,

it:

3. _ i * , ,,

 

This is to certify that the
thesis entitled

Antecedents and Outcomes of Cerebral Ventricular
Enlargement in the Absence of Prior Genninal Matrix -
lntraventricular Hemorrhage in Low Birth Weight Infants

presented by

Isoken Nicholas Olomu

has been accepted towards fulfillment
of the requirements for the

Master of Science degree in Epidemiology

 

 

 

Major Professor’s Signature

 

/2‘ /?f/)a
’ 7
Date

MSU is an Affinnative Action/Equal Opportunity Employer

 

LIBRARY
Michigan State
University

 

 

 

 

 

 

 

 

 

 

 

 

 

PLACE IN RETURN BOX to remove this Checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:/Proi/Acc&Pres/ClRC/Date0ueindd

ANTECEDENTS AND OUTCOMES OF CEREBRAL VENTRICULAR
ENLARGEMENT IN THE ABSENCE OF PRIOR GERMINAL MATRIX —
INTRAVENTRICULAR HEMORRHAGE IN LOW BIRTH WEIGHT INFANTS

By

Isoken Nicholas Olomu

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTERS OF SCIENCE

EPIDEMIOLOGY

2010

ABSTRACT

ANTECEDENTS AND OUTCOMES OF CEREBRAL VENTRICULAR
ENLARGEMENT IN THE ABSENCE OF PRIOR GERMINAL MATRIX -
INTRAVENTRICULAR HEMORRHAGE IN LOW BIRTH WEIGHT INFANTS

By
Isoken Nicholas Olomu

Background and methods: Cerebral ventricular enlargement in low birth weight
infants predicts poor neurodevelopmental outcome. The etiology or outcomes of
ventricular enlargement (VE) in absence of intraventricular hemorrhage (lVH) are
unclear. In the neonatal brain hemorrhage (NBH) study, timed, cranial ultrasound
(US) scans were performed in 1,105 infants with birth weights $2 kg. We
searched the NBH database to determine the prevalence, etiology and outcomes
of infants with VE.

Results: VE was identified in 23 of 1088 (2.1%) infants with US scans. Infants
with VE had similar gestational age, birth weight and head circumference as
infants with normal scans. Mothers of VE infants were more likely to be
unemployed and to use drugs. VE and VE+HE were significantly associated with
increased risk of overall mortality and of disabling cereme palsy in survivors.
Conclusion: Maternal social factors may play a role in V5; VE and VE+HE are

associated with increased risk of death and disabling cereme palsy.

DEDICATION

To my wife and friend Adesuwa, and our sons Nosa, Noma, Etin, Osa and IK,

for your love, patience, prayers and encouragement.

ACKNOWLEDGEMENTS

I am grateful to Nigel Paneth for allowing me access to the Neonatal Brain
Hemorrhage (NBH) Study data and for his help and guidance as Chairman of my
thesis committee. I am also grateful to Claudia HoIzman and David Todem for
serving on my thesis committee. I appreciate and am grateful for the help I
received from Ting Hong and Steven Korzeniewski with data handling and

analysis.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................ vii
LIST OF FIGURES ............................................................................... ix
CHAPTER 1
VENTRICULAR ENLARGEMENT IN THE FETUS AND LOW BIRTH WEIGHT
PRETERM INFANT ............................................................................... 1
Introduction ........................................................................................... 1
Ventricular Enlargement in the Fetus .......................................................... 2
Definition of Fetal Ventricular Enlargement ......................................... 2
Epidemiology and Neurologic Outcome of Fetal Ventricular
Enlargement ............................................................................ 5
Ventricular Enlargement in Low Birth Weight Infants ..................................... 8
Post Hemorrhagic Ventricular Enlargement in Low Birth Weight |nfants....8
Normal Pressure Ventricular Enlargement ............................... 11
Post Hemorrhagic Hydrocephalus .......................................... 15
Ventricular Enlargement Without Prior GM-IVH ................................. 17
CHAPTER 2
i. THE NEONATAL BRAIN HEMORRHAGE STUDY ................................... 21
Brief Description .................................................................................. 21
Infant Enrolment and Study Sites ................................................... 22
Maternal interviews, Maternal and Neonatal Chart Abstraction .............. 23
Cranial Ultrasound Procedures and Interpretation .............................. 24
Head circumference measurements ................................................ 26
Follow—up Examination ................................................................ 26
Objective of Present Analysis ................................................................. 27
Suitability of the NBH Data Set for the Study of Isolated Ventricular
Enlargement ............................................................................... 29
ii. METHODS ....................................................................................... 30
Identification of Infants with Isolated Ventricular Enlargement and
Selection of Comparison Group From the NBH Dataset ................ 30
Infant Classification .................................................................... 30
Data Collection .......................................................................... 31
Statistical Methods ...................................................................... 31

CHAPTER 3
RESULTS - ANTECEDENTS AND OUTCOMES OF ISOLATED
VENTRICULAR ENLARGEMENT IN THE NEONATAL HEMORRHAGE (NBH)

STUDY .............................................................................................. 33
Characteristics of Cohorts by Cranial Ultrasound Findings ................... 36
Rate of Detection of Ventricular Enlargement in the VE and VE+HE
Groups ..................................................................................... 38
Maternal Characteristics ............................................................... 39
Mortality and Neurodevelopmental Outcomes at 2years ..................... 45
CHAPTER 4

VENTRICULAR ENLARGEMENT, PERIVENTRICULAR WHITE MATTER
INJURY AND NEUROLOGIC OUTCOMES IN THE NEWBORN HEMORRHAGE

(NBH) STUDY —- DISCUSSION ............................................................... 49
Summary of Findings .................................................................... 49
Significance of VE in LBW Infants ................................................... 50
Frequency of Isolated Ventricular Enlargement in Low Birth Weight
Infants ....................................................................................... 51
Intrauterine Growth and VE ........................................................... 52
Apgar Score and VE .................................................................... 53
Hypertensive Disorders of Pregnancy and VE ................................... 54
Chorioamnionitis and VE ............................................................... 56
Timing of VE in LBW Infants .......................................................... 56
Neonatal and Two Year Outcomes of Infants with VE ......................... 58

BIBLIOGRAPHY .................................................................................. 61

vi

Table 1.1

Table 1.2

Table 1.3

Table 1.4

Table 1.5

Table 1.6

Table 1.7

Table 1.8

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 3.5

Table 3.6

Table 3.7

LIST OF TABLES

Ventricular width according to gestational age in 427 normal
pregnancies ........................................................................ 3

Lateral Ventricular Atrium Measurements in Different
Populations of Fetuses Without Abnormalities ............................ 4

Neurologic outcome of isolated mild fetal ventriculomegaly ........... 7

Papile Classification of Genninal Matrix-lntraventricular
Hemorrhage in Premature Infants ............................................ 9

Grading of Severity of Gerrninal Matrix-lntraventricular
Hemorrhage by Ultrasound Scan ('Volpe classification’) ............... 9

Frequency of normal pressure ventricular enlargement in low
birth weight infants with germinal matrix — intraventricular

hemorrhage ...................................................................... 14
Frequency of GM-IVH, GM-IVH associated mortality and PHH
following GM-IVH in very low birth weight infants ...................... 16

Prevalence of ventricularenlargement in absence of germinal
matrix - intraventricular hemorrhage ...................................... 18

Frequency distribution of infants in the NBH study by cranial
ultrasound results .............................................................. 33

Some characteristics of infants with ventricular enlargement in the
absence of germinal matrix intraventricular hemorrhage (VE) ...... 34

Some characteristics of infants with initial IVE, but later developed

germinal matrix-intraventricular hemorrhage ............................ 35
Low birth weight infants with isolated ventricular enlargement:
comparison of singletons and products of twin gestations ........... 36
Characteristics of all cohort infants with IVE, VE+HE, HE and
Normal CUS scans ............................................................. 37
Frequency of isolated ventricular enlargement (VE) and ventricular
enlargement associated with GM-IVH (VE+HE) at each CUS ............. 39
Maternal characteristics: demographic data ............................. 41

vii

Table 3.8

Table 3.9

Table 3.10

Table 3.1 1

Table 3.12

Table 3.13

Table 3.4

Maternal characteristics: medical and pregnancy complications
(entire cohort) .................................................................... 42

Maternal characteristics: mode of delivery and delivery
complications (entire cohort) ................................................. 43

Unadjusted odds ratios (OR) and 95% confidence intervals (CI) for
the influence of different pregnancy and delivery characteristics on

the occurrence of VE and VE+HE compared to infants with Normal
CUS scans ....................................................................... 44

Adjusted odds ratios (OR) and 95% confidence intervals (CI) for the
influence of different pregnancy and delivery characteristics on the
occurrence of VE and VE+HE compared to infants with Normal
CUS scans ........................................................................ 45

Mortality and neurodevelopmental outcomes at age 2years in
infants with ventricular enlargement (VE, VE+HE) compared with
infants with Normal CUS scans ............................................. 46

Unadjusted odds ratios (OR) and 95% confidence intervals (CI) for
the risk of outcome measures in infants with ventricular
enlargement (VE, VE+HE) compared to infants with normal CUS
scans .............................................................................. 47

Adjusted odds ratios (OR) and 95% confidence intervals (CI) for the
risk of outcome measures in infants with ventricular enlargement
(VE, VE+HE) compared to infants with normal CUS scans .......... 48

viii

Figure 1.1

Figure 1.2

Figure 1.3
Figure 2.1
Figure 2.2
Figure 3.1

LIST OF FIGURES

Ventricular enlargement in very low birth weight infants ............. 10

Natural history of germinal matrix — intraventricular hemorrhage

in very low birth weight infants .............................................. 12
Evolution of ventricular dilatation following GM-IVH .................. 13
New Jersey counties served by NICUs in the NBH study ........... 21
Data collected in the neonatal period in the NBH study ............. 24
Prevalence of VE and VE+HE at each CUS scan ..................... 40

CHAPTER 1
VENTRICULAR ENLARGEMENT IN THE FETUS AND LOW BIRTH WEIGHT
PRETERM INFANT
Introduction
Cereme ventricular enlargement (VE) in very low birth weight (birth weight <
1,500 g) newborn infants is always cause for concern as it is often associated
with an increased risk of motor, sensory and cognitive impairments‘. VE often
occurs as a sequel to germinal matrix-intraventricular hemorrhage (GM-IVH)2,
which is the most common form of intracranial injury in very low birth weight
(VLBW) infants3. The germinal matrix is the site of hemorrhage in most cases of
GM-IVH‘; it is also the site of origin of oligodendrocytes and astrocytes
precursors", cells that are critical to normal cerebral white matter and cortical
development. However, while GM-IVH is an important risk factor for VE in VLBW
infants, ventricular enlargement is also seen in the absence of intraventricular
hemorrhage. Enlargement of the cerebral ventricles, whether or not associated
with GM-IVH, may reflect damage to the periventricular white matter and
adjacent neural tissues. Enlargement of the cereme ventricles also causes injury
to the periventricular microvasculature further increasing the risk of adverse
motor, sensory and cognitive outcomes in VLBW infants with GM—IVH.
In this introductory chapter, I will discuss:
1L Ventricular enlargement in the fetus

> Definition

> Epidemiology and neurologic outcomes of ventricular enlargement

in the fetus

4 Ventricular enlargement in the VLBW infant
> Post hemorrhagic ventricular enlargement
9 ‘Normal pressure’ ventricular enlargement
9 Post hemorrhagic hydrocephalus
> Ventricular enlargement without prior GM-IVH (‘isolated ventricular

enlargement)

Ventricular enlargement in the fetus

Enlargement of one or both lateral ventricles is one of the most common fetal
abnormalities detected by prenatal ultrasound 6. Therefore assessment of the
size of the fetal lateral cerebral ventricles is an important part of routine prenatal
ultrasound evaluation of the developing fetus. F urthermore, 3% to 12% of
fetuses with VE have chromosomal anomalies while 30% to 60% have other
anomalies 6. Detection of fetal VE therefore, calls for careful evaluation of the
fetus for other congenital anomalies, and for investigation of the fetus and mother

for possible explanations.

Definition of fetal ventricular enlargement
The width of the atria of the lateral ventricles measured in the axial plane remains

the easiest and most reproducible method of detecting fetal VE7. The lateral

ventricles are relatively large compared to the size of the cereme hemispheres
during late embryonic and early fetal life; however, the lateral ventricular width
relative to that of the cerebral hemispheres decreases from approximately 70%
at 15 weeks gestation to about 30% at 20 weeks. The absolute diameter of the
lateral ventricles at the level of the atrium remains stable through most of second

and third trimesters of pregnancy 8' 9 (Table 1.1).

Table 1.1: Ventricular width according to gestational age in 427 normal
pregnancies (Almog et al, 20039)

 

 

 

 

Gestational Number Ventricular Width (mm)
Age (weeks) of Percentile
Cases Mean (SD) 5‘“ 95IF 4SD>Mean

20-21 38 5.89 (1.12) 4.79 8.50 10.37
22-23 106 5.66 (0.72) 4.50 7.06 8.54
24-25 77 6.04 (0.97) 4.60 7.77 9.92
26-27 35 6.01 (1.21) 3.90 9.00 10.85
28-29 52 6.38 (1.26) 4.35 9.11 11.42
30-31 30 6.57 (1.02) 4.28 7.80 10.65
32—33 26 6.77 (1.46) 4.20 8.90 12.60
34-35 31 7.22 (1.12) 5.46 9.28 11.71
36-40 32 6.92 (1.50) 4.00 9.80 12.92

 

 

 

 

 

 

 

 

A measurement of 10 mm, which represents 2.5 to 4 SDs above the mean, is
accepted as the upper limit of normal 7' 1°' 1‘. In a prospective cohort study to

establish the range of cereme ventricular size in normal gestation, Almog et aI9

measured fetal ventricular size at the level of the atrium in utero, in 427
pregnancies at gestational ages ranging from 20 to 40 weeks and found a mean
ventricular width of 6.2 :t 1.2 mm. The ventricular size did not change significantly
from 20 to 40 weeks of gestation (Table 1.1). The authors, in a review of 8
previous studies, reanalyzed additional 7789 fetal ventricular measurements and
found that the pooled mean of 6.4 mm 1 3 SDs correspond to 10 mm, the widely

accepted upper limit of normal fetal ventricular atrial diameter (T able 1.2).

