INTRODUCTION — Postnatal human immunodeficiency virus (HIV) transmission occurs in utero, peripartum, and postnatally via breastfeeding; without interventions the risk of perinatal HIV-1 transmission is 20 to 45 percent [1].
Over the past two decades, clinical trials have demonstrated that the risk of vertical transmission can be significantly reduced with the administration of antiretroviral medications during pregnancy, delivery, and the postpartum period. Breastfeeding is important for the nutritional and overall health of the infant, and maternal antiretroviral therapy and infant antiretroviral prophylaxis have each been shown to substantially decrease HIV transmission during breastfeeding [2,3].
This topic will discuss clinical data from resource-limited settings on the prevention of HIV transmission during the breastfeeding period. Advice on the use of antepartum and intrapartum antiretroviral prophylaxis for prevention of vertical HIV transmission in resource-limited settings are discussed in detail elsewhere in UpToDate and in the World Health Organization consolidated guidelines on HIV prevention, testing, treatment, service delivery, and monitoring. (See "Prevention of vertical HIV transmission in resource-limited settings".)
EPIDEMIOLOGY OF HIV TRANSMISSION THROUGH BREASTFEEDING
Risk of HIV transmission through breast milk — Breastfeeding contributes to a substantial proportion of new infant HIV infections [4]. Early evidence for HIV transmission via breast milk included detection of virus in breast milk, observational studies in which breastfed infants had higher transmission than non-breastfed infants, and cases in which women acquired HIV after pregnancy and then transmitted to their breastfeeding infant [1]. There have also been reports of infection in infants born to mothers without HIV but breastfed by surrogates with HIV [5]
HIV detection in infants during the first two weeks of life may represent transmission that has occurred in utero, during delivery, or within the early days of breastfeeding [6]. This complicates estimation of early breast milk HIV transmission risk during the first month of life. To try to distinguish potential transmission routes, most studies that have evaluated the risk of infant acquisition of HIV through breastfeeding have defined breastfeeding transmission as infant HIV infection detected at one month of life or later following a negative HIV virologic test.
Timing of HIV transmission through breast milk — Data from clinical trials on prevention of vertical transmission of HIV suggest that the risk of transmission through breastfeeding is highest in early months of the infant's life [7-9]:
●In the Nairobi trial, which evaluated rates of HIV infant transmission among infants randomized to either breastfeeding or replacement feeding, 67 percent of the risk difference between the two trial arms occurred by two months while 75 percent occurred within six months [10].
●In a study of a Malawian cohort of 672 mother-infant pairs, HIV transmission risk declined as breastfeeding continued: 0.7 percent/month during months 1 through 5, 0.6 percent/month during months 6 through 11, and 0.3 percent/month during months 12 through 17 [11].
One meta-analysis (Breastfeeding and HIV International Transmission Study) suggested that the risk of late breast milk transmission (postnatal transmission occurring after one month of age) appeared relatively constant at 8.9 transmissions per 100 person-years of breastfeeding [12].
Risk factors for HIV transmission through breast milk — Clinical and laboratory data suggest that the most important factor affecting HIV transmission to the infant is the level of viremia in the mother [13]. Other risk factors that may affect HIV transmission to the infant include the mother's immunologic status, presence of maternal HIV drug resistance, presence of mastitis, and patterns of feeding.
●Maternal HIV viral load − Higher maternal plasma HIV ribonucleic acid (RNA) and breast milk HIV deoxyribonucleic acid (DNA) and RNA levels are associated with increased transmission risk [13-15]. In a case-control study of 92 infants with HIV and 187 infants without HIV, higher maternal viral RNA levels were associated with an increased risk in HIV transmission to the infant [14]. Nevertheless, some evidence suggests that plasma HIV suppression does not completely eliminate the risk of transmission through breast milk. (See 'Transmission risk despite viral suppression' below.)
In a longitudinal assessment of 275 Kenyan mothers with HIV, a log increase of breast milk HIV RNA levels was associated with twofold higher risk of HIV transmission [16]. Several studies have suggested that levels of cellular HIV DNA in breast milk is more predictive of transmission than levels of cell-free HIV RNA in breast milk [8,17].
