Strategic Approaches to Decrease Breast Milk Transmission of HIV-1: The Importance of Small Things

Research on the prevention of mother-to-child transmission (PMTCT) of human immunodeficiency virus type 1 (HIV-1) has yielded a number of effective interventions resulting in extremely high programmatic success in developed countries, where a transmission risk of <1% is attainable. The cornerstones of this success are the use of combination antiretrovirals and the avoidance of breast-feeding. Importantly, the prevalence of maternal HIV-1 infection is low in developed countries, and there is programmatic ability to deliver integrated, multifaceted care to infected women

Worldwide, the majority of infant HIV-1 infections occur in Africa. A decade ago, maternal-infant single-dose nevirapine was shown in the HIVNET 012 study to halve mother-to-child transmission of HIV-1 [1]. The single-dose nevirapine regimen, which was unexpectedly efficient in decreasing perinatal HIV-1 transmission, was also eminently feasible. At a cost of ∼$3 and requiring only a single maternal and infant dose, the treatment seemed possible to implement rapidly, but implementation proved to be slower than predicted. Despite the simplicity of single-dose nevirapine, the need to systematically provide antenatal HIV-1 testing during pregnancy proved to be an obstacle. Six years after the results of HIVNET 012 were released, only an estimated 11% of HIV-1-infected pregnant women received antenatal HIV-1 testing and single-dose nevirapine [2]. Intensification of efforts over the past 5 years has produced promising results in some countries. In low- and middle-income countries, however, HIV-1 testing during pregnancy remains dismally low (∼18%) [3]. Thus, it remains critical to establish HIV-1 testing systems during pregnancy (within which PMTCT interventions will be delivered), and the importance of preventing breast milk transmission of HIV-1 will become increasingly evident. Integration of highly active antiretroviral therapy (HAART) for treatment-eligible mothers into these programs is another effective way to prevent breast milk transmission of HIV-1

In this issue of the Journal, Taha et al [4] report the results of an important study conducted to determine the benefit of maternal HAART (based on maternal treatment indications) in the prevention of late breast milk transmission of HIV-1. This study provides several notable findings relevant to expanding PMTCT programs. First, the authors found that identifying and treating eligible mothers with HAART was highly effective in preventing late breast milk transmission. Second, infant prophylaxis regimens to prevent earlier breast milk transmission of HIV-1 had neither extended benefit nor increased “rebound” late transmission after their discontinuation. Finally and more speculatively, among HAART-ineligible women infant prophylaxis and maternal HAART were estimated to be comparably effective in preventing breast milk transmission of HIV-1

During the study, which was based in Malawi and was conducted from 2004 through 2007, maternal HAART became more available and was administered through an HIV treatment program separate from the research study. Maternal HAART use was recorded during follow-up as part of research procedures. In the cohort of 2318 mother-infant pairs, HAART-eligible mothers who did not receive HAART had a high risk of late breast milk transmission of HIV-1 after 14 weeks post partum (10.56 cases per 100 person-years), significantly higher than that for HAART-eligible women who received HAART (1.79 cases per 100 person-years). In HAART-eligible mothers, maternal HAART reduced the risk of infant infections due to breast-feeding by ∼82%. In addition, only 3% of HAART-eligible women received treatment before 14 weeks post partum. Thus, the estimated benefit of maternal HAART in preventing HIV-1 infection in infants would be even higher if HAART-eligible women received HAART earlier

The findings of Taha and colleagues provide a clear estimate of the impact of maternal HAART on breast milk HIV-1 transmission in immunosuppressed mother-infant pairs and highlight its importance. Earlier studies of late postnatal transmission consistently noted markedly increased risks of breast milk transmission of HIV-1 in women with low CD4 cell counts, probably a result of high systemic and breast milk HIV-1 levels. In a large meta-analysis of individual patients, women with CD4 cell counts <200 cells/μL had an ∼8-fold increased risk of late postnatal HIV-1 transmission, compared with nonimmunosuppressed women [5]. Similarly, Taha and colleagues noted that HAART-eligible women (CD4 cell count, <250 cells/μL) who did not receive treatment had a significantly higher transmission risk (∼10%) than did HAART-ineligible women (CD4 cell count, >250 cells/μL) (∼4%). The contribution of immunocompromise to mother-to-child transmission of HIV-1 is often overlooked, but there are marked differences in estimates of transmission risk between studies that provided HAART to immunocompromised women and those that did not. Despite the profound influence that maternal HAART has on transmission estimates, data on maternal HAART provision often have limited visibility in PMTCT studies; as a result, maternal HAART has been underconsidered as an intervention in its own right. It is important to understand that tiered regimens, in which HAART-eligible women receive HAART and others receive short-course antiretrovirals, have a relatively low overall transmission risk, largely owing to treatment of immunocompromised women [6]. As PMTCT programs expand, this component needs to receive increased attention, for both optimal maternal care and increased PMTCT efficacy, particularly in breast-feeding women

