British Journal of Haernatology, 1990, 75, 537-542

Hereditary deficiency of all vitamin IS-dependent procoagulants and anticoagulants w.

B. B R E N N E R , s. T A V O R I , A. Z I V E L I N , * c. B. KELLER,? J. S U T T I E , t I. TATARSKY A N D u. S E L I G S O H N * Hematology Institute, Rambam Medical Center, Haifa, *Hematology Institute, Tel-Aviv Medical Center, Tel-Aviv, Israel, and TDepurtment of Biochemistry, University of Wisconsin, Madison, Wisconsin, U.S.A. Received 4 December 1989; accepted for publication 29 March 1990

Summary. Hereditary combined deficiency of vitamin Kdependent factors is a rare entity. We report a 7-year-oldgirl of Arab origin with hereditary deficiency of the procoagulants factors 11. VII. IX and X and the natural anticoagulants proteins C and S. The patient is the tenth offspring of a consanguinous marriage and presented at 6 weeks with spontaneous intracerebral haemorrhage. Symptoms improved following plasma infusion. A sibling died at 5 d from uncontrollable umbilical bleeding. Blood coagulation workup at 6 years showed: factor II:C (activity) 12 U/dl, factor 1I:Ag (antigen)40 U/dl: factor VI1:C 12 U/dl; factor 1X:C 36 U/dl, factor 1X:Ag 57 U/dl; factor X:C 1 7 U/dl, factor X:Ag 54 U/dl; protein C activity 43 U/dl; protein C:Ag 45 U/dl;

protein S :Ag 34 U/dl; levels of factors V :C and VlII :C were normal. Assays of coagulation factors in the parents and five of the siblings were within the normal range. Following acute infection and dilantin therapy procoagulant activity levels were reduced further and were partially increased after vitamin K infusion. Crossed immunoelectrophoresis of prothrombin in the presence of calcium lactate revealed a population of des-carboxyprothrombin. Serum vitamin K epoxide levels were undetectable. The data suggest that the defect in our patient stems from abnormal carboxylation of the vitamin K-dependent proteins and that the mode of inheritance is autosomal recessive.

Vitamin K is a necessary cofactor for the hepatic carboxylation of glutamic acid residues in a number of proteins including the procoagulants factors 11, VII, IX, and X. This carboxylation is required for normal haemostasis since it enables calcium binding and attachment of the procoagulants to phospholipids (Sperling et al. 1978; Esmon et al, 1975). Acquired deficiency of vitamin K-dependent factors is common innewborns (Sutherland et al. 1967),in malabsorption and in hepatic obstruction (Blanchard et al. 1981) and has been associated with a bleeding tendency which is usually responsive to vitamin K administration. The natural anticoagulants protein C and protein S are also vitamin Kdependent (Stenflo, 1976; Discipio et al, 1977) and their plasma levels are reduced in patients with vitamin K deficiency or following oral anticoagulant therapy. Hereditary deficiency of vitamin K-dependent procoagulants is a rare bleeding disorder which has been reported in only six patients (McMillan & Roberts, 1966; Fischer & Zweymuller, 1966: Johnson et al, 1980; Goldsmith et al, 1982; Vicente rJt al, 1984: Pauli et al, 1987). We report an Arab Israeli girl who presented with a severe bleeding

diathesis in infancy. Laboratory workup revealed deficiency of all vitamin K-dependent factors including protein C and protein S . Further investigation suggested that the defect stems from abnormal carboxylation of the vitamin Kdependent procoagulants and anticoagulants. MATERIALS AND METHODS

Blood collection. Following informed consent, blood samples were obtained by venipuncture through a 21 gauge needle into 0.4% sodium citrate in a ratio of 9 :1.Samples were then centrifuged at 3000 g for 20 min at 4°C and plasma aliquots were immediately frozen at - 70°C until further analysis was performed. Factors 11, V. VII, VIII. IX and X activities were assayed by one-stage tests (Seligsohn et al, 1984). Factors II:Ag, IX:Ag, X: Ag and vWF: Ag were determined by electroimmunoassay (Seligsohn et al, 1984). Protein C:Ag (PC:Ag) and protein S: Ag (PS:Ag) were analysed by electroimmunoassay, employing the following antibody solutions: For protein C 2.5% polyethylene glycol (PEG) 8000 and 0.2% rabbit polyclonal anti-protein C antibodies and for protein S Tris Tricine (0.08 M-Tris, 0.02 M Tricine) containing 0.2% goat anti-protein S antibodies which recognize total protein S. A

Correspondence: Dr B. Brenner, Institute of Hematology, Rambam Medical Center, P.O. Box 9602, Bat-Galim. Haifa, Israel.