●Acute versus chronic HIV infection − Observational data suggest acute maternal HIV infection during breastfeeding is associated with greater transmission risk to the infant than chronic maternal HIV infection. In meta-analyses of several observational studies, the risk of postnatal HIV transmission was significantly higher among women with acute versus chronic HIV infection (29 versus 14 percent in absence of antiretrovirals; 2.9-fold increased risk in studies including antiretrovirals); this observation may be related to the much higher level of HIV RNA seen in acute HIV infection [7,18].
●Maternal immunologic factors − Several studies have noted increased transmission among immunosuppressed females with low CD4 cell counts [1]. This observation may be related to the higher viral burden in patients with advanced HIV disease, but maternal or infant immune status may confer some independent influence on transmission as well.
HIV specific immunoglobulins (IgA and IgM), antibody-dependent cell cytotoxicity, and HIV-specific cytotoxic T cells are detected in breast milk and may influence HIV transmission through breastfeeding [15,19-21].
Detection of HIV-specific T cells in maternal breast milk was also associated with decreased HIV transmission [22]. Secretory leukocyte protease inhibitor (SLPI) is a protein found in saliva, breast milk, and genital secretions, which is capable of inhibiting HIV in vitro. Higher levels of SLPI in infant saliva have been associated with a lower risk of HIV acquisition [23], while SLPI levels in breast milk do not appear to modify transmission rates [24-26].
●Maternal presence of HIV drug resistance − Presence of HIV drug resistance in the mother may be associated with increased risk of postnatal HIV transmission through breast milk, independent of maternal viral load. In a case-control study of 148 breast fed infants (37 with postnatally acquired HIV and 111 without HIV), presence of HIV drug resistance in the mother was associated with postnatal transmission, independent of maternal HIV viral load (odds ratio 4.45) [27].
●Breast infections − Mastitis and breast abscess have been associated with an increased risk of HIV transmission, which may be due to the increased recruitment of HIV-infected inflammatory cells in an area of infection [14,28-32].
●Patterns of feeding − Mixed feeding (breastfeeding along with ingestion of solids/liquids) during the first three to six months of life is associated with an increased risk of infant HIV transmission compared with breastfeeding alone in infants of mothers not receiving antiretroviral therapy. This is further discussed elsewhere. (See 'Efficacy of alternative feeding strategies' below.)
Mechanism of breast milk transmission — The mechanism of transmission of HIV through breast milk has not been fully elucidated. HIV RNA can be detected in colostrum and breast milk [7,33], but some studies suggest that HIV-infected cells within breast milk may play a more important role in infant transmission than cell-free virus [8,9]. Additionally, the portal of entry of HIV in the breastfed infant has not been defined but may include the intestine or tonsillar tissues.
Breaches in intestinal epithelial integrity or compromise in intestinal cellular tight junctions (eg, between epithelial or dendritic cells) may allow entry of infectious virions [26,34-36]. Infants who become infected via breastfeeding have higher levels of lipopolysaccharide, suggesting that compromised gut integrity may facilitate transmission [37]. The possibility of HIV entry into tonsillar lymphocytes has been suggested from studies of a related virus (ie, simian immunodeficiency virus) in infant macaques [38-40].
RECOMMENDATIONS TO PREVENT TRANSMISSION DURING BREASTFEEDING — The use of antiretroviral agents is a critical component of prevention of vertical transmission during antepartum, peripartum, and early postpartum periods (see "Prevention of vertical HIV transmission in resource-limited settings"). During breastfeeding, maternal combination antiretroviral therapy (ART) and infant antiretroviral prophylaxis strategies are each comparably effective at reducing the risk of transmission [3]. Efforts to diagnose, treat, and retain individuals on ART during pregnancy and breastfeeding and efforts to prevent new infant HIV infections after birth will all be necessary to attain elimination of infant HIV. The World Health Organization (WHO) recommends maternal ART to reduce transmission during breastfeeding, as this benefits maternal health and decreases risk of infant HIV and simplifies antiretroviral management of pregnant and breastfeeding females through the entire period of transmission risk [41,42]. WHO guidelines recommend lifelong ART for all individuals with HIV, including pregnant and breastfeeding females [41]. (See "When to initiate antiretroviral therapy in persons with HIV", section on 'Introduction'.)
