Archives ofDisease in Childhood 1993; 68: 297-302 297
Relationship between vitamin K dependent Arch Dis Child: first published as 10.1136/adc.68.3_Spec_No.297 on 1 March 1993. Downloaded from coagulation factors and anticoagulants (protein C and protein S) in neonatal vitamin K deficiency
Tetsuo Matsuzaka, Hiroya Tanaka, Masafumi Fukuda, Mikihiro Aoki, Yoshiro Tsuji, Hisayoshi Kondoh
Abstract which are functionally defective. Dysfunction To determine the relationship between of these coagulation factors and inhibitors are vitamin K dependent coagulation factors clinically well known and seen in such dis- and natural anticoagulants, namely pro- eases as haemorrhagic disease of the new- tein C and protein S, in various degrees born8-10 and thromboembolic disorders of vitamin K deficiency, plasma values including neonatal purpura fulminans.1 1-5 As for clotting activity, protein induced by vitamin K deficiency in the early neonatal vitamin K absence (PIVKA-II), protein C period causes haemorrhagic diseases rather antigen, gammacarboxy protein C anti- than thromboembolic disorders, vitamin K gen, and protein S antigen including total deficiency has been viewed exclusively from and free fractions and activity of protein its relevance to coagulation factors. C were measured in 66 full term and We report our study of the changes of pro- healthy breast fed neonates who did not tein C and protein S as well as of coagulation receive vitamin K supplement at birth. factors from the occurrence of vitamin K The 66 neonates were divided into a con- deficiency to its correction with vitamin K trol group (17 cases) and a low group administration. Furthermore, the relation- (49 cases) according to their values for ship between the breast milk intake and clotting activity-that is, 220% or <20% vitamin K deficiency in neonates was also during the first six days of life-and vita- investigated. min K was immediately given when the neonates showed values <20%. In the low group clotting activity gammacarboxy Subjects and methods protein C, free protein S, and protein C SUBJECTS activity was significantly decreased to a Sixty six healthy, full term (range 38-41 minimum on day 2 or 3, and increased in weeks) neonates with a normal birth weight
parallel after vitamin K administration. (range 2755-3970 g) aged 0 to 6 days were http://adc.bmj.com/ Furthermore, they were positively corre- studied. There was no history of maternal lated with one another and inversely cor- anticonvulsant or anticoagulation treatment. correlated with the PIVKA-II concen- They fed exclusively on breast milk except for trations. These findings suggest that supplements of 5% glucose solution, and did simultaneous gammacarboxylation of not receive vitamin K prophylaxis at birth. coagulation factors and proteins C and S Informed consent was obtained from all par-
acts to maintain both coagulation and ticipating parents. on October 1, 2021 by guest. Protected copyright. anticoagulation activities in parallel at various concentrations of vitamin K. The breast milk intake in the group with low SCREENING FOR HYPOPROTHROMBINAEMIA AND Nagasaki University values of clotting activity was significant- VITAMIN K ADMINISTRATION School ofMedicine, ly lower than that in the control group A test for clotting activity (Normotest, Nagasaki, Japan, during the first three days of life. Nyegaard and Co) was used as the screening Department of Paediatrics Therefore, mothers should be advised to test for hypoprothrombinaemia because this Tetsuo Matsuzaka try to increase the breast milk secretion, test accurately reflects total coagulant activity Hiroya Tanaka for example, by providing frequent suck- of factors II, VII, and X and requires only Masafumi Fukuda lings to infants in the first days oflife. 