Using genome-wide gene expression profiles for early identification of encephalopathic babies at risk of adverse neurological outcomes

Aim:
I will examine if the body (host) response (activity of specific genes in blood) in babies who are born in a poor condition (perinatal asphyxia), can be used to identify those who are likely to have long-term neurodisability.

Background:
We know that babies who suffer from a lack of blood flow and restricted oxygen to the brain around the time of birth (perinatal asphyxia), are at a high risk of long term brain damage. Lowering the infant's body temperature (cooling therapy) soon after birth is currently used as standard care for these babies, but it does not work in all of them. Up to half the cooled babies still develop long-term disability, and this is not always directly related to the severity of perinatal asphyxia, or how unwell the babies are at the time of birth. Rapid identification of the babies who will not respond to cooling therapy, and understanding why they do not, is fundamental for the development of future neuroprotective therapies. My preliminary research has shown that a perinatal asphyxia results in unique changes to genetic activity in the blood, which correlates with brain injury on MRI. I now wish to test the hypothesis that gene activity in the blood at the time of birth, can accurately identify babies with perinatal asphyxia who are at risk of long-term neurodisability.

Design and Methods:
I will collect a small amount of blood soon after birth, from a total of 130 term babies with perinatal asphyxia recruited from five NHS hospitals, over a one-year period. These babies will have a detailed medical examination at 18 months of age, to examine for any neurodisability. In the first part of my project, I will use specific statistical methods to identify a minimal set of activated or deactivated genes that are associated with long-term disability (gene signature). In the second part, I will test the accuracy of this signature in predicting long-term disability. Finally, I will examine the specific biological function of the genes (pathways) activated in babies with long-term neurodisability.

Benefits:
If successful, this work will eventually lead to a paradigm shift in the way we identify and treat babies with perinatal asphyxia, and will open up a new avenue for developing individualised neuroprotective therapies. Once the specific gene expression signature is trimmed down to a handful of genes, it can then be easily developed as a rapid bedside diagnostic test, using polymerise chain reaction (PCR). In the first instance, this test can identify babies who may not respond to cooling therapy alone, and hence may need additional therapies, but also infants who do not meet the current cooling criteria (for example mild encephalopathy), but still develop adverse outcomes. The biological pathways identified in this work will help the development of future neuroprotective treatments. Finally, at present we are able to identify only a small proportion of babies at the time birth, who later develop cerebral palsy. Once developed as a bedside test, it may have a role in universal neonatal screening for such infants. Given that perinatal asphyxia can cost £750,000 for each child with cerebral palsy, any reduction in these numbers would have substantial health and economic benefits.

Patient and Public Involvement:
Discussions with parent groups and the Bliss charity during the project development helped me to focus more on patient centred outcomes - i.e. an early prediction of adverse outcome from a blood test by itself was not particularly useful for parents, and they were much more interested in how this could modify any current or future treatments available. Quarterly PPI meetings will be conducted during the course of this project, assisting with dissemination of the results and guiding future steps of the research. I have set up an advisory group of parents whose children had perinatal asphyxia and they helped me with designing study documents.

Background:
Rapid identification and risk stratification of babies with perinatal asphyxia is required for targeted neuroprotective therapies. Minimal host transcript signature from genome wide expression is currently being developed as a point of care test in several diseases, including bacterial infections. I have recently identified a host transcript signature of perinatal hypoxic-ischemic insult that correlates with brain injury on magnetic resonance imaging.

Aims:
To examine if a host transcript signature at birth can accurately predict adverse neurological outcomes at 18 months, after neonatal encephalopathy.

Methods:
I will collect 0.5 ml blood for gene expression (Next Generation Sequencing) within six hours of birth, from a prospective multi-centric cohort of 130 encephalopathic babies. In the discovery part of my project, I will identify a minimal host transcript that is associated with adverse neurological outcomes at 18 months (training set [n= 88]) and examine the prognostic accuracy of this transcript (test set [n= 22]). In the validation part of my project, the most significant differentially expressed genes will be further validated in a different platform (real-time reverse transcription polymerase chain reaction) (n= 20 babies). Finally, I will examine the gene expression pathways associated with long-term neurodisability.

Outcome variables and potential benefits:
Sensitivity and specificity (with confidence intervals) of the minimal host response transcript will be reported, in addition to specific up and down regulated pathways. This work will underpin the development of a rapid diagnostic test for early detection and risk stratification in neonatal encephalopathy, thus enabling individualised neuroprotective therapies. Pathway analysis may identify specific therapeutic targets.

If I am able to identify a minimal host transcript signature of long-term neurodisability, this will result in a paradigm shift in neonatal neuroprotection. Prof Levin's group has recently found a minimal host transcript signature for bacterial infections, and this is currently being developed as a rapid diagnostic test based on polymerise chain reaction. The only limiting factor is automated bedside RNA extraction from blood. A number of researchers at Imperial College and elsewhere are currently working on this issue, and are expected to develop a point of care RNA extraction device within the next five years.

An immediate clinical benefit of this test would be the identification of encephalopathic infants who do not fulfil the cooling therapy criteria (for example mild encephalopathy), but who later develop adverse outcomes. Secondly, this test can identify the babies who may not respond to cooling therapy and may need specific neuroprotective therapies, for example anti-oxidant, or anti-inflammatory therapy. Thirdly, this research can also open newer therapeutic avenues by shedding light on the biological pathways involved in the most severe forms of brain injury. This may lead to further drug development and evaluations, eventually benefitting all babies with neonatal encephalopathy. The therapeutic targets identified would also be of considerable interest to preclinical researchers and pharmaceutical companies. Fourthly, we are only able to identify a minority of babies at the time birth, who later develop cerebral palsy, as many of these babies may be clinically asymptomatic. It is possible that this test may evolve into a universal screening tool (umbilical cord blood) for identification all babies at risk of adverse outcome. Finally, this study will have a major impact for low and middle-income countries where the burden of neonatal encephalopathy is high, once the host transcript signature is developed as a low cost point of care test.

The NHS support cost for infants following neonatal encephalopathy is estimated to be £2.25 billion per year. Therapeutic hypothermia is estimated to have saved the NHS £0.5 billion since its introduction as a standard clinical practice. Therefore, a further 30% improvement in hypothermic neuroprotection by new drug therapies will result in an additional 50% saving in health care costs. Moreover, any resultant drug development and manufacturing will have a positive impact on the UK industry and on the profile of UK clinical trials. This technology has the potential to become an affordable, point of care diagnostic test. Hence, the intellectual property generated in this work would be of substantial interest to the health care industry, which will facilitate commercialisation and clinical use in the UK, and worldwide.