Infant antiretroviral use remains important as short-term postexposure prophylaxis after delivery and in settings in which maternal antiretroviral use is delayed or interrupted during breastfeeding.
Details on the duration of breastfeeding and the timing and selection of antiretroviral regimens are discussed below.
Options for infant feeding — Recommendations on infant feeding must balance the risk of HIV transmission through breast milk on one hand and the risk of malnutrition and serious infections through unsafe feeding practices on the other (see 'Replacement feeding' below). The WHO recommends that for mothers known to have HIV, public health authorities should focus either on promoting breastfeeding and providing antiretroviral interventions to prevent transmission or on avoiding all breastfeeding and providing alternate sources of safe nutrition [41,43]. Promoting breastfeeding with antiretroviral interventions is the better strategy in most resource-limited settings. In an analysis of pooled data from studies in Asia and Africa, lack of breastfeeding was associated with mortality in HIV-exposed but uninfected infants [44]. Maternal ART during breastfeeding use was associated with lower child mortality rates.
Exclusive breastfeeding protocol — For females with HIV in those countries that have chosen to support breastfeeding with antiretroviral interventions, the WHO recommends exclusive breastfeeding for six months, in combination with maternal ART and a short-term period of infant prophylaxis, to minimize HIV transmission from the mother while optimizing the health benefits of breastfeeding for the infant [41,45]. Subsequently, breastfeeding, along with maternal ART (with provision of ART adherence support) and appropriate complementary feeding, should continue without restrictions for up to 24 months or longer, similar to the general population [46].
Earlier studies suggested that exclusive breastfeeding was difficult to implement given cultural or logistic constraints with <10 percent of females implementing exclusive breastfeeding despite counseling [47,48]. With expanded counseling and home visits, higher levels of exclusive breastfeeding are possible, with some studies reporting >80 percent of females with HIV exclusively breastfeeding at three months postpartum [49-51]. Breastfeeding support at delivery has been associated with increased prevalence of exclusive breastfeeding at six weeks [52].
Maternal antiretroviral use
When to initiate an antiretroviral regimen — The WHO guidelines recommend initiation of ART for all pregnant or breastfeeding females with HIV, regardless of their CD4 cell count, to reduce the risk of transmission to their infants and for their own clinical benefit [41]. Breastfeeding females already taking ART should continue it. For other females, an ART regimen should be initiated as soon as possible (eg, as soon as a female with HIV is found to be breastfeeding or as soon as HIV is diagnosed in a breastfeeding female) [41]. Maternal ART has multiple benefits, including prevention of HIV transmission to the infant, improved long-term infant HIV-free survival, maternal health, and prevention of HIV transmission to sexual partners [41,42,44,53].
Data from trials of maternal antiretroviral use to prevent transmission during breastfeeding suggest that earlier initiation of combination antiretroviral drug regimens during or before pregnancy (rather than at delivery) is associated with a lower risk of infant transmission [54]. This is because of the time that it takes for antiretroviral agents to lead to complete viral suppression. (See 'Efficacy of antiretroviral drugs to prevent transmission during breastfeeding' below.)
Regimen selection — The WHO guidelines recommend initiation of tenofovir disoproxil fumarate, lamivudine (or emtricitabine), and dolutegravir in pregnant or breastfeeding females with HIV [41]. Although rare neural tube defects have been observed among females receiving dolutegravir at the time of conception in observational cohorts [55,56], further accumulation of data demonstrates no difference in risk of neural tube defects between dolutegravir-based and non-dolutegravir based regimens [57]. (See "Safety and dosing of antiretroviral medications in pregnancy", section on 'Dolutegravir'.)