As in real-world settings, implementation of maternal HAART in the study by Taha and colleagues required CD4 cell count testing to determine eligibility for HAART and referral to HIV treatment programs for receipt of therapy. Women who had eligible low CD4 cell counts throughout follow-up were more likely to start HAART (50.5%) than those who had initially high and then low counts (23.3%). However, the majority of eligible women did not start HAART; the reasons included waiting periods to be on the treatment roster, skipped visits before CD4 cell counts were available, guardian consent issues, and unwillingness to start treatment. More efficient administration of HAART to eligible women will require a better understanding of these factors. Many maternal-child health (MCH) PMTCT programs refer women for HAART and experience attrition issues similar to those observed in the study by Taha and colleagues. Efforts either to provide HAART in MCH programs or to increase effective transfers between MCH programs and treatment programs could increase prompt HAART initiation. Innovative ways of assessing CD4 cell counts in MCH programs would be useful for prompt identification of HAART-eligible women. Unfortunately, surrogate measures for CD4 cell count (clinical, hemoglobin, or body mass index) fare poorly in identifying women who would benefit most immediately from treatment [7]. If HAART were provided universally without the requirement for pretreatment CD4 cell counts, the question of when and in whom to stop HAART after breast-feeding would remain, because HAART would reconstitute CD4 cell counts, making it difficult to determine eligibility for indefinite treatment post hoc. Making CD4 cell counts available as point-of-care assays would rapidly identify HAART-eligible women. For PMTCT, point-of-care rapid HIV tests revolutionized service delivery by decreasing the inevitable attrition that arose from the need to go to another location for a test or wait for results. Similar approaches regarding CD4 cell count testing will be important for increasing the use of HAART in MCH settings without an existing laboratory infrastructure

A variety of approaches may decrease breast milk transmission of HIV-1, including avoidance of breast milk, early weaning, maternal HAART, maternal short-course antiretrovirals, and infant antiretroviral prophylaxis. Avoidance of breast-feeding or shortened breast-feeding is possible in some settings in Africa, but for many women breast-feeding remains the best option to optimize infant survival because of inadequate sanitation or lack of a safe water supply. Presently, there is no clear evidence that universal maternal HAART would be superior to infant antiretroviral prophylaxis for mother-infant pairs who receive tiered PMTCT (ie, maternal HAART if eligible and combination short-course antiretrovirals if ineligible for HAART). In the study by Taha and colleagues, HAART-ineligible women comprised ∼70% of the cohort; even with a higher CD4 cell count cutoff (CD4 cell count, <350 cells/μL), it is likely that >50% of women would be ineligible for HAART and have a lower transmission risk than HAART-eligible women. HAART-ineligible women in the study had a transmission risk of 3.66 per 100 person-years after 14 weeks, and the authors speculate that infant prophylaxis may decrease this by 67% (as they demonstrated for earlier breast milk transmission of HIV-1 in the parent clinical trial), resulting in a transmission risk similar to that for maternal HAART. Comparable transmission risks for infant prophylaxis and maternal HAART were noted in the BAN (Breastfeeding, Antiretroviral, and Nutrition) study from Malawi [8]. Determining comparative effectiveness and risks of infant prophylaxis versus maternal HAART in HAART-ineligible mothers will be critical in formulating guidelines for preventing late breast milk transmission of HIV-1

In summary, PMTCT programs need to accelerate many incremental steps to successfully prevent HIV-1 infections in infants. Say that HIV-1 infection is diagnosed during pregnancy in only 30% of HIV-1-infected women; in this scenario, even the most efficacious treatment will not benefit the 70% of mother-infant pairs whose infection is left undiagnosed. Thus, expanded antenatal HIV-1 testing is the first priority and is facilitated by rapid opt-out testing strategies. Delivering the most effective PMTCT antiretroviral regimen is the next step. Efforts to expand delivery of antiretroviral therapy have been accelerated by repackaging nevirapine, offering nevirapine doses for home deliveries, and using generic, simplified HAART regimens. Analogously, identification of HAART-eligible women and delivery of maternal HAART will require rapid CD4 cell count testing and effective integrated transfers between antenatal and HIV treatment programs. With more efficient CD4 cell count testing and prompt initiation of HAART in eligible women, antenatal HIV-1 testing will result in both increased maternal survival and decreased infant infections. As research informs new PMTCT policy recommendations, it is critical to improve the logistics of CD4 cell count testing and HAART integration to increase the effectiveness of PMTCT and lay the foundation for even more effective programs

• Received July 31, 2009.
• Accepted July 31, 2009.

• © 2009 by the Infectious Diseases Society of America