Table 1.2: Lateral Ventricular Atrium Measurements in Different Populations of
Fetuses Wrthout Abnormalities 9

 

 

 

Authors Number Ventricular Vtfidth (mm)
(Year) of Cases Mean (SD) 380 > Mean 480 > Mean
Hilpert et al (1995) 608 6.5 (1.5) 11.0 12.5
Cardoza et al (1988) 100 7.6 (0.6) 9.4 10.0
PiIu et al (1989) 171 6.9 (1.3) 10.8 12.1
Heiserrnan et al (1991) 52 6.5 (1 .3) 10.4 11.7
Achiron et al (1993) 5400 6.6 (1.2) 10.2 11.4
Patel et al (1995) 219 6.1 (1 .3) 10.0 11.3
Farrell et al (1994) 739 5.4 (1 .2) 9.0 10.2
Alagappan et al (1994) 500 6.6 (1.4) 10.8 12.2
Almog et al (2003) 427 6.2 (1 .2) 9.8 11.0
Calculated Average 8216 6.4 (1 .2) 10.0 11.2

 

 

 

 

 

 

 

Ventricular enlargement in the fetus is therefore defined as a transverse diameter
of fire lateral ventricle, measured at the level of the ventricular atrium, which is
mter than 10 mm. Fetal ventricular enlargement is mild when the atrial width is
10 - 15 mm; and severe when greater than 15 mm.

A ventricular atrial width greater than 10 mm at the time of delivery of a LBW,
preterm infant may represent fetal ventricular enlargement. Ventricular
enlargement without evidence of prior GM-IVH is reported in up to 16% of LBW
infants delivered at or less than 34 weeks gestation 12‘ ‘3. Infants with ventricular
enlargement on initial scans in these studies were delivered at significantly
earlier gestational ages and were of smaller birth weights. Ventricular
enlargement detected in the immediate postnatal period in LBW infants likely
represents fetal ventricular enlargement in some cases. The prevalence of
ventricular enlargement reported in these studies however appear higher than
that in earlier studies described below; this may be because the studies below

defined ventricular enlargement as an axial diameter greater than 8 mm“ 13.

Epidemiology and neurologic outcome of fetal ventricular enlargement
Mild fetal ventricular enlargement is the most common abnormality found on
prenatal ultrasound scans 6. The reported prevalence vary widely: 1.48 per
1000 births in a low risk population, 22 per 1000 births in a high-risk population
1" 1‘ and about 7 per 1000 in unselected pregnancies 3.

Mild fetal ventricular enlargement is described as ‘isolated’ when no associated

cerebral or extracranial ultrasound abnormalities are detected at the time of initial

presentation; however, cases may subsequently be shown to have other
anomalies. Isolated mild fetal ventricular enlargement may be unilateral, bilateral
or asymmetrical; it is mostly non-progressive and the fetal head is usually of
normal size. Isolated mild fetal ventricular enlargement on prenatal ultrasound is
often a dynamic process - about 30% resolve spontaneously, 55% - 60% remain
stable and in about 10% -15% of cases, progressive ventricular enlargement
develops 1525.
The neurologic outcomes in infants with isolated mild fetal ventricular
enlargement have only been reported in case series, case reports and a couple
of case control studies. The studies involved small numbers of infants and
utilized mostly non-standardized methods of developmental assessments done
at widely varying postnatal ages. A systematic review of such studies with a total
of 577 cases, found an overall survival rate of 92.7%. Of 485 cases followed, 413
(85.2%) were developmentally normal, 38 (7.8%) had mild developmental delay
and 34 (7.0%) had moderate or severe delay 26. Another review of studies with
more than 20 survivors to follow-up indicated the overall incidence of
developmental delay range from 0% to 36%; 88.2% of the evaluated infants had
normal development [Table 1.3] 6. The neurologic outcome is more favorable

when fetal ventricular enlargement is unilateral 22 and when there is prenatal

resolution 23' 27. However, the risk of developmental delay is higher in infants with

Table 1.3: Neurologic outcome of isolated mild fetal ventriculomegaly [adapted
from Kelly et al 61

 

 

Age at
Authors IMV Survived Followed Screening Testing Normal
(Year) (n) (n) (n) Test (Months) (rt/o)”
Arora et al 30 27 27 DDS 1.5-72 19 (70)
(1 998)
Bloom et al 62 60 22 BSID 21 61:17.4 14 (64)
(1 997)
Bromley et 27 26 26 Unknown 3-18 21 (81)
al (1 991 )
Lipitza et aj 26 26 26 NA Unknown 25 (96)
(1 998)
Patel et al 42 37 34 Unknown 1.5-70.3 28 (82)
(1994)
Pilu et al 27 27 21 Interview / 21-72 21 (100)
(1999) Phone
Robson et 50 37 37 NA Newborn 37 (100)
al (1998)
Vergani et 45 45 45 M-C&G 3-72 45 (100)
al (1 998)
Totals 309 285 238 -— 0-72 210 (88)

 

 

 

 

 

 

 

 

 

a Unilateral ventriculomegaly; b Percentage of cases followed;

IMV — Isolated mild ventriculomegaly, BSID - Bayley Scales of Infant Development;
DDS - Denver Development Screening; M-C&G - Milani-Comparetti &Gidon;
NA - Not available

atrial width greater than 12 mm 2‘ and when other anomalies are detected on
subsequent evaluations 22' 23.

Ventricular enlargement in low birth weight infants

Enlargement of the cerebral ventricles is present in 10 to 30% of VLBW infants
evaluated with cranial ultrasound or magnetic resonance imaging in the first 2 to
4 days of life 1:12:13'28. An even higher proportion of VLBW infants have
ventricular enlargement when imaging studies are repeated at term equivalent
age ‘2' ‘3' 29. In up to 50% of these infants, no radiologic evidence of prior GM-IVH
is demonstrable. Factors associated with increased risk of VE in VLBW infants
include extreme prematurity (28 weeks gestation or less), severe intraventricular
hemorrhage, chronic lung disease and white matter injury ”'13. Ventricular

enlargement in low birth infants may or may not be preceded by germinal matrix

— intraventricular hemorrhage (GM-IVH).

Post hemorrhagic ventricular enlargement in low birth weight infants
Enlargement of the cereme ventricle in the low birth weight infant occurs most
often in infants with GM-IVH. A direct relationship exists between the severity of
GM-IVH and the risk of ventricular enlargement; infants with moderate to severe
GM-IVH have a higher risk of ventricular enlargement than infants with minor or
no bleeds. While there is currently no universally accepted or satisfactory
classification of the severity of GM-IVH, the Papile and Volpe classifications are

rriost commonl y used (Tables 1.4 8. 1.5).

Table 1.4: Papile Classification of Gerrninal Matrix-lntraventricular Hemorrhage

in Premature Infants3°

 

 

SEVERITY DESCRIPTION
Grade 1 Gerrninal matrix hemorrhage
Grade 2 Gerrninal matrix-intraventriventricular hemorrhage with

blood within the lateral ventricles but no distention
Grade 3 Gerrninal matrix-intraventriventricular hemorrhage with
blood filling and distending the lateral ventricles
Grade 4 Gerrninal matrix-intraventriventricular hemorrhage with

parenchymal ‘extension’

 

 

 

 

Table 1.5: Grading of Severity of Gerrninal Matrix-lntraventricular Hemorrhage

by Ultrasound Scan (‘Volpe classification’)31

 

SEVERITY DESCRIPTION

 

Grade 1 Gerrninal matrix hemorrhage with no or minimal intraventricular
hemorrhage (10% of ventricular area on parasagittal view)
Grade 2 lntraventricular hemorrhage (10%-50% of ventricular area on
parasagittal view)

Grade 3 lntraventricular hemorrhage (>50% of ventricular area on

parasagittal, view usually distends lateral ventricle)

 

Separate Periventricular echodensity (location and extent)

notation

 

 

 

Figure 1.1: Ventricular enlargement in very low birth weight infants

 

 

Very Low Birth Weight Infants

 

 

 

 

 

 

 

V 1
Gerrninal Matrix- No Gerrninal Matrix-
lntraventricular Hemorrhage lntraventricular Hemorrhage

 

 

 

 

 

 

 

 

 

 

 

 

 

II

Go to Early Ventricular Late Ventricular
Figure 2 Enlargement Enlargement

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The natural history of ventricular enlargement following GM—IVH has been well
characterized (Figure 1.2) and extensively reviewed 31‘33. Infants with moderate
to severe GM-IVH may experience an immediate, transient distension of the
lateral ventricles by intraventricular blood — the so-called ‘grade III IVH’ on the
Papile classification 3°; this may or may not be followed by progressive
ventricular dilatation within days to weeks.

In about 65% of cases, a period of progressive ventricular dilatation often ensues

after the initial diagnosis of GM-IVH; this is followed by spontaneous arrest and
subsequent complete or partial resolution of ventricular dilatation. In these cases
ventricular dilatation is not associated with excessive increase in occipito-frontal

circumference or increased intracranial pressure (‘normal pressure ventricular

enlargement’).

In another 30% of cases, an initial phase of slow ventricular dilatation is followed,
If! about 2 - 4weeks, by the development of post-hemorrhagic hydrocephalus

10

characterized by accelerated ventricular distension associated with raised
intracranial pressure and greater than normal growth in occipito-frontal
circumference requiring surgical and/or nonsurgical intervention for management.
Lastly, in about 5% of cases, rapidly progressive ventricular enlargement ensues
within days of the initial detection of GM-IVH in infants with the most severe

grades of hemorrhage.

‘Normal preesure’ ventricular enlargement

In her original autopsy study of intraventricular hemorrhage in preterm low birth
weight infants, Larroche found several infants with ventricular enlargement and
normal head growth 3". This phenomenon was later confirmed in vivo by several
investigators who demonstrated the presence of ventricular dilatation by CT scan
or ultrasonography prior to any increase in occipito-frontal head circumference or
the onset of clinical features indicative of raised intracranial pressure in infants
with intraventricular hemorrhage 3537. This phenomenon may occur after any
grade of intraventricular hemorrhage and probably results from partial obstruction
to the normal flow and absorption of cerebrospinal fluid (CSF) within the
ventricular system (Figure 1.3). The obstruction is probably caused in part by an
obliterative arachnoiditis affecting principally the posterior fossa; less commonly,
the obstruction may also be due to blood clot and other cellular debris 3‘.
Studies of the natural history of GM-IVH indicate that normal pressure ventricular

enlargement occur in 15% to 30% of LBW infants with GM-IVH (Table 1.6).

11

Figure 1.2: Natural history of germinal matrix - inrtraventricular hemorrhage in

very low birth weight infants

 

 

Gerrninal Matrix-
lntraventricular Hemorrhage

 

 

 

 

I
I
I
I
I

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

Transient Ventricular _' Progressive No Ventricular
Distension Ventricular Dilatation Dilatation
NoIlCP, No' OFC |
|
|
, ., I
Spontaneous Surgical or Non-Surgical Rapidly Progessive
Arrest Therapy Ventricular Dilatation

 

 

 

 

 

  

 

 

 

Partial or Complete Resolution of
Ventricular Enlargement

 

 

ICP- lntracranial pressure, OFC — Occipito-frontal circumference

 

Normal pressure ventricular enlargement (Figure 1.3) is thought to result from:
1) Increased compliance of the periventricular tissue due to the immature state of
development and possible prior hypoxic-ischemic or compressive injury to the

I'm mature brain.

12

Figure 1.3: Evolution of ventricular dilatation following GM-IVH [Modified from
Hill et al 198333]

 

l GM-IVH I

 

Obstruction to CSF Flow & Absorption

 

 

EXCESSIVE ACCUMULATION OF CSF'J'

 

Compliant eriventricular Tissues

 

VENTRICULOMEGALY

 

Large Suba chnoid Space
Decreased Brain rowth / Cereme Atrophy
White Matt Injury

 

l NORMAL PRESSURE HYDROCEPHALY

 

 

 

 

 

New Equilibrium Betwein Disequilibrium:
CSF Formation & Limits of Compliance
Absorption Exceeded
ARREST :l: RESOLUTION PROGRESSIVE VENTRICULAR
NORMAL TICP DILATATION, 'ICP, [*rcv

 

 

 

 

 

-CSF — Cerebrospinal fluid, TICP — lntracranial pressure, *ICV — lntracranial

volume

13

Table 1.6: Frequency of normal pressure ventricular enlargement in low birth

weight infants with germinal matrix - intraventricular hemorrhage

 

 

Author GM—IVH No Ventricular NP Ventricular
(Year) (n) enlargement (%) enlargement (%)
Hill et al (1981) 87 47 (54) 20 (23)
Ahmann et al (1980) 77 35 (45) 12 (16)
Dykes et (1989) 409 254 (62) 125 (30)
Levene et al (1981) 68 24/39 (62) 1 1/39 (28)
Enzmann et al (1985) 1 15 90 (78) 22 (19)

 

 

 

 

 

 

NP — normal pressure, GM-IVH — germinal matrix - intraventricular hemorrhage

2) The presence of a large subarachnoid space over the cerebral convexities in
preterm infants may also be contributory. The continued accumulation of CSF
within the ventricular system leads to progressive ventricular enlargement with
normal intracranial pressure (ICP) until the limits of periventricular compliance
are exceeded (Figure 1.3).

