, Zimmerman TS, Collins CJ, Brown M, Morin MJ, Ling EH, Livingston DM.(1985) Molecular cloning of cDNA for human von Willebrand factor: authentication by a newmethod. Cell 41, 49–56

81

Maas C, Renné T. (2018) Coagulation factor XII in thrombosis and inflammation. Blood 131,1903-1909.

MacFarlane RG. (1964) An enzyme cascade in the blood clotting mechanism, and its functionas a biological amplifier. Nature 202, 498-9.

Makris M, Federici AB, Mannucci PM, Bolton-Maggs PH, Yee TT, Abshire T, Berntorp E. (2015)The natural history of occult or angiodysplastic gastrointestinal bleeding in von Willebranddisease. Haemophilia 21, 338-42.

Mancuso DJ, Tuley EA, Westfield LA, Lester-Mancuso TL, Le Beau MM, Sorace JM, Sadler JE.(1991) Human von Willebrand factor gene and pseudogene: structural analysis anddifferentiation by polymerase chain reaction. Biochemistry 30, 253-69.

Mann G, Brummel-Ziedins K, Orfeo T, Butenas S. (2006) Models of blood coagulation. BloodCells, Molecules, and Disease 36, 108-117.

Mannucci PM, Federici AB. (1995) Antibodies to von Willebrand factor in von Willebranddisease. Adv Exp Med Biol. 386, 87–91.

Mannucci PM. (2004) Treatment of von Willebrand's Disease. N Engl J Med. 2004 351, 683-94.

Mannucci PM: Italian Working Group (1977) Spectrum of von Willebrand disease: a study of100 cases. Br J Haematol 1977; 35: 101-112.

Mannucci PM, Tenconi PM, Castaman G, Rodeghiero F. (1992) Comparison of four virus-inactivated plasma concentrates for treatment of severe von Willebranddisease: a cross-overrandomized trial. Blood 79, 3130-7.

Mannucci PM, Federici AB. (1995) Antibodies to von Willebrand factor in von Willebranddisease. Adv Exp Med Biol. 386, 87-92.

Mannucci PM, Tamaro G, Narchi G, Candotti G, Federici A, Altieri D, Tedesco F. (1987) Life-threatening reaction to factor VIII concentrate in a patient with severe von Willebrand diseaseand alloantibodies to von Willebrand factor. Eur J Haematol. 39, 467-70.

Marcus AJ, Broekman MJ, Drosopoulos JH, Olson KE, Islam N, Pinsky DJ, Levi R. (2005) Role ofCD39 (NTPDase-1) in thromboregulation, cerebroprotection, and cardioprotection. SeminThromb Hemost. 31, 234-246.

Mast AE. (2016) Tissue Factor Pathway Inhibitor: Multiple Anticoagulant Activities for a SingleProtein. Arterioscler Thromb Vasc Biol. 36, 9-14

Mazurier C, Dieval J, Jorieux S, Delobel J, Goudemand M. (1990) A new von Willebrand factor(vWF) defect in a patient with factor VIII (FVIII) deficiency but with normal levels andmultimeric patterns of both plasma and platelet vWF. Characterization of abnormal vWF/FVIIIinteraction. Blood. 75, 20-6.

Meijer P, Haverkate F.(2006) An external quality assessment program for von Willebrandfactor laboratoryanalysis: an overview from the European concerted action on thrombosis anddisabilities foundation. Semin Thromb Hemost. 32, 485-491.

82

Michaux G, Abbitt KB, Collinson LM, Haberichter SL, Norman KE, Cutler DF. (2006) Thephysiological function of von Willebrand's factor depends on its tubular storage in endothelialWeibel-Palade bodies. Dev Cell 10, 223-232

Millar CM, Riddell AF, Brown SA, Starke R, Mackie I, Bowen DJ, Jenkins PV, van Mourik JA.(2008) Survival of von Willebrand factor released following DDAVP in a type 1 von Willebranddisease cohort: influence of glycosylation, proteolysis and gene mutations. Thromb Haemost.99, 916-24.

Moenen FCJI, Nelemans PJ, Schols SEM, Schouten HC, Henskens YMC, Beckers EAM. (2018)The diagnostic accuracy of bleeding assessment tools for the identification of patients withmild bleeding disorders: A systematic review. Haemophilia 2018 24, 525-535.

