Maternal Administration of Vitamin K Does Not Improve the Coagulation Profile of Preterm Infants

(1)

Maternal

Administration

of Vitamin

K Does

Not

Improve

the Coagulation

Profile

of Preterm

Infants

Nadya

J. Kazzi,

MD,

Nestor

B. Ilagan,

MD, Keh-Chyang

Liang,

MD,

George

M. Kazzi,

MD, Ronald

L Poland,

MD,

Lucille

A. Grietsell,

MD,

Yukihiko

Fujii,

MD, and Yves

W. Brans,

MD

From the Departments of Pediatrics and Obstetrics and Gynecology, Wayne State University School of Medicine, Hutzel Hospital and Grace Hospital, Detroit, Michigan

ABSTRACT. The effect of maternal administration of

vitamin K1 on cord blood prothrombin time, activated

partial thromboplastin time, activity of factors II, VII, and X, and antigen levels of factors II and X in infants

<35 weeks’ gestation was evaluated. Pregnant women in

preterm labor were randomly assigned to receive 10 mg

of vitamin K1 intramuscularly or no injection. If delivery did not occur in 4 days, the dose of vitamin K1 was

repeated. Women who continued their pregnancy 4 days

beyond the second dose received 20 mg of vitamin K1

orally daily until the end of the 34th week of gestation.

The birth weights of infants ranged from 370 to 2550 g

and gestational age ranged from 22 to 34 weeks. The

prothrombin time, activated partial thromboplastin time,

factors II, VII, and X activity, and factors II and X antigen levels were not statistically different in either

group of infants. Intraventricular hemorrhage occurred

in 25 of 51 control infants and 25 of 47 vitamin K-treated infants. More control infants had grade III

intraventric-ular hemorrhage on day 1 (P = .032), but on day 3 and

14 oflife, the severity of intraventricular hemorrhage was

comparable in both groups. Infants in whom an intraven-tricular hemorrhage developed were significantly smaller,

younger, and more critically ill than infants without

intraventricular hemorrhage. Administration of vitamin

K1to pregnant women at <35 weeks’ gestation does not

improve the hemostatic defects nor does it reduce the

incidence or severity of intraventricular hemorrhage in

their infants. Pediatrics 1989;84:1045-1050; antenatal vi-tamin K1, vitamin K dependent coagulation factors, pre-term infant, intraventricukir hemorrhage.

Intraventricular hemorrhage continues to be a

major cause of neurodevelopmental morbidity and

mortality in infants <35 weeks’ gestation.’

Inter-ventions aimed at preventing or decreasing the

Received for publication Mar 9, 1989; accepted Apr 20, 1989. Reprint requests to (N.J.K.) Hutzel Hospital, 4707 St Antoine, Detroit, MI 48201.

PEDIATRICS (ISSN 0031 4005). Copyright C 1989 by the American Academy of Pediatrics.

severity of this problem have not been consistently

successful.2 Several investigators suggested that

de-fects in coagulation, particularly in critically ill

premature infants, may contribute to the

occur-rence, as well as the progression of intraventricular

hemorrhage. Some of these coagulation defects

improve following the administration of vitamin K,

after birth. There is, however, a lag period of 4 to

24 hours before this effect of the vitamin is noted,6

whereas echoencephalograms have documented the

occurrence of intraventricular hemorrhage as early

as 6 hours postnatally.7 Efforts aimed at prevention

of intraventricular hemorrhage have concentrated

on the antenatal period because many events during

the birth process, such as labor and fetal distress,

may affect the occurrence of these hemorrhages.

Two reports have been published to date

sug-gesting that maternal antenatal administration of

vitamin K, improves the prothrombin and partial

thromboplastin activities and reduces the incidence

and severity of intraventricular hemorrhage in the

infants.8’9

Prothrombin time and partial thromboplastin

time are nonspecific tests for coagulation activity.

