Breast Milk CMV and the risk of feeding the VLBW infant

The recently published article by Josephson and colleagues confirms that serious infections due to postnatally acquired CMV in very low birth weight infants are a real concern. The study documented that properly screened (CMV seronegative) and filtered blood and blood components effectively blocks transmission of CMV from these previously documented sources and in turn confirmed that the primary source of transmission is maternal milk from seropositive mothers. Of the infants who were exposed to mother’s breast milk that was positive for CMV, 15 % developed serologic evidence of CMV disease and less than 3% developed significant clinical disease, including NEC, with a mortality of 60% (3/5).

Of importance to note was the fact that the study was performed primarily by a team of hematologists and transfusion experts and unfortunately lack any details as to the clinical course of the infected infants — there were no data on birth weight or gestational age data, no indication as to quantity of milk ingested, when breastfeeding was initiated, the percentage of raw milk ingested versus frozen thawed milk, post natal age of onset if disease, bowel biopsy or post mortem findings. Furthermore not all the mothers had their milk tested for presence of CMV. Thus, these significant methodological limitations preclude accurate mathematical calculations as to actual risk of feeding human milk to the VLBW infant. Furthermore, the absence of any basic clinical data precludes identifying who are the truly high-risk infants.
From the literature what do we know?

Approximately 75% percent are CMV seropositive at the time of delivery and by the end of the first month postpartum approximately 75% of these mothers will be excreting CMV in their milk. The percentage of mothers who excrete and the viral load increases as the postpartum age increases. Thus the risk of clinically significant disease increases with the number of days the infant is fed and his/her age. There is no risk from a seronegative mother or if the milk is free of CMV.

Given the above what should the clinician do?

1. The absolute minimum is to test all mothers of VLBW infants at the time of delivery. If they are seronegative there should be no hesitation to feed fresh raw beast milk ad lib.
2. If the mother is seropositive the benefits of feeding colostrums and fresh milk in the first 2 weeks postpartum outweigh the risk of developing clinical disease.
3. Ideally, if there are resources to test the mother’s milk with PCR and the milk continues to be CMV negative then continue to feed raw milk ad lib.
4. If the breast milk cannot be tested, I suggest that from 2 weeks till 2 months feeding frozen thawed milk is preferred as this will reduce, but not eliminate, the CMV dose inoculum. Donor pasteurized milk by definition is CMV negative and can be used as lib.
5. Most importantly, if the infant who is receiving non-pasteurized human milk from a seropositive mother and has any clinical deterioration, including sign/symptoms of NEC, rule out sepsis diagnosis, or an unexplained pneumonitis, the infant should be tested for CMV in addition to the standard bacterial cultures. If CMV positive by urine PCR, consultation with an infectious expert is in order so as to decide on the possibility of beginning anti viral therapy such as gancyclovir. CMV IgM is not a valuable tool in identifying infected infants given its high false negative rate

Dr. Arthur I Eidelman is a Professor of Pediatrics at Shaare Zedek Medical Center, Jerusalem, Israel. He is a Fellow of the Academy of Breastfeeding Medicine and past president of ABM.

Posts on this blog reflect the opinions of individual ABM members, not the organization as a whole.