3) Ventricular enlargement associated with neonatal intracranial hemorrhage
may not be explained by obstruction to CSF flow and absorption alone.
Decreased brain growth and cereme atrophy as a result of injuries to the
cerebral white matter, cortical and deep cortical gray matter have been described
even in infants with uncomplicated 38 and mild grades (I & II) of IVH 39. The
ventricles enlarge to replace tissue loss from parenchymal brain injury, hence

ventricular enlargement.

14

Poet hemorrhagic hydrocephalus

Post hemorrhagic hydrocephalus (PHHC) has been defined as “the progressive
accumulation of cerebrospinal fluid (CSF) in the ventricles and/or subarachnoid
space resulting from obstruction to the normal CSF pathways directly due to
hemorrhage into the CSF space” 32. Thus, in PHHC the ventricles are distended
(not merely enlarged) by accumulation of CSF (not simply by large volume of
intraventricular blood). These changes may be accompanied by clinical signs of
raised intracranial pressure, excessively rapid increase in occipito-frontal head
circumference associated with rapidly progressive enlargement of the cerebral
ventricles necessitating medical and I or surgical intervention. These changes
may be delayed for days to weeks following the initial detection of intraventricular
hemorrhage. In a review of studies of GM-IVH in preterm low birth weight infants,
the rate of PHHC ranged from 3 - 15%; when all the studies were pooled, the
rate of PHHC following any GM-IVH was about 10% (Table 1.7).

Factors associated with increased risk of PHH in these studies include:

1) Gestational age of the infant at birth - the more premature the infant, the
higher the risk of posthemorrhagic ventricular enlargement;

2) Timing of the bleed — the earlier bleeds, especially bleeds in the first 24 hours
of life, tend to be more severe and are associated with higher rates of
posthemorrhagic ventricular enlargement.

3) The severity of initial bleed - this is the most important determinant of PHHC;
bleeds of grades III or worse are more likely to be complicated by PHHC than

grades I and Il bleeds.

15

Table 1.7: Frequency of GM-IVH, GM-IVH associated mortality and PHH

following GM-IVH in very low birth weight infants.

 

 

 

GM-IVH
Author Study & Infant GM-IVH Mortality
(Year) Characteristics Rate Rate *PHH Rate
Ahmann PA, N=191, GA < 35 wks, 77/191 22/77 8/77
(1980) Single center, (40%) (28%) (10%)
Smith WL et N=92, Bt wt < 2kg, Single 29/92 8/29 1/29
al, (1983) center (31%) (28%) (3%)
Ment LR, N: 438, Bt Wt 512509, 95/228** 27/133 5/95***
(1984) Single center, (42%) (20%) (5%)
Enzmann D, N=377, Bt Wt 515009, 1 15/377 19/1 15 22/377
(1985) Prospective, single center (30.5%) (16%) (6%)
Murphy BP, N=1 127, Bt Wt $15009, 248/1 127 54l248 37/248
(2002) MuIti-hospital study (22%) (22%) (15%)
Larroque B N=2667, GA 23-32wks, 613/2667 84/613 80l§fl
(2003) Multi-hospital study (23%) (14%) (13%)
Total 1177/4682 214/1 21 5 145/1439
(25%) (17%) (10%)

 

 

 

 

 

** - Survived first 36 postnatal hours

16

*** - Survived first postnatal week

 

Ventricular enlargement without prior GM-IVH (‘ieolated ventricular
enlargement)

Data from cranial ultrasound 4°, computed tomography scans 4‘, and more
recently, magnetic resonance imaging (MRI) studies ‘2 indicate that ventricular
enlargement occurs in preterm low birth weight infants without prior GM-IVH.

In a recent survey of very low birth weight infants born between 23 and 30 weeks
gestation, 119 infants were studied with serial MRI scans 12. Thirty-six infants
(30%) had ventricular enlargement identified on the initial scan carried out on the
second day of life, raising the possibility that some cases of ventricular
enlargement are prenatal in origin. Twenty of the 36 infants (55%) with
ventricular enlargement had no prior IVH. The proportion of infants with
ventricular enlargement increased from 30% to 39% (36/119 to 47/119) on the
final scan done at term equivalent age; 22 (47%) of these infants did not have
prior IVH ‘2 indicating that VE may also develop postnatally in VLBW infants in
the absence of a history or neuroradiographic evidence of GM-IVH. The
explanation for the higher estimate of VE reported by Dyet et al compared to the
other reports listed in Table 1.8 is not exactly clear. It may be related to several
factors including:

a) The use of MRI in their study - the other studies listed in the table utilized CT
or ultrasound scans. MRI scans have been shown to be more sensitive than
ultrasound scans in the identification of white matter injury and ventricular

enlargement in VLBW infants”.

17

Table 1.8 - Prevalence of ventricular enlargement in absence of germinal matrix

- intraventricular hemorrhage

 

 

 

 

 

 

Author Infant & Study Criteria for
(Year) Characteristics Prevalence diagnosis of VE
Ancel P, N=1929, GA 22-32 wk, 9811929 (5.1%) Judgment
(2006) CUS, Multicenter

cohort
Larroque B, N=2667. GA 23-32 wk. 133/2667 (5.0%) Judgment
(2003) CUS, Multicenter

hospital study
De Felice c, N=483, Bt Wt 590- 35/483 (7.2%) Measurement”
(2001) 38509, CUS, Single

center
Kuban K, N=1310, Bt Wt 500— 10/1310 (0.8%) Judgment
(1999) 15009, CUS,

Multicenter hospital

study
Smith WL, N=92, Bt wt < 2kg, 4/92 (4.3%) Judgment
(1933) CUS, Single center
AZIZ K N=669, Bt Wt 500- 8/669 (1.2%) Measurement“
(1995 12499, Survivors to

1yr, CUS, Population

based study

N=228, Bt Wt s 12509. 5/228 (2.2%)
Uggt‘lI-R’ CUS, Single center
Dyet, LE et al N=1 19 20,119 (16.8%) Measurementf
(2006)
Total N=7454 313/7454 (4.2%)

 

CUS — Cranial ultrasound; GA — Gestational age

*Lateral ventricle width greater than 97th centile for postconceptual age

“Lateral ventricle greater than 2 SDs above the mean for gestational age‘3

TAxial diameter >8mm at GA 525 wk; axial diameter >10mm at GA >25 wk12

18

 

b) Dyet et al used lateral ventricle axial diameter > 8mm at GA 525 week or
diameter >10mm at GA >25 week 12as criteria for VE; this is different from the
criteria used in the other studies listed,

c) A large proportion of the infants evaluated in data reported by Dyet had diffuse
white matter abnormalities and widened extracereme space. The pathogenesis
of these changes is likely multi-factorial and the lesions could have contributed to
the higher prevalence of VE in the cohort,

d) The study was hospital based and was conducted in a quaternary medical
center; there may have been an unduly high representation of very sick VLBW
infants in the study population.

It has been suggested that ventricular enlargement without prior GM-IVH may be
as common as ventricular enlargement associated with GM-IVH ‘1'“. Ventricular
enlargement in VLBW infants without documented GM-IVH has not been
systematically studied. In a case control follow-up study of outcomes in newborns
with abnormal cranial ultrasound findings, Garfinkel et al 4° performed serial
cranial ultrasounds at less than 24 hours of age and on the third and seventh day
of life. They identified a subset of infants with ventricular enlargement with no
evidence of hemorrhage. The outcomes for infants with ventricular enlargement
and no prior GM-IVH were comparable to the outcomes in infants with grades III
and IV periventricular - intraventricular hemorrhage. Several cranial US and
more recently MRI studies indicate the prevalence of VE in VLBW infants range
from 1 to 5 percent 455‘. From these studies, factors that appear to contribute to

VE include degree of prematurity; perinatal asphyxia, severity of initial and later

19

respiratory illness, multiple gestation and perinatal infection. The numbers of
infants with VE in these studies have been small and the maternal, prenatal and
perinatal risk factors for VE were not determined. This is important because VE is
widely accepted as a marker of white matter injury (WMI) and there is a direct
correlation between WMI on brain imaging studies and the risk of cereme palsy
in VLBW infants. Furthermore, there have been few population based studies of
the long-term cognitive and neuromotor outcomes of infants with VE.

The Newborn Brain Hemorrhage (NBH) study, a large population-based study of
intraventricular hemorrhage in preterm low birth weight infants (birth weights
2000 9), presents an opportunity for in-depth study of VE as a number of infants
in the study were noted to have ventricular dilatation prior to the documentation

of any intraventricular hemorrhage.

20

CHAPTER 2
i) THE NEONATAL BRAIN HEMORRHAGE STUDY
Brief description

Figure 2.1: New Jersey counties served by the NICUs in the NBH study”.

 

   

 

Study Enrolled

25 Miles Centers Subjects

 

 

 

 

1 St. Peters Medical Center 525

2 Monmouth Medical Center 311

3 Jersey Shore Medical 269
1,105

 

 

 

 

The Central New Jersey Neonatal Brain Hemorrhage (NBH) Study was designed

21

to determine the antewdents and consequences of germinal matrix
intraventricular hemorrhage (GM-IVH) in a large cohort of unselected low birth
weight infants. The study also explored correlations of cranial US and
neuropathologic findings in infants that succumbed in the neonatal period. Details

of the NBH study have been described in several previous publications 1' 5255.

Infant enrolment and study sites

Over a period of 34 months from August 1984 to June 1987, live-bom infants that
met the study birth weight criteria of 501 to 20009 were enrolled prospectively
into the NBH study at three participating neonatal intensive care units. The three
units are located at Jersey Shore Medical Center (JSMC) in Neptune, Monmouth
Medical Center (MMC), Long Branch, and St Peter’s Medical Center (SPMC)
New Brunswick in the three New Jersey counties of Middlesex, Monmouth and
Ocean respectively [see Figure 2.1]. Prior to the study, an analysis of hospital
admission data along with New Jersey birth certificates for the above three
counties, indicated that 85 % of infants with birth weight 501 — 2000 9 were born
at or cared for at the three study sites. Moreover, 90% of infants with birth weight
of 1500 g or less were cared for at the study sites”. However, while 7-8% of the
three counties population were noted to be African-Americans in the 1980 and
1990 census data, 26% of the study population were African-Americans”.
Furthermore, about 25% of infants enrolled were products of multiple gestations
but not all such infants entered the study as some co—twins were above the birth

weight criterion or were still bom52. The high proportions of African-American

22

infants and products of multiple gestation in the study population is in line with
the known contributions of race and multiple gestation to the high rates of low

birth weights“.

Maternal interviews, maternal and neonatal chart abstraction

The design of the NBH study has been described in detail elsewhere 52' 53. In
summary, 1105 infants with birth weights 501 - 20009 managed in the study
hospitals were enrolled into the study. Of this number, 982 (89%) were delivered
at the study hospitals while 123 (11%) were transferred to the study hospitals
from elsewhere. Variables relevant to brain hemorrhage in low birth weight
infants were obtained from maternal interviews conducted by trained nurses soon
after delivery, and from abstraction of the mothers’ and infants’ medical records.
At the maternal interview, information was requested on reproductive and
contraceptive history; history of the index pregnancy including illnesses,
medication, tobacco, alcohol and drug use by trimester of pregnancy, and
feelings and attitudes towards the pregnancy and delivery”.

Trained research nurses abstracted maternal prenatal, labor and delivery records
onto data forms designed for the study. Data obtained from the infants’ medical
records included neonatal resuscitation data, initial physical examination,
postnatal physical signs, physiologic measurements, diagnoses and pertinent
laboratory findings. The later data were separately recorded for each interval

between cranial ultrasound scans [see Figure 2.2] 53.

23

Figure 2.2: Data collected in the neonatal period in the NBH studysz.

 

 

‘—

 

 

e Time >
Time period Prenatal Intrapartum Birth 4 hrs 24 hrs 7 days Discharge
Ultrasound Scans 1 2 3 4/5
Data forms Maternal Labour/ 1 2 3 4
interview, delivery Neonatal resuscitation, initial physical
prenatal record examination, neonatal records for inter-scan
records intervals
Anthropometry Head circumference: at initial physical
examination, with each scan, weekly until
discharge
Weight: at initial physical examination, weekly
until discharge

 

—

Cranial ultrasound procedures and interpretation

All the CUS scans in the NBH study were performed by trained, hospital based
ultrasound technologists (sonographers). The sonographers participated in a
special training session, prior to recruitment of study subjects, to become familiar
with neonatal neuroanatomy and Ieam the proper techniques for obtaining the
protocol images. Details of the cranial ultrasound protocol and interpretation in
the NBH study have been published elsewhere 52' 57.

In the NBH study, cranial ultrasound scans (US) were performed prospectively
on 1,088 of the 1,105 (98.5%) infants in the cohort. The scans were performed
as closely as possible to ages 4 hours, 24 hours and 7days 53. Scans were
obtained using Diasonic DFR sector scanners (Diasonics Inc., San Francisco)
equipped with 5 - 7.5 MHz transducers. Ultrasound technologists obtained six
coronal and six parasagittal views through the anterior fontanel using acoustic
gel to couple the transducer to the skin to improve transmission of the sound

beam.