The tenofovir, lamivudine (or emtricitabine), and dolutegravir regimen aligns with recommendations for all adults in resource-limited settings. While neither the tenofovir, lamivudine (or emtricitabine), and dolutegravir nor tenofovir, lamivudine (or emtricitabine), and efavirenz has been explicitly studied in trials of preventing HIV transmission during breastfeeding, both result in viral suppression in the majority of individuals [58,59], which is the most important clinical factor affecting the risk of postnatal transmission (see 'Risk factors for HIV transmission through breast milk' above). A multicenter, randomized trial compared dolutegravir versus efavirenz in combination with lamivudine (or emtricitabine) and either tenofovir disoproxil fumarate or tenofovir alafenamide and found that dolutegravir-containing regimens had superior virologic efficacy at delivery compared with efavirenz [60]. Although the tenofovir alafenamide (TAF) regimen (compared with tenofovir disoproxil fumarate) had a better safety profile with fewer composite adverse pregnancy outcomes, TAF is not widely available in resource-limited settings. Other randomized trials found more rapid viral decline with dolutegravir-based then efavirenz-based regimen and persistent maternal viral suppression during breastfeeding [61,62].
Any maternal ART regimen that results in virologic suppression should result in significantly decreased perinatal and postnatal transmission. For example, a protease inhibitor-based regimen used in one large trial resulted in an 18-month postnatal transmission rate of only 0.6 percent [3]; in a separate study comparing an efavirenz-based regimen with a lopinavir-ritonavir-based regimen, infant 12-month transmission rates were similarly low [3,63].
Discussion of alternative WHO-recommended first-line or second-line regimens is also found elsewhere. (See "Use and impact of antiretroviral therapy for HIV infection in resource-limited settings", section on 'Regimen selection'.)
Acute HIV infection postpartum — In regions with high HIV seroprevalence, females may be at risk for acquiring HIV during the breastfeeding period. Females who acquire HIV during breastfeeding have a high risk of transmitting HIV to their breastfeeding infant [18]. Thus, acute or early HIV in the postpartum period warrants immediate ART with the same regimens discussed above, in addition to at least 12 weeks of enhanced antiretroviral prophylaxis for breastfeeding infants. (See 'Regimen selection' above and 'Infant antiretroviral use' below.)
Maintaining retention and adherence to antiretrovirals during breastfeeding — Adherence counseling approaches may need to be adapted for this period. Specific counseling may be needed to help females continue antiretrovirals during breastfeeding, particularly after nine months when visits to maternal child health centers may decrease. A variety of other approaches have been evaluated to improve retention and adherence, including lay counselors, cash transfers, engagement of partners, and health interventions; some have had positive results [64-68].
Infant antiretroviral use
Breastfeeding infants — For breastfeeding infants, the purpose of antiretroviral prophylaxis is to prevent transmission from exposure to HIV during delivery and the early breastfeeding period, as well as during any periods of breastfeeding when the mother may not have virologic suppression. All breastfeeding infants born to mothers with HIV should receive antiretroviral prophylaxis [41]. The recommended regimen depends on the infant’s risk of infection, as determined by the timing of maternal infection and ART use (algorithm 1):
●For infants born to mothers with established HIV who, by the time of delivery, had achieved viral suppression on ART or had been regularly taking ART for more than four weeks, we recommend daily nevirapine for six weeks.
●For infants born to mothers with HIV who had a viral load >1000 copies/mL within the four weeks prior to expected delivery, who received no ART or less than four weeks of ART by the time of delivery, or who acquired HIV during pregnancy or breastfeeding, we recommend daily nevirapine PLUS twice-daily zidovudine for six weeks, followed by an additional six weeks of the same combination regimen or nevirapine alone.
●For infants identified after birth as having been exposed to HIV, a virologic diagnostic test should be performed. In addition to initiation of maternal ART, the infant should start daily nevirapine plus twice-daily zidovudine immediately and continue for at least six weeks. This should be followed by an additional six weeks of either nevirapine plus zidovudine or nevirapine alone. If the virologic test identifies that the infant has HIV, infant prophylaxis should be stopped and combination ART initiated. The WHO recommends that a raltegravir-based antiretroviral regimen should be first-line therapy for all children with HIV under the age of one month [41]. Dolutegravir is approved for children older than one month who weigh more than 3 kg.
