Impact of Breastfeeding on Cytomegalovirus Transmission in HIV-Exposed Infants: Systematic Review and Meta-Analysis

Abstract

Background  

In the setting of maternal HIV-1, infant CMV infection is associated with impaired growth and development. HIV-1/CMV co-infected infants have a high risk of mortality, neurologic deficits, and HIV-1 disease progression. Infants may acquire CMV in utero, during delivery, or postnatal through breast milk or saliva. Maternal CMV antibodies protect against congenital disease and infection; but postnatal protection wanes rapidly. In sub-Saharan Africa, 80% of children acquire CMV during the first year of life, and acquisition may occur earlier if mothers have HIV-1. HAART started during the third trimester may decrease infant CMV infections, by mechanisms independent of breast milk CMV levels. Preventing or delaying CMV infection may represent a novel strategy to improve the health of both HIV-infected and HIV-exposed uninfected infants in sub–Saharan Africa, but requires a better understanding of CMV replication and transmission in the setting of maternal HIV-1. This review is focus on those issues and highlights different interactions between cytomegalovirus, HIV, antiretroviral therapy and breastfeeding.

        Objectives  

To investigate whether breastfeeding increase the risk of cytomegalovirus infection in HIV-exposed infants.

To evaluate whether antiretroviral therapy prophylaxis could decrease cytomegalovirus breastfeeding infection in postpartum.

Search methods  

We searched Cochrane Central Register of Controlled Trials (CENTRAL), Scorpus, PubMed,LILACS, Web of Science and Clinicaltrials. Moreover, we performed hand searches in different website conferences, such as the International AIDS Conferences and the annual Conferences on Retroviruses and Opportunistic infections. JT and LM extracted data in the study eligibility form.

Selection criteria  

We selected randomized controlled trials (RCTs), non -randomized control trials, quasi-randomized control trials and prospective cohort studies assessing cytomegalovirus transmission in HIV-exposed infants through breastfeeding.

Data collection and analysis  

JT and EM independently identified and assessed eligible studies that met inclusion criteria. Study design, characteristics of study populations, interventions, controls and results were extracted by JT and EM. Moreover, the risk of bias was assessed independently by JT and EM. We conducted meta-analysis when the population, the intervention, the outcomes and the study design were as similar as possible. We reported the odds ratio, the risk ratio and the mean difference with their respective 95% confidence intervals for the different outcomes.

Main results

 The review has shown that breastfeeding HIV-exposed infants were 77% in risk of developing CMV infection compared to no breastfeeding and/or formula feeding (OR 0.23 95% CI (0.04 to 1.32), 5 studies, 1007 infants, P=0.10), this result was not statistically significant. Antiretroviral therapy was effective in reducing 83% of CMV infection compared to no antiretroviral therapy or antiretroviral prophylaxis (OR 0.17 95% CI 0.03 to 0.94, 3 studies, 560 infants, P= 0.04).

The calculated mean difference of time to CMV infection between breastfeeding HIV-exposed infant group and control was 5.61 months 95% CI [5.34, 5.88]. We can conclude that breastfeeding increase the mean difference of time to CMV infection high as 5.34 months and as low as 5.88 months compared the control group. This result was highly statistically significant with P < 0.00001. In breastfeeding HIV-exposed infants, cytomegalovirus screening was 47% effective in >= 4 months compared to < 4 months of age (RR 1.47 95%CI 1.13 to 1.93, 1 study, 146 infants, P= 0.005). CMV DNA and CMV PCR were statistically significant in breastfeeding HIV-exposed infants, with 63% (OR 0.37 95% CI (0.20 to 0.67), 5 studies, 1823 infants, P=0.001) and 50% (OR 0.50 95% CI 0.30 to 0.84, 2 studies, 365 participants, P=0.009), respectively compared to the control group. Lastly, the review has illustrated that breastfeeding HIV-exposed infants who were positive in CMV PCR were 42% in risk of dying compared to the compared group (RR 1.42 95% CI 0.57 to 3.57, 3 studies, 1426 infants, P=0.46). The overall evidence was graded very low for CMV infection, CMV DNA and all cause of mortality. However, CMV screening was graded low; time to CMV infection CMV IgG were graded moderate.

Authors' conclusions  

This review has shown CMV infection occurs frequently in breastfeeding HIV-exposed infants. This infection is more likely to happen in the last months of exclusive breastfeeding. However, maternal antiretroviral therapy could decrease significantly CMV infection in HIV-exposed children. In addition, this study illustrated that CMV should be screened above four months postnatal because CMV infection is acquired around five months post natal. Then, we urge further research in this field so that CMV infection could be minimized in HIV-exposed infants in the last months of exclusive breastfeeding. Therefore, this review should be taken in a context of several limitations.