番茄社区

Small brain bleeds in very premature infants don't appear to hurt their early neurodevelopment, a longitudinal study suggested.

Low-grade hemorrhage among babies born before 27 weeks' gestation wasn't associated with poorer cognitive, language, or motor scores or higher incidence of cerebral palsy at 18 to 22 months, Allison Payne, MD, of Rainbow Babies and Children's Hospital in Cleveland, and colleagues found.

Severe hemorrhage, though, was associated with some deficits in language and cognition at that point, the group reported online in JAMA Pediatrics.

"This does not imply, of course, that these low-grade hemorrhages will not affect outcome at later ages," Martha Sola-Visner, MD, of Boston Children's Hospital and Harvard, cautioned in an accompanying editorial.

Longer-term follow-up is needed to look for behavioral problems, attention deficit-hyperactivity disorder, and other more subtle impacts, the editorialist and researchers agreed.

Nevertheless, the findings were reassuring compared with the poorer early outcomes seen with low-grade hemorrhage in three older series, Sola-Visner pointed out.

"The finding of any hemorrhage by use of cranial ultrasonography is a cause of great parental anxiety, but as we convey the news of a grade 1 or 2 intraventricular hemorrhage to parents, the data presented here support a less pessimistic view of the effect of this finding on an infant's neurodevelopmental outcome, at least at 18 to 20 months," she concluded.

The difference between studies might reflect better outcomes with contemporary practice, particularly the shift toward antenatal steroids and away from postnatal steroids for preemies in the 1990s, the editorial suggested.

The study followed children born at less than 27 weeks' gestational age from 2006 through 2008 at the 16 centers of the National Institute of Child Health and Human Development's Neonatal Research Network.

The analysis was limited to the 1,472 infants examined for periventricular-intraventricular hemorrhage by ultrasonography within the first 28 days of life who survived to ages 18 to 22 months.

A total of 31% had periventricular-intraventricular hemorrhage, including 140 with grade 1, 130 with grade 2, and 181 with grade 3 or 4.

Severe, grade 3 or 4 hemorrhage was associated with worse outcomes on almost every count compared with low-grade hemorrhage after adjustment for estimated gestational age at delivery, sex, chorioamnionitis, antenatal and postnatal steroid use, and other factors. The odds were (all significant):

• 3.44-fold higher for any cerebral palsy
• 3.79-fold higher for a gross motor function score more than 2
• 2.05-fold higher for severe language impairment
• 2.04-fold higher for severe neurodevelopmental impairment

The odds were 76% higher for mild cognitive impairment and 79% higher for mild neurodevelopmental impairment.

Severe hemorrhage predicted a 3.91-point lower score on cognitive tests on average and 3.19-point lower score on language tests, both statistically significant.

The grade 1 or 2 cases weren't associated with any poor neurodevelopmental outcomes at follow-up in the unadjusted or adjusted models, when compared against children with no hemorrhage.

Low-grade brain bleeds didn't independently predict any outcome.

The study didn't show any difference between grade 1 and 2 either but wasn't powered to do so, the researchers noted.

Another limitation was the highly operator-dependent technique and interpretation of cranial ultrasound in detecting periventricular-intraventricular hemorrhage, "particularly at the lowest grades."

Exclusion of infants with an early porencephalic cyst or hydrocephalus requiring shunt might have underestimated severe impairment from grade 3 or 4 bleeds, but was done to provide "the clearest possible picture of low-grade periventricular-intraventricular hemorrhage," the group added.

Comment