Maternal valacyclovir and infant cytomegalovirus acquisition: a randomized controlled trial among HIV-infected women

Abstract

Background: Studies in HIV-1-infected infants and HIV-1-exposed, uninfected infants link early cytomegalovirus (CMV) acquisition with growth delay and cognitive impairment. We investigated maternal valacyclovir to delay infant acquisition of CMV.

Methods: Pregnant women with HIV-1, HSV-2 and CD4 count >250 cells/µl were randomized at 34 weeks gestation to 500 mg twice-daily valacyclovir or placebo for 12 months. Maternal CMV DNA was measured in plasma at 34 weeks gestation, in cervical secretions at 34 and 38 weeks gestation, and in breast milk at 7 postpartum timepoints; infant CMV DNA was measured in dried blood spots at 8 timepoints including birth.

Results: Among 148 women, 141 infants were compared in intent-to-treat analyses. Maternal and infant characteristics were similar between study arms. Infant CMV acquisition did not differ between study arms, with 46/70 infants (66%) in placebo arm and 47/71 infants (66%) in the valacyclovir arm acquiring CMV; median time to CMV detection did not differ. CMV DNA was detected in 92% of 542 breast milk specimens with no difference in CMV level between study arms. Change in cervical shedding of CMV DNA between baseline and 38 weeks was 0.40-log greater in the placebo arm than the valacyclovir arm (p = 0.05).

Conclusions: In this cohort of HIV-1-seropositive mothers, two-thirds of infants acquired CMV by one year. Maternal valacyclovir had no effect on timing of infant CMV acquisition or breast milk CMV viral loads, although it modestly reduced cervical CMV shedding. Maternal prophylaxis to reduce infant CMV acquisition warrants further evaluation in trials with antiviral agents.

Trials registration: ClinicalTrials.gov NCT00530777.

Conflict of interest statement

Competing Interests: Please note that GlaxoSmithKline donated valacyclovir and matched placebo to the study. This does not alter our adherence to all the PLOS ONE policies on sharing data and materials.

Figures

Figure 1. Study flow chart.

Figure 2. Valacyclovir and infant CMV acquisition.

A) Kaplan-Meier survival curves show cumulative probability of CMV DNA detection in infant dried blood spots (DBS) in all infants by randomization arm. P value for log-rank test. B) Loess curves fitted to mean CMV DNA level. HIV-infected infant CMV DNA levels are indicated by Xs, HIV-exposed uninfected infant CMV DNA levels are indicated by closed circles; black indicates placebo and red valacyclovir arm. Dotted line indicates assay limit of detection (100 copies/million cells).

Figure 3. Valacyclovir and maternal CMV levels.

A) Loess curves fitted to mean CMV DNA levels in breast milk samples collected at 2, 6 and 14 weeks postpartum, and 6, 9 and 12 months postpartum. Red line and markers indicate women randomized to valacyclovir. Dotted line indicates assay limit of detection (100 copies/ml). Note: few women were breastfeeding after 180 days postpartum. B) Bars show proportion of women with detectable CMV DNA in the cervix at pre-randomization (34 weeks gestation) and 4 weeks post-randomization (38 weeks gestation). P values below the graph compare the proportion of detectable responses between arms, p values above graph compare the proportion of detectables within each arm, over the two time-points. C) Plots show individual CMV DNA levels for women at baseline and post-randomization, red middle bars show group medians, red whiskers show upper and lower quartiles. P values below the graph show comparison of CMV DNA levels between arms, p values above graph compare CMV DNA levels within each arm, over the two time-points.