The dilemma of breastfeeding very low birth weight infants at risk for symptomatic cytomegalovirus infection

Lombardi et al. [1] recently reported the risk of breast milk-acquired human cytomegalovirus (HCMV) infection in very low birth weight infants. The authors correctly state that there exists no consensus, but various recommendations for breastfeeding particularly preterm infants of less than 30 weeks of gestational age. Literature reports a wide range of data on the rate of breast milk-acquired HCMV infection, and the authors present a true range of 5.7–58.6%. Nevertheless, the important message is that symptomatic HCMV disease occurs in significantly less cases (rate 0–34.5% with a median rate of 3.7%), and severe sepsis-like syndrome in 0–13.8% (median 0.7%) of cases [2]. Since the early 1970s, 18 cases have been reported with symptomatic breast milk-acquired HCMV infection of whom 11 (61%) had sepsis-like syndrome [3]. The latter all had a gestational age equal to or below 28 weeks. I fully agree with Lombardi et al. [1] that the suggestion of feeding preterm newborns with pasteurized milk only obviously exceeds the potential of “clinical deterioration” by HCMV disease against the advantage of fresh milk feeding. Holder pasteurization or short-term heating are the methods of choice that eliminate the risk of HCMV transmission. In contrast, the method of freezing/thawing does not completely eliminate transmission and disease in the preterm infant but preserves nutritional and immunologic components of breast milk. Thus, I feel better to recommend pasteurization of human milk in an extremely preterm infant in order to reduce the risk of HCMV disease to zero (3). It seems to be quite inconsequent being in fear of HCMV disease in the preterm infant and otherwise recommending a virus elimination method that does not work perfectly. Despite this academic dispute human milk breastfeeding in preterm infants remains highly important, and the risk of severe HCMV acquisition via breastfeeding seems to be generally overestimated.