Mohl A, Marschalek R, Masszi T, Nagy E, Obser T, Oyen F, Sallai K, Bodó I, Schneppenheim R.(2008) An alu-mediated novel large deletion is the most frequent cause of type 3 vonWillebrand disease in Hungary. J Thromb Haemost. 6, 1729–35

Nachman R, Levine R, Jaffe EA. (1977) Synthesis of factor VIII antigen by cultured guinea pigmegakaryocytes. J Clin Invest. 60, 914-921.

Nichols WL, Hultin MB, James AH, Manco-Johnson MJ, Montgomery RR, Ortel TL, Rick ME,Sadler JE, Weinstein M, Yawn BP. (2008) von Willebrand disease (VWD): evidence-baseddiagnosis and management guidelines, the National Heart, Lung, and Blood Institute (NHLBI)Expert Panel report (USA). Haemophilia 14, 171–232.

Nightingale T, Cutler D. (2013) The secretion of von Willebrand factor from endothelial cells;an increasingly complicated story. J Thromb Haemost. 11, 192-201.

Nilsson IM, Blomback M, Jorpes E, Blomback B, Johansson SA. (1957) Von Willebrand's diseaseand its correction with human plasma fraction 1-0. Acta Med Scand. 29, 179-88.

Nilsson IM. (1999) The history of von Willebrand disease. Haemophilia 5, 7-11.

Ott, H.W., Griesmacher, A., Schnapka-Koepf, M., Golderer, G., Sieberer, A., Spannagl, M.,Scheibe, B., Perkhofer, S., Will, K. & Budde, U. (2010) Analysis of von Willebrand factormultimers by simultaneous high- and low-resolution vertical SDS-agarose gel electrophoresisand Cy5-labeled antibody high-sensitivity fluorescence detection. American Journal of ClinicalPathology, 133, 322–330.

Pannekoek H, Voorberg J. (1989) Molecular cloning, expression and assembly of multimericvon Willebrand factor. Baillieres Clin Haematol. 2, 879-896.

Pasi KJ, Collins PW, Keeling DM, Brown SA, Cumming AM, Dolan GC, Hay CRM., Hill FGH, LaffanM, Peake IR. Management of von Willebrand disease: a guideline from the UK HaemophiliaCentre Doctors’ Organization. Haemophilia 2004, 10, 218–231.

Pergantou H, Xafaki P, Adamtziki E, Koletsi P, Komitopoulou A, Platokouki H. (2012) Thechallenging management of a child with type 3 von Willebrand disease and antibodies to vonWillebrand factor. Haemophilia 18, e66-7.

Philipp CS, Faiz A, Heit JA, Kouides PA, Lukes A, Stein SF, Byams V, Miller CH, Kulkarni R. (2011)Evaluation of a screening tool for bleeding disorders in a US multisite cohort of women withmenorrhagia. Am J Obstet Gynecol. 204, e1-e7.

83

Pike GN, Cumming AM, Thachil J, Hay CR, Burthem J, Bolton-Maggs PH. (2017) Evaluation ofthe use of global haemostasis assays to monitor treatment in factor XI deficiency. Haemophilia2017 23. 273-283.

Pikta M, Zemtsovskaja G, Bautista H, Nouadje G, Szanto T, Viigimaa M, Banys V. (2018).Preclinical evaluation of a semi-automated and rapid commercial electrophoresis assay for vonWillebrand factor multimers. Journal of clinical laboratory analysis. In press

Pimanda JE, Annis DS, Raftery M, Mosher DF, Chesterman CN, Hogg PJ. (2002) The vonWillebrand factor-reducing activity of thrombospondin-1 is located in the calcium-binding/Cterminal sequence and requires a free thiol at position 974. Blood 100, 2832-2838.

Quinsey NS, Greedy AL, Bottomley SP, Whisstock JC, Pike RN. (2004) Antithrombin: in controlof coagulation. I Int J Biochem Cell Biol. 36, 386-389.

Quiroga T, Goycoolea M, Muñoz B, Morales M, Aranda E, Panes O, Pereira J, Mezzano D.(2004) Template bleeding time and PFA-100 have low sensitivity to screen patients withhereditary mucocutaneous hemorrhages: comparative study in 148 patients. J ThrombHaemost. 2, 892–8.

Rastegarlari G, Pegon JN, Casari C, Odouard S, Navarrete AM, Saint-Lu N, van Vlijmen BJ,Legendre P, Christophe OD, Denis CV, Lenting PJ. (2012) Macrophage LRP1 contributes to theclearance of von Willebrand factor. Blood 119, 2126–34.

