The effect of prenatal highly active antiretroviral therapy on the transmission of congenital and perinatal/early postnatal cytomegalovirus among HIV-infected and HIV-exposed infants.

BACKGROUND

Before highly active antiretroviral therapy (HAART), congenital cytomegalovirus (CMV) rates were higher among human immunodeficiency virus (HIV)-exposed infants than unexposed infants. This study examines congenital and perinatal/early postnatal (P/EP) CMV among HIV-exposed infants pre- and post- HAART.

METHODS

Infants born to HIV-infected women were evaluated for congenital CMV (CMV-positive culture in first 3 weeks of life) and P/EP CMV (positive culture in first 6 months of life). Prenatal maternal HAART was defined as triple antiretroviral therapy (ART) with at least 1 nonnucleoside reverse-transcriptase inhibitor or protease inhibitor.

RESULTS

Among 414 infants evaluated, 1678 CMV assessment days were completed (mean = 3 assessment days per infant). Congenital CMV rates did not differ by time period, HAART use, or infant HIV infection status. P/EP CMV rates were greater for the 1988-1996 birth cohort (17.9%) compared with the 1997-2002 birth cohort (8.9%) (P < .01), HIV-infected versus uninfected infants (P < .01), and infants with no maternal ART versus those with ART (P < .01). Controlling for potential confounders, P/EP CMV was associated with no maternal ART (odds ratio = 4.7; P < .01), and among those with no maternal ART, P/EP CMV was associated with maternal CD4 count ≤200 cells/μL (P < .01). For HIV-uninfected infants with P/EP CMV, symptoms including splenomegaly, lymphadenopathy, and hepatomegaly were associated with no maternal HAART versus those with HAART (41% vs 6%; P < .05).

CONCLUSIONS

Although congenital CMV rates did not change, the post-HAART era showed reduced P/EP CMV and occurrence of related clinical symptoms. These findings underscore the importance of prenatal HAART for all HIV-infected pregnant women.