Cytomegalovirus (CMV) Infection Acquired through Breast Milk in Preterm Infants : Short- and Long-Term Outcome

BACKGROUND: Shedding of CMV into breast milk is the main source of CMV postnatal infection in early life. Many studies have shown a higher risk for premature infants, compared to term infants, of acquiring a symptomatic infection via this route. However, few studies have investigated the long-term neurological outcome of preterm infants with breast milk-acquired CMV infection. OBJECTIVE: To evaluate short- and long-term outcomes of CMV infection acquired by breast milk in preterm infants. DESIGN/METHODS: During a 36-month period we assessed maternal CMV serological status after preterm delivery ≤ 34 wks; vaginal swabs were taken on the day of delivery. We collected saliva and urine samples from all breastfed infants at 3 days of life, weekly until discharge and then monthly until babies were breastfed. At the same time breast milk samples were taken from the mothers. All samples were tested for CMV by “shell vial” culture (SVC) and by nested PCR. Children's blood was tested by PCR on day 1 of life to identify congenital CMV infection. CMV infected infants were included in a follow-up study until age 6 years. RESULTS: 78/94 (83%) women were CMV-seropositive; vaginal swabs were positive in 10/94 (10.6%) women who underwent caesarean section. CMV was detected in milk samples by PCR in 47 (66.2%) and by SVC in 43 (60.6%) of the 71 seropositive breastfeeding mothers. No infant born to CMV-seronegative mothers became infected. 96 neonates were born to the 71 CMV-seropositive lactating mothers: median GA 31 wks, median BW 1240 g. Congenital infection was diagnosed in 1 newborn. The mother to child transmission rate by breast milk was 29.8% (14 of 47 mothers), resulting in postnatal infection in 17/62 (27.5%) infants who received CMV-positive milk. In these 17 infants viral DNA or virus was detected in saliva or urine samples at a mean age of 49.7 days. Four (23.5%) of the 17 infants had clinical signs of infection: sepsis-like symptoms in 2 babies and sensorineural hearing loss at follow-up evaluation in the other 2 babies. CONCLUSIONS: In our study, the proportion of infants who acquired a CMV infection by breast milk was high and the incidence of neurological sequelae was far from negligible. It could be recommended to treat human milk from seropositive women to prevent CMV transmission to extremely small preterm infants.