24

While no explicit criteria were developed for the identification of ventricular
enlargement in the NBH study, study ultrasound technicians were requested to
judge enlargement separately for the third, fourth and each lateral ventricle and
to grade any enlargement as mild, moderate or severe on each cranial
ultrasound scan.

All study cranial US scans were initially read by one of five radiologists based at
the hospital from which the film originated. All films were independently read
again either by one of two outside readers who were consultants to the study or
by another of the participating radiologists but from a different hospital. Both
readers recorded ultrasound data on a form that reported observations (‘normal’,
“abnormal echodensity‘, ‘abnormal echolucency’ and ‘structure not visualized’)
and interpretations (‘ventricular enlargement”, ‘germinal matrix hemorrhage’ and
‘ventricular hemorrhage’) separately. Concordance on presence or absence,
location and laterality of an observed and interpreted lesion or ventricular
enlargement constituted agreement and the accepted interpretation. Scans were
submitted to a third reader in case of disagreement as to the presence or
absence or time of onset of a lesion for a consensus as described elsewhere 53.
The radiologists were provided the infants’ birth weight but were blinded to the
infants’ clinical information and to previous ultrasound reading(s).

Several months after initiation of the study, a predischarge CUS scan was added
to the protocol; infants hospitalized for long enough periods also had CUS at 5

weeks and at about monthly intervals until discharged 1. This allowed for

25

improved detection of white matter damage and monitoring of ventricular

enlargement in the study infants.

Head circumference measurements

Study neurologists trained the ultrasound technologists to measure head
circumferences of study infants. The technologists measured the Infants’ head
circumference at each cranial ultrasound scan. The attending neonatologist or
pediatric resident physician also measured the infants’ head circumference on
admission. The reliability of head circumference measurements by ultrasound
technologists has been reported elsewhere“. The Pearson correlation coefficient
between measurements obtained by technologists and admitting physicians on

initial physical examination was 0.93.

Follow-up examination

Follow-up assessment was obtained at 2 years corrected age (based on
mother‘s last menstrual period) on 777 (86.2%) of the 901 survivors: 721 (80%)
by examination while clinical information was obtained by mail or phone interview
in 56 (6.2%) children who had moved out of state. Details of the results and
procedures of the 2—year follow-up evaluation have been published elsewhere 1.
In brief, the 2-year assessment focused on detection of major developmental
handicaps especially cereme palsy (CP). Initial assessment of motor
development, hearing, vision, and neurologic status were performed by a

pediatric nurse practitioner specially trained for the project. Study infants’ motor

26

tone, extrapyramidal movements and tendon reflexes in all limbs were assessed
quantitatively and scored on an ordinal scale. The nurse also documented the
preservation of primitive reflexes and measured the range of hip abduction and
extension, popliteal extension and ankle dorsiflexion. When abnormalities were
suspected, infants were referred to one of four consultant child neurologists who

were unaware of the nurse’s specific findings, for further evaluation and

classification of CP by subtype 1.

Objective of the present analysis

While the NBH study was designed to study the prenatal and perinatal
antecedents and outcomes of GMIIVH in infants with birth weights 501 to 2000 9,
some infants were found with ventricular enlargement with or without GMIIVH. Of
the 1019 infants with cranial US diagnoses in the neonatal period, 171(16.8%)
had isolated GMIIVH with no ventricular enlargement. One hundred and twelve
infants (11%) had parenchymal echodensities / echolucencies or ventricular
enlargement, in either use with or without GMIIVH. This latter group of infants
likely had sonographic appearances consistent with damage to the white matter
I. At the two-year follow-up evaluation, this group of infants had a 15-fold
increased risk of disabling cerebral palsy (DCP) compared to infants with normal
scans 1. However, infants with VE in the absence of GMIIVH in the NBH study
have neither been identified nor studied separately and little is known of the
perinatal risk factors, approximate time of onset and long-term outcomes of low

birth weight infants with VE without prior GMIIVH. This is an important group of

27

infants to study because VE is widely regarded as a sonographic evidence of

59, 60

focal and / or diffuse white matter injury , and white matter injury is

associated with a significantly increased risk of poor neurodevelopmental and

cognitive outcomes 1.

Ventricular enlargement without prior GMIIVH is becoming an increasingly
recognized pattern of brain injury in low birth weight and very low birth weight
infants. For example, in a recent brain MRI study of a sequential cohort of VLBW
infants at a median age of 2 days, about 15% of the infants had VE without prior
GMIIVH 28. A clear understanding of the perinatal risk factors associated with this
complication of prematurity is imperative in order to plan effective preventive
strategies.

This analysis of the NBH study data was therefore undertaken with the following
objectives:

1) To identify the prevalence of ventricular enlargement without prior
intraventricular hemorrhage (‘isolated ventricular enlargement’) in a
population of low birth weight infants

2) To identify the prenatal and perinatal antecedents of isolated ventricular
enlargement in low birth weight infants

3) To determine the 2-year neurodevelopmental outcomes in infants with

isolated ventricular enlargement.

28

Suitability of die NBH data set for the study of isolated ventricular
enlargement

The NBH data was obtained from a large population study carried out in three
centers located in the three New Jersey counties of Middlesex, Monmouth and
Ocean. The study enrolled a large number of infants with birth weights 501 to
20009m. From previous census data, it was estimated that approximately 85% of
infants that met the birth weight criterion for enrolment into the study in the three
study counties, were delivered and I or cared for at the study centers. Enrolment
of infants from the three study centers therefore provided for a good opportunity
to obtain a representative sample of the target study population.

Secondly, the NBH study utilized a rigorous CUS protocol that paid close
attention to infants’ age at time of study. The first protocol scan was obtained at
close to 4 hours of age. This is relevant to this analysis as ventricular
enlargement detected at this age is likely to be related to prenatal and [or
perinatal factors. Furthermore, accurate timing of the initial and subsequent
scans allowed for rational evaluation of the antecedents and factors associated
with the progression or non-progression of early cranial ultrasound abnormalities
in the study infants.

Thirdly, the antecedents, risk factors and neurocognitive outcomes of ventricular
enlargement following GMIIVH (post hemorrhagic ventricular enlargement) have
been studied extensively and are well described 3"33. There are so far, very
limited numbers of studies of low birth weight infants with ventricular enlargement

in the absence of a prior GMIIVH. A good number of infants in the NBH study

29

also had ventricular enlargement following GMIIVH. Therefore, the NBH data set
provides an excellent opportunity to study the prenatal, perinatal and postnatal
correlates and long-term neurocognitive outcomes in infants with isolated
ventricular enlargement; it also provides us with the opportunity to compare and

contrast infants with VE and GMIIVH and infants with VE without prior IVH.

it) METHODS
Identification of infants with isolated ventricular enlargement (cases) and

selection of comparison groups (controls) from the NBH dataset

We performed a search of the archived, electronic NBH database to identify
infants in whom any ventricular enlargement (VE) had been documented. VE
was said to be present when there was menlargement (mild, moderate or
severe) of at least one lateral ventricle based on the consensus of two study
radiologists. The assessment of ventricular enlargement in the NBH study was

therefore by judgment and not by an objective measurement.

Infant classification

Infants with ventricular enlargement were classified as having isolated ventricular
enlargement (VE) if there was no associated GMIIVH prior to, or at the time of,
initial identification of ventricular enlargement.

Infants were classified as ventricular enlargement with hemorrhage (VE+HE) if
CUS demonstrated any GMIIVH prior to or at the time of initial detection of VE.
On the other hand, infants with GMIIVH and no ventricular enlargement at any
time were classified as hemon'hage only (HE).

30

Infants with neither ventricular enlargement nor GMIIVH at any time were

classified as normal scans (Normal) and served as another comparison group.

Data collection

We pulled and reviewed the original data forms of the infants with VE including
prenatal forms, pregnancy and delivery history, neonatal data forms and
ultrasound interpretation forms. Data extracted from the sheets included:

1) Maternal history and demographic information - including age, maximum
educational level attained, history of smoking, alcohol and illicit drug use;

2) History and complications of pregnancy — including infections, hypertensive
disorders of pregnancy, vaginal bleeding;

3) Labor and delivery data including mode of delivery, duration of membrane
rupture, duration of labor, Apgar scores at 1 and 5 minutes, Chorioamnionitis,
placental and umbilical cord abnormalities; and

4) Infant data including birth weight, head circumference, CUS results and death;
we also reviewed the follow-up records at 2 years of age for neuromotor and
cognitive functions.

The same data were extracted from the archived electronic NBH database for the

three comparison groups.

Statistical Methods
Characteristics of mothers and infants in the VE, VE+HE, HE and Normal groups

were compared and p-values calculated. Means and standard deviations were

31

calculated for continuous variables and the groups compared with student t-test
while categorical variables were compared with chi-square and Fishers exact
test. Conditional logistic regression was used to estimate effect of perinatal
factors on the occurrence of VE, VE+HE and HE for matched pairs. Mortality,
cerebral palsy (CP) and disabling cereme palsy (DCP) rates were compared
among the three groups. Risks of death, CP and DCP by exposure to VE,
VE+HE or HE were estimated. Statistical analysis was with SAS statistical

software (SAS Institute Inc., Cary, NC,).

32

CHAPTER 3

RESULTS — ANTECEDENTS AND OUTCOMES OF ISOLATED

VENTRICULAR ENLARGEMENT IN THE NEONATAL BRAIN HEMORRHAGE

(NBH) srqu

A search of the archived electronic NBH database, and review of the stored

forms, showed that 23 of 1088 infants (2.1%) that had at least one protocol CUS

scan met our definition of isolated ventricular enlargement (VE). These

constituted the study subjects. Eighty-seven infants (8.0%) had ventricular

enlargement noted either at the time of initial identification of GM-IVH or at

subsequent cranial US scans (VE+HE). One hundred seventy infants (15.6%)

Table 3.1: Frequency distribution of infants in the NBH study by cranial

ultrasound results

 

 

 

 

Cranial Ultrasound Results Frequency (%)
Isolated ventricular enlargement (VE) 23 (2.1)
Isolated germinal matrix — intraventricular

hemorrhage (HE) 170 (15.6)
Gerrninal matrix — intraventricular hemorrhage

and ventricular enlargement (HE+VE) 87 (8.0)
Normal scans (Normal) 808 (74.3)
Total 1088 (100)

 

33

 

Table 3.2: Some characteristics of infants with ventricular enlargement in the

absence of germinal matrix intraventricular hemorrhage (VE)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Best Birth Head 5 min
Gestational Weight Circumference Apgar
Infant ID Gender Age (wks) (gm) (cm)* Score
84102700 Female 27.3 800 24 7
84300500 Male 28.3 1035 27 7
84301300 Male 30.7 1295 26.5 6
84303900 Male 28.0 1020 26 7
85105500 Male 27.6 1280 27.5 4
85107100 Male 26.6 1070 25.2 9
85109600 Male 30.3 1800 29.5 8
85110000 Male 33.9 1765 28 9
85112100 Male 31.6 1795 29 5
85207500 UTDT 30.0 1332 23.5 2
85313400 Female 36.4 1850 30 6
86116000 Male 30.4 1500 29 0
861 16300 Male 29.1 1 134 27 8
86118700 Female 29.9 1430 28 8
86119500 Female 27.7
87121600 Male 36.6 1380 30.2 5
85110602 Male 35.4 1775 33.8 8
8531 1702 Male 23.0 590 21 3
86119802 Male 33.1 1580 30.5 7
86217102 Female 28.9 1430 28 6
86336202 Male 29.0 1600 30 7
87344702 Female 31.7 1820 31 9
87345802 Female 31.6 1500 29.5 8

 

 

 

 

 

 

 

 

*Head circumference measured at initial physical exam
TUTD - Unable to determine

34

had germinal matrix - intraventricular hemorrhage without documented
ventricular enlargement (HE) in any CUS scan. Eight hundred and eight infants
(74.3%) had neither ventricular enlargement nor GMIIVH at any time (Normal
Scans). The latter three groups constituted our comparison groups (Table 3.1).
Of the 23 infants with VE, 15 (65.2%) were male and 8 (34.8%) had 5-minute
Apgar score less than 7 (Table 3.2). Three of the 23 infants with VE
subsequently developed GM-IVH (late onset GM-IVH) after the identification of
ventricular enlargement; the other 20 infants remained free of GMIIVH. The
infants with ‘late onset GM-IVH’ were all males, and their gestational age, birth

weight and 5-minute Apgar scores are shown in Table 3.3.

Table 3.3: Some characteristics of infants with initial IVE, but later developed
germinal matrix intraventricular hemorrhage

 

 

 

 

Best Birth Head
Gestational Weight Circumference 5 min
Infant ID Gender Age (wks) (gm) (cm)* Apgar
861 16000 Male 30.4 1500 29 0
85311702 Male 23.0 590 21 3
86336202 Male 29.0 1600 30 7

 

 

 

 

 

 

 

 

Note: Head circumference measured at initial physical exam.

Seven (30.4%) of the cases of VE were products of multiple gestation; they were
all the 2"“ of a pair of twins. There were no higher order multiples among the
cases. The cases that were products of twin gestation had similar birth weight
and gestational age to the singleton cases (Table 3.4). Among the infants with

35

normal scans, 22.8% were products of twin gestation while 2.4% were of higher
order multiple gestations, giving a multiple gestation rate of 25.1%. On the other
hand, of the infants with HE+VE, 19.5% were twins while 4.6% were higher order
multiples, with a multiple gestation rate of 24.1%. The multiple gestation rate was

similar in the cases and comparison groups.