Rehemtulla A, Kaufman RJ. (1992) Preferred sequence requirements for cleavage of pro-vonWillebrand factor by propeptide-processing enzymes. Blood 79, 2349-2355.

Rendu F, Brohard-Bohn B. (2001) The platelet release reaction; granules´ constituents,secretion and function. Platelets 12, 261-273

Roberts HR, Monroe DM, Escobar MA. (2004) Current concepts of hemostasis. Anesthesiology10, 722-730.

Robertson JD, Yenson PR, Rand ML, Blanchette VS, Carcao MD, Notley C, Lillicrap D, James PD.(2011) Expanded phenotype–genotype correlations in a pediatric population with type 1 vonWillebrand disease. J Thromb Haemost. 9, 1752–60

Rodeghiero F, Castaman G, Dini E. (1987) Epidemiological investigation of the prevalence ofvon Willebrand’s disease. Blood 69, 454-459.

Rodeghiero F, Castaman G, Tosetto A, Batlle J, Baudo F, Cappelletti A, Casana P, De Bosch N,Eikenboom JC, Federici AB, Lethagen S, Linari S, Srivastava A. (2005) The discriminant power ofbleeding history for the diagnosis of type 1 von Willebrand disease: an international,multicenter study. J Thromb Haemost. 3, 2619-2626.

Rodeghiero F, Tosetto A, Abshire T, Arnold DM, Coller B, James P, Neunert C, LillicrapD; ISTH/SSCjoint VWF and Perinatal/Pediatric Hemostasis Subcommittees Working Group.(2010) ISTH/SSC bleeding assessment tool: a standardized questionnaire and a proposal for anew bleeding score for inherited bleeding disorders. J Thromb Haemost. 8, 2063-2065.

Rodgers SE, Wong A, Gopal RD, Dale BJ, Duncan EM, McRae SJ. (2014) Evaluation of pre-analytical variables in a commercial thrombin generation assay. Thromb Res. 134, 160-4.

84

Rugeri L, Beguin S, Hemker C, Bordet JC, Fleury R, Chatard B, Negrier C, Dargaud Y. (2007)Thrombin-generating capacity in patients with von Willebrand's disease. Haematologica 92,1639-46.

Ruggeri ZM (2002). Platelets in atherothrombosis. Nat Med 8, 1127-1234

Ruggeri ZM, Ciavarella N, Mannucci PM, Molinari A, Dammacco F, Lavergne JM, Meyer D.(1979) Familial incidence of precipitating antibodies in von Willebrand’s disease: a study offour cases. J Lab Clin Med. 94, 60–75.

Ruggeri ZM. (2003) Von Willebrand factor, platelets and endothelial cell interactions. J ThrombHaemost. 1, 1335-1342.

Rydz N, Swystun LL, Notley C, Paterson AD, Riches JJ, Sponagle K, Boonyawat B, MontgomeryRR, James PD, Lillicrap D. (2013) The C-type lectin receptor CLEC4M binds, internalizes, andclears von Willebrand factor and contributes to the variation in plasma von Willebrand factorlevels. Blood 121, 5228–37

Sadler JE, Budde U, Eikenboom JC, Favaloro EJ, Hill FG, Holmberg L, Ingerslev J, Lee CA, LillicrapD, Mannucci PM, Mazurier C, Meyer D, Nichols WL, Nishino M, Peake IR, Rodeghiero F,Schneppenheim R, Ruggeri ZM, Srivastava A, Montgomery RR, Federici AB; Working Partyon von Willebrand Disease Classification. (2006) Update on the pathophysiology andclassification of von Willebrand disease: a report of the Subcommittee on von WillebrandFactor. J Thromb Haemost. 4, 2103-14.

Sadler JE, Gralnick HR. (1994) Commentary: a new classification for von Willebrand disease.Blood 84, 676-679.

Sadler JE, Mannucci PM, Berntorp E, Bochkov N, Boulyjenkov V, Ginsburg D, Meyer D, Peake I,Rodeghiero F, Srivastava A. (2000) Impact, diagnosis and treatment of von Willebrand disease.Thromb Haemost. 84, 160-74.

Sadler JE, Shelton-Inloes BB, Sorace JM, Harlan JM, Titani K, Davie EW. (1985) Cloning andcharacterization of two cDNAs coding for human von Willebrand factor. Proc Natl Acad Sci U SA 82, 6394–6398

Sadler JE. (1998) Biochemistry and genetics of von Willebrand factor. Annu Rev Biochem. 67,395-424.

