Open access Original research bmjpo: first published as 10.1136/bmjpo-2020-000648 on 30 July 2020. Downloaded from Pilot study to establish a prospective neonatal cohort: Study of Preterm Infants and Neurodevelopmental Genes (SPRING) Hilary S Wong ,1 Lucinda Hopkins,2 Michael C O'Donovan,2 Anita Thapar,2 Neena Modi3 To cite: Wong HS, Hopkins L, ABSTRACT What is known about the subject? O'Donovan MC, et al. Background Genetic risk variants and preterm birth Pilot study to establish a are early and potent risk factors for later neuropsychiatric ► Infants born very preterm are at increased risk of prospective neonatal cohort: disorders. To understand the interrelationships between Study of Preterm Infants and developing neuropsychiatric conditions such as these factors, a large- scale genetic study of very preterm Neurodevelopmental Genes attention-deficit hyperactivity disorder and autism (VPT, <32 weeks gestation) infants with prospective (SPRING). BMJ Paediatrics Open spectrum disorder. follow- up is required. In this paper, we describe a 2020;4:e000648. doi:10.1136/ ► A large number of genetic risk variants have been streamlined study approach, using efficient processes for bmjpo-2020-000648 identified for neuropsychiatric disorders. biological and clinical data collection, to feasibly establish ► Additional material is such a cohort. published online only. To view, Methods We sought to recruit 500 VPT families within a please visit the journal online 1 year period from neonatal units. Treating clinical teams What this study adds? (http:// dx. doi. org/ 10. 1136/ recruited eligible participants, obtained parent consent, bmjpo- 2020- 000648). collected blood samples and posted specimens to the ► A large- scale efficient and cost- effective cohort research laboratory. We extracted all clinical data from study of preterm infants for genetic investigations in Received 30 January 2020 the National Neonatal Research Database, an existing UK the UK is feasible. http://bmjpaedsopen.bmj.com/ Revised 30 April 2020 resource that captures daily patient- level data on all VPT ► Longitudinal health and education outcome ascer- Accepted 15 May 2020 infants. tainment of very preterm infants through routine Results Between May 2017 and June 2018, we data linkage are acceptable to parents. established a cohort of 848 VPT infants and their parents from 60 English neonatal units. The study population (median (IQR), gestation: 28.9 (26–30) weeks; birth weight: 1120 (886–1420) g) represented 18.9% of eligible infants behavioural and mental health problems, born at the study sites during the recruitment period with approximately a quarter of children (n=4491). From the subset of 521 complete family trios, born extremely preterm meeting neuropsy- 5 we successfully completed genotyping for 510 (97.9%) chiatric diagnostic criteria by childhood. trios. Of the original 883 infants whose parents consented However, there is considerable heteroge- on September 26, 2021 by guest. Protected copyright. to participate, the parents of 796 (90.1%) infants agreed to neity in outcomes, the causes of which are future data linkage and 794 (89.9%) agreed to be recalled. unknown. © Author(s) (or their Conclusion We demonstrate the feasibility and employer(s)) 2020. Re- use Previous epidemiological studies have permitted under CC BY. acceptability of streamlined strategies for genetic, neonatal suggested family and genetic factors may Published by BMJ. and longitudinal data collection and provide a template for contribute to the link between preterm birth 1 future cost- effective and efficient cohort development. Department of Paediatrics, and later neuropsychiatric disorders. Many University of Cambridge School, of these studies have observed an associa- Cambridge, UK 2 BACKGROUND tion between maternal mental health prob- MRC Centre for 6 7 Neuropsychiatric Genetics and Globally, approximately 15 million infants lems and risk of preterm birth. Positive Genomics, Cardiff University, are born preterm (<37 weeks gestation) family history of psychiatric disorder further Cardiff, UK each year and the prevalence is rising.1 The contributes to the risk of adverse mental 3 Section of Neonatal Medicine, causes of spontaneous preterm births and health outcomes among individuals born Imperial College London the conditions justifying ‘medically indicated’ preterm.8 9 Genomic studies have identified Department of Medicine, London, UK preterm birth are complex and poorly under- large numbers of genetic risk variants for stood, with both environmental factors2 3 and major neuropsychiatric disorders, including 4 Correspondence to genetic components implicated. Preterm those for attention- deficit hyperactivity Professor Neena Modi; n. modi@ infants experience significant