In the unusual circumstance of a mother who cannot tolerate or declines ART, the infant should continue nevirapine prophylaxis throughout the duration of breastfeeding, until one week following breastfeeding cessation (if nevirapine is not tolerated, daily lamivudine can be used).
Nevirapine and zidovudine dosing are based on the weight of the infant (table 1).
A disadvantage of infant prophylaxis is the risk of developing HIV drug resistance if the infant does acquire infection during breastfeeding. Additionally, drug resistance has also been observed in infants who become infected despite maternal combination drug use during breastfeeding [69-71]. However, since the effectiveness of nevirapine-based prophylaxis in preventing transmission among breastfeeding infants is high, the overall risk of drug resistance among infants receiving prophylaxis is low.
Infants on replacement feeding — For infants who receive replacement feeding instead of their mothers' breast milk, the purpose of infant antiretroviral prophylaxis is to prevent transmission from exposure to HIV during delivery. All replacement feeding infants born to mothers with HIV should receive antiretroviral prophylaxis [41]. The recommended regimen depends on the infant’s risk of infection, as determined by the timing of maternal infection and ART use (algorithm 1). This is discussed in detail elsewhere. (See "Prevention of vertical HIV transmission in resource-limited settings", section on 'Infant postnatal prophylaxis'.)
Infants who are identified more than 24 hours after birth as having been exposed to HIV should have a virologic diagnostic test performed. Uninfected infants who are using replacement feeding do not need antiretroviral prophylaxis because there is no ongoing risk of transmission through breast milk, and the period for effective prophylaxis of intrapartum transmission may be passed (eg, infant prophylaxis initiated 24 to 48 hours or more after birth is unlikely to affect intrapartum transmission). Combination ART should be initiated if the child is diagnosed with HIV. The WHO recommends that a raltegravir-based antiretroviral regimen should be first-line therapy for all children with HIV under the age of one month [41]. Dolutegravir is approved for children older than one month who weigh more than 3 kg [41].
EFFICACY OF ANTIRETROVIRAL DRUGS TO PREVENT TRANSMISSION DURING BREASTFEEDING — Clinical trials have demonstrated that the use of antiretroviral medications during pregnancy, labor, and delivery effectively reduces intrauterine and intrapartum HIV transmission from 24 percent to approximately 1 percent [53,72,73] (see "Prevention of vertical HIV transmission in resource-limited settings"). However, without ongoing prophylaxis after delivery, one meta-analysis demonstrated that an additional 9 percent of infants who are born to mothers with HIV will become infected after one month of age through breastfeeding [74]. Fortunately, several randomized trials and observational studies in resource-limited settings have shown that use of antiretroviral medications by the mother or infant during breastfeeding can substantially reduce rates of late postnatal HIV transmission to the infant to less than 3 percent [13,75-77], and in a more recent study, down to 0.6 percent at 18-months of follow-up [3].
The discussion below highlights some of the main concepts that have emerged within the last several years regarding the most effective use of antiretroviral medications for the prevention of HIV during breastfeeding. Drugs that have been studied include nevirapine, which is a nonnucleoside reverse transcriptase inhibitor (NNRTI); zidovudine, abacavir, and lamivudine, which are nucleoside reverse transcriptase inhibitors (NRTI); lopinavir/ritonavir, which are protease inhibitors (PI); dolutegravir, which is an integrase inhibitor. (See "Safety and dosing of antiretroviral medications in pregnancy".)
Efficacy of maternal antiretroviral drug use — Several studies have demonstrated that pregnant and breastfeeding females who achieve virologic suppression on antiretroviral agents have low rates of HIV transmission to their infants, with reported rates ranging from <1 to 5 percent [3,13,75,76,78-80]. This supports the observation that there is a significant association between the level of the mother's viral load and HIV infant transmission (see 'Risk factors for HIV transmission through breast milk' above). Since several weeks may lapse before viral suppression can be achieved following initiation of antiretroviral agents, drugs should be initiated before or during pregnancy, rather than at delivery, to maximally reduce infant HIV transmission as illustrated in the studies discussed here [81].