Sadler JE. (2003) Von Willebrand disease type 1: a diagnosis in search of a disease. Blood 101,2089-2093

Salinas V, Carmona R, Mohammed BM, Martin EJ, Brophy DF, Young G. (2015) Is some betterthan none: are TEG and TGA profiles different in severe FVIII-deficient patients with inhibitors?Haemophilia 21, 398-404.

Sanders YV, Groeneveld D, Meijer K, Fijnvandraat K, Cnossen MH, van der Bom JG, Coppens M,de Meris J, Laros-van Gorkom BA, Mauser-Bunschoten EP, Leebeek FW, Eikenboom J; WiNstudy group. (2015) von Willebrand factor propeptide and the phenotypic classification of vonWillebrand disease Blood 125, 3006-3013;

Sanders YV, van der Bom JG, Isaacs A, Cnossen MH, de Maat MP, Laros-van Gorkom BA,Fijnvandraat K, Meijer K, van Duijn CM, Mauser-Bunschoten EP, Eikenboom J, Leebeek FW;

85

WiN Study Group. (2015) CLEC4M and STXBP5 gene variations contribute to von Willebrandfactor level variation in von Willebrand disease. J Thromb Haemost. 13, 956-66.

Sanders YV, Fijnvandraat K, Boender J, Mauser-Bunschoten EP, van der Bom JG, de MerisJ, Smiers FJ, Granzen B, Brons P, Tamminga RY, Cnossen MH, Leebeek FW WiN Study Group.(2015) Bleeding spectrum in children with moderate or severe von Willebrand disease:Relevance of pediatric-specific bleeding. Am J Hematol. 90, 1142-8.

Santos KF, Battisti V, Corrêa Mde C, Mann TR, Pereira Rda S, Araújo Mdo C, BrülêAO, Schetinger MR, Morsch VM. (2010) Enzymes that hydrolyze adenine nucleotides inplatelets and polymorphisms in the alpha2 gene of integrin alpha2beta1 in patients with vonWillebrand disease. Mol Cell Biochem. 340, 249-56.

Schmidt DE, Bruzelius M, Majeed A, Odeberg J, Holmström M, Ågren A. (2017) Whole bloodristocetin-activated platelet impedance aggregometry (Multiplate) for the rapid detectionof Von Willebrand disease. Thromb Haemost. 117, 1528-1533.

Schmidt DE, Majeed A, Bruzelius M, Odeberg J, Holmström M, Ågren A. (2017) A prospectivediagnostic accuracy study evaluating rotational thromboelastometry andthromboelastography in 100 patients with von Willebrand disease. Haemophilia 23, 309-318.

Schneppenheim R, Krey S, Bergmann F, Bock D, Budde U, Lange M, Linde R, Mittler U, Meili E,Mertes G, et al. (1994) Genetic heterogeneity of severe von Willebrand disease type III in theGerman population. Hum Genet. 94, 640–52.

Shelton-Inloes BB, Chehab FF, Mannucci PM, Federici AB, Sadler JE. (1987) Gene deletionscorrelate with the development of alloantibodies in von Willebrand disease. J Clin Invest. 79,1459–65.

Smith NL, Chen M-H, Dehghan A, Strachan DP, Basu S, Soranzo N, et al. (2010) Novelassociations of multiple genetic loci with plasma levels of factor VII, factor VIII, and vonWillebrand factor: The CHARGE (Cohorts for Heart and Aging Research in GenomeEpidemiology) Consortium. Circulation 121, 1382–1392.

Sobel M, McNeill PM, Carlson PL, Kermode JC, Adelman B, Conroy R, Marques D. (1991)Heparin inhibition of von Willebrand factordependent platelet function in vitro and in vivo. JClin Invest. 87, 1787-1793.

Stites DP, Hershgold EJ, Perlman JD, Fudenberg HH. (1971) Factor 8 detection byhemagglutination inhibition: hemophilia A and von Willebrand’s disease. Science 171, 196–7.

Stockley J, Nisar SP, Leo VC, Sabi E, Cunningham MR, Eikenboom JC, et al. (2015) Identificationand Characterization of Novel Variations in Platelet G-Protein Coupled Receptor (GPCR) Genesin Patients Historically Diagnosed with Type 1 von Willebrand Disease. PLoS ONE 10(12):e0143913.

Stockschlaeder M, Schneppenheim R, Budde U. (2014)Update on von Willebrand factor multimers: focus on high-molecular-weight multimers andtheir role in hemostasis. Blood Coagul Fibrinolysis 25, 206-16.