0 01 ml whole blood. The test was Mikihiro Aoki capillary Yoshiro Tsuji performed every day from 0 to 6 days of age. (Arch Dis Child 1993;68:297-302) As the values for clotting activity in patients Scientific Data Centre with vitamin K deficiency haemorrhage were for the A-Bomb Disaster Vitamin K is required for the post-transla- less than 10%,16 we regarded values <20% as Hisoyoshi Kondoh tional gammacarboxylation of glutamic an indication of being at risk of haemorrhagic acid residues in the vitamin K dependent disease of the newborn and immediately gave Correspondence to: Dr Tetsuo Matsuzaka, proteins,' including not only coagulation fac- 2 mg of vitamin K2 syrup (Eisai Ltd).Sixty six Department of Paediatrics, tors II, VII, IX, and X but also protein C2 neonates were divided into a low (49 cases) Nagasaki University School of Medicine, and protein S,3 which are the important and a control group (17 cases) according to 7-1 Sakamoto-machi, inhibitors of coagulation.4-7 Vitamin K defi- whether their values were <20% or .20%. Nagasaki 852, Japan. ciency results in the biosynthesis of abnormal The low group was further divided into three Accepted 18 September 1992 non-gammacarboxy forms of these proteins subgroups: low-1 (12 cases), low-2 (19 298 Matsuzaka, Tanaka, Fukuda, Aoki, Tsuji, Kondoh
cases), and low-3 (18 cases) showing values were discarded, and the samples in the low-I <20% on day 1, 2, and 3 of life, respectively. group were excluded from the following Arch Dis Child: first published as 10.1136/adc.68.3_Spec_No.297 on 1 March 1993. Downloaded from In order to overcome the effect ofpacked cell measurements because the number of samples volume on coagulation activity, the clotting on day 1 was only 1. factor value was corrected using a correction factor: (100-Hn)/(100-Hs), where Hn is the ASSAY METHODS normal mean packed cell volume (haemato- Protein induced in vitamin K absence crit) value (0-45 is used here) and Hs, the (PIVKA-II) was measured by enzyme linked packed cell volume ofa sample. We use the cor- immunosorbent assay (ELISA) using mono- rected value of the clotting activity throughout clonal antibody against PIVKA-II (Eitest the study except for the screening test. monoP-II; Eisai Ltd).'7 The lowest limit of plasma PIVKA-II is 0-06 arbitrary units (AU)/ml (1 AU corresponds to 1 gg of puri- PREPARATION OF PLASMA SAMPLES fied prothrombin). Blood sampling for the following assays was Total protein C antigen was quantitated by performed on days 0, 3, 4, and 6 in the con- sandwich ELISA (Asserachrom protein C, trol group, whereas in the three low groups it Boehringer Mannheim Yamanouchi). The was performed on day 0, immediately before lowest limit of this assay is 0 1% of the normal (on day 1 in low-i, on day 2 in low-2, and on adult concentration. day 3 in low-3 group) and one day after vita- Gammacarboxy protein C was also mea- min K administration, and on day 6 of life. sured by one step sandwich ELISA (TD-82; Venous blood was collected into siliconised Teijin Ltd)18 using two monoclonal anti- tubes containing 1/10th the volume of 3-8% protein C antibodies, one of which recognised sodium citrate, centrifuged at 2500 g for 20 a gammacarboxyglutamic acid domain related minutes at 4°C, and the supernatant plasma conformational change induced by metal ions, was stored at -70°C. The clotted samples and the other was directed against the activa- tion peptide. Briefly, diluted plasma samples, peroxidase-conjugated antibody against gam- 80 macarboxyglutamic acid domain and anti- body (against the activation peptide) coated 70 polystyrene beads were mixed and incubated. After washing, the beads were immersed in a 60 substrate solution containing hydrogen 50 peroxide. Then, the reaction was stopped by addition of acetic acid, and absorbance of g 40 650 nm was read using a spectrophotometer (Shimadzu). The lowest limit of this assay is 30 10% of the normal adult concentration. Protein C activity was measured by an 20 activated partial thromboplastin time