537

B. Brenner et a1

538 I

II

8

f PROBAND

@

DECEASED

-w ASYMPTOMATIC -STUDIED, NORMAL

Fig 1. Pedigree of an Israeli Arab family with

hereditary deficiency of vitamin K-dependent factors.

@ DIED OF BLEEDING

Table 1. Vitamin K-dependent factor levels (U/dl) in the propositus and her parents Factor levels

1I:C

1I:Ag

VII:C

1X:C

1X:Ag

-

13 36 25 7

57 57

-

-

17

7

13 12 16 2

10

40 58 100

81

2

54 51 45

45 58 -

Father Mother

109 112

76 82

107 106

100 76

100 96

132 136

100 100

Normal range* + 2 SD

77-125

63-139

63-155

63-139

55-160

69-137

Patient At 6 weeks At 6 years Dilantin treatment Acute infection and dilantin

0 12 3 I

86-146

X:C

11

X:Ag

PC:Ag

PC:Act

43

PS:Ag

-

34 43

-

-

100 100

-

100 82

60-136

60-146

74-126

PC:Act =protein C activity: PS =protein S. *Determined in at least 30 healthy subjects.

1% agarose (Seakem, Marine Colloids) was used in all electroimmunoassays and gels were run at room temperature. Polyclonal antibodies used were either commercial (Stago, France) or self made (proteins C and S). Protein C activity was assayed by chromogenic substrate (Strachromprotein C. Stago. France). BaC12 precipitation. To 600 pl of plasma 50 p1 1 M BaClz were added and mixed for 30 min at 4°C. After centrifugation the supernatant was removed and the precipitate was eluted with 250 p1 30 mM EDTA at pH 7.5. Procoagulant and anticoagulant antigen levels in plasma supernatant and eluate were then measured by electroimmunoassay using Tris Tricine PEG buffer (0.08 M Tris, 0.02 M Tricine, 2.5% PEG) for factors IX, X, and protein C and barbital buffer for prothrombin. Crossed immunoelectrophoresis (CIE) ofprothrombin. CIE was performed using a 1%seakem agarose in a Tris Tricine buffer.

The conditions for the horizontal run were: voltage 170 V, current 20 mA, interplate voltage 3.5 V/cm. running time 3-4 h. The vertical run conditions were: current 10 mA. voltage 60-80 V, interplate voltage 2 V/cm. and running time 1 6 h. Serum vitamin K and KO. Serum phylloquinone and phylloquinone epoxide assays were carried out essentially as described by Kindberg & Suttie (1989). Serum samples were thawed in a 37°C water bath, and 50 pl of trans 3Hphylloquinone (8000 dpm, 0.2 ng) was added to 1.0 ml of serum, mixed, and incubated at room temperature for 1 h. 6 ml of ethano1:ether (1: 2. v/v) was added, and the solution was mixed on a vortex mixer for 1 min. After 10 min, 5 ml of hexane was added, the suspension was mixed for 1min more, and centrifuged for 10 min at 1000g. The hexane layer was removed, dried under filtered air, the residue was dissolved in 250 pl hexane, and the total sample was injected onto a

Deficiency of all vitamin K-dependent Factors Waters pPorasil high performance liquid chromatography (HPLC) column (30 cm x 3.9 mm i.d.) and eluted with 25% dichloromethane in hexane with a flow rate of 1.5 ml/min. The eluate was monitored at 254 nm, and the fraction containing trans-phylloquinone (quinone and epoxide) was collected and dried under N2. This residue was dissolved in methanol. Recovery of both phylloquinone and phylloquinone epoxide was assessed from the recovery of 3H-phylloquinone. Two hundred pl of the methanol solution was injected onto a reverse-phase Zorbax ODS analytical HPLC column (25 c m x 4 . 6 mm i.d.; DuPont) and eluted with methanol containing 15% dichloromethane, 0.01 M ZnC12, 0.005 M Na acetate, and 0.005 M acetic acid at a flow rate of 1.0 ml/min. The vitamin peaks were detected fluorometrically ( E x= 3 30 nm, Em=430 nm) after post-column reduction using a zinc reducer column as described by Haroon et al(1986).

