Protecting the tiniest lives: how biological signals can improve care for premature babies
Metabolomics
When a baby is born too soon, every minute counts – and so does every decision. To reduce uncertainty in those critical moments, biopharmaceutical scientist Manchu Umarani Thangavelu dedicated her PhD to identifying molecular markers of preterm birth and neonatal infection.
In the neonatal intensive care unit, treatments for premature babies are often life-saving but also carry risks. Because timely intervention is crucial, mothers at risk of preterm birth are often given corticosteroids to accelerate the baby’s lung development. Premature newborns showing possible signs of infection are given preventive antibiotics.
However, every treatment requires careful consideration. Unecessary antibiotic exposure may disrupt the newborn’s developing microbiome and contribute to antimicrobial resistance, while repeated corticosteroid use could affect growth and brain development.
Reducing uncertainty in preterm care
‘Many decisions in perinatal care are made under considerable uncertainty,’ says Umarani Thangavelu. Her PhD project initially focused on improving the diagnosis of neonatal sepsis, a severe bloodstream infection. For this, she used metabolomics, the study of small molecules in the body that reflect ongoing biological processes.
She soon realised that uncertainty affects many aspects of preterm care. ‘Doctors have to act quickly, but the available diagnostic tools are not always specific or fast enough. That gap really motivated me to look at the bigger picture - from predicting preterm birth to deciding on antibiotic use for preterm newborns.’
Measuring molecular signals in mothers and babies
Umarani Thangavelu measured tiny signaling molecules in urine from pregnant women at risk of early delivery (before 37 weeks of pregnancy) and blood from preterm infants with suspected infection. These advanced analyses revealed distinct molecular patterns, some linked to different triggers of premature birth, others reflecting inflammation related to infection after birth.
9-HODE: a potential clue for preterm birth
One molecule stood out: 9-HODE, a fatty acid linked to inflammation. Although an infection can trigger an inflammatory reaction, 9-HODE was consistently detected in preterm births without infection as well, regardless of the underlying cause.
‘Preterm birth is a complex process with many possible causes,’ Umarani Thangavelu explains, ‘such as infection, but also placental dysfunction, cervical insufficiency, or tension or stretching in the uterus.’
This suggests that 9-HODE may reflect a general inflammatory response that the body activates in preterm birth, rather than a specific marker of infection. The molecule may potentially serve as an indicator of preterm birth, while also highlighting the importance of context, Umarani explains. ‘If we interpret every inflammatory signal as infection, we risk overtreating babies with antibiotics.’
‘With modern analysis techniques, we can obtain reliable data from the tiniest drop of blood.’
Just a few tiny drops of blood
Because premature infants can safely provide only tiny blood volumes, Umarani Thangavelu also evaluated a new technique called Volumetric Absorptive Microsampling (VAMS). It involves a small polymer tip that collects precise amounts of blood — sometimes as little as 10 microliters. Umarani Thangavelu: ‘With modern molecular analysis, we can obtain reliable molecular data from these drops. That makes research safer and much more feasible for fragile newborns.’
Predicting risk for personalised treatment
Umarani Thangavelu’s findings offer early evidence that metabolomics could help personalise neonatal care. Molecular signatures in the urine or blood may help doctors decide when antibiotics are truly necessary and potentially even distinguish between different types of infection. She also observed that boys and girls respond differently to sepsis, which could guide future strategies.
Could all pregnant women eventually be routinely tested for these molecules to predict preterm birth risk? ‘Potentially yes, larger studies are needed,’ Thangavelu explains. ‘If validated, these molecular markers could eventually help identify women at increased risk earlier and contribute to safer, more tailored care for babies born too soon.’
Umarani Thangavelu defended her thesis ‘Survival of the Littlest: Improving Preterm Outcomes through Metabolomics and Microsampling’ on 26 February in the Academic building. Her work was a collaboration between the Erasmus Medical Centre in Rotterdam and the Metabolomics and Analytics Centre of the Leiden Academic Centre for Drug Research. Her promoters are Professor Irwin Reiss, Professor Thomas Hankemeier and Dr Bert Wouters, and Dr Alida Kindt-Dunjko was her supervisor.