Major Finding: Very low-birth-weight infants with symptomatic, postnatally acquired CMV infection had high rates of pneumonitis (73%) and late-onset sepsis (43%).
Data Source: A retrospective cohort study of 34 infants with symptomatic CMV infection.
Disclosures: None was reported.
VANCOUVER, B.C. — Postnatally acquired cytomegalovirus infection can cause severe illness in very low-birth-weight infants in the short term, based on findings from a retrospective study.
Infants in the study who became infected and symptomatic with cytomegalovirus (CMV) infection in the postnatal period had high rates of complications.
In fact, their clinical and laboratory findings were similar to those of congenitally infected infants.
“The biggest take-home message is that postnatal CMV infection can cause significant morbidity, and it can potentially lead to [poor] long-term outcomes,” lead investigator Dr. Sarah A. Meyer said in an interview.
“It is often something that we don't think about a lot, and we just need to keep it in our mind that if we have babies that present with some of these symptoms, we should be testing them and following their outcomes,” she said.
Much is known about congenitally acquired CMV, according to Dr. Meyer of Children's Hospital Boston. But comparatively little is know about CMV acquired in the postnatal period through breast milk.
Using hospital records for the years 1997-2009, she and her colleagues retrospectively studied 34 infants who had symptomatic, culture-positive CMV infection and were cared for in a neonatal intensive care unit.
Of the infants (all but 1 of whom met criteria for very low birth weight [VLBW]), 22 had been infected postnatally, whereas the other 12 infants (having a range of birth weights) had been infected congenitally.
Compared with their congenitally infected counterparts, the postnatally infected infants had a lower median birth weight (688 vs. 1,500 g), had a younger median gestational age (26 vs. 32 weeks), were older on the day of diagnosis (52.5 vs. 3.5 days of life), and were more likely to have been breastfed (100% vs. 67%).
The proportions delivered by cesarean section were similar, she reported in a poster at the meeting.
Among those infected in the postnatal period, the time to CMV diagnosis was correlated with the length of exposure to breast milk (r = 0.84), indicating that the risk of viral transmission persisted with continued exposure. In contrast, the time to CMV diagnosis was not correlated with the day of life on which infants were first fed breast milk.
The most common complications with postnatal infection were pneumonitis (present in 73% of infants), colitis (50%), hepatosplenomegaly (36%), and intracranial findings (27%).
Relative to their congenitally infected counterparts, postnatally infected infants had generally similar clinical findings, but were more likely to have pneumonitis (73% vs. 0%) and less likely to have petechiae and purpura (10% vs. 50%) and retinitis (0% vs. 25%).
The two groups were also similar in rates of hematologic and cerebrospinal fluid laboratory abnormalities, presence of cerebrospinal fluid CMV positivity by polymerase chain reaction testing, and median blood CMV viral load.
Among the 15 infants overall with neurologic follow-up, the rate of hearing loss was 71% in those congenitally infected, compared with 13% in those postnatally infected, with numbers too small to permit statistical comparison.
Rates of developmental delay and cerebral palsy were similar, although these sequelae in the postnatally infected infants also could have been related to their prematurity, noted Dr. Meyer.
A final analysis did suggest that symptomatic postnatal CMV infection may add substantial morbidity above and beyond that due to having a very low birth weight, she said.
Compared with 1,226 infants from the general VLBW population, the VLBW infants with postnatal cytomegalovirus infection had a higher rate of bronchopulmonary dysplasia (81% vs. 16%) and late-onset sepsis (43% vs. 18%).
Rates of necrotizing enterocolitis and intraventricular hemorrhage did not differ significantly between the two groups.
Dr. Meyer said that teasing out causal associations is difficult in the VLBW population.
“Whether or not these babies are immunosuppressed to begin with and that predisposes them to late-onset sepsis and acquiring CMV, or acquiring CMV reduces their immunity level and that predisposes them to late-onset sepsis—which one came first is not exactly clear,” she explained.
“Hopefully, that can be separated out in future studies.”
Monitoring infants with postnatally acquired CMV infection long term, while important, is just an initial step, according to Dr. Meyer.
“An area of study that is still really needed is to look at how treating babies with antivirals affects their outcomes,” she said.