method (Staclot protein C; Diagnostica http://adc.bmj.com/ 10 Stago) using Protac, an activator of protein C. 0 The lowest limit is 50/o of the normal adult 0 2 3 4 5 6 value. Total protein S antigen was determined by conventional Laurell rocket immunelec- trophoresis using antiprotein S antibody as pre- 16 Free S was viously reported.'9 protein antigen on October 1, 2021 by guest. Protected copyright. 14 measured by one step sandwich ELISA (TD- 84; Teijin Ltd),20 using polyclonal antiprotein 12 S antibody and peroxidase-conjugated mono- clonal antiprotein S antibody that recognised 10 the C4bp binding site. The lowest limit of this assay is 5% of the normal adult value. E 8 MEASUREMENT OF BREAST MILK INTAKE IN 6 NEONATES The volume of breast milk intake was mea- 4 sured every day from 0 to 6 days after birth by 2 weighing neonates before and after feeding and a 1 g gain in weight was assumed to be 0I 1 ml of milk intake. 0 1 2 3 4 5 6 STATISTICAL ANALYSIS Days All data analyses were performed using IBM
Figure 1 Serial clotting activity and PIVKA-IIfor the control group and two groups 9377 and SAS Statistical Software. Student's t shown to have low values for clotting activity on day 2 or 3 (low-2 group and low-3 group test and Wilcoxon's rank sum test were used respectively); results are mean (SE). Note the rise in clotting activity andfall in PIVKA-II to analyse normally distributed and non- in the groups with low values after vitamin K administration. The control group had no vitamin K, whereas low-2 and low-3 groups had it on day 2 and 3, respectively; normally distributed data, respectively. Re- **p method of least squares, and the correlation increased to the maximum on day 2 (mean coefficient (r) was calculated. 8d1 AU/ml, range 0-3-22d1 AU/ml) in the Arch Dis Child: first published as 10.1136/adc.68.3_Spec_No.297 on 1 March 1993. Downloaded from low-2 group and on day 3 (mean 1 12 AU/ml, Results range 0-7-29 1 AU/ml) in the low-3 group, TOTAL COAGULATION ACTIVITIES AND PIVKA-II and decreased gradually with a half life of The values for clotting activity in the control about 48 hours. group decreased slightly to a minimum (mean 4322%, range 28-75%) on day 2, and then rose gradually (fig 1). In the low-2 and low-3 groups, they decreased markedly on day 2 and CHANGES IN CONCENTRATIONS OF on day 3, respectively, and rose rapidly after ANTICOAGULANTS (PROTEIN C AND PROTEIN S) oral administration of vitamin K2. The mean The changes in protein C activity and gam- value in the low group immediately before macarboxy protein C values were essentially vitamin K adminstration was 14-5% (range the same as those found for the clotting 10-24%). activity. In the low group, they decreased to a All infants in the control group showed minimum on day 2 (mean 10/9% in protein C PIVKA-II concentrations of 1-5 AU/ml or less activity, 12-8% in gammacarboxy protein C) during the first six days of life (fig 1). or on day 3 (mean 70% in protein C activity, However, the concentrations in the low group 1600% in gammacarboxy protein C), while in were 0-6 AU/ml or less at birth, then the control group no significant decreases were observed (fig 2). The values for protein C antigen showed no significant changes in 40 1 Protein C activity any of the groups. Both total and free protein S concentrations in the control group gradually increased with 30 age but the ratio of free to total protein S remained almost the same, ranging from 62-0 to 67-8% (fig 3). In the low group, there were % 20 significant decreases in free protein S and in the ratio of free to total protein S on day 2 (mean 14-8% and 50 9%) in the low-2 group 10 Control (n = 8) Low-2 (n = 11) and on day 3 (mean 16-0% and 49%) in the ****~* Low-3 (n =7) low-3 group, but there was no significant decrease in total protein S (fig 3). 