Table 3.4: Low birth weight infants with isolated ventricular enlargement:

comparison of singletons and products of twin gestations

 

 

 

 

 

 

Singletons Twin Gestations
Characteristics (N=15) (N=7) P-Value
Best gestational age (wks) 30.3130 30413.9 NS
Birth weight (gm) 1 3651326 1470141 2
Head circumference (cm) 27.3120 29113.9
Males (%) 11/15 (73) 4/7 (57)
5 min Apgar (median) 7 7

 

 

 

 

 

 

Characteristics of cohort infants by cranial ultrasound findings

The characteristics of the four groups of cohort infants that had CUS scans are
shown in Table 3.5. The mean gestational age, birth weight and head
circumference at birth were significantly different among the 4 groups: the
VE+HE group had the lowest gestational age, birth weight and smallest head
circumference at birth (P<0.0001). The frequency of small weight for gestational
age (SGA) was also different between the groups. Of the infants with ventricular
enlargement, small weight for gestation was more frequent among infants with

36

GM-IVH than among infants with isolated ventricular enlargement. Of note,

infants with Normal US scans had the highest rate of SGA.

Table 3.5: Characteristics of all cohort infants with IVE, VE+HE, HE and Normal

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CUS scans
Normal
VE VE+HE HE
Scans
Characteristics (n=23) (n=87) (n=170) P-value
(n=808)
Gestational age 30.3 1 3.0 28.7 1 3.8 29613.3 31.5 1 3.4 <0.0001
(wk8)
Birth wt (gm) 1399 1 349 1126 1 364 12481382 1464 1 392 <0.0001
Opcficm) 27.9 1 2.8 26.1 1 2.7 26512.8 28.1 1 2.8 <0.0001
Percent *SGA 13.0 24.1 18.8 31.2 <0.0001
5 Median 7.0 7.0 7.0 8.0 <0.0001
m'm‘te < 7 (%) 40.9 46.9 28.8 18.2 <0.0001
Apgar
Male Gender (%)T 65.2 59.7 55.3 49.6 0.007
Ventilation days 3 9 4 O
. 0.01
(medran)*
0 en da s 5 10 6 1
xyg. y 0.0007
(medlan)*
Cord pH 5 7.25 36.4 36.3 27.2 29.2
0.45
(%)
First CO; 5 30 (%) 19.0 33.7 35.6 35.4 0.48
"BPD (%) 13.0 24.1 16.4 7.8 <0,0001

 

 

 

 

 

 

 

Fisher’s Exact Test, 1 Among infants with durations greater than 0, *Small for

gestational age " Bronchopulmonary dysplasia, fOccipito-frontal circumference

37

 

Infants with VB tends to be more mature, have higher birth weight and head
circumference at birth compared to infants with VE+HE and HE only. All these
measurements tended to be greater in infants with Normal scans compared to
the other 3 groups. There was a slight male preponderance among infants with
abnormal CUS scans compared with infants with normal scans. The proportion of
infants with 5—minute Apgar score less than 7 was higher among infants with VE
and VE+HE than in infants with HE and Normal scans.

Infants with VE were more likely to be males compared to the other groups but
infants with VE+HE were more likely to require assisted ventilation and
supplemental oxygen for more days compared to the other groups. They also
had a higher rate of bronchopulmonary dysplasia compared to the other groups.
However, there was no significant difference between cord pH and initial PaCOz

in all 4 groups.

Rate of detection of ventricular enlargement in the VE and VE+HE groups

Proportion of infants with ventricular enlargement was higher at each of the first
three scans in the HE+VE group compared with infants with VE. (Table 3.6).
There was a sharp increase in the proportion of infants with ventricular
enlargement on the 3rd scan in the group with VE+HE (Figure 3.1). The
frequency of detecting new onset ventricular enlargement was similar in both
groups at the 4‘“ and later scans. The second CUS obtained at 13 — 48 hours of
life yielded only one new VE case; however, there was a sharp increase in
detection of new cases on the 3rd scan. In infants with GM-IVH, the rate of

detection of new ventricular enlargement was similar on the 1“ and 2"d scans,
38

followed by a significant increase on the 3'“ scan. By the 4th and subsequent
scans, the rate of detection dropped off in the VE+HE group but continued to rise

in the infants with isolated ventricular enlargement.

Table 3.6: Frequency of isolated ventricular enlargement (VE) and ventricular

enlargement associated with GM-IVH (VE+HE) at each CUS

 

 

 

 

 

 

 

Total Scanned Rate of Detection

CUS Scans (N) IVE (%) VE+HE (%)
W Scan 1015 5 (0.5) 17 (1.7)

(0-12hrs)

2"" Scan 1001 1 (0.1) 17 (1.7)

(13-48hrs)

3rd Scan 926 10 (1 .1) 45 (4.8)

(49hrs — 7days)

24‘" Scan 517 7 (1.4) 8 (1.5)

( >7daYS)

Totals 1015 23 (2.3) 87 (8.5)

 

 

 

 

 

 

Maternal characteristics

Maternal characteristics including age, highest educational attainment, smoking
and alcohol use were similar in the 4 groups (Table 3.7). However, mothers of
infants with Normal scans were least likely to use illicit drugs compared to
mothers in the other groups. The proportion of mothers with addiction to social

(alcohol, tobacco, marijuana) or hard (cocaine, heroin) drugs was similar in all 4

39

groups; however, the proportion of unemployed parents was highest among the

infants with VE.

Figure 3.1: Prevalence of VE and VE+HE at each ultrasound scan.

 

 

% Ventrlcular Enlargement
w

    

15t Scan 2nd Scan 3rd Scan 2 4th Scan
Cranial Ultrasound Scans

+VE *VE-FHE

 

 

 

Hypertensive disorders of pregnancy including pregnancy induced hypertension,
preeclampsia and chronic hypertension were significantly more frequent in
mothers whose infants had Normal CUS scans compared to mothers whose
infants had ventricular enlargement with or without hemorrhage (Table 3.8). The

proportion of mothers with vaginal bleeding and infectious complications was

40

similar in all 4 categories. However, Chorioamnionitis was more frequent among
mothers of infants with VE+HE and HE than in mothers of infants with Normal
scans (Table 3.9). There was also a trend towards less frequent rupture of fetal

membranes for greater than 18 hours before delivery among mothers of infants

Table 3.7: Maternal characteristics: demographic data

 

 

 

 

 

 

 

 

 

 

 

 

VE VE+HE HE Normal

Characteristics (N=23) (N=87) (N=170) (N=808) P-Value
Maternal age (yrs) 26.55 26.92 25.77 26.91 0.2907
Education (yrs) 12.20 12.74 12.98 12.99 0.4157
Smoking (%) 30.43 16.09 16.47 19.78 0.3328
Alcohol use (%) 43.48 26.44 31 .18 33.01 0.3985
Drug1 use‘(%) 4.35 4.60 5.88 2.18 0.0327
Any addiction2(%) 65.22 42.53 46.47 46.84 0.2864

White (%) 69.57 62.07 62.72 69.33
Race Black (%) 17.39 34.48 30.18 25.00 0.1672

Other (%) 13.04 3.45 7.10 5.67
Unemployed3 (%) 34.78 10.34 12.35 13.47 0.0198

 

 

 

 

 

 

 

 

1 Drug includes any use of marijuana, cocaine or heroine.

2Any addiction includes addiction to tobacco, alcohol, marijuana, cocaine or herein.
3 Unemployed: both mother and father are unemployed.

4 Fisher exact test used for P—value.

with VE compared to the other 3 groups. Rates of cesarean section delivery and

the frequency of placental and umbilical cord abnormalities were similar in the 4

41

groups of mothers. We explored the influence of different pregnancy and delivery
complications on the risk of developing ventricular enlargement in the absence or

presence of GMIIVH (Table 3.10). Pregnancy induced hypertension was

Table 3.8: Maternal characteristics: medical and pregnancy complications

 

 

 

 

 

 

(entire cohort)
VE VE+HE HE Normal
Complications (N=23) (N=87) (N=170) (N=808) P-
value

Pregnancy induced 4.35 8.05 12.35 22.45 0.0001
hypertension (%)

Preeclampsia (%) 4.35 3.45 2.94 9.59 0.0087
Chronic hypertension (%) 4.35 3.45 2.94 8.86 0.0208
Vaginal bleeding1 (%) 39.13 25.29 33.53 35.07 0.3038
Infectious complications2 (%) 60.87 51.72 57.65 58.86 0.6294

 

 

 

 

 

 

 

 

1Includes placenta previa and abruptio placenta

2Include all respiratory and urinary tract infections

significantly associated with a decreased risk of ventricular enlargement in
infants with VE+HE but not in infants without hemorrhage. Preeclampsia was
also associated with a tendency towards reduction of the risk for ventricular
enlargement in infants with GM-IVH; this however did not achieve statistical
significance. On the other hand, chorioamnionitis was associated with a nearly 3-
fold increased risk of ventricular enlargement in infants with GM-IVH. The role of
chorioamnionitis in VB could not be estimated, probably because of the small

42

Table 3.9: Maternal characteristics: mode of delivery and delivery complications

(entire cohort)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

VE VE+HE HE Normal P-value
Characteristics (N=23) (N=87) (N=170) (N=808)
Cesarean section 47.83 40.23 45.88 48.06 0.5605
(%)
‘ROM > 18hours 8.70 22.99 30.59 23.06 0.0585
(%)
TMSAF' (%) 4.35 6.90 4.71 5.70 0.9109
Oligohydramnios 0 0 0.59 1 .21 0.6175
(%)
Placental 8.70 5.75 9.41 9.47 0.7218
abnormalities (%)
Cord abnormalities 4.35 8.05 11.18 10.68 0.6598
(%)
Chorioamnionitis" 0 9.20 8.82 3.52 0.0044
(%)

 

*Fishers exact test for P-value iFRupture of membranes '

fluid

Meconium stained amniotic

number of infants in this category. Maternal smoking or alcohol use during

 

pregnancy, placental abnormalities, prolonged rupture of membranes for greater
than 18 hours before delivery were not associated with increased risk of

ventricular enlargement in low birth weight infants with or without GM-IVH.

43

After adjusting for gestational age and birth weight, only chorioamnionitis

remained significantly associated with increased risk of ventricular enlargement

in infants with GM—IVH (Table 3.11); pregnancy induced hypertension was no

longer associated with increased risk of ventricular enlargement in this group of
infants. None of the factors evaluated was associated with increased risk of VE in

infants with no GM-IVH after adjusting for gestational age and gender.

Table 3.10: Unadjusted odds ratios (OR) and 95% confidence intervals (CI) for

the influence of different pregnancy and delivery characteristics on the

occurrence of VE and VE+HE compared to infants with Normal CUS scans

 

 

 

 

 

 

 

 

 

 

 

 

 

VE vs. Normal P- VE+HE vs. Normal P-

Risk factors OR (95% CI) values OR (95% CI) values
PIH 0.15 (0.02, 1.17) 0.071 0.30 (0.13, 0.66) 0.003
Preeclampsia 0.42 (0.05, 3.22) 0.410 0.33 (0.10, 1.09) 0.069
Smoking 1.77 (0.71, 4.38) 0.214 0.77 (0.42, 1.41) 0.409
Alcohol use 1.56 (0.67, 3.6) 0.297 0.72 (0.44, 1.20) 0.214
Placenta 0.91(0.21, 3.95) 0.900 0.58 (0.23, 1.48) 0.257
abnormality

ROM > 18 hrs 0.31 (0.07, 1.36) 0.123 0.99 (0.58, 1.68) 0.998
Chorioamnionitis UTE UTE 2.77 (1.22, 6.28) 0.014

 

PIH = Pregnancy induced hypertension

ROM = Rupture of membranes

UTE = Unable to estimate

44

 

Mortality and neurodevelopmental outcomes at 2 years

The overall mortality and mortality within 28 days were significantly higher in
infants with VE and VE+HE compared to infants with Normal CUS scans (Table
3.12). While the rates of non disabling cereme palsy was similar in all three
groups, infants with VE and VE+HE had significantly higher rates of disabling

cerebral palsy compared with infants with normal CUS scans.

Table 3.11: Adjusted* odds ratios (OR) and 95% confidence intervals (CI) for
the influence of different pregnancy and delivery characteristics on the

occurrence of VB and VE+HE compared to infants with Normal CUS scans

 

 

Risk factors VE vs. Normal VE+HE vs. Normal
OR (95% Cl) P-Value OR (95% Cl) P-Value
PIH 0.18 (0.02, 1.38) 0.099 0.46 (0.20, 1.05) 0.065

 

Preeclampsia 0.52 (0.06, 3.98) 0.529 0.54(0.16, 1.79) 0.317

 

Smoking 1.79 (0.72, 4.45) 0.206 0.78 (0.42, 1.44) 0.436

 

Alcohol use 1.61 (0.69, 3.74) 0.262 0.77 (0.46, 1.29) 0.324

 

Placental 0.94 (0.21, 4.13) 0.943 0.65 (0.25, 1.69) 0.383
abnormality

 

ROM > 18 0.30 (0.07, 1.31) 0.112 0.93 (0.54, 1.61) 0.819
hours

 

Chorioamnion UTE UTE 2.43 (1.03, 5.77) 0.042
itis

 

 

 

 

 

 

 

*Adjusted for best gestational age and gender; UTE = Unable to estimate

45

Using a multivariate logistic regression model, VE and VE+HE were associated
with increased risk of both overall mortality and mortality in the first 28 days of life
(Table 3.13). While neither VE nor VE+HE were associated with nondisabling

cerebral palsy, both lesions were significant risk factors for disabling cerebral

Table 3.12: Mortality and neurodevelopmental outcomes at age 2years in
infants with ventricular enlargement (VE, VE+H E) compared with infants with

Normal CUS scans

 

 

 

 

 

 

 

 

 

 

 

 

Independent Variables
VE VE+HE Normal

Outcome Measures N (%) N (%) N (%) *P-Values
Mortality within 28 9 (39.1) 23 (26.4) 95 (11.5) < 0.0001
days
Overall Mortality 10 (43.5) 35(40.2) 120 (14.6) < 0.0001
Non Disabling 2 (8.7) 5 (5.8) 39 (4.8) 0.5119
Cerebral Palsy
Disabling Cerebral 6 (26.1) 25 (28.7) 17 (2.1) < 0.0001
Palsy

* Exact test

palsy. Following adjustment for best gestational age and birth weight, VE but not

VE+HE remained a significant risk factor for mortality within 28days (Table

3.14); both lesions remained significant risk factors for disabling cerebral palsy

but not for nondisabling cereme palsy.

