Skip to main content
Premium Trial:

Request an Annual Quote

Study Shows Preemies' Oral Feeding Readiness Predicted By Salivary Biomarkers

Premium
premature infant

NEW YORK (GenomeWeb) – Using whole-transcriptome microarray analysis, mathematical modeling, and RT-qPCR, a team of researchers at Tufts University and Harvard Medical School have discovered a panel of salivary biomarkers that can predict when a baby born prematurely is ready to begin oral feeding.

Validation of the panel in 400 neonatal saliva samples suggested the method could be an improvement over current "best-guess" techniques that neonatologists use to tell when a preterm infant is neurobiologically mature enough to attempt oral feeding, a prerequisite for the baby to leave the hospital.

In a study published last week in the Journal of Pediatrics, the group said the panel could reduce hospital expenses, and may also help clinicians tailor therapies.

Oral feeding is a significant milestone in development. The current gold standard is an educated clinical guess about whether or not an individual baby can feed, but "at the end of the day, if that baby can't, you have no idea why," Jill Maron, an assistant professor of pediatrics at Tufts University and lead author on the study, told GenomeWeb in an interview.

"If you wait long enough, most of these babies will learn to eat," Maron explained, but unnecessarily delaying oral feeding could be costly.

"It's a big contributor to length of stay in the hospital, and every day a baby stays in the hospital increases costs," she said. Conversely, correctly identifying babies who are able to feed, and individually addressing the source of the problem for the ones who aren't, "could be a huge healthcare savings."  

There are a few assessment tools that might help show when an infant is ready, but a recent Cochrane review found studies of them are currently insufficient to guide clinical practice.

Babies who are not mature enough for oral feeding are sustained by nasal gastric tube feeding, Maron explained. Babies can be fed orally while the tube is still in place, but they need to have proper development of appetite and facial musculature.

Besides the stress to caregivers, if babies are fed orally before they are ready there is risk of their developing a feeding aversion, which could in turn lead to a feeding disorder or readmission for failure to thrive. "If they don't get the calories, they don't grow, and their brain doesn't grow either," Maron explained.

The new biomarker assay could provide an objective measure to tell clinicians whether an individual baby is physically able to make the switch, or whether it would be best to wait.

The group chose saliva for the study in part because blood testing is constrained by neonatal blood volumes, which, in a one- to two-pound premature baby, can be on the order of a can of soda, Maron said.

"We can't keep sticking them, so we really needed to think about alternative ways to gain access to this information," she said.

Saliva is "a biofluid that for years so many of us have been trying to bring to the bedside, and for these babies it is an enormous opportunity because they don't have blood volumes to give," she added.

The initial biomarker panel had 21 genes and three reference genes. All three of these controls needed to amplify for a sample to be included, which led to a 10 percent loss from the data set. After controlling for confounders like age and gender in the remaining samples, only five of 21 genes stayed in the group's model, and these showed good accuracy in determining feeding success. The panel may ultimately include other genes in the future, but, Maron explained, "Our bar is pretty low … it isn't like we need a 99 percent accurate test; we need something that's better than flipping a coin."

Maron's group had previously shown that one of the five genes in the panel, neuropeptide Y receptor 2 (NPY2R), was associated with oral feeding in neonates. Because NPY2R expression is developmentally regulated in circuits within the hypothalamus that control feeding behavior, the authors argued that this could be a biomarker of hunger signaling. Another of the genes, AMPK, is associated with energy expenditure, while WNT3 is related to facial development.

These markers could potentially indicate points of therapeutic intervention. If, for example, clinicians discovered a particular baby's oral feeding is prevented by insufficient facial motor development, they could try different physical therapies, including pacifier-like devices that use biofeedback to coordinate low-pressure suction and behavior to help a baby learn to suck.

But, if a baby is not feeding because circuits that control appetite aren't mature, "It isn't that the baby doesn't yet know how to do it, it doesn't know it is supposed to eat," Maron explained. Physical therapy might be an unnecessary expense in these instances.

Salivary biomarkers of brain development are, at the least, correlative. However, there is reason to believe that some of the mRNA in saliva of premature babies could be brain-derived. "We know that the salivary gland is a direct filtrate of blood," Maron said, "so elements that are in the bloodstream are readily in saliva."

In premature infants, the salivary gland and the blood-brain barrier are also less developed than in adults.

Additionally, NPY2R is expressed in the arcuate nucleus in the brain, which is one of the only areas permeable to the blood-brain barrier. The receptor is expressed in other parts of the body as well, and Maron reiterated that the study is "not showing mechanism; we're showing plausibility and, through repeated measures, the feasibility of this approach."

The group currently has a patent pending on the biomarker panel and it is seeking collaborators to commercialize the assay. Ideally, it would be developed into a point-of-care test, because lab-based tests with two-day turnaround times are less useful. The assay also requires somewhat complicated sample prep, since genomic DNA needs to be removed and salivary mRNA samples, which tend to be degraded, have to be converted to cDNA before RT-qPCR. "All these things take time, so finding a platform that is going to work best for that is probably going to be the biggest obstacle for implementation," Maron said.

The group may also look at how this panel correlates with long-term development. "Studies have shown if you don't eat on time, you don't sit, walk, or talk on time either, so developmentally, this is really an important milestone," she said.