A treatment for a deadly condition in newborns fails to improve outcomes, study finds

What do you do when a child is born with poor consciousness, not crying, not sucking, and not breathing on their own? For babies born with hypoxic-ischemic encephalopathy, commonly called birth asphyxia, doctors have limited options. And a treatment that many had hoped would be effective turns out not to be.

A new study, published Wednesday in the New England Journal of Medicine, found that the hormone erythropoietin did not improve survival and other outcomes in newborns with hypoxic-ischemic encephalopathy. Unexpectedly, the newborns who received the hormone were also more likely to experience a serious adverse event, potentially including death. The clinical trial of 500 infants was the culmination of decades of study, including prior work that suggested erythropoietin was safe and had promise in mice, macaques, and, in a smaller study, humans. Another clinical trial of erythropoietin is ongoing in Australia.

“There is actually a very strong and robust body of literature that shows that erythropoietin provides neuroprotection in newborn brains,” said Hannah Glass, a pediatric neurologist at the University of California, San Francisco, founding co-director of its Neuro-Intensive Care Nursery, and author of the new paper. “So it was a big surprise and disappointment that this trial was negative.”

Erythropoietin, which some athletes have used illicitly to boost performance, tells your body to make red blood cells. But it can also do more than that. In 1993, an influential publication reported that the erythropoietin receptor was found on mouse neurons. “That was totally out of the box,” said Sandra Juul, senior author of the study, clinical trialist, and professor of pediatrics in the Division of Neonatology at the University of Washington and Seattle Children’s Hospital. “People at the time thought erythropoietin had one job, and that was to increase red blood cell production.” “But actually, no, it has direct effects on brain cells,” added Yvonne Wu, a pediatric neurologist at UCSF, and first author of the new study. The erythropoietin receptor is expressed in the brain and, in mice, erythropoietin has been shown to improve neurological function after neonatal stroke and spinal cord trauma.

Based on previous positive results, some physicians were already prescribing erythropoietin to newborns with hypoxic-ischemic encephalopathy. “We heard, just anecdotally, that there were some practitioners, in the United States and in other countries, who decided to just go ahead and start doing this treatment,” said Adam Hartman, a program director at the National Institute of Neurological Disorders and Stroke who helped oversee the implementation of this study. Hypoxic-ischemic encephalopathy is the second largest cause of neonatal deaths, behind premature birth, and is relatively common, afflicting 1.5 babies out of 1,000 live births.

The exact reason that erythropoietin failed is unclear. Researchers believe one possible factor relates to a change in the standard of care for hypoxic-ischemic encephalopathy. Therapeutic hypothermia, a technical term for cooling the baby to 33.5 degrees Celsius for 72 hours, initiated within the first 24 hours of life, has become widespread in the past decade. In preclinical studies, hypothermia was not yet common and most animal models did not include cooling. A more recent study performed in sheep did not find erythropoietin to provide an additional benefit on top of hypothermia. The study authors point out, though, that they had finished enrolling the babies before that result came out.

“Even though the study was negative,” said Juul, “we’ll learn a lot.” A different, earlier study she was involved in found that erythropoietin did not help preterm babies. Despite that, the data that team collected has allowed them to publish 20 papers, she said. The researchers have a wealth of data to further examine, including histological sections of placentas, brain MRI scans from newborns, and records on adverse events. This could help answer questions about biomarkers of the condition, how to predict outcomes, and the timing of symptoms. “It shows the importance of doing Phase 3 studies. It’s important to do these large trials to see what the real answer is,” added Juul.

“It was a negative trial, and so people always ask, ‘Was it adequately designed, was it adequately executed?’” said Hartman. “People need to know that this was well-executed and, in this case, no really means no.”

Though meaningful, the result was still a disappointment for a field that has had few advances in the past decades. “It’s very hard to get parents of sick newborns to enroll in trials,” said Christopher Elitt, a pediatric neurologist at Boston Children’s Hospital who was not involved in the study. “These types of studies don’t come along very often because they are expensive and difficult to do.”

Sometimes there is an indication that physicians should be concerned a newborn could develop hypoxic-ischemic encephalopathy. In about a quarter of cases, there is something known as a “sentinel event.” Sometimes, the umbilical cord is wrapped around the baby’s neck, or the baby’s shoulder gets stuck while coming out, or the placenta separates from the uterus prematurely. “But even then, it’s just right before delivery,” said Fernando Gonzalez, a neonatologist at UCSF, the other co-director of its Neuro-Intensive Care Nursery, and author of the study. “It’s almost always a surprise when they’re born with concern for HIE.” Ultrasounds and electronic fetal monitoring are commonly normal, too, he added. Gonzalez leads a team that treats between 35 to 50 cases of hypoxic-ischemic encephalopathy every year, and personally saw around 15 babies with the condition last year.

Outcome notwithstanding, researchers said, the fact that this study has been done is in itself a positive, given how little study there is of potentially helpful treatments for newborns. “There are so many things that we do off-label with babies, because there just isn’t any evidence,” said Gonzalez. “It makes sense to use it, it just hasn’t been studied.”

Experts said the new results highlight the importance of running larger trials, which study enough people to minimize unintentional mistakes. The Phase 2 study found erythropoietin was beneficial. However, two babies in that trial had genetic disorders that were associated with abnormal neurological development; they were both assigned to the placebo group, dragging down the overall outcomes among newborns who did not receive erythropoietin. After re-analyzing the data without those two infants, “the difference that we saw at age 12 months went away,” said Wu. “It was no longer significant statistically.”

Some experts said the new result was not as surprising, given that diseases often have many pathways. Perhaps too frequently, scientists pin their hopes on a single magical pill or silver bullet, expecting it to work. “You often hope it, but you often see that’s not the case,” said Frank van Bel, a professor emeritus of neonatology at the University Medical Center Utrecht in the Netherlands, who was not involved in the study. “I think it’s more the combination of the right medicine rather than one single therapy.”

The current treatment, hypothermia, has its own drawbacks. To cool the baby, physicians have to place them on a cooling blanket away from their parents. This may lead to bonding issues, Wu said, as the baby cannot interact with their mother in the first few days of life. And in other, more measurable outcomes, hypothermia is not a cure-all either. Of the babies who did not receive erythropoietin, 11.5% died. “The adverse outcomes, if you look at the placebo arm, is still over 40%,” said Floris Groenendaal, a neonatologist at Wilhelmina Children’s Hospital in the Netherlands who did not participate in the trial. “There is definitely a need for additional therapies on top of therapeutic hypothermia.”

In low- and middle-income countries, hypothermia is not used because a large-scale trial found it to be ineffective. Experts have speculated that the effectiveness varies by geography because of differences in the underlying risk factors of hypoxic-ischemic encephalopathy, which potentially include infection, hypertension, and type 2 diabetes. And Gonzalez pointed out that because of quirks in how different cooling systems measure body temperature — UCSF uses rectal probes, others use esophageal ones — the babies may have been kept at too low a temperature. “It was a surprising finding,” said Gonzalez. “But that’s why we study things like EPO and other things, to see, maybe there’s other medications or other strategies that would hopefully benefit in low- to middle-income countries.”

Notwithstanding the conclusion that erythropoietin did not provide a benefit and may actually have caused harms, Hartman, of the National Institute of Neurological Disorders and Stroke, sees an important take-home message that connects with other government-funded studies. “This is the reason why we do clinical trials. Because there are a lot of really good ideas out there, and until you test them, you don’t know how the result’s going to turn out,” said Hartman. “In this case, there is enough equipoise, that you really needed to ask the question.” There are trials that have an inconclusive result, and he is glad this is not one of them.