a

9

3

5 39

25t

b

+ 0

a

IA

RESULTS The patient was a female neonate who presented with multiple ecchymoses and bleeding from puncture sites shortly after birth. Her prothrombin time was longer than 120 s (normal range 13-14 s) and the activated partial thromboplastin time (aPTT) was above 180 s (normal range 28-35 s). No response to vitamin K was observed and symptoms subsided followingplasma transfusion. The patient was readmitted at 6 weeks with spontaneous intracerebral haemorrhage. Coagulation workup revealed prothrombin activity (factor 1I:C) of 0 u/dl, factor V1I:C 1 3 u/dl, factor IX: C 1 3 u/dl and factor X:C 10 u/dl (for normal ranges see Table I). Plasma levels of factors V. VIII and XI were normal. Liver function tests were normal. Again improvement was achieved only after plasma therapy. The patient was the tenth offspring of an Arab-Israeli family (Fig 1).The parents were first cousins. An older female sibling died at 3 d from uncontrollable bleeding. Vitamin K procoagulant and anticoagulant levels were normal in the parents (Table I) and all six siblings available for study. For the next several years the patient suffered mostly from mucocutaneous bleeding with one episode of haemarthrosis. Vitamin-K dependent factor levels at 6 years revealed moderate deficiency of factors 11, VII and X and mild deficiency of factor IX with factor activity levels 1.5-3-fold lower than factor antigen levels (Table I). Proteins C and S antigen levels were also reduced. On physical examination no skeletal abnormalities were noted. Plain X-rays of the spine and palms were normal. At 7 years the patient was reexamined while under dilantin treatment for a convulsive disorder-a sequela of the intracerebral haemorrhage. At this time factor 11:C level was particularly decreased (TableI), probably indicating interference with carboxylation by dilantin. Shortly afterwards the patient was admitted with an acute viral infection and severe epistaxis necessitating tamponade. Activity levels of all procoagulants were severely depressed (Table I). After intravenous administration of 10 mg vitamin K, procoagulant activity levels increased rapidly, reached a peak at 4.5 h and remained stable at 48 h (Fig 2).

0

1

48

5

Time after vitamin K infusion (h) Fig 2. Procoagulant activity levels before and after intravenous infusion of 10 mg vitamin K.

Table It. Antigen levels (U/dl) of factors before and after BaCl2 precipitation Sample

1I:Ag

IX:Ag

X:Ag

PC:Ag

Normal Plasma Supernatant Eluate

100 0 100

100 0 100

100 0 100

100 0 100

Propositus Plasma Supernatant Eluate

56 5 50

57 0 65

51 0 65

58 0 49

100 30 80

150 45 100

50 0 50

68 11 58

Patient on warfarin Plasma Supernatant Eluate

Specific tests When the patient was 7 years old and in a stable condition, several specific tests were carried out as follows: (1)Factor antigen levels were measured before and after barium chloride precipitation (Table 11). In normal pooled plasma which contains carboxylated protein, 100%of the vitamin K-dependent factors precipitated with BaCI2 and were fully retrieved in the eluate. In contrast, significant amounts remained in the supernatant of a warfarin treated patient reflecting the presence of des-carboxy proteins, Interestingly, in the proband’s plasma most of the vitamin

540

B. Brenner et al

Fig 3. CIE of prothrombin (factor 11) in the propositus, a patient on warfarin and normal pool plasma (NP), done in the presence of EDTA (left panel) and calcium lactate (right panel).

K-dependent factors precipitated with barium chloride with only about 10% of prothrombin remaining in the supernatant (Table 11). ( 2 ) CIE offactor II in the presence of 1 mM EDTA revealed a similarly migrating single peak in normal plasma, warfarin plasma and proband’s plasma (Fig 3 , left panel). However, when the CIE was performed in the presence of 1 mM Ca lactate, the proband’s factor I1 migrated to a more anodic position than the normal factor 11 (Fig 3 , right panel). After vitamin K administration the propositus’ plasma revealed an additional shoulder at the site of the normally migrating peak. ( 3 ) Serum phylloquinone levels were somewhat increased in the proband (2.1 ng/ml) and were normal in her parents and three siblings (ranging from 0.38 to 0.98 ng/mI) (Table 111). Serum vitamin K epoxide levels were below the level of detection (0.1ng/ml) in the proband, her parents and three of her siblings.

Table 111. Serum phylloquinone (K) and phylloquinone epoxide (KO) Vitamin concentration (ns/mU Subject

K

Proband Mother of proband Father of proband Sister of proband Sister of proband Sister of proband

2.10 0.98 0.38 0.96 0.87 0.62

Normal range

0.3-1.0

KO 40.10

< 0.10