In the Mma Bana trial in Botswana, 560 women with CD4 cell counts ≥200 cells/microL were randomly assigned to receive either triple NRTIs (zidovudine, lamivudine, and abacavir) or a PI-containing regimen (lopinavir-ritonavir with zidovudine and lamivudine) [13,82]. The regimen was initiated between 26 and 34 weeks of gestation and continued through weaning or six months postpartum (whichever occurred first). All infants also received short-term postnatal antiretroviral prophylaxis; 97 percent breastfed for a median of 5.8 months. Women with advanced immunosuppression (CD4 cell counts <200 cells/microL) were placed in the observational group and initiated on chronic antiretroviral therapy (ART) with nevirapine, zidovudine, and lamivudine for their own health. High rates of maternal viral suppression to <400 copies/mL were seen in both groups (96 percent in the NRTI group versus 93 percent in the PI group). The overall rate of mother-to-child transmission was 1.1 percent. This rate is similar to that seen among treated females in resource-rich settings who did not breastfeed [13,82,83]. Because of the low number of maternal-infant transmissions (eight cases total), the effect of specific immunologic or virologic thresholds on the risk of transmission could not be analyzed. Rates of adverse events were similar in the NRTI and PI groups (6 percent each). However, there was a significantly higher rate of preterm delivery in the PI group than the NRTI group (23 versus 15 percent, respectively). (See "Safety and dosing of antiretroviral medications in pregnancy", section on 'Preterm birth'.)
Similarly, the multisite PROMISE trial, conducted in sub-Saharan Africa and India, demonstrated a transmission rate of 0.3 percent at six months and 0.6 percent at 12 months postpartum among breastfeeding women who continued PI-based ART for the duration of breastfeeding [3].
Observational studies have found somewhat higher overall transmission rates of approximately 3 to 5 percent with use of combination antiretroviral regimens in breastfeeding women [75,79]. For example, an observational study in Mozambique of mothers with HIV who were provided antenatal and postnatal antiretroviral medications for six months (mainly zidovudine or stavudine in combination with lamivudine and nevirapine) demonstrated an overall HIV infant transmission rate of 2.8 percent at 12 months [76].
In the Breastfeeding, Antiretrovirals, and Nutrition (BAN) trial, the postnatal risk of HIV transmission was 2.9 percent in infants of mothers who were randomly assigned to receive a combination antiretroviral regimen started at two weeks postpartum delivery until weaning at 26 to 28 weeks [2]. In this study, there was a delay of 12 weeks in the effect of maternal antiretroviral use, which reflects the time needed to achieve virologic suppression. This suggests that earlier initiation of antiretrovirals during pregnancy would further decrease the risk of infant HIV transmission. The BAN trial is discussed in greater detail elsewhere. (See 'Efficacy of maternal versus infant antiretroviral prophylaxis' below.)
Efficacy of infant antiretroviral prophylaxis — Several trials have demonstrated the success of infant antiretroviral prophylaxis in decreasing the risk of HIV acquisition during breastfeeding [2,84-86]. Nevirapine is effective as a single prophylactic agent and does not appear to have significant safety concerns when used in infants; zidovudine may not add any additional benefit for prevention of transmission and may increase the risk of anemia [2,85].
In a pooled analysis of data from 5,396 mother-infant pairs included in five randomized trials that evaluated the prevention of HIV transmission during breastfeeding, a longer duration of infant nevirapine prophylaxis was associated with a greater reduction in the risk of HIV infection [87]. This finding was confirmed in a subsequent clinical trial of 1,527 breastfeeding infants born to mothers with HIV in four African countries that evaluated the incremental benefit of extended nevirapine prophylaxis [86,88]. All infants received nevirapine from birth to six weeks; subsequently the infants were randomly assigned to either extended nevirapine prophylaxis or placebo until six months of follow-up or until breastfeeding cessation, whichever came first. Compared with six weeks of infant nevirapine followed by placebo, extended nevirapine prophylaxis reduced postnatal transmission of HIV by 54 percent at six months (2.4 percent placebo versus 1.1 percent extended nevirapine) but not subsequently [88]. More prolonged infant prophylaxis is also feasible and effective, as illustrated by a randomized trial of over 1,000 HIV-exposed infants, in which the postnatal transmission rate at 50 weeks was 1.4 to 1.5 percent with infant prophylaxis with daily lopinavir-ritonavir or lamivudine during breastfeeding for a median of 41 weeks [89].
