We are pleased to announce the selected performers.
Every 16 seconds one baby is stillborn. That amounts to more than two million stillborn babies globally every year.i Stillbirths have long-lasting personal and psychological consequences for parents, as well as substantial costs for wider society.ii
“Experiencing a stillbirth during pregnancy or childbirth is a tragedy insufficiently addressed in global agendas, policies and funded programmes. There are psychological costs to women, especially women, and their families, such as maternal depression, financial consequences and economic percussions, as well as stigma and taboo.”
Early recognition of emerging complications in utero, coupled with timely and safe delivery, is estimated to have the potential to reduce the number of stillborn babies by half. Yet progress to reduce stillbirth remains stubbornly slow. In sub-Saharan Africa headway in reducing stillbirth rates has been outpaced by growth in the total number of births, so stillbirth numbers are actually rising.i In the USA stillbirth rates have been static for more than a decade, which amounts to a total of 12,000 stillborn babies each year. Every child’s death is heartbreaking, and this number of stillbirths is ten times higher than the annual number of deaths from childhood cancer.v
Worldwide, great strides are being made in reducing the number of baby deaths that occur after birth, but reductions in baby deaths that occur before birth, (stillbirths) are lagging behind.i Globally, in the year 2000 there was 1 stillbirth for every 3 newborn deaths in the first month of life. By 2019, in nearly 50 countries that ratio was more than 1 to 1. For some babies, remaining in utero is higher risk than being born, largely because in utero, life-threatening complications can develop and progress undetected. Our goal is to be able to measure, model and predict gestational development, with a primary focus to reduce stillbirth rates by half. To achieve this we need non-invasive, scalable ways to assess gestational development in utero.
Stillbirth is the endpoint of a number of different processes that involve the mother, baby, or the placenta – or a combination of the three.v The placenta is the life support system of the developing baby. In humans, the placenta couples the separate maternal and developing baby’s circulations to allow transfer of oxygen and nutrients from mother to baby. This placental transfer function is influenced not only by placental size and structure but also by the integrity of the maternal circulation through the uterine vasculature; and adequate circulation through the umbilical cord for the baby. The placenta is also itself metabolically active and secretes bioactive substances and hormones that influence the maternal response to accommodate pregnancy, whilst acting as a barrier to substances, such as viruses and certain drugs, that may damage the developing baby.
The lack of methods to assess gestational development in utero limits our ability to predict the risk of stillbirth. Today 25-50% of stillbirths are unexplained – meaning that no conditions that affect the mother, baby, or placenta that could contribute to the baby’s death are identified. Even in cases where possible contributory conditions are found, it is extremely rare to have sufficient resolution on the sequence, timing and exact mechanisms leading to stillbirth.iv Such inadequate basic understanding restricts opportunities to advance preventative treatments. By developing new measures and models of gestational development, we also will identify new opportunities to prevent stillbirths.
To date, characterisation of gestational development has relied on intermittent, indirect measures. For example, ultrasound assessment of the baby’s growth and/or Doppler assessment of blood flow in the umbilical cord, are often performed weeks to months apart. These tests have poor predictive performance for the risk of stillbirth. Being able to measure and integrate maternal, baby and placental signals, daily or even more frequently, is central to characterising gestational development and is likely key to preventing stillbirth.
Advances in mobile sensing technologies and optical imaging, coupled with advances in data analytics, provide opportunities to assess placental function, maternal response and baby’s behaviour in utero, in real time, at greater resolution than ever before. For example – moving from weekly subjective assessment by healthcare practitioners in clinics, to remote, hourly objective assessments of in utero activity could detect acute reductions in oxygen supply to the baby, that if not acted on, may cause stillbirth within a few hours.
In addition, rapid non-invasive analysis of material from the placenta and the developing baby is becoming a reality through analysis of cell-free nucleic acids, placental vesicles, and exosomes that circulate in the mother’s blood; the use of ‘omic’ platforms can detect novel biomarkers of gestational health and disease; and advanced high-resolution in vivo (e.g. MRI) and ex vivo (e.g. microCT) imaging coupled with mathematical modelling, can add new insights into the characterisation of placental transfer functions.