46

 

Table 3.13: Unadjusted odds ratios (OR) and 95% confidence intervals (CI) for

the risk of outcome measures in infants with ventricular enlargement (VE,

VE+HE) compared to infants with normal CUS scans.

 

 

 

 

 

 

 

 

 

 

VE VE+HE

Outcome OR OR
Measures (95% Cl) P-Value (95% CI) P-Value
Mortality within 4.93 0.0003 2.75 0.0001
28 days (2.07, 11.71) (1.63, 4.64)
Overall mortality 4.51 0.0005 3.94 < 0.0001

(1.93, 10.52) (2.46, 6.31)
Non Disabling 1.91 0.3905 1.22 0.6751
Cerebral Palsy (0.43, 8.46) (0.47, 3.20)
Disabling 16.75 < 0.0001 19.14 < 0.0001
Cerebral Palsy (5.87, 47.75) (9.81, 37.33)

 

47

 

Table 3.14: Adjusted" odds ratios (OR) and 95% confidence intervals (CI) for
the risk of outcome measures in infants with ventricular enlargement (VE,

VE+HE) compared to infants with normal CUS scans.

 

 

 

 

 

VE VE+HE

Outcome OR (95% Cl) P-Value OR (95% Cl) P-Value

Measures

Mortality within 28 4.83 0.0013 1.25 0.4473

days (1.84, 12.63) (0.69, 2.26)

Mortality 4.28 0.0025 1.98 0.0122
(1.67, 10.98) (1.16, 3.38)

Non Disabling 1.80 0.4396 1.07 0.8939

ce'eb'a' Pa'sy (0.40, 8.01) (0.39, 2.89)

Disabling Cerebral 15.67 < 0.0001 21.83 < 0.0001

Pa'sy (5.43, 45.19) (10.56, 45.11)

 

 

 

 

 

 

 

*Adjusted for best gestational age and gender

48

CHAPTER 4

VENTRICULAR ENLARGEMENT, PERIVENTRICULAR WHITE MATTER
INJURY AND NEUROLOGIC OUTCOMES IN THE NEWBORN BRAIN
HEMORRHAGE (NBH) STUDY - DISCUSSION

Summary of findings

Ventricular enlargement without prior germinal matrix - intraventricular
hemorrhage (VE) was identified in 23 of 1088 (2.1%) infants by protocol cranial
ultrasound (CUS) scans in the NBH study. Pregnancy induced hypertension
(PIH) was significantly more frequent in mothers of infants with normal CUS
scans compared mothers of infants with abnormal scans. Furthermore, PIH was
associated with decreased risk of ventricular enlargement in infants with GM-IVH.
The protective effect of maternal high blood pressure against the development of
ventricular enlargement in infants with GM-IVH was lost after adjustment for best
gestational age (BGA) and gender. On the other hand, chorioamnionitis was
more frequently seen in infants with VE+HE and HE compared to infants with
Normal scans; chorioamnionitis was also associated with increased risk of
ventricular enlargement in infants with GM-IVH, this association persisted after
adjustment for BGA and gender. Overall mortality, death within 28days and
disabling cereme palsy (DCP) were more frequent in infants with ventricular
enlargement with or without GM-IVH compared to infants with Normal scans.
These abnormal CUS findings were also associated with increased risk of

mortality and DCP; however, after adjustment for BGA and gender, VE+HE was

49

no longer associated with increased risk of mortality within 1" 28 days.
Ventricular enlargement associated with GM-IVH was associated with a greater
risk of DCP compared to VE, this relationship persisted after adjustment for BGA

and gender.

Significance of VE in LBW infants

Advances in obstetric and neonatal care in the last three decades, including the
use of antenatal steroids, exogenous surfactant therapy, improved monitoring
technologies and ventilatory methods have led to increased survival of smaller
and less mature VLBW infants“ 62. These advances also contributed to a steady
decline in the prevalence and severity of GM-IVH, the most common form
cereme abnormality detected on cranial ultrasound scan in the VLBW premature
infant 63' 6". However, the improved survival rates of VLBW infants in the 1990s
was not accompanied by improvement in the rates of neurodevelopmental
impairment“, and some studies of long-tenh outcomes reported increased rates
of neurodevelopmental problems including cerebral palsy in surviving VLBW
infants%' 67. Cerebral white matter injury (WMI) is now widely regarded as the
major form of brain injury and the leading cause of long-term neurologic disability
in surviving VLBW infants“. Cranial ultrasound (CUS)50 ‘8 and magnetic

‘2' 69 studies indicate VE is a frequent marker of WMI

resonance imaging (MRI)
and may be seen in association with focal, cystic and diffuse white matter lesions

as well as with intraventricular hemorrhage in VLBW infants.

50

Frequency of isolated ventricular enlargement in low birth weight infants
The NBH study presents one of the first opportunities to take a close look at VE
in a cohort of unselected, population based study of LBW infants. There are no
comparable population studies of GM-IVH that utilized timed cranial ultrasound
scans starting from very early after delivery. The study provided an opportunity to
estimate the approximate time of onset of intracranial abnormalities detected by
CUS in LBW infants.

However, a review of some other studies of unselected low birth weight infants
for the frequency of VE on early cranial US scans revealed a wide range with
figures ranging from 1.2% to 27%‘2' 51. In a study of a Provincial cohort of very
preterm infants, birth weight 500 — 1250 g, Aziz et al51 found VE by CUS scan in
8 of 669 (1 .2%) infants that survived to 1-year adjusted age. Lower rates of
cranial ultrasound (CUS) abnormalities in this study may be explained by the fact
this was an unselected population study and not a study of sick infants in tertiary
centers. Furthermore, the study radiologists only reviewed and reported scans
that were initially read as abnormal by hospital radiologists that did not participate
in the study. Therefore, lesions that could have been considered abnormal by
study radiologists may have been missed at the initial, non-protocol reading
stage, contributing to the rather low frequency of abnormal CUS scans. On the
other hand, Dyet et al12 carried out a hospital based study of 119 consecutively
recruited preterm infants born at < 30 weeks gestational age. They obtained
serial cranial MRI scans, starting as soon as possible after birth, at a median age

of 2 days. Scan timing was however, not standardized and depended on infants’

51

clinical stability. Twenty (16.8%) and 22 (18.5%) of 119 infants with no
intraventricular hemorrhage had VE on the first and final scans respectively.
Ancel et al‘8 found a frequency of VE in 98 of 1929 (5%) evaluated by cus
scans in a population study of preterm infants of 22 to 32 weeks gestational age
(the EPIPAGE Study). In that study, the observed frequency of VE was not
related to gestational age but was higher in products of multiple than singleton
gestations; infants with intrauterine growth restriction were not at a higher risk of
VE". However, in the EPIPAGE study, there was no standardized protocol for
obtaining cranial ultrasound scans; the scans evaluated were those obtained for
routine surveillance of perinatal brain injury during routine neonatal care. The
timing, technique, quality and reading of the scans were not standardized. This
calls to question the accuracy of the reported frequency of VE in the study; a
recent audit of clinical neonatal ultrasound scan interpretations indicate low
frequencies of accurate identification of cerebral abnormalities including

ventricular dilatation and white matter lesions".

Intrauterine growth and VE

Infants with VE in our study were significantly less likely to be small for
gestational age than infants with Normal CUS scans. No consistent relationship
between VE and intrauterine growth in preterm infants has been established. A
review of the relevant literature reveals studies that have variously reported
increased frequency", no change in frequency72 and decreased frequency” 73 of

ventricular enlargement in infants with intrauterine growth restriction when

52

compared with gestational age matched, appropriate for gestational age preterm
infants.

In a small retrospective review, Kriss and Decal71 found a significantly higher
frequency of non-pathologic, mild isolated ventriculomegaly on CUS scans in
asymmetric IUGR infants compared to appropriate for gestational age (AGA),
matched controls. In the report, ventricular enlargement was found to have
resolved without intervention on follow-up CUS scans obtained 4 to 12 weeks
after the initial scan. Furthermore, IUGR infants with mild isolated ventricular
enlargement were asymptomatic and remained normal on follow up evaluations
at ages ranging from 3 to 5 years.

On the other hand, Inder et al69 carried out a qualitative MRI study of 100
consecutive VLBW preterm infants at term equivalent age. They found IUGR had
a significant protective effect on the risk of moderate to severe white matter injury
and VE in preterm infants. This protective effect was lost after controlling for
gestational age and mode of delivery. The finding in this analysis of the highest
frequency of SGA among infants with Normal CUS scans may be reflective of the

protective effect of IUGR on ventricular enlargement in infants without GM—IVH.

Apgar scores and VE

Compared to infants with normal CUS scans, a higher proportion of infants with
VE have a 5-munite Apgar score less than 7 in this report. Although the reason
for this is not obvious from this analysis, chronic sublethal hypoxia in an

I74

experimental animal mode and perinatal asphyxia in human neonates with or

53

without intracranial hemorrhage“1 are associated with cereme ventriculomegaly.
Low birth weight preterm infants are at a higher risk of cardiopulmonary
depression and therefore frequently require active neonatal resuscitation and

have lower Apgar scores compared to their term counterparts”.

Hypertensive disorders of pregnancy and VE

Hypertensive disorders of pregnancy including pregnancy induced hypertension,
preeclampsia and chronic hypertension were significantly more frequent in
mothers with Normal CUS scans compared to mothers of infant with ventricular
enlargement, with or without prior GM-IVH.

The placenta in hypertensive disorders of pregnancy is characterized by
inadequate invasion and replacement of the endothelial layers of the uterine
spiral arteries by cytotrophoblast cells 76. The process of cytotrophoblast invasion
results in the transformation of the spiral arteries from thick walled muscular
vessels into saclike, low resistance flaccid vessels that eventually accommodate
the 10-fold increase in uterine blood flow required to meet the demands of the
growing fetus 77. Failure of vascular remodeling of the spiral arteries leads to a
state of progressive placental ischemia and hypoxia believed to be the trigger for
the systemic maternal vascular endothelial dysfunction and inflammation seen in
preeclampsia as pregnancy progresses 78' 79. Inadequate remodeling of the
uterine spiral arteries in preeclampsia may lead to a state of chronic placental

underperfusion and fetal hypoxia that could contribute to impaired fetal growth.

54

In this analysis, infants with Normal CUS scans also had a significantly higher
frequency of SGA than infants with VE. Furthermore, pregnancy induced
hypertension was associated with a reduced risk of ventricular enlargement in
infants with GM-IVH. Although this association was lost after adjusting for
gestational age and gender, the data suggest that hypertensive disorders of
pregnancy are associated with intrauterine growth restriction and protective
against ventricular enlargement in low birth weight infants with GM-IVH.

Several reports indicate that very low birth weight preterm infants born to women
with preeclampsia and pregnancy induced hypertension are at a lower risk of

°°' 8‘ and intraventricular hemorrhage 82' 83. However, there are no

cerebral palsy
direct studies of the influence of preeclampsia on the size of the cereme
ventricles in preterm infants, and reports on the association between white matter
injury and preeclampsia have been conflicting: while some investigators report
that preeclampsia is associated with a reduced risk of ventriculomegaly and

white matter damage in VLBW infantsn' 73' “'86

, others have found preeclampsia
leclampsia is associated with a higher risk of cystic PVL in infants of 33 to 35
weeks GA87 and yet other investigators found no influence of preeclampsia I
eclampsia on the risk of brain injury in VLBW infants“. A neuropathologic and
cranial ultrasound study found no association between preeclampsia and
prenatal ischemic brain injury but a significant association between preeclampsia

and postnatal ischemic cereme damage”. These varied results indicate the

etiology of VE and white matter injury in VLBW infants is multifactorial and likely

55

includes factors that are operative in the prenatal, perinatal and postnatal

periods.

Chorioamnionitis and VE

The relationship between chorioamnionitis and VE could not be estimated from
the data because of the absence of chorioamnionitis from this group. However,
chorioamnionitis was associated with increased risk of ventricular enlargement in
infants with GM-IVH. This association remained statistically significant after
controlling for gestational age and gender. This finding is in agreement with what
has been demonstrated by several investigators that chorioamnionitis is
associated with increased risk of GM-IVH, white matter injury and ventricular

enlargement ‘9' 90""

Timing of VE in LBW infant

The distribution in the rate of detection of ventricular enlargement in infants with
no prior GM-IVH is different from that noted in infants with prior GM-IVH. In both
groups of infants, ventricular enlargement was noted on the 13t scan, but with a
higher frequency in infants with GM-IVH. In the both groups, this may include
contribution from fetal ventriculomegaly; however, the higher rate in the VE+HE
suggest ventricular dilatation associated with early IVH - the so-called grade III
IVH on the Papille classification. In this regard, it has been shown in several
studies that up to 40% of intraventricular hemorrhage in very low birth weight

preterm infants occurs within the first 24 hours of life”: 93. In the NBH study, over

56

35% of IVH occurred on the 13t U/S scan obtained at the age of 4.9 1 2.2 hours“.
The 2"d scan obtained at 13 - 48 hours of life (mean age 25.5 1 4.8 hours)”
revealed only 1 new case of VE, while the rate of detection in the VE+HE group
was still as high as in the first scan. The timing of the 2"d scan corresponds to
the age at which low birth weight infants are very susceptible to developing
intraventricular hemorrhage as a number of the risk factors for this complication,
including respiratory distress syndrome, pneumothoraces, cardiovascular and
hemodynamic instability are quite prevalent at this age. The persistently high rate
of detection of ventricular enlargement in the VE+HE group on the second scan
likely represent ventricular dilatation associated with new onset GM-IVH at this
age.