The PROMISE trial, which evaluated even longer infant prophylaxis, reported very low transmission rates (0.57 percent) with daily infant nevirapine prophylaxis for 18 months and no maternal ART during breastfeeding [3]. A similar low rate of transmission was observed with maternal ART and only six weeks of infant nevirapine during breastfeeding, as discussed in detail below. (See 'Efficacy of maternal versus infant antiretroviral prophylaxis' below.)
Efficacy of maternal versus infant antiretroviral prophylaxis — Maternal combination ART or infant nevirapine prophylaxis decreases the risk of HIV transmission during breastfeeding with similar efficacy. However, given the WHO recommendation of lifelong ART for all individuals with HIV, including pregnant and breastfeeding females, maternal ART is the preferred strategy for prevention of breast milk transmission. (See 'Recommendations to prevent transmission during breastfeeding' above.)
The PROMISE trial was a large, randomized controlled trial conducted in 14 sites in sub-Saharan Africa and India that demonstrated no difference in postnatal HIV transmission rates with maternal ART versus infant prophylaxis during 24 months of breastfeeding [3]. Among 2,431 total mother-infant pairs, postnatal HIV infection occurred in 7 of 1,219 infants in the maternal treatment arm (0.57 percent) and 7 of 1,211 infants in the nevirapine prophylaxis arm (0.58 percent, hazard ratio 1.0, 95% repeated CI 0.3-3.1). Infant HIV-free survival was high in both arms (97.1 and 97.7 percent with maternal treatment and infant prophylaxis, respectively). The overall low postnatal transmission rates in this trial reflect that all women enrolled in the PROMISE study had received maternal antepartum antiretroviral drugs.
The similar HIV transmission rates with the two prevention strategies in the PROMISE trial are consistent with findings from the earlier BAN trial, which was the first trial to assess the relative effectiveness of maternal or infant antiretroviral interventions compared with controls for prevention of HIV transmission to infants [2]. In this trial, mother-infant pairs were randomly assigned at delivery to a maternal regimen group, an infant regimen group, or neither intervention (the control group). Mothers who were enrolled had a CD4 cell count ≥ 250 cells/microL. The infant regimen included daily nevirapine alone for 28 weeks; the maternal combination drug regimen varied during the trial due to emerging safety data related to individual agents (eg, hypersensitivity reactions and nevirapine use in women with CD4 counts >250 cells). Most of the women in the trial received zidovudine, lamivudine, and ritonavir-boosted lopinavir. Perinatal antiretroviral prophylaxis (single-dose nevirapine plus one week of zidovudine and lamivudine) for the prevention of vertical transmission was given at delivery to all mothers and infants according to the contemporaneous standard of care. Women were advised to exclusively breastfeed during the study until 24 to 28 weeks when they were then instructed to rapidly wean. Maternal antiretrovirals were discontinued one week after weaning. The primary efficacy endpoint was the rate of detection of HIV infection at 28 weeks postpartum among infants who were uninfected at two weeks of life between the two intervention arms compared to the control arm. The trial demonstrated that either a maternal regimen of combination antiretroviral medications or an infant regimen of daily nevirapine were both effective in reducing the risk of HIV transmission (cumulative postnatal risk 2.9 percent and 1.7 percent, respectively) compared with controls (5.7 percent). The effectiveness of infant prophylaxis was evident by six weeks, while the relative effectiveness of maternal ART was only first apparent at 12 weeks. After 18 weeks of age, the risk of infection was similar between the two arms. This delayed effect in the maternal ART arm may be related to the time needed to attain maternal viral suppression, since ART was first initiated at delivery. The study did not assess the association between HIV RNA detection and risk of infant transmission.