Their values can be affected by a change in activity

or antigen level of several coagulation factors that

are not vitamin K dependent.’#{176} In this study, we evaluated specifically the effect of maternal

admin-istration of vitamin K, on the activities of vitamin

K-dependent factors, II, VII, and X and antigen

levels

of factors II and X in preterm neonates. The

incidence and severity of intraventricular hemor-rhage were also studied.

MATERIALS AND

METhODS

This prospective randomized trial was conducted

(2)

Hutzel Hospitals of the Detroit Medical Center

during 2 years, from December 1986 through

De-cember 1988. The study protocol was approved by

the Human Investigation Committee at Wayne

State University and by the Research Committees

at Grace and Hutzel Hospitals. Pregnant women in

premature labor or with premature rupture of the

membranes at <35 weeks’ gestation were recruited

for the study. Patients with diagnosed congenital

fetal anomalies or pregnancies complicated by a

bleeding diathesis, severe preeclampsia, or fetal

dis-tress necessitating imminent delivery at the time

of recruitment were excluded. After an informed

consent was obtained, pregnant women were

ran-domized to a treatment or control group by

consec-utively drawing cards preceded from a table of

random numbers. Women in the treatment group

received 10 mg (1 mL) of vitamin K1

(Aquamephy-ton; Merck, Sharpe and Dohme, West Point,

Penn-sylvania) intramuscularly. If delivery did not occur

within 4 days, the dose of vitamin K, was repeated.

If delivery did not occur 4 days after the second

dose, the women received 20 mg of vitamin K1 orally

daily until the end of the 34th week of gestation or

delivery, if it occurred earlier. Patients in the

con-trol group did not receive a placebo.

Chorioamnion-itis was diagnosed if any two of the following

ma-ternal findings were present: fever, an increase in

white blood cell count above the baseline value

obtained at the time of hospital admission, and abdominal tenderness.

At the time of delivery, cord blood was obtained

for measurements of prothrombin time, partial

thromboplastin time, activities of coagnlatio

fac-tors II, VII, X, and antigen levels of factors II and

x.

Cord blood samples were obtained by

venipunc-ture of the umbilical vein with an 18-gauge needle

after clamping the cord and before delivery of the

placenta. Blood was placed in glass citrated tubes, mixed gently, and immediately centrifuged; the

plasma was frozen at -20#{176}C for later analysis.

Plasma prothrombin time, activated partial

throm-boplastin time, and the coagulation factor activities

were determined in an Automated Coagulation

Laboratories (ACL-810, Instrumentation

Labora-tories, Lexington, Massachusetts) using reagents

(plasmas deficient in the measured factor and

con-trol plasmas). The antigen levels were determined

by immunoelectrophoresis (Laurell Rockett

technique”) using antisera from

Calbiochem-Behr-ing, La Jolla, California.

All infants received the standard dose of vitamin

K1 intramuscularly after birth (0.5 mg for infants

with birth weights s1000 g and 1.0 mg for infants

with birth weights >1000 g). Intraventricular

hem-orrhage was diagnosed by sonographic

examina-tions done on days 1, 3, and 14 whenever possible

and subsequently as clinically indicated. The

echoencephalograms were done at the bedside with

a portable real-time sector scanner (ATL Mark IV,

Advanced Technology Laboratories, Bothell,

Washington) using a 5-MHz transducer. Static

im-ages in various planes were obtained on x-ray films

which were reviewed by a radiologist unaware of

the infant’s treatment group. Papile et al

classification’2 was used for grading

intraventricu-lar hemorrhage. Autopsy results, when available,

were used for infants who died before an

echoen-cephalogram could be obtained. Apnea was defined

as cessation of breathing for greater than 15

sec-onds associated with bradycardia (heart rate <100

beats per minute). Patent ductus arteriosus was

diagnosed clinically (cardiac murmur, bounding

pulses, increasing respiratory support when

im-provement was expected) and confirmed by

two-dimensional echocardiography and Doppler studies.

Data relating to the various risk factors

associ-ated with intraventricular hemorrhage in low birth

weight infants were recorded prospectively for all

infants. Considering the incidence of

intraventric-ular hemorrhage in infants <35 weeks’ gestation to

be about 40% and hoping for a 50% reduction in

the incidence of intraventricular hemorrhage in the

treated group, we estimated the number of infants

required to detect such an effect to be 46 pairs

accepting an a error = .05 and 9 error =

.20.