86

Stufano F, Baronciani L, Mane-Padros D, Faraudo S, Peyvandi F. (2018) A comparativeevaluation of a new fully- automated assay for von Willebrand factor collagen binding activityto an established method. Haemophilia 24, 156–161.

Szántó T, Joutsi-Korhonen L, Deckmyn H, Lassila R. (2012) New insights into von Willebranddisease and platelet function. Semin Thromb Hemost. 38, 55-63.

Timm A, Hillarp A, Philips M, Goetze JP. (2015) Comparison of automated von Willebrandfactor activity assays. Thromb Res. 135, 684-91

Tosetto A, Castaman G, Plug I, Rodeghiero F, Eikenboom J. (2011) Prospective evaluation ofthe clinical utility of quantitative bleeding severity assessment in patients referred forhemostatic evaluation. J Thromb Haemost. 9, 1143-1148.

Tran HT, Tjønnfjord GE, Holme PA. (2014) Use of thromboelastography and thrombingeneration assay to predict clinical phenotype in patients with severe FVII deficiency.Haemophilia 20, 141-6.

Turecek PL, Mitterer A, Matthiessen HP, Gritsch H, Varadi K, Siekmann J, Schnecker K,Plaimauer B, Kaliwoda M, Purtscher M, Woehrer W, Mundt W, Muchitsch EM, Suiter T,Ewenstein B, Ehrlich HJ, Schwarz HP. (2009) Development of a plasma- and albumin-freerecombinant von Willebrand factor. Hamostaseologie 29, S32-8.

Valarche V, Desconclois C, Boutekedjiret T, Dreyfus M, Proulle V. (2011) Multiplate wholeblood impedance aggregometry: a new tool for von Willebrand disease. J Thromb Haemost. 9,1645-7.

Valentijn KM, Eikenboom J. (2013) Weibel-Palade bodies: a window to von Willebrand disease.J Thromb Haemost. 11, 581-592.

Valentijn KM, van Driel LF, Mourik MJ, Hendriks GJ, Arends TJ, Koster AJ, Valentijn JA. (2010)Multigranular exocytosis of Weibel-Palade bodies in vascular endothelial cells. Blood 116,1807-1816.

Valentijn KM, Sadler JE, Valentijn JA, Voorberg J, Eikenboom J. (2011) Functional architectureof Weibel-Palade bodies. Blood 117, 5033-43.

van der Meer JH, van der Poll T, van 't Veer C. (2014) TAM receptors, Gas6, and protein S: rolesin inflammation and hemostasis. Blood 123, 2460-9.

van Galen KP, Sanders YV, Vojinovic U, Eikenboom J, Cnossen MH, Schutgens RE, van der BomJG, Fijnvandraat K, Laros-Van Gorkom BA, Meijer K, Leebeek FW, Mauser-Bunschoten EP; WiNStudy Group. (2015) Joint bleeds in von Willebrand disease patients have significant impact onquality of life and joint integrity: a cross-sectional study. Haemophilia 21, e185-92.

van Loon J, Dehghan A, Weihong T, Trompet S, McArdle WL, Asselbergs FW, Chen MH, LopezLM, Huffman JE, Leebeek FW, Basu S, Stott DJ, Rumley A, Gansevoort RT, Davies G, Wilson JJ,Witteman JC, Cao X, de Craen AJ, Bakker SJ, Psaty BM, Starr JM, Hofman A, Jukema JW, DearyIJ, Hayward C, van der Harst P, Lowe GD, Folsom AR, Strachan DP, Smith N, de Maat MP,O'Donnell C. (2016) Genome-wide association studies identify genetic loci for low vonWillebrand factor levels. Eur J Hum Genet. 24, 1035–40.

87

van Schooten CJ, Shahbazi S, Groot E, Oortwijn BD, van den Berg HM, Denis CV, Lenting PJ.(2008) Macrophages contribute to the cellular uptake of von Willebrand factor and factor VIIIin vivo. Blood 112 1704-1712.

van Veen JJ, Gatt A, Makris M. (2008) Thrombin generation testing in routine clinical practice:are we there yet? Br J Haematol. 142, 889-903.

Veyradier A, Fressinaud E, Sigaud M, Wolf M, Meyer D. (1999) A new automated method forvon Willebrand factor antigen measurement using latex particles. Thromb Haemost. 81, 320-1.