An analysis of 48 week outcomes of the BAN trial demonstrated that, although the cumulative risk of HIV transmission still remained less in the maternal and infant prophylaxis arms (4 percent in both) than in the control group (7 percent), the transmission events that occurred after 28 weeks (when all prophylaxis had been discontinued) were evenly distributed across all groups [90]. Half of the late infections occurred at least 42 days after the reported cessation of breastfeeding. Furthermore, overall infant morbidity increased from 0.7 adverse events per 1000 person-weeks during the intervention period to 1.1 events per 1000 person-weeks after week 28, which correlated with the time when women weaned their infants. Although the study was not designed to evaluate the effect of early weaning, its results suggest that in resource-limited settings, instructing women to cease breastfeeding early at 24 to 28 weeks may not be an effective strategy against further HIV transmission and, as seen in other studies, may be associated with increased infant morbidity and mortality. (See 'Shortened breastfeeding' below.)
Transmission risk despite viral suppression — Although there is strong evidence that HIV is not sexually transmitted when the viral load is undetectable, results from breastfeeding cohorts suggest that the concept that undetectable equals untransmissible ("U=U") may not apply to breast milk transmission [91-93]. This may be due to higher persistent exposure, local activation, or other factors.
In a systematic review among women on ART during breastfeeding, the pooled transmission risk at 12 months was 2.93 percent [92]. It is likely that not all breastfeeding women in this review were actually virally suppressed, despite receiving ART. However, in the PROMISE trial, two of seven HIV transmissions during breastfeeding occurred among women on ART with viral loads <40 copies/mL near the time of transmission [93]. Further evidence in this area will be useful for females in other settings to make informed decisions regarding breast or replacement feeding.
PENETRATION OF ANTIRETROVIRAL MEDICATIONS INTO BREAST MILK — The ability of an antiretroviral drug to pass from a mother receiving antiretroviral therapy into the breast milk could potentially affect efficacy of the drug for preventing postnatal infection. While drugs with more penetration into breast milk may have greater efficacy, such penetration could also be associated with a risk of toxicity in the breastfeeding infant.
A systematic review of pharmacokinetic studies that measured drugs in breast milk of women with HIV taking antiretrovirals identified 24 relevant studies and suggested that infants would be exposed to quantitatively important levels of some nucleoside reverse transcriptase inhibitors (NRTIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs) through ingestion of breast milk [94]. While there was substantial variability in the clinical and laboratory methods used, NRTIs such as lamivudine and zidovudine showed accumulation in breast milk relative to maternal plasma, with breast milk to maternal plasma ratios ranging from 0.89 to 1.21. For NNRTIs such as nevirapine and efavirenz, ratio estimates ranged from 0.71 to 0.94, and for protease inhibitors such as lopinavir-ritonavir, they ranged from 0.17 to 0.21. The review estimated that, relative to recommended pediatric doses, a breastfed infant may ingest 8.4, 12.5, and 1.1 percent of lamivudine, nevirapine, and efavirenz, respectively, via breast milk.
In an intensive pharmacokinetic evaluation of samples from maternal and infant blood and breast milk, emtricitabine and lamivudine accumulated in breast milk and were detected in breastfeeding infants, whereas tenofovir had limited penetration into breast milk and was undetectable in breastfeeding infants [95]. Similar findings related to tenofovir were observed in a study of infants of women without HIV receiving tenofovir for pre-exposure prophylaxis, with tenofovir undetectable in 94 percent of infant plasma specimens [96]. Dolutegravir levels in breast milk have been observed to be 3 percent of plasma concentration, with significant infant levels (3 to 8 percent of maternal exposure); maternal dolutegravir dosing that is stopped at delivery is estimated to result in subsequent exposure through breastfeeding equivalent to 4.5 days of infant prophylaxis [61,97,98].
DRUG TOXICITY CONCERNS
Maternal toxicity — The antiretroviral regimens used in clinical trials of preventing vertical transmission of HIV were relatively well tolerated. Nevirapine has been associated with the most adverse effects. In the Breastfeeding, Antiretrovirals, and Nutrition (BAN) trial, the rate of neutropenia with nevirapine was 6.2 percent, and 6 of 39 women who initially received a nevirapine-containing regimen developed a hypersensitivity reaction, including two