The data were analyzed using nonparametric

tests because of the non-normal distribution of

many of the variables. The Mann-Whitney U test

was used to compare the coagulation profiles of the

vitamin K and control groups of infants. x2 test and

Fisher’s exact test were used for categorical

van-ables. A two-tailed P value of <.05 was required to

reject the null hypothesis, ie, that antenatal therapy

with vitamin K1 had no effect.

RESULTS

The study population consisted of 112 women

(56 in the vitamin K group and 56 in the control

group) in preterm labor or with premature rupture

of the membranes. There were 22 women (12 in the

vitamin K group and 10 in the control group) who were excluded because their pregnancies continued beyond 34 weeks of gestation. Women in both

groups were comparable with respect to

complica-tions during the pregnancy, prolonged rupture of

the membranes for more than 24 hours, the

devel-opment of chorioamnionitis, prenatal care, and race

distribution (Table 1). None of the women were

receiving long-term anticonvulsant therapy. In the

vitamin K group (n = 44), 11 women were receiving

vitamin K1 orally before delivery. The median

(3)

deliv-No. of Control Infants (n = 46)

No. of Vitamin K-Treated

Infants (n = 44)

* The median time between enrollment and delivery was 25 hours (range 2 to 672 hours)

for the control group and 41 hours (1 to 648 hours) for the vitamin K-treated group.

t Placenta praevia, abruptio placenta, chronic hypertension, preeclampsia, and diabetes.

ery was 41 hours (range 1 to 648 hours). There were

seven twin gestations in the vitamin K group and

one triplet and four twin gestations in the control

group. No maternal or neonatal complications were

noted from vitamin K, administration.

Some selected neonatal characteristics and

din-ical variables among infants in both groups are

given in Table 2. No significant differences were

found in birth weight, gestational age, gender

dis-tnibution, cord pH, route of delivery, and 5-minute

Apgar score 5. The need for mechanical

ventila-tion for >24 hours and the incidence of air leaks

were similar in both groups. Volume expanders with

or without vasopressors were used as frequently in

both groups to treat hypotension. Apneas and

bra-dycardias necessitating treatment with methylxan-thines occurred significantly more often in the

vi-tamin K group (P = .03). Median cord plasma

TABLE 1. Maternal Characteristics*

vitamin K1 levels were two and a half times as high

in the vitamin K group as compared with the

con-trol group (0.024 vs 0.010 ng/mL, P = .046).

The coagulation profiles ofboth groups of infants

are shown in Table 3. Cord blood samples were

obtained from 39 infants in the control group and

40

infants in the vitamin K group. The median

values for prothrombin time, activated partial

thromboplastin time, coagulation factors II, VII,

and X activities, and antigen levels were not

statis-tically different in the vitamin K group as compared

with the control group. Cord blood plasma samples

from 6 infants in the control group and 9 infants

in the vitamin K group failed to coagulate. It is

possible that some specimens had some microclots

that could not be detected with the naked eye. Some

of these specimens had a low factor II activity and!

or antigen levels. However, when these infants’ data

Race

Black 40 37

White 6 7

Prenatal care 38 38

Pregnancy complicationst 8 10

Premature rupture of membrane >24 h 34 25

Chorioamnionitis 17 ii

Multiple gestation 5 7

Antenatal corticosteroids 0 2

Tocolytic agents 21 18

Oral vitamin K, 11

TABLE

2.

Selected Neonatal Clinical Data

Clinical Data Control Infants (n=52)

Vitamin K-Treated Infants (n = 51)

Birth weight (median g [range]) 1520 (370-2340) 1490 (420-2550)

Gestational age (median wk [range]) 32 (23-34) 31 (22-34)

Appropriate for gestational age (No.) 45 48

Sex: male/female (No.) 26/26 26/25

Cord pH (median [range]) 7.37 (7.21-7.47) 7.34 (7.15-7.48)

Cord plasma vitamin K (median ng/ 0.010 (0.000_0.271)* 0.024 (0.000-6.798)

mL [range])

Delivery (No.)