adversities and disorder and autism spectrum disorder, which imperial. ac. uk are at elevated risk of intellectual disability, are the neuropsychiatric disorders most Wong HS, et al. BMJ Paediatrics Open 2020;4:e000648. doi:10.1136/bmjpo-2020-000648 1 Open access bmjpo: first published as 10.1136/bmjpo-2020-000648 on 30 July 2020. Downloaded from strongly associated with preterm birth in childhood.10 for trio- based genome analysis, single-parent families were However, the relevance of these genetic risk variants for also included. Similarly, we recruited infants where one those born preterm has not been investigated. or both parents declined to participate in the study them- A large preterm birth cohort with genomic data and selves. We did not recruit parents known to be genetically prospective clinical phenotyping would be required unrelated to the child, for instance if the pregnancy had to investigate the relationship between genetic risks, arisen from a donor gamete. Potential participants could preterm birth and neuropsychiatric outcomes. There is be approached and recruited at any point during their no sufficiently powered existing prospective cohort of neonatal hospitalisation at the discretion of the local preterm infants for such investigations and the costs, research staff. Informed written consent was obtained time and resource requirements for recruitment and data from each parent (for themselves) and their offspring at collection by established methods are likely to be prohib- the local site. We sought to recruit 500 VPT infants over itive. Our aims were to examine the feasibility of large- the 1 year recruitment phase. scale recruitment and DNA collection in a multicentre setting, by capitalising on existing research infrastructure Biological specimen collection and laboratory processing in the UK to streamline operational efficiencies, and the Blood samples from participants were collected into acceptability to parents of prospective follow- up through EDTA specimen bottles. For infants, 0.5 to 1 mL of blood routine data linkage and direct recall for detailed pheno- were collected by heel lance, venepuncture or from an typing. Our secondary aim was to conduct a prelimi- indwelling arterial catheter during blood sampling for nary investigation to explore the hypothesis that the routine clinical care to avoid an additional intervention very preterm (VPT, <32 weeks gestation) population is for the infant. If the target volume was not achieved, enriched for rare pathogenic copy number variants; the a second aliquot was obtained, also collected during results of this investigation were reported in a separate routine sampling. Five millilitre of blood was collected paper.11 from each parent. The blood specimens were labelled using barcodes and a unique participant study number, packaged in a designated mailing box for biological spec- METHODS imens, in accordance with UN3373 regulations and P650 Study sites and recruitment packaging instructions, and posted to the research labo- We invited expressions of interest in participation as study ratory at the MRC Centre for Neuropsychiatric Genetics sites from all 43 neonatal intensive care units (NICUs) and Genomics (MRC CNGG), Cardiff University within and 77 local neonatal units (LNUs) in England. In the 7 days of sample collection using UK standard postal http://bmjpaedsopen.bmj.com/ UK, neonatal care is provided within managed clinical services. No laboratory processing of the samples took networks where LNUs provide the full range of care, place at the recruitment study sites. including short periods of intensive care, for babies deliv- DNA was extracted from the blood specimens in-house ered in the nominal catchment area of the hospital.12 at the MRC CNGG, from full sets of parent–infant trios in Most babies born over 27 weeks gestation will usually the first instance, using GE Healthcare DNA extraction receive their full care in their LNU. Babies who are more kits. DNA sample quantification was determined using preterm or those who require complex or longer-term Quant- iT PicoGreen dsDNA assay kits and samples intensive care are transferred to a NICU, where tertiary were genotyped using Illumina OmniExpress BeadChip specialist care is provided. From the units that expressed arrays. Samples with concentrations below 50 ng/μL interest, we selected study sites based on the potential were excluded from genotyping. DNA was considered of on September 26, 2021 by guest. Protected copyright. number of eligible participants, geographical coverage adequate quality on successful completion of array- based and record of accomplishment of research activities in genotyping. the unit. The study was included in the National Insti- tute for Health Research Clinical