Around 50 - 70% of Australian women will have already been infected with cytomegalovirus (CMV) when they become pregnant. It is a little known fact that the majority of these women will naturally excrete cytomegalovirus into their breast milk and that a proportion of babies will become infected though milk ingestion. Although this is generally of no medical concern, in the extraordinary setting of the neonatal intensive care unit, cytomegalovirus infection can be difficult to diagnose and to manage. Common symptoms are hepatitis and sepsis like syndromes and the long term consequences of infection for later developmental outcomes are not clear. 

We collected breast milk from 29 mothers of very preterm babies (<32 weeks gestation) on days 2-4, 8-12 and 26-30 post partum and evaluated the milk for the presence of CMV by PCR analysis and long term culture. We also evaluated breast milk and maternal plasma for the presence of CMV-specific IgG. Whilst maternal plasma was not a useful indicator of viable CMV in breast milk, higher titres of CMV-specific IgG in breast milk were associated with viral isolation from samples collected at later times post partum.  In most cases these samples were detected by quantitative PCR, but the virion copy number was very small.  Other possible biological indicators of infection such as the presence of cytokines and other immunomodulatory factors such as lactoferrin and lysozyme were also assessed, but no other correlates to potential infection were determined.

Although 60% of women were positive for CMV (determined by maternal plasma titre and CMV-specific IgG present in breast milk and/or PCR positive breast milk or CMV detected in culture), only one baby was determined to have clinically significant CMV infection. This underscores the difficulty with evaluating post-partum infection in this setting, and suggests that any interventions will need to be well justified and carefully evaluated.