0 I 1 0 1 2 3 4 5 6 50- RELATIONSHIP BETWEEN COAGULATION Gammacarboxy protein C FACTORS AND PROTEIN C AND PROTEIN S All data from the three groups (excluding the 40- data after vitamin K administration) were L ;T...... used for the analyses of relationship between http://adc.bmj.com/ 30 a-~~~~~~~~~~~~~~~~~...... {'...... coagulation and anticoagulation factors. Values for clotting activity, protein C activity, 20 gammacarboxy protein C, and free protein S ** -.- ~Control(n=8) in all groups were positively correlated with 10 - **~ ~~O-*-0. Low-2 (n = 11) one another, and they were inversely corre- L~Low-3 (n = 7) lated with logarithmic values of PIVKA-II Namely, the values for clotting activity on October 1, 2021 by guest. Protected copyright. 0 (table). 2 3 4 5 6 and anticoagulant (protein C) activity were 0 1 correlated not only with the degree of gam- macarboxylation of each precursor (PIVKA-II 50 - Protein C antigen and gammacarboxy protein C, respectively), but also with each other (fig 4). The free pro- 40 - tein S concentration was also correlated with all of them. 30- 20 BREAST MILK INTAKE AND VITAMIN K Control (n = 8) DEPENDENT COAGULATION ACTIVITIES 10 -O-- ~~~~~~~~~Low-3~n =7) The cumulative breast milk intake in both 10o ... Low-2 n = 1 1) control and low groups was very low until 2 days of age (mean volume on day 2 was 90 ml 0 1 2 3 4 5 6 in the control, and 39 ml in the low group) Days and increased rapidly on day 3 (mean 318 ml in the control and 161 in the low group; fig Figure 2 Decrease in values ofprotein C activity and gammacarboxy protein C in the 5). The breast milk intake in the low group low-2 and low-3 groups and their recovery after vitamin K administration; results are was significantly lower than that in the control mean (SE). The control group had no vitamin K, whereas low-2 and low-3 groups had it group during the first three days of life on day 2 and 3, respectively; *p 40 Total protein S the newborn. Breast feeding is considered to be an important aetiological factor in idio- Arch Dis Child: first published as 10.1136/adc.68.3_Spec_No.297 on 1 March 1993. Downloaded from 30 pathic disease because of the relatively low concentration of vitamin K22 and low excre- tion of breast milk during the first few days % 20 of life.23 We used the following criteria for vitamin K deficiency: (1) PIVKA-II concentration was 10 Control (n = 8) more than 1-3 AU/m124 and (2) the PIVKA-II Low-2 (n = 1 1) concentration and value -''l"Low-3 (n = for clotting activity 7) corrected 0 after the administration of vitamin K. We found -% of neonates in 0 1 2 3 only9 (1/11) 4 5 6 the control group and 78-3% (18/23) of neonates in the low group showed vitamin K 50 deficiency from two to three days after birth. Furthermore, vitamin K deficiency was relat- 40 ed to the amount of breast milk intake. These findings seem to represent the typical features 30 of classic haemorrhagic disease of the new- born and confirm the previous observations that vitamin K storage in newborn infants is 20 inadequate to maintain their requirement after birth and that the diet is the main source of 10 vitamin K in this period.23-27 The values for protein C antigen and pro- 0 tein C activity in the control group were 0 1 2 3 4 5 6 notably lower than those found in adults, as previous investigators reported.28-30 Further- more, in vitamin K deficiency, protein C 80 Free:total activity and gammacarboxyl protein C decreased to near the values observed in 60 ...... patients with homozygous protein C defi- ciency,'2-14 while the decrease in