Veyradier A, Boisseau P, Fressinaud E, Caron C, Ternisien C, Giraud M, Zawadzki C, TrossaertM, Itzhar-Baïkian N, Dreyfus M, d'Oiron R, Borel-Derlon A, Susen S, Bezieau S, DenisCV, Goudemand J, French Reference Center for von Willebrand disease. (2016) A LaboratoryPhenotype/Genotype Correlation of 1167 French Patients From 670 Families With vonWillebrand Disease: A New Epidemiologic Picture. Medicine (Baltimore). 95, e3038.

von Willebrand EA. (1926) Pseudohaemophilia. Finska Läkaresällskapet Handlingar (FLS Handl)LXVII, 87–112.

Wahlberg T. (1982) Carriers and noncarriers of haemophilia A. Evaluation of bleedingsymptoms registered by a self-administered questionnaire with binary (no/yes) questions.Thromb Res. 25, 415-422.

Wang JW, Valentijn KM, de Boer HC, Dirven RJ, van Zonneveld AJ, Koster AJ, VoorbergJ, Reitsma PH, Eikenboom J. (2011) Intracellular storage and regulated secretion of vonWillebrand factor in quantitative von Willebrand disease. J Biol Chem. 2011 286, 24180-8.

Waters EK, Hilden I, Sørensen BB, Ezban M, Holm PK. (2015) Thrombin generation assay usingfactor XIa to measure factors VIII and IX and their glycoPEGylated derivatives is robust andsensitive. J Thromb Haemost. 13, 2041-52.

Wellington K, Wagstaff AJ. (2003) Tranexamic acid: a review of its use in the management ofmenorrhagia. Drugs 63, 1417-33.

Welsh JD, Poventud-Fuentes I, Sampietro S, Diamond SL, Stalker TJ, Brass LF. (2017)Hierarchical organization of the hemostatic response to penetrating injuries in the mousemacrovasculature. J Thromb Haemost. 15, 526-537.

Werner EJ, Broxson EH, Tucker EL, Giroux DS, Shults J, Abshire TC. (1993) Prevalence of vonWillebrand disease in children: a multiethnic study. J Pediatr. 123, 893-898.

Whiting D, DiNardo JA. (2014) TEG and ROTEM: technology and clinical applications. Am JHematol. 89, 228-32.

Young G, Blain R, Nakagawa P, Nugent DJ. (2006) Individualization of bypassing agenttreatment for haemophilic patients with inhibitors utilizing thromboelastography. Haemophilia12, 598-604.

Young G, Sørensen B, Dargaud Y, Negrier C, Brummel-Ziedins K, Key NS. (2013)Thrombin generation and whole blood viscoelastic assays in the management of hemophilia:current state of art and future perspectives. Blood 121, 1944-50.

88

Zannettino AC, Holding CA, Diamond P, Atkins GJ, Kostakis P, Farrugia A, Gamble J, To LB,Findlay DM, Haynes DR. (2005) Osteoprotegerin (OPG) is localized to the Weibel-Palade bodiesof human vascular endothelial cells and is physically associated with von Willebrand factor. JCell Physiol. 204, 714-723.

Zhang ZP, Blomback M, Egberg N, Falk G, Anvret M. (1994) Characterization of the vonWillebrand factor gene (VWF) in von Willebrand disease type III patients from 24 families ofSwedish and Finnish origin. Genomics 21, 188–93

Zhang ZP, Blombäck M, Nyman D, Anvret M. (1993) Mutations of von Willebrand factor genein families with von Willebrand disease in the Aland Islands. Proc Natl Acad Sci U S A. 90, 7937-40.

Zhou YF, Eng ET, Nishida N, Lu C, Walz T, Springer TA. (2011) A pH-regulated dimeric bouquetin the structure of von Willebrand factor. EMBO J 30, 4098-4111.

Zhou YF, Eng ET, Zhu J, Lu C, Walz T, Springer TA. (2012) Sequence and structure relationshipswithin von Willebrand factor. Blood 120, 449-458.

Zhukov O, Popov J, Ramos R, Vause C, Ruden S, Sferruzza A, Dlott J, Sahud M. (2009)Measurement of von Willebrand factor-FVIII binding activity in patients with suspected vonWillebrand disease type 2N: application of an ELISA-based assay in a reference laboratory.Haemophilia 15, 788–796

Zimmerman TS, Ratnoff OD, Powell AE. (1971) Immunologic differentiation of classichemophilia (factor 8 deficiency) and von Willebrand’s disease, with observations on combineddeficiencies of antihemophilic factor and proaccelerin (factor V) and on an acquired circulatinganticoagulant against antihemophilic factor. J Clin Invest 50, 244–54.