Novel diagnostic markers for necrotising enterocolitis in the premature infant

Necrotising enterocolitis is the most prevalent acute gastro-enterological
disease in the Neonatal Intensive Care Unit (NICU). The disease often
progresses rapidly, with potentially life-threatening complications, e.g. bowel
perforation necessitating laparotomy. Morbidity and mortality are therefore
high, in some series up to 40%. However, diagnosis is often difficult as signs
are often non-specific, and to date specific markers for the disease or its
complications are not available. Early diagnosis allows for early treatment,
which may improve outcome in patients with NEC.
Promising markers for the diagnosis of NEC are Fatty Acid Binding Proteins
(FABPs). FABPs are proteins which are important in the fat-metabolism of the
enterocyte. When enterocytes are damaged, FABPs are released in the circulation
and subsequently secreted by the kidney. This also holds true for parts of the
tight junctions, the bridges between enterocytes. When these are damaged,
Claudin-3 is released. Claudin-3 can also be assessed in urine. Preliminary
data suggest that both urine FABP and Claudin levels (expressed as ratios to
urine Creatinine) are early indicators for loss of gut integrity and therefore
possible early indicators for NEC and its complications. Although I-FABP and
Claudin have been suggested as early markers for NEC, little is known about the
location and quantification of FABP*s in the bowel wall during the development
of (complications of) NEC.

As ischemia and reperfusion are supposedly involved in the pathophysiological
mechanisms underlying NEC, early identification of ischemia of the gut can be
of vital importance.This is possible using near-infrared spectroscopy (NIRS).
NIRS is a continuous, non-invasive portable technique allowing for the
determination of regional tissue oxygenation and thus the presence of
ischemia. NIRS is increasingly being used to monitor cerebral oxygenation, but
can also be used to monitor oxygenation of abdominal tissues. As the cerebral
oxygenation of the brain remains relatively constant, the ratio between local
oxygenation and cerebral oxygenation has been shown to provides an adequate
estimate of local tissue oxygenation.

It is known that neonates with NEC have a disturbed intestinal flora. As yet,
it is unknown whether this is cause or effect of the disease. Although regular
culture is able to identify most pathogens, there are pathogens which can not
be cultured using the routine techniques. Recently a new technique has been
developed to investigate the presence of bacteria on the DNA/RNA level using
the so-called denaturing gradient gel electrophoresis (DGGE)
GGE-16SrRNA-technique. This offers the possibility to investigate pathogens
which can not be identified using routine culturing techniques.

Finally, cytokines may allow for early diagnosis of NEC. Increased serum levels
of cytokines are found in early phases of inflammatory conditions. As yet, many
assays for cytokines can only be performed on serum samples. Recently, new
techniques have become available which makes it possible to analyse urine
cytokine profiles. When this method would also be possible in neonates,
obtaining bloodsamples for these assays will not be necessary anymore.

The overall objective of this study is an improvement in (early) diagnosis of
NEC and its life-threatening complications, thus to be able to detect children
who are *at risk* for NEC and its complications as early as possible.

To this end we will as a first objective assess urine FABP/Claudin to
Creatinine ratio as early markers for NEC and its possible complications and
compare them with regional tissue oxygenation as measured by NIRS. This should
give insight in the relation between tissue oxygenation and enterocyte damage.
It also should provide insight in the use of NIRS and FABP/Claudin levels as
diagnostic markers for (complicated) NEC and might offer new modalities to
guide therapy based on these results.

Second objective is to correlate the FABP/Claudin levels and NIRS data with the
other serum markers of the inflammatory and metabolic response such as serum
IL1, IL6, IL10, TNFa and lactate and base excess.

Third objective is - when surgery has become necessary - to analyse resectional
specimens for the exact location and quantification of FABPs.

Fourth objective is to validate the Biosource* Multiplex Assay (Invitrogen), a
novel assay to test multiple cytokines and chemokines in small volumes of
urine, by comparing urine results with results obtained in serum.

Fifth objective is to gain insight in the inflammatory and metabolic process,
including gut wall integrity and cytokine profiles during the development of
NEC and its complications.

Sixth objective is is to assess faecal bacterial flora using both routine
cultures and the DGGE-16SrRNA-technique.

Prospective cohort study

When there is a suspicion of NEC, urine samples and blood will be collected
every six to eight hours and abdominal X-rays will be performed upon the first
suspicion of NEC until confirmation of the diagnosis by either clinical or
surgical means. This is, except for the urine samples, according to routine
clinical practice. The diagnosis NEC will be established using the Bells
criteria, with the presence of NEC defined as signs of pneumatosis intestinalis
on the abdominal X-ray as assessed by an independent pediatric radiologist.(see
appendix 1)
For the serum analysis 100µl of blood will suffice. This can be obtained during
the daily routine bloodtests without any further burden for the child. Urine
(at least 3 ml) is collected via a small cotton wool placed into the diaper for
at least one hour. Urine is subsequently removed from the cotton wool by
compressing the cotton wool in a syringe.
At the same time as the interventions mentioned above, regional tissue
oxygenation of the gut and the brain is evaluated using the NIRS system.
Daily 30 grams of faeces will be collected via the diaper for culture.
Specimens will be stored at -80°C and subsequently analysed in the MUMC.

After the confirmation of NEC, children will be examined every eight hours
using the parameters mentioned above until the time of first resumption of
enteral feeding or surgery, whichever comes first. When surgery will become
necessary, patients will also be assessed every eight hours postoperatively,
until the first enteral feeding is resumed.
After enteral feeding has been started, blood/urine and NIRS data will be
collected daily until full enteral feeding is resumed. This holds true for both
the surgical and non-surgical NEC patients. In patients in whom the diagnosis
NEC can be discarded, urine/blood/NIRS data will be obtained daily until
resumption of full enteral feeding.

Surgery

When surgery becomes necessary and bowel is resected, small parts of the
resected bowel and adjacent normal tissue will be collected. One part is snap
frozen, the other part is stored in Formalin. We will use routine
Hematoxin-Eosin staining for histologic evaluation and use immunohistochemistry
for the determination of the localisation of the FABPs. These proteins will
also be assessed on the mRNA and/or protein level. We will also perform
immunohistochemistry for carbohydrases to examine the gut maturity. All
resectional specimens will be analysed in the Maastricht University Medical
Center.

Urine (at least 3 ml) will be collected via tissues placed in the diaper.
Patient will therefore not undergo extra procedures. Regarding the bloodtests:
100µl will be enough, which can be obtained during the daily routine bloodtests
without any further burden for the child. NIRS measurements are non-invasive
and do not interfere with routine care.
Data from this study can not be obtained in another population, as (premature)
neonates are the only children who develop NEC. The results from this study
might offer new and non-invasive tools for the early diagnosis of NEC, or
complications of NEC, thereby improving allowing for early treatment and
improving outcome.