total protein 50 C antigen was less. These findings indicate % 40 that only gammacarboxy protein C is func- 30 tionally active, and decarboxylated protein C is in 20 Control (n = 8) (PIVKA-PC) present vitamin K defi- . ----! - ~~~~~~~~~~~Low-2(n = 11) ciency. Thus, the measurement of gamma- 10 - Low-3 (n = 7) carboxy protein C is useful in the estimation of protein C activity and for evaluation of 0 1 2 3 4 5 6 patients with vitamin K deficiency. http://adc.bmj.com/ Days Protein S is synthesised not only in liver cells, but also in endothelial cells31 and Figure 3 Serial changes in values of total andfree protein S and ratio offree to total megakaryocytes,32 and serves as a cofactor of protein S; results are mean (SE). The control group had no vitamin K, whereas low-2 and activated C. In low-3 groups had it on day 2 and day 3, respectively; *p<0 05, **p 50% of total protein S is in the free form and on October 1, 2021 by guest. Protected copyright. Correlation between clotting activity, PIVKA-II, protein C, and protein S (figures are the remaining 50 to 60% is complexed to C4b correlation coefficients) binding protein (C4bp), a regulatory compo- PIVKA-II Protein C Gammacarboxy Free nent of the complement system. Only the free activity protein C protein S form can act as a cofactor of activated protein Clotting activity -0-837**** 0 793**** 0-710**** 0 554**** C.'2 The ratio of free to total protein S in PIVKA-II -0-661 **** -0-541**** -0 335** Protein C activity 0799**** 0-648**** neonates is considered to be high because of Gammacarboxy protein C 0-684**** the low values of C4bp,34 35 although the total Free protein S protein S concentration is much lower than ****p 100 n 84 50 n 67 r=-0*837 45 r=0-799 Arch Dis Child: first published as 10.1136/adc.68.3_Spec_No.297 on 1 March 1993. Downloaded from 0001p 20Q5 O 0 10 20 30 40 50 60 70 80 90 5 10 1 20 2 40 45 Clotting activity (%) Gammacarboxy protein(%C ) 50 50 n=67 n 6 45 r 0648 wer4mdeusngr=a 0-793 v~p0OO ~-S 0 <00 40 p<0*OOOl0 4- 35 20 ~~~~~~~~20 * * *i 15 1 * * 2 10 0 0 0 1020 3040 5060 7080 90 0 10 20 30 40 50 60 Clotting activity (%) Free protein S (%) Figure 4 Pairwoise correlation analyses ofclotting activity, PIVYA-II, protein C, and protein S. Correlation analyses were made using all data from the three groups, exceptfor the data after vitamin K administration. status and a low intake, which was frequently seen in normal breast fed neonates during the first days of life, caused a deficiency of pro- 300 -Control (n = 16) ..Low (n=45) tein C and protein S as well as hypopro- thrombinaemia. Therefore, we must take this 200 into consideration in assessing the activities of these proteins, especially in making a diag- S A nosis of congenital protein C or protein 100 deficiency. Moreover, mothers should be advised to suckle infants frequently in the first 0 1..2.3 days of life in order to increase the breast http://adc.bmj.com/ milk secretion.37 Days Figure 5 Cumulative breast milk intakefrom 0 to 3 days oflife; results are mean (SE). The milk intake in the group 1 Stenflo J, Fernlund P, Egan W, Roepstorff P, Vitamin K with the low valuesfor clotting activity is significantly lower dependent modifications of glutamic acid residues in pro- than that in the control group; *p<005, **p 13 Sills RH, Marlar RA, Montgomery RR, Deshpande GN, 25 Gellis SS, Lyon RA. The influence of the diet of the new- Humbert JR. Severe homozygous protein C deficiency. bom infant on the prothrombin index. Jf Pediatr 1941; Arch Dis Child: first published as 10.1136/adc.68.3_Spec_No.297 on 1 March 1993. Downloaded from JPediatr 1984;105:409-13. 13:495-502. 