Science Spotlight

Reducing opportunity: Antivirals reduce CMV vertical transmission

Boeckh Lab (Vaccine and Infectious Disease Division)

Cytomegalovirus (CMV) is an opportunistic infection with HIV that can lead to morbidity and mortality. In infants with HIV and CMV there is an increased risk of disease progression, neurologic disease and death. In Kenya most HIV- exposed children acquire CMV within the first year of life, mostly through the mother’s breast milk. In the context of HIV infection maternal immunosuppression leads to higher CMV loads in breast milk as well as higher rates of infant CMV transmission. Therefore highly active antiretroviral therapy (HAART) may affect CMV load and transmission in these populations through controlling immunosuppression. Currently data on co-infection in the presence of antiretrovirals is unclear and with the increase use of HAART in pregnant women there should be studies to look at the effects of HAART on other co-infections. A collaboration between Fred Hutch (Vaccine and Infectious Disease Division) scientists and the University of Washington was designed to look at antiretroviral use and CMV transmission in a cohort of mothers and children in Kenya from 2003 to 2005. Two different treatment groups consisting of 51 women total were followed longitudinally, one treated with a multiagent HAART regimen starting at 32 weeks and for 6 months after delivery and the other group received single agent zidovudine for 34 weeks during pregency plus a dose of nevirapine during delivery to both mother and infant (ZDV/dsNVP). 

Graphical representation showing probability of CMV infection in the two groups over time.
Graphical representation of the probability of CMV infection of infants comparing mothers either on HAART treatment or ZDV/dsNVP. Image provided by Dr. Slyker.

During the course of the study 3 infants acquired HIV (1 ZDV/dsNVP and 2 HAART) and CMV was detected in 84% of infant plasma by 1 year of age. Comparing the Kaplan-Meier survival curves for CMV DNA, both groups showed similar curves until three months. At three months the curves diverged and the probability of CMV infection was found to be significantly lower in the HAART groups compared to the ZDV/dsNVP arm at six and twelve months of age. The median time to CMV infection also differed between groups with the HAART and ZDV/dsNVP groups being 3.1 and 3 months respectively. In regards to the health of the mother, there was no association between risk of CMV infection and early breast milk viral load, baseline plasma HIV load, or CD4 T cells percent. In sampling the breast milk the researchers found that CMV was present by 1 week postpartum in 72% of women and in 98% of milk samples by 2 weeks. Interestingly, at 1 week there were more frequent and higher levels of CMV DNA found in breast milk from the HAART treated mothers than the ZDV/dsNVP group. Overall CMV viral load peaked at 3-4 weeks then slowly declined, showing slower decline in the HAART group. Contrary to what the researchers thought, HAART treatment lead to higher levels of CMV in breast milk even though the risk for CMV transmission was reduced in this group. Therefore, the researches conclude that HAART does decrease risk of infant CMV transmission however this risk reduction is not mediated through CMV load decrease in breast milk.

One of the researchers, Dr. Slyker explained “We were surprised to see an effect on CMV transmission but not on CMV DNA levels in breast milk, which we think is the major mode of CMV transmission in the first year of life. Other groups have also observed this in observational studies. Our next step is to explore maternal and infant immune mechanisms of protection.” The group hypothesizes that CMV load may be increased due to the changes in the inflammatory environment following HAART initiation and that earlier initiation of HAART may reduce this effect. They also hypothesize that maternal immune reconstitution due to HAART could contribute to the lower risk by improving passive transfer of anti-CMV antibodies from the mother to the infant thus reducing risk without affecting breast milk viral levels.


Funding for this study was provided by the National Institutes of Health.


Slyker JA,Richardson BA,Chung M,Atkinson C,Asbjornsdottir KH,Lehman D,Boeckh M,Emery V,Kiarie J,John-Stewart G. (2016). Maternal highly active antiretroviral therapy reduces vertical CMV transmission but does not reduce breast milk CMV levels. AIDS Res Hum Retroviruses.