No animal models replicate the large size and unique structure of the human placenta. However, animal studies have helped to elucidate a developing baby’s responses to oxygen and nutrient restriction. These studies indicate there may be opportunities in human pregnancy for recognition of evolving complications, providing opportunities for intervention. In particular, a developing baby’s response to inadequate placental oxygen and nutrient transfer includes changes in heart rate, blood flow patterns, growth trajectory and behaviour. In response to acute lack of oxygen (triggered, for example, by compression of the umbilical cord), there is a reduction in a developing baby’s heart rate from the normal baseline (in humans at term, 120 -150 beats per minute) of ≥15 beats per minute; which may be followed by compensatory increases in heart rate if oxygen restriction persists. In response to a lack of oxygen, the developing baby also stops performing breathing-like movements (rapid, 1-4Hz episodic movements occurring 30-40% of the time after 30 weeks gestation)vii and the normal cycle of in utero sleep and wakeful periods is affected, with a resulting reduction in the baby’s movements.
Tragically, in up to 55% of stillbirths, mothers report a decrease in their baby’s movements in the week before their baby died. Attempts to use the subjective maternal perception of reduced baby movements as an opportunity to increase monitoring and/or expedite birth to reduce stillbirth have been unsuccessful.viii Accurate and frequent objective measures of in utero behaviour of the developing baby, in combination with other measures, hold more promise.
Large observational studies in humans have shown other insights. Pregnant women who fall asleep on their back have a 2.6-fold increased risk of stillbirth.ix When a pregnant woman lies flat, the uterus can fall backwards compressing her aorta, and inferior vena cava (one of the main veins returning blood to the heart) affecting in utero blood flow. This can happen during maternal sleep and has led to recommendations that pregnant women fall asleep on their side rather than their back. There are no proven interventions to promote this practice.
Together these observations and technological advances indicate that there is potential for a step-change in our ability to reduce stillbirths, with integrated measures of maternal, developing baby and placental function that accurately model gestational progression. These could allow opportunities for timely and safe delivery— based on individual risks— to prevent stillbirth.
Effective predictive models of gestational development would also minimise unnecessary, potentially harmful interventions in healthy pregnancies. As the birth of a baby removes their risk of stillbirth, a strategy of intentionally delivering babies before their due date — by induction of labour or planned caesarean section — is increasingly used in attempts to reduce stillbirths. There has been a 40%-60% increase in such healthcare provider-initiated births over the past decade in a variety of settingsx, with the result that in many countries only around half of births are preceded by spontaneous onset of labour. In the absence of alternative approaches, this trend is understandable.
However, any benefit in reducing stillbirth needs to be carefully balanced against health risks of the infant being born early, even in births close to term. A lack of precision in current risk-based approaches to stillbirth reduction means that many babies are unnecessarily delivered ‘’just in case” of late pregnancy complications. For example, it has been estimated that to prevent a single stillbirth with the now common strategy of offering provider-initiated birth at 39 weeks gestation, more than a thousand women will undergo induction of labour or Caesarean section rather than awaiting spontaneous labour.xi These provider-initiated births have a substantial burden on maternity services, and may also be harmful to children in the long term, as childhood need for special educational support and behavioural problems are all lowest in babies born at or after their due date.xii
Our goal is to create the scalable capacity to measure, model and predict gestational development, with sufficient accuracy to reduce stillbirth rates by half, without increasing provider-initiated delivery.
Sarah Stock, MD, PhD is a practicing Professor and Consultant in Maternal and Fetal Health, with expertise in stillbirth and preterm birth. With a laboratory science background, she now focuses on clinical trials and international data-driven studies. She earned her MD from Manchester University Medical School and PhD in Reproductive Biology from University of Edinburgh and completed her specialist and subspecialist Maternal and Fetal Medicine clinical training at Edinburgh, with periods at Glasgow, London and Australia.
Who are eligible Wellcome Leap program performers?
Performers are from universities and research institutions: small, medium and large companies (including venture-backed); and government or non-profit research organizations. We encourage individuals, research labs, companies, or small teams to apply in program areas best aligned with their expertise and capabilities. It is not necessary to form a large consortium or a single team to address all thrusts or an entire program goal in an abstract or proposal. Indeed, one of the benefits of our programs is that we actively facilitate collaboration and synergies dynamically among performers as we make progress together toward the program’s goals.