Brain network analysis in preterm infants using electro-encephalography

In the Netherlands 1-2% of live-born infants are born very preterm (<32 weeks
of gestation). Multiple factors, such as hypoxia, infection and inflammation,
can damage the extremely vulnerable brain of preterm infants and cause brain
injury such as periventricular leukomalacia (PVL) and intraventricular
haemorrhage (IVH). This brain injury is associated with compromised long term
developmental outcome and social behaviour in preterm children. Although the
neurological complications of very preterm birth can be visualised, to a
certain extent, using cranial ultrasound and MRI of the brain in early stages,
it remains difficult to give an accurate prognosis of neurodevelopment for the
individual patient.

In addition to structural investigation of brain damage related to preterm
birth, brain activity can be measured using electro-encephalography (EEG),
magnetic-encephalography (MEG) or functional MRI (fMRI). From this activity,
functional brain networks can be derived by measuring the strength, and
sometimes direction, of the functional coupling between different brain
regions. The architecture of functional brain networks is associated with
cognitive performance in adults, especially in complex cognitive functions,
such as attention, executive functioning and memory.Recent studies show that
during development in healthy children the organisation of these functional
networks mature from more random to efficient *small world* network structures,
necessary for optimal cognitive functioning, and that in preterm born children
differences in these networks can be demonstrated at the age of 6 years.

Better insight in the pattern of functional brain network disruption associated
with very preterm birth can lead to a better understanding of the underlying
processes and may lead to improved individual prognostication and eventually,
may serve as a marker for new intervention strategies in this patient group.

We hypothesize that functional brain network measures may serve as a marker for
(disrupted) neurodevelopment in very preterm born infants.

We hypothesize that functional brain network measures may serve as a marker for
(disrupted) neurodevelopment in very preterm born infants.

Objective: we propose to perform two studies aiming to answer the following
questions.

Phase 1: Retrospective study:
Detecting and characterizing functional brain networks in a spectrum from very
preterm to term born infants (cross-sectional). How can we optimally perform
functional brain network analyses in this type of patients; which network
measures can be optimally used in these young infants? Can we estimate sample
sizes needed for future studies based on the observed results? How do
functional brain networks relate to gestational age?


Phase 2: Prospective study:
1 Detecting and characterizing functional brain network differences in very
preterm born infants at TEA and at 2 years in comparison to term born controls
(cross-sectional).
2 Detecting and characterizing longitudinal functional brain network
development changes between baseline and the (corrected) age of 2 years.
Questions that are addressed are: is there a change in brain network topology
over time. And if that is the case is there a difference in development of
brain networks between the preterm and term born children? (longitudinal)
3 Relating functional brain network changes to specific neurodevelopmental
deficits. If patterns of change in functional brain network topology are found,
can a relationship with neurological performance be established at the
(corrected) age of 2 years? And are there region specific changes related to
certain neurodevelopmental deficits?

Phase 1: Explorative retrospective study
In the database of the department of clinical neurophysiology EEGs of
approximately 75 neonates performed from 2012 until 2015 are available for
exploring the optimal methods for network analyses in the very young. These
EEGs were performed in routine clinical practice in infants admitted at the
NICU of the VUMC (from 25 weeks of gestational age to term born patients). All
EEG data and clinical information will be analysed anonymously.

Phase 2: Prospective study

Preterm born infants from the NICU at the VUMC (n=10-20, gestational age: 28-32
weeks) and a healthy, term born, control group (n=10-20) from the maternity
ward at the VUMC (healthy term borns who are admitted because of maternal
indication only) will be investigated. These are relatively small numbers,
after we have conducted the retrospective study we might be able to better
estimate accurate sample sizes for the current study. We estimate that at the
NICU of the VUmc inclusion will take approximately 12 months. At baseline (TEA
for preterm born infants and in the first week after birth for the term borns)
and at 2-year follow-up an EEG will be recorded. At the age of 2 years
neurodevelopment will be assessed using the Bailey Scales of Infant Development
III (BSIDIII). Preterm born infants will be investigated at TEA and at the
corrected age of 2 years. Term born infants will be investigated within the
first week of life and at the age of 2 years. To evaluate possible brain
injury, which could influence network development, all patients will undergo a
cerebral ultrasound at TEA for preterm infants and in the maternity ward for
term born infants.

EEGs are routinely used diagnostic procedures in standard clinical care at the
NICU at the VUmc. EEG measurements are non-invasive, taking approximately 45-60
minutes. The procedure is not considered to be painful or difficult and has
negligible risks. Regular clinical follow-up of the preterm born infants
includes a BSIDIII assessment at the corrected age of 2 years and we will ask
parents for permission to register a second EEG in addition. The term born
infants are invited for a second EEG and a BSIDIII assessment at the VUMC at
the age of 2 years, meaning one extra trip to the hospital. All participants
will be examined using cerebral ultrasound, which is also a routine diagnostic
procedure in infants. It is a non-invasive, pain-free examination with
negligible risks. There is no individual benefit from the EEG or the cerebral
ultrasound. Unexpected findings which could lead to medical treatment will be
discussed with caregivers of participants.