Vaginal 34 33

Cesarean 18 18

5-mm Apgar score 5 (No.) 7 5

Mechanical ventilation >24 h (No.) 25 25

Air leaks (No.) 1 3

Volume expansion (No.) 20 17

Apneas and bradycardias (No.) 5* 14

Patent ductus arteriosus (No.) 3 5

(4)

TABLE

3.

Coagulation Profile*

Coagulation Factor P Value

Prothrombin time (s) 17.1 (10.2-NC)

Activated partial thromboplastin time (s) 70.6 (36.0-NC)

Factor II activity (%) 20.5 (0-45.0)

Factor II antigen (%) 27.5 (0-58.0)

Factor VII activity (%) 45.0 (6.0-273.0)

Factor X activity (%) 28.0 (8.0-120.0)

Factor X antigen (%) 49.0 (0-350.0)

.937 .448 .269 .911 .842 .567 .397

* Values are expressed as percentages of the normal adult subject, reference plasma =

100%. NC, noncoagulable. Statistical significance determined by Mann-Whitney U test.

Control Infants Vitamin

K-(n = 39) Treated

Infants (n = 40)

17.7 (9.7-NC) 65.6 (30-NC) 25.5 (1.0-42.0) 29.0 (13.0-52.0) 46.5 (13.0-254.0) 34.5 (5.0-87.0) 55.0 (27.0-82.0)

were excluded from the data analysis, the lack of

difference between the two groups persisted.

Neither echoencephalogram nor autopsy were

available in one control and 4 vitamin K-treated

neonates. Intraventricular hemorrhage was

diag-nosed in 25 of 51 control infants and 25 of 47

vitamin K-treated infants (P = 1.00). The severity

of intraventricular hemorrhage in the two groups is

shown in Table 4. More infants in the control group

had grade III intraventricular hemorrhage on day 1

(P = .032), but the severity of intraventricular

hemorrhage was subsequently comparable in both

groups of infants. The coagulation profile of infants

in whom intraventricular hemorrhage developed

within each ofthe study groups was not statistically

different from the coagulation profile of infants in

whom intraventricular hemorrhage did not develop.

However, infants in whom intraventricular

hem-orrhage developed were significantly smaller,

younger, and required more ventilatory support and

volume expansion than infants without

intraven-tricular hemorrhage (Table 5).

DISCUSSION

The results of the present study suggest that

administration of vitamin K, to pregnant women in

preterm labor does not improve prothrombin time

and partial thromboplastin time or increase the

activities and antigen levels of factors II, VII, and

x

in their neonates. Similarly, Larsen et al’3

re-ported that administration of vitamin K, to

preg-nant women at term did not affect the activities of

factors II, IX, and X in their infants.

Two recent investigations,8’9 however, have

indi-cated an improvement in prothrombin time, partial

thromboplastin time, and prothrombin activity in

preterm infants whose mothers received vitamin

K,. The prothrombin time and partial

thrombo-plastin time are nonspecific tests for the extrinsic

and intrinsic coagulation cascade, respectively.

Their values can be affected by the activity and/or

antigen level of several coagulation factors that are

TABLE 4.

Severity of Intraventric ular Hemorrhage*

Intraventricular Hemorrhage

Grade

Control Infants (n = 52)

Vitamin K-Treated Infants

(n = 51)

Day it Day 3 Day 14 Day 1 Day 3 Day 14

0 I II III IV

:1:

24 28 17

4 5 4

2 6 4

9 7 7

0 0 0

13 6 20

25 23 15

7 5 8

5 6 6

1 6 5

1 3 3

12 8 14

* Results are numbers of infants.

t

P < .05 for vitamin K-treated infants vs control infants.

:1:

Encephalogram not performed.

not dependent on vitamin K for their activity.’0

Furthermore, the wide range of accepted values for

prothrombin time and activated partial

thrombo-plastin time reported in preterm infants may not

reflect moderate changes in the activity or antigen

level of vitamin K-dependent coagulation factors.’4

The prothrombin time, activated partial

thrombo-plastin time, and factor II activity values reported

by Morales et al9 and Pomerance et al8 in their

study and control groups are well within the range

reported in sick preterm infants.

In term infants, activities of factors II, VII, IX,

and X as reflected by the thrombotest have been

shown to improve following postnatal

administra-tion of vitamin K,.6 In low birth weight infants,

vitamin K, administration has been shown to

im-prove the activity and antigen level of factor II

during the first week of life.’5 Vitamin K, has been

shown to cross the placenta in term pregnancies.’6

However, the cord blood plasma levels achieved

following parenteral administration of the vitamin

to pregnant women at term are several

hundred-fold lower than those observed following direct

administration of the vitamin to the neonate

post-natally.’7 These findings were confirmed in this

study by measurements of plasma vitamin K, levels

(5)

TABLE 5. Characteristics of Infants With and Without Intraventricular Hemorrhage

Characteristic Control Infants Vitamin K-Treated Infants

Intraventricular P Value Intraventricular Hemorrhage P Value Hemorrhage

Yes No Yes No

(n=25) (n=26) (n=25) (n=22)

Gestational age (median wk [range]) 30 32

(23-34) (27-34)

.017 31 32

(27-34) (26-34)

.091

Birth weight (median g [range]) 1330 1710

(370-2280) (840-2340)

.022 1290 1700

(720-2550) (800-2390)

.013

pH (median [range]) 7.37 7.38

(7.21-7.47) (7.23-7.46)

.511 7.33 7.34 (7.15-7.48) (7.27-7.44)

.399

Route of delivery (No.)

Vaginal 16 18 17 15

Cesarean 9 8 .761 8 7 .753

Ventilation (No.) 16 8 .022 16 7 .007

Volume expansion (No.) 15 4 .001 13 2 .0005

Patent ductus arteriosus (No.) 2 1 .555 4 1 .158

Apneas and bradycardias (No.) 3 2 .640 12 2 .0014

to those reported in healthy term infants at birth by other investigators.’6”9

The median interval between the first dose of

vitamin K, and delivery was 41 hours, including 11

mothers who were receiving oral vitamin K, before

delivery. Thus, according to Wefring,6 there was

ample time for vitamin K, to cross the placenta and

exert its effects on the vitamin K-dependent

pre-cursor proteins. The discrepancy between antigen

and activity levels for factors II and X observed

suggests that the higher plasma vitamin K levels

attained in the vitamin K group of infants were

inadequate to improve the activities of these

fac-tors. Alternatively, the hepatic microsomal enzyme

system responsible for the activation and synthesis

of vitamin K precursor proteins may have been

immature and unable to respond adequately. In

some infants, we observed a double peak for factor

II antigen on immunoelectrophoresis, indicating

the presence of two prothrombin molecules with

markedly different electrophoretic mobility. It is

possible that one of these factor II proteins was

dysfunctional, similar to previously described “fetal

fibrinogen.”#{176} The median values for the antigens

and activities of vitamin K-dependent factors II, VII, and X in our control population were

compa-rable to those reported by other investigators in

infants of similar gestational age, birth weight, and

severity of illness.’4’’ The maturation of the

vita-mm K-dependent coagulation factors has been

shown to improve with advancing gestation and

postnatal age.2’ The mean values for the vitamin

K-dependent coagulation factors observed in

pre-term infants at birth are between 25% and 70% of

adult values. By 6 months of age, most of these

coagulation factors achieve adult levels. This

mat-uration process does not seem to be accelerated or

induced by the administration of vitamin K, as

observed in this study.

Vitamin K functions as a cofactor for the

con-version of the precursor proteins of factors II, VII, IX, and X into proteins with coagulant activity.10

There is no reference in the literature indicating

the tissue or plasma vitamin K levels at which the carboxylation of precursor proteins occurs in

tie-sues. Mean plasma vitamin K levels reported in

normal adults range between 0.2 and 0.3 ng/mL.’6

Similar plasma levels are achieved in normal term

infants within the first week of life.

Measure-ments of hepatic tissue content of vitamin K both

in infants and adults indicate that small amounts

of the vitamin are stored in the liver.ss

Maternal administration of vitamin K did not

reduce the incidence nor the severity of

intraven-tricular hemorrhage in their preterm infants. The

incidence of intraventricular hemorrhage was ap-proximately 48% in neonates of both groups as

might be expected. The more frequent occurrence

ofgrade III intraventricular hemorrhage on the first day of life in the control group of infants suggests

that maternal administration of vitamin K may

provide a transient protective effect which later on

is offset by other facthrs that come into play in the

etiology of intraventricular hemorrhage. The

mech-anism of such a protection is unclear because

ma-ternal administration of vitamin K failed to

im-prove the activities of the vitamin K-dependent

coagulation factors in preterm infants. Premature

infants in whom an intraventricular hemorrhage

developed within each of the experimental groups

were younger, smaller, and more ill as compared

with infants without intraventricular hemorrhage.

The lack of correlation between the activity and

(6)

occurrence of intraventricular hemorrhage

mdi-cates that hemostatic defects normally present in

preterm infants do not play a major role in the

etiology of intraventricular hemorrhage. Three

other plasma proteins activated by vitamin K

in-dude proteins C, 5, and Z. When activated, protein

C, in the presence of protein 5, has a potent

anti-coagulant effect. The function of protein Z has not

been established.24 Several vitamin K-dependent

proteins other than plasma coagulation proteins

have been reported in several tissues of the body

including kidney, bone, spleen, pancreas, lung,

pla-centa, testes, thyroid, thymus, cartilage, and

uterns.’ Such a protein has not been identified in

brain tissue or its cerebral vessels. Thus, maternal

administration ofvitamin K could not have affected

the occurrence of intraventricular hemorrhage for

other reasons than improvement in the activities of

vitamin K-dependent coagulation factors measured

in this study.

In previous investigations,8’9

echoencephalo-grams performed to diagnose intraventricular

hem-orrhage were obtained during the first 72 to 96

hours of life. Even though the great majority of

intracranial hemorrhages in preterm infants occur

by 96 hours of life,’26 intraventricular hemorrhage

develops in 8% of these infants between the 4th and the 14th day of life. Furthermore, progression

of the severity of intraventricular hemorrhage

be-yond the first 96 hours and until day 14 of life has

been documented by several investigators.7’24 In

this study, intraventricular hemorrhage was

diag-nosed after the first 72 hours oflife in 4 of9l (4.4%)

surviving infants and progression of

intraventricu-lar hemorrhage was noted in 5 of 91 (5.5%) infants.

Thus, reports of incidence and severity of

intraven-tricular hemorrhage should include

echoencephal-ogram results obtained after the first 96 hours of

life. Failure to do so may have contributed to the

more optimistic results of other investigators.

ACKNOWLEDGMENT

This work was supported, in part, by a grant from

Children’s Hospital of Michigan Research Committee,

Detroit.

REFERENCES

1. Papile LA, Munsick.Bruno G, Schaefer A. Relationships of cerebral intraventricular hemorrhage and early childhood neurologic handicaps. J Pediatr. 1983;103:273-277

2. Ment LR, Ehrenkranz RA, Duncan CC. Intraventricular hemorrhage of the preterin neonate: prevention studies.

Semin PerinatoL 1988;12:359-372

3, Gray OP, Ackerman A, Fraser AJ. Intracranial hemorrhage and clotting defects in low birth weight infants. Lancet.

1968;1:545-548

4. McDonald MM, Johnson ML, Rumack CM, et a!. Role of

coagulopathy in newborn20intracranial hemorrhage. Pedi-atrics. 1984;74:26-31

5. Beverley DW, Chance GW, Inwood MJ, et a!. Intraventric-ular haemorrhage and haemostasis defects. Arch Dis Child J. 1984;59:444-448

6. Wefring KW. Hemorrhage in the newborn and vitamin K prophylaxis. J Pediatr. 1962;61:686-692

7. Dolfin T, Skidmore MB, Fong KW, et al. Incidence, severity, and timing of subendymal and intraventricular hemorrhages in preterm infants born in a perinatal unit as detected by serial real time ultrasound. Pediatrics. i983;71:54i-546 8. Pomerance JJ, Teal JG, Gogolok JF, et al. Maternally

administered antenatal vitamin K1: effect on neonatal

pro-thrombin activity, partial thromboplastin time, and intra-ventricular hemorrhage. Obstet GynecoL 1987;70:295-299 9. Morales WJ, Angel JL, O’Brien WF, et al. The use of

antenatal vitamin K in the prevention of early neonatal intraventricular hemorrhage. Am J Obstet Gynecol.

1988;159:774-779

10. Hathaway WE, Bonnar J. Perinatal Coagulation. New York, NY: Grune & Stratton; 1978:53-80

11. Laurell CB. Quantitative estimation of proteins by

electro-phoresis in agarose gel containing antibodies. Anal Biochem..

1966;15:45-52

12. Papile LA, Burstein J, Burstein R. Incidence and evolution of subependymal and intraventricular hemorrhage: a study

of infants with birth weights less than 1500 gm. J Pediatr.

1978;92:529-534

13. Larsen IF, Jacobsen B, Holm HH, et al. Intrauterine

injec-tion of vitamin K before the delivery during anticoagulant

therapy of the mother. Acts Obstet Gynecol Scand.

1978;57:227-230

14. Barnard DR, Simmons MA, Hathaway WE. Coagulation studies in extremely premature infants. Pediatr Res. 1979;13:1330-1335

15. Ogata T, Motohara K, Endo F, et al. Vitamin K effect in

low birth weight infants. Pediatrics. 1988;81:423-427

16. Shearer MJ, Barkhan P, Rahim S, et aL Plasma vitamin K1 in mothers and their newborn babies. Lancet. 1982;2:460-463

17. McNinch AW, Upton C, Samuels M, et al. Plasma concen-trations after oral or20intramuscular vitamin K1 in

neo-nates. Arch Dis Child. 1985;60:814-818

18. Kazzi NJ, Ilagan NB, Liang KC et al. Transfer of vitamin

K1 across the placental barrier in preterm pregnancies. Ob-stet GynecoL In press

19. Hiraike H, Kimura M, Itokawa Y. Distribution of K vita-mine (phylloquinone and menaquinones) in human placenta and maternal and umbilical cord plasma. Am J Obstet Gyn-ecoL 1988;158:564-569

20. Witt I, Muller H, Kunzer W. Evidence for the existence of foetal fibrinogen. Thromb Dio,th Haemorrh. 1969;22:101-109

21. Andrew M, Pars B, Milner R, et al. Development of the human coagulation system in the healthy premature infant.

Blood. 1988;72:1651-1657

22. Greer FR, Mummah-Schendel LL, Marshall S, Suttie 1W. Vitamin K1 (phyiloquinone) and vitamin K2 (menaquinone)

status in newborns during the first week of life. Pediatrics.

1988;81:137-140

23. Khayata S, Kindberg C, Greer FR, Suttie 1W. Vitamin K1 and vitamin K2 in infant human liver. Pediatr Res.

1988;23:486A. Abstract

24. Partridge JC, Babcock DS, Steichen JJ et al. Optimal timing for diagnostic cranial ultrasound in low birth-weight infants: detection of intracranial hemorrhage and ventricular dila-tion. J Pediatr. 1983;102:281-287

25. Perlman JM, Volpe JJ. Intraventricular hemorrhage in

ex-tremely small premature infants. Am J Diii Child. 1986;140:1122-1124

(7)

1989;84;1045

Pediatrics

Lucille A. Grietsell, Yukihiko Fujii and Yves W. Brans

Nadya J. Kazzi, Nestor B. Ilagan, Keh-Chyang Liang, George M. Kazzi, Ronald L. Poland,

Preterm Infants