The 3rd sans obtained between ages 49 hours to 10 days (mean age of 7.2 1
0.8 days) shows an increase in the rate of detection of ventricular enlargement in
both the VE and VE+HE groups, with a sharp increase in the VE+HE group. In
the VE+HE group, the ventricular enlargement likely represent a combination of
progressive posthemorrhagic ventricular dilatation and cereme white matter and
cortical injury” 5°. In the VE cases, ventricular enlargement detected on the 3"
scan most likely represents diffuse white matter injury. In the NBH study,
neuropathologic evaluation of infants that came to autopsy, and in whom three or
more CUS scans had been performed before death, showed that ventricular
enlargement on ultrasound was a marker of white matter damage in infants that

survived six days or longer”.

57

The 4th and later scans were predischarge studies added to the protocol several
months into the NBH study. In keeping with the pattern of posthemorrhagic
ventricular dilatation after GM-IVH, fewer instances of new onset ventricular
enlargement were noted in the VE+HE group. On the other hand, there was still .a
modest increase in the proportion of infants in the VE group that had new onset
ventricular enlargement at the 4th and later scans. This is in keeping with serial
CUS and MRI studies which show increases in the proportion of VLBW infants

with ventricular enlargement without prior GM-IVH at term equivalent age'z' 9‘.

Neonatal and 2 year outcomes of infant with VE

Infants with VE and VE+HE had significantly increased risk of overall mortality
and mortality in the 1"t 28days of life compared to infants with normal CUS scans.
This increased risk persisted for infants with VE but not in those with VE+HE
after adjusting for gestational age and gender. The increased risk of early
mortality in infants with VE suggests the underlying cause of ventricular
enlargement in this group of infants may have been present prenatally. This may
be related to factors other than hypertensive disorders of pregnancy which has
been shown to be associated with a lower risk of ventriculomegaly“.

However, infants in both groups had significantly higher risks of disabling
cereme palsy (DCP) compared to infants with normal scans, even after adjusting
for gestational age and gender.

The higher frequency of cereme palsy in the VE and VE+HE groups is in

keeping with the neuropathologic changes of ventricular dilatation that occur in

58

the context of cerebral white matter injury. Cereme white matter injury includes a
spectrum of cerebral pathologic changes that range from focal injury to diffuse,
extensive cereme white matter lesions.

The focal lesions are characterized by well circumscribed necrotic or cystic
lesions that occur commonly in the subventricular zone adjacent to the lateral
ventricle and involve injury to all cellular elements (periventricular white matter
injury, PVWMI). The lesions are often accompanied by reduction in white matter
volume secondary to cyst formation or ventricular dilatations' 95. PVWMI results
from selective vulnerability of the periventricular white matter in the preterm
neonate to ischemic, infectious or metabolic insults and are closely correlated to
spastic cerebral palsy”.

Diffuse cerebral white matter injury, on the other hand, is characterized by areas
of poor myelination and diffuse gliosis in the deep cereme white matter. The
injured cell type that provokes the gliosis in diffuse white matter injury is
presumably the ologidendrocyte progenitors? The later cells are destined to
develop into mature oligodendrocytes which form the myelin of the cerebral white
matter. Thus the the principal neuropathologic sequela of periventricular white
matter injury is diminution of white matter volume and ventriculomegaly.

The ultrasonographic finding of ventricular dilatation in the absence of
intraventricular blood is a risk factor for cerebral palsy97 and has been shown to
increase the risk of neurodevelopmental delay 4 to 5 fold”. In this analysis,
infants with VE+HE were at a higher risk of DCP compared to infants with VE.

This is consistent with result of cranial ultrasound studies in VLBW infants which

59

indicate that GM-IVH have a 5 to 9 fold increased risk of WMD regardless of size,
Iaterality, or extent of lesion”.

Isolated dilatation of the lateral ventricles identified by early or late cranial
ultrasound scans in the neonatal period is a risk factor for later development of
cerebral palsy 9°' 97. Furthermore, late onset ventricular enlargement without
periventricular-intraventricular hemorrhage is strongly associoated with white
matter injury in the forms of periventricular Ieukomalacia or cereme cortical
atrophy or both”. Because PVWMI is by far the most important brain lesion
associated with later development of neurologic and cognitive impairments 1,
sonographic evidence of white matter damage is now regarded as a critical piece
of information in the counseling of families on the likely long-term

neurodevelopmental outcomes of their LBW infants.

60

BIBLIOGRAPHY

Pinto-Martin JA, Riolo S, Cnaan A, HoIzman C, Susser MW, Paneth N.
Cranial ultrasound prediction of disabling and nondisabling cereme palsy
at age two in a low birth weight population. Pediatrics. Feb
1995;95(2):249-254.

Allan WC, Holt PJ, Sawyer LR, Tito AM, Meade SK. Ventricular dilation
after neonatal periventricuIar-intraventricular hemorrhage. Natural history
and therapeutic implications. Am J Dis Child. Jul 1982;136(7):589-593.

Ahmann PA, Lazzara A, Dykes FD, Brann AW, Jr., Schwartz JF.
lntraventricular hemorrhage in the high-risk preterm infant: incidence and
outcome. Ann Neurol. Feb 1980;7(2):118—124.

Enzmann D, Murphy-lnlvin K, Stevenson D, Ariagno R, Barton J, Sunshine
P. The natural history of subependymal germinal matrix hemorrhage. Am
J Perinatal. Apr 1985;2(2):123—133.

Kinney HC, Back SA. Human oligodendroglial development: relationship
to periventricular Ieukomalacia. Semin Pediafi Neurol. Sep 1998;5(3):180-
1 89.

Kelly EN, Allen VM, Seaward G, Vlfindrim R, Ryan G. Mild
ventriculomegaly in the fetus, natural history, associated findings and
outcome of isolated mild ventriculomegaly: a literature review. Prenat
Diagn. Aug 2001 ;21(8):697-700.

Hilpert PL, Hall BE, Kurtz AB. The atria of the fetal lateral ventricles: a
sonographic study of normal atrial size and choroid plexus volume. AJR
Am J Roentgenol. Mar 1995;164(3):731-734.

Cardoza JD, Goldstein RB, Filly RA. Exclusion of fetal ventriculomegaly
with a single measurement: the width of the lateral ventricular atrium.
Radiology. Dec 1988;169(3):711-714.

Almog B, Gamzu R, Achiron R, Fainaru O, Zalel Y. Fetal lateral ventricular
width: what should be its upper limit? A prospective cohort study and
reanalysis of the current and previous data. J Ultrasound Med. Jan
2003;22(1):39-43.

61

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Farrell TA, Hertzberg BS, Kliewer MA, Harris L, Paine SS. Fetal lateral
ventricles: reassessment of normal values for atrial diameter at US.
Radiology. Nov 1994;193(2):409-411.

Alagappan R, Browning PD, Laorr A, McGahan JP. Distal lateral
ventricular atrium: reevaluation of normal range. Radiology. Nov
1 994; 1 93(2):405-408.

Dyet LE, Kennea N, Counsell SJ, et al. Natural history of brain lesions in
extremely preterm infants studied with serial magnetic resonance imaging
from birth and neurodevelopmental assessment. Pediatrics. Aug

2006;1 18(2):536-548.

Miller SP, Ferriero DM, Leonard C, et al. Early brain injury in premature
newborns detected with magnetic resonance imaging is associated with
adverse early neurodevelopmental outcome. J Pediatr. Nov

2005; 147(5):609-616.

Achiron R, Schimmel M, Achiron A, Mashiach S. Fetal mild idiopathic
lateral ventriculomegaly: is there a correlation with fetal trisomy?
Ultrasound Obstet Gynecol. Mar 1 1993;3(2):89-92.

Hudgins RJ, Edwards MS, Goldstein R, et al. Natural history of fetal
ventriculomegaly. Pediatrics. Nov 1988;82(5):692-697.

Goldstein RB, La Pidus AS, Filly RA, Cardoza J. Mild lateral cerebral
ventricular dilatation in utero: clinical significance and prognosis.
Radiology. Jul 1990;176(1):237-242.

Nicolaides KH, Berry S, Snijders RJ, Thorpe-Beeston JG, Gosden C. Fetal
lateral cereme ventriculomegaly: associated malformations and
chromosomal defects. Fetal Diagn Ther. 1990;5(1):5-14.

Bromley B, Frigoletto FD, Jr., Benacerraf BR. Mild fetal lateral cerebral
ventriculomegaly: clinical course and outcome. Am J Obstet Gynecol. Mar
1991 ;164(3):863-867.

Patel MD, Filly AL, Hersh DR, Goldstein RB. Isolated mild fetal cerebral
ventriculomegaly: clinical course and outcome. Radiology. Sep
1994;192(3):759—764.

Tomlinson MW, Treadwell MC, Bottoms SF. Isolated mild

ventriculomegaly: associated karyotypic abnormalities and in utero
observations. J Matem Fetal Med. Jul-Aug 1997;6(4):241-244.

62

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

Arora A, Bannister CM, Russell 8, Rimmer S. Outcome and clinical course
of prenatally diagnosed cereme ventriculomegaly. Eur J Pediatr Surg.
Dec 1998;8 Suppl 1:63-64.

Lipitz S, Yagel S, Malinger G, Meizner I, Zalel Y, Achiron R. Outcome of
fetuses with isolated borderline unilateral ventriculomegaly diagnosed at
mid-gestation. Ultrasound Obstet Gynecol. Jul 1998;12(1):23—26.

Robson S, McCormack K, Rankin J. Prenatally detected mild/moderate
cerebral ventriculomegaly: associated anomalies and outcome. Northern
Congenital Abnormality Survey Steering Group. EurJ Pediatr Surg. Dec
1998;8 Suppl 1:70-71.

Vergani P, Locatelli A, Strobelt N, et al. Clinical outcome of mild fetal
ventriculomegaly.[comment]. American Journal of Obstetrics &
Gynecology. 1 998; 1 78(2):21 8-222.

Senat MV, Bernard JP, Schwarzler P, Britten J, Vllle Y. Prenatal diagnosis
and follow-up of 14 cases of unilateral ventriculomegaly. Ultrasound in
Obstetrics & Gynecology. 1999;14(5):327-332.

Laskin MD, Kingdom J, Toi A, Chitayat D, Ohlsson A. Perinatal and
neurodevelopmental outcome with isolated fetal ventriculomegaly: a
systematic review. J Matem Fetal Neonatal Med. Nov 2005;18(5):289-
298.

Bloom SL, Bloom DD, DellaNebbia C, Martin LB, Lucas MJ, Twickler DM.
The developmental outcome of children with antenatal mild isolated
ventriculomegaly. Obstet Gynecol. Jul 1997;90(1):93—97.

Maalouf EF, Duggan PJ, Rutherford MA, et al. Magnetic resonance
imaging of the brain in a cohort of extremely preterm infants. J Pediatr.
Sep 1999;135(3):351-357.

Mirrniran M, Barnes PD, Keller K, et al. Neonatal brain magnetic
resonance imaging before discharge is better than serial cranial
ultrasound in predicting cerebral palsy in very low birth weight preterm
infants. Pediatrics. Oct 2004;114(4):992-998.

Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of
subependymal and intraventricular hemorrhage: a study of infants with
birth weights less than 1,500 gm. J Pediatr. Apr 1978;92(4):529-534.

Volpe JJ. Neurology of the Newborn. 4th ed. Philadelphia: WB Saunders
Company; 2001.

63

32.

33.

34.

35.

36.

37.

38.

39.

41.

42.

du Plessis AJ. Posthemonhagic hydrocephalus and brain injury in the
preterm infant: dilemmas in diagnosis and management. Semin Pediatr
Neural. Sep 1998;5(3):161-179.

Hill A. Ventricular dilation following intraventricular hemorrhage in the
premature infant. Can J Neural Sci. May 1983;10(2):81-85.

Larroche JC. Sub-ependymal pseudo-cysts in the newborn. Biol Neonate.
1972;21(3):170-183.

Volpe JJ, Pasternak JF, Allan WC. Ventricular dilation preceding rapid
head growth following neonatal intracranial hemorrhage. Am J Dis Child.
Nov 1977;131(11):1212-1215.

Allan WC, Dransfield DA, Tito AM. Ventricular dilation following
periventricular-intraventricular hemorrhage: outcome at age 1 year.
Pediatrics. Feb 1984;73(2):158-162.

Hill A, Volpe JJ. Seizures, hypoxic-ischemic brain injury, and
intraventricular hemorrhage in the newborn. Ann Neural. Aug
1981 ;10(2):109-121.

Vasileiadis GT, Gelman N, Han VK, et al. Uncomplicated intraventricular
hemorrhage is followed by reduced cortical volume at near-temi age.
Pediatrics. Sep 2004;114(3):e367-372.

Patra K, Vlfllson-Costello D, Taylor HG, Mercuri-Minich N, Hack M. Grades
l-ll intraventricular hemorrhage in extremely low birth weight infants:
effects on neurodevelopment. J Pediatr. Aug 2006;149(2):169-173.

Garfinkel E, Tejani N, Boxer HS, et al. Infancy and early childhood follow-
up of neonates with periventricular or intraventricular hemorrhage or
isolated ventricular dilation: a case controlled study. Am J Perinatal. Jul
1988;5(3):214-219.

Flodmark O, Scotti G, Harwood-Nash DC. Clinical significance of
ventriculomegaly in children who suffered perinatal asphyxia with or
without intracranial hemorrhage: an 18 month follow-up study. J Comput
Assist Tomogr. Oct 1981;5(5):663-673.

Debillon T, N‘Guyen S, Muet A, Quere MP, Moussaly F, Roze JC.
Limitations of ultrasonography for diagnosing white matter damage in
preterm infants. Arch Dis Child Fetal Neonatal Ed. Jul 2003;88(4):F275-
279.

47.

49.

50.

51.

52.

53.

Levene MI. Measurement of the growth of the lateral ventricles in preterm
infants with real-time ultrasound. Arch Dis Child. Dec 1981;56(12):900-
904.

Pellicer A, Cabanas F, Garcia-Alix A, Perez Rodriguez J, Quero J. Natural
history of ventricular dilatation in preterm infants: prognostic significance.
Pediatr Neural. Mar-Apr 1993;9(2):108-1 14.

Smith WL, McGuinness G, Cavanaugh D, Courtney S. Ultrasound
screening of premature infants: longitudinal follow-up of intracranial
hemorrhage. Radiology. May 1983;147(2):445-448.

Ment LR, Duncan CC, Scott DT, Ehrenkranz RA. Posthemonhagic
hydrocephalus. Low incidence in very low birth weight neonates with
intraventricular hemorrhage. J Neurosurg. Feb 1984;60(2):343-347.

Larroque B, Marret S, Ancel PY, et al. White matter damage and
intraventricular hemorrhage in very preterm infants: the EPIPAGE study. J
Pediatr. Oct 2003; 143(4):477-483.

Ancel PY, Livinec F, Larroque B, et al. Cereme palsy among very preterm
children in relation to gestational age and neonatal ultrasound
abnormalities: the EPIPAGE cohort study. Pediatrics. Mar

2006;1 17(3):828-835.

De Felice C, Toti P, Laurini RN, etal. Early neonatal brain injury in
histologic chorioamnionitis. J Pediatr. Jan 2001;138(1):101-104.

Kuban K, Sanocka U, Leviton A, et al. White matter disorders of
prematurity: association with intraventricular hemorrhage and
ventriculomegaly. The Developmental Epidemiology Network. J Pediatr.
May 1999; 1 34(5):539-546.

Aziz K, Vlckar DB, Sauve RS, Etches PC, Pain KS, Robertson CM.
Province-based study of neurologic disability of children weighing 500
through 1249 grams at birth in relation to neonatal cerebral ultrasound
findings. Pediatrics. Jun 1995;95(6):837-844.

Paneth N, Rudelli R, Kazam E, Monte W. Brain damage in the preterm
infant Cambridge University Press; 1994.

Pinto-Martin J, Paneth N, Witomski T, et al. The central New Jersey

neonatal brain haemorrhage study: design of the study and reliability of
ultrasound diagnosis. Paediatr Perinat Epidemiol. Apr 1992;6(2):273-284.

65

55.

56.

57.

58.

59.

60.

61.

62.

63.

65.

Paneth N, Pinto-Martin J, Gardiner J, et al. Incidence and timing of
germinal matrix/lntraventricular hemorrhage in low birth weight infants. Am
J Epidemiol. Jun 1 1993;137(11):1167-1176.

Qiu H, Paneth N, Lorenz JM, Collins M. Labor and delivery factors in brain
damage, disabling cerebral palsy, and neonatal death in Iow-birth-weight
infants. Am J Obstet Gynecol. Oct 2003;189(4):1143-1149.

Heron M, Sutton PD, Xu J, Ventura SJ, Strobino DM, Guyer B. Annual
summary of vital statistics: 2007. Pediatrics. Jan; 125(1):4—1 5.

Pinto J, Paneth N, Kazam E, et al. lnterobserver variability in neonatal
cranial ultrasonography. Paediatr Perinat Epidemiol. Jan 1988;2(1):43-58.

Bhushan V, Paneth N. The reliability of neonatal head circumference
measurement. J Clin Epidemiol. 1 991 ;44(1 0): 1 027-1 035.

Leviton A, Paneth N. White matter damage in preterm newboms—an
epidemiologic perspective. Ear/y Hum Dev. Oct 1990;24(1):1-22.

Paneth N, Rudelli R, Monte W, et al. White matter necrosis in very low
birth weight infants: neuropathologic and ultrasonographic findings in
infants surviving six days or longer. J Pediatr. Jun 1990;116(6):975—984.

Hack M, Fanaroff AA. Outcomes of children of extremely low birthweight
and gestational age in the 1990‘s. Early Hum Dev. Jan 1999;53(3):193-
218.

Fanaroff AA, Hack M, Walsh MC. The NICHD neonatal research network:
changes in practice and outcomes during the first 15 years. Semin
Perinatal. Aug 2003;27(4):281-287.

Philip AG, Allan WC, Tito AM, Wheeler LR. lntraventricular hemonhage in
preterm infants: declining incidence in the 1980s. Pediatrics. Nov
1 989;84(5):797-801 .

Changing outcome for infants of birth-weight 500-999 g born outside level
3 centres in Victoria. The Vlctorian Infant Collaborative Study Group. Aust
N Z J Obstet Gynaecol. Aug 1997;37(3):253-257.

Wood NS, Marlow N, Costeloe K, Gibson AT, Wilkinson AR. Neurologic
and developmental disability after extremely preterm birth. EPICure Study
Group. N Engl J Med. Aug 10 2000;343(6):378-384.

\Nllson-Costello D, Friedman H, Minich N, Fanaroff AA, Hack M. Improved
survival rates with increased neurodevelopmental disability for extremely

66

67.

69.

70.

71.

72.

73.

74.

75.

76.

77.

low birth weight infants in the 19903. Pediatrics. Apr 2005;115(4):997-
1 003.

Lorenz JM, Wooliever DE, Jetton JR, Paneth N. A quantitative review of
mortality and developmental disability in extremely premature newborns.
Arch Pediatr Adolesc Med. May 1998;152(5):425435.

Volpe JJ. Cerebral white matter injury of the premature infant-rnore
common than you think. Pediatrics. Jul 2003;112(1 Pt 1):176-180.

Inder TE, Wells SJ, Mogridge NB, Spencer C, Volpe JJ. Defining the
nature of the cereme abnormalities in the premature infant: a qualitative
magnetic resonance imaging study. J Pediatr. Aug 2003;143(2):171-179.

Reynolds PR, Dale RC, Cowan FM. Neonatal cranial ultrasound
interpretation: 8 clinical audit. Arch Dis Child Fetal Neonatal Ed. Mar
2001 ;84(2):F92-95.

Kriss VM, Desai NS. Nonpathologic ventriculomegaly in the premature
intrauterine growth retarded infant. J Perinatal. Jul-Aug 1997;17(4):292-
295.

Baud O, Zupan V, Lacaze-Masmonteil T, et al. The relationships between
antenatal management, the cause of delivery and neonatal outcome in a
large cohort of very preterm singleton infants. Bjag. Jul 2000;107(7):877-
884.

Zupan V, Gonzalez P, Lacaze-Masmonteil T, et al. Periventricular
Ieukomalacia: risk factors revisited. Dev Med Child Neural. Dec
1996;38(12):1061-1067.

Ment LR, Schwartz M, Makuch RW, Stewart WB. Association of chronic
sublethal hypoxia with ventriculomegaly in the developing rat brain. Brain
Res Dev Brain Res. Dec 7 1998;111(2):197-203.

Yu W, Wood C. Perinatal asphyxia and outcome of very low birthweight
infants. Med J Aust. Dec 16 1978;2(13):578-581.

Naicker T, Khedun SM, Moodley J, Pijnenborg R. Quantitative analysis of
trophoblast invasion in preeclampsia. Acta Obstet Gynecol Scand. Aug
2003;82(8):722-729.

Kam EP, Gardner L, Loke YW, King A. The role of trophoblast in the

physiological change in decidual spiral arteries. Hum Reprod. Aug
1999;14(8):2131-2138.

67

78.

79.

80.

81.

82.

83.

84.

87.

Roberts JM, Taylor RN, Goldfien A. Clinical and biochemical evidence of
endothelial cell dysfunction in the pregnancy syndrome preeclampsia. Am
J Hypertens. Aug 1991;4(8):700—708.

Redman CW, Sacks GP, Sargent IL. Pmeclampsia: an excessive
maternal inflammatory response to pregnancy. Am J Obstet Gynecol. Feb
1999;180(2 Pt 1):499-506.

Collins M, Paneth N. Preeclampsia and cereme palsy: are they related?
Dev Med Child Neural. Mar 1998;40(3):207-211.

Dammann O, Allred EN, Veelken N. Increased risk of spastic diplegia
among very low birth weight children after preterm labor or prelabor
rupture of membranes. J Pediatr. Mar 1998;132(3 Pt 1):531-535.

Kuban KC, Leviton A, Pagano M, Fenton T, Strassfeld R, Wolff M.
Maternal toxemia is associated with reduced incidence of germinal matrix
hemorrhage in premature babies. J Child Neural. Jan 1992;7(1):70-76.

Leviton A, Pagano M, Kuban KC, Krishnamoorthy KS, Sullivan KF, Allred
EN. The epidemiology of germinal matrix hemorrhage during the first half-
day of life. Dev Med Child Neural. Feb 1991;33(2):138-145.

Ancel PY, Marret S, Larroque B, et al. Are maternal hypertension and
small-for-gestational age risk factors for severe intraventricular
hemorrhage and cystic periventricular Ieukomalacia? Results of the
EPIPAGE cohort study. Am J Obstet Gynecol. Jul 2005;193(1):178-184.

Leviton A, Paneth N, Susser M, et al. Maternal receipt of magnesium
sulfate does not seem to reduce the risk of neonatal white matter damage.
Pediatrics. Apr 1997;99(4):E2.

McEIrath TF, Allred EN, Boggess KA, et al. Matemal antenatal
complications and the risk of neonatal cereme white matter damage and

later cerebral palsy in children born at an extremely low gestational age.
Am J Epidemiol. Oct 1 2009; 1 70(7):81 9-828.

Resch B, Vollaard E, Maurer U, Haas J, Rosegger H, Muller W. Risk
factors and determinants of neurodevelopmental outcome in cystic
periventricular Ieucomalacia. EurJ Pediatr. Sep 2000;159(9):663-670.

Friedman SA, Schiff E, Kao L, Sibai BM. Neonatal outcome after preterm

delivery for preeclampsia. Am J Obstet Gynecol. Jun 1995;172(6):1785-
1 788; discussion 1 788-1792.

68

90.

91.

92.

93.

94.

95.

96.

97.

99.

Murphy DJ, Squier MV, Hope PL, Sellers S, Johnson A. Clinical
associations and time of onset of cerebral white matter damage in very
preterm babies. Arch Dis Child Fetal Neonatal Ed. Jul 1996;75(1):F27-32.

Allan WC, Vohr B, Makuch RW, Katz KH, Ment LR. Antecedents of
cereme palsy in a multicenter trial of indomethacin for intraventricular
hemorrhage. Arch Pediatr Adolesc Med. Jun 1997; 1 51 (6):580-585.

McEIrath TF, Allred EN, Leviton A. Prolonged latency after preterm
premature rupture of membranes: an evaluation of histologic condition and
intracranial ultrasonic abnormality in the neonate born at <28 weeks of
gestation. Am J Obstet Gynecol. Sep 2003;189(3):794—798.

Shaver DC, Bada HS, Korones SB, Anderson GD, Wong SP, Arheart KL.
Early and late intraventricular hemorrhage: the role of obstetric factors.
Obstet Gynecol. Nov 1992;80(5):831-837.

Szymonowicz W, Yu W. Timing and evolution of periventricular
haemorrhage in infants weighing 1250 g or less at birth. Arch Dis Child.
Jan 1984;59(1):7-12.

Inder TE, Warfield SK, Wang H, Huppi PS, Volpe JJ. Abnormal cereme
structure is present at term in premature infants. Pediatrics. Feb
2005;1 15(2):286-294.

Volpe JJ. Neurobiology of periventricular Ieukomalacia in the premature
infant. Pediatr Res. Nov 2001;50(5):553-562.

Flodmark 0, Roland EH, Hill A, Whitfield MF. Periventricular Ieukomalacia:
radiologic diagnosis. Radiology. Jan 1987;162(1 Pt 1):119-124.

Murphy DJ, Hope PL, Johnson A. Neonatal risk factors for cereme palsy
in very preterm babies: case-control study. ij. Feb 8
1 997;314(7078):404-408.

Stewart AL, Reynolds EO, Hope PL, et al. Probability of
neurodevelopmental disorders estimated from ultrasound appearance of
brains of very preterm infants. Dev Med Child Neural. Feb 1987;29(1):3—
1 1.

Inder TE, Huppi PS, Warfield S, et al. Periventricular white matter injury in

the premature infant is followed by reduced cerebral cortical gray matter
volume at term. Ann Neurol. Nov 1999;46(5):755-760.

69