14 Seligsohn U, Bertina A, Abind M, Rubin L, Attias D, 26 Keenan WJ, Jewett T, Glueck HI. Role of feeding and vita- Zivelin A, Rapaport SI. Homozygous protein C de- min K in hypoprothrombinemia of the newbom. Am J ficiency manifested by massive venous thrombosis in Dis Child 1971;121:271-7. the newborn. N EnglJ_7Med 1984;310:559-62. 27 Von Kries R, Becker A, Gobel U. Vitamin K in the new- 15 Mahasandana C, Suvatte V, Marlar RA, Manco-Johnson bom: influence of nutritional factors on acarboxy- M, Jacobson L, Hathaway HE. Neonatal purpura fulmi- prothrombin detectability and factor II and VII clotting nans associated with homoxygous protein S deficiency. Lancet activity. EurJPediatr 1987;16:123-7. 1990;i:61-2. 28 Polack B, Pouzol P, Amiral J, et al. Protein C level at birth. 16 Matsuzaka T, Yoshinaga M, Tsuji Y. Prophylaxis of Thromb Haemost intracranial hemorrhage due to vitamin K deficiency in 1984;52:188-90. infants. Brain Dev 29 Schettini F, de Mattia D, Altomare M, et al. Postnatal 1987;9:305-8. development of protein C in full-term newboms. Acta 17 Motohara K, Kuroki Y, Kan H, Endo F, Matsuda I. Paediatr Detection of vitamin K deficiency by use of an enzyme- Scand 1985;74:226-9. linked immunosorbent assay for circulating abnormal 30 Karpatkin M, Mannucci P, Bhogal M, et al. Low protein C prothrombin. Pediatr Res 1985;19:354-7. in the neonatal period. BrJHaematol 1986;62:137-42. 18 Wakabayashi K, Sakata Y, Aoki N. Conformation-specific 31 Fair DS, Marlar RA, Levin EG. Human endothelial cells monoclonal antibodies to the calcium-induced structure synthesize protein S. Blood 1986;67:1168-71. of protein C. J Biol Chem 1986;261:11097-105. 32 Ogura M, Tanabe N, Nishioka J. Biosynthesis and secre- 19 Kamiya T, Sugihara T, Ogata K, et al. Inherited deficiency tion of functional protein S by a human megakaryocytic of protein S in a Japanese family with recurrent venous cell line (MEG-01). Blood 1987;70:301-6. thrombosis: a study of three generations. Blood 1986; 33 Dahlback B, Stenflo J. High molecular weight complex in 67:406-10. human plasma between vitamin K-dependent protein S 20 Takahashi H, Tatewaki W, Wada K, Shibata A. Plasma and complement component C4b-binding protein. Proc protein S in disseminated intravascular coagulation, liver Nad Acad Sci USA 1981;78:2512-6. disease, collagen disease, diabetes mellitus, and under 34 Malm J, Bennhagen R, Holmberg L, Dahlback B. Plasma oral anticoagulant therapy. Clin Chim Acta 1989;182: concentration of C4b-binding protein and vitamin K- 195-208. dependent protein A in term and preterm infants: low 21 Lane PA, Hathaway WE. Vitamin K in infancy. J Pediatr levels of protein S-C4b-binding protein complexes. Br Jf 1985;106:351-9. Haematol 1988;68:445-9. 22 Haroon Y, Shearer MJ, Rahim S, Gum WG, McEnery G, 35 Schwarz HP, Muntean W, Watzke H, Richter B, Griffin Barkhan P. The content of phylloquinone (vitamin K1) JH. Low total protein S antigen but high protein S in human milk, cow's milk and infant formula foods activity due to decreased C4b-binding protein in determined by high-performance liquid chromatography. neonates. Blood 1988;71:562-5. J Nutr 1982;112:1105. 36 Mikami S, Ueda N, Imanaka Y, Sugimoto M, Hashimoto 23 Motohara K, Matsukane I, Endo F, KiyotaY, Matsuda I. Re- K, Takamiya 0. PIVKA-protein C and PIVKA-protein S lationship ofmilk intake and vitamin K supplementation to in hemorrhagic diseas of the newborn (in Japanese). vitamin K status in newborns. Pediatrics 1989;84:90-3. Journal of Obstetrics and Gynaecology and Neonatal 24 Motohara K, Matsukane I, Endo F, Matsuda I, Kiyota Y, Hematology 1989;13:87-92. Takemoto J. Role of milk intake in early neonatal vitamin 37 Yamauchi Y, Yamanouchi I. Breast-feeding frequency dur- K deficiency (in Japanese). Nihon Shonikagakkai Zasshi ing the first 24 hours after birth in full term neonates. (Tokyo) 1987;91:1435-41. Pediatrics 1990;86: 171-5. http://adc.bmj.com/ on October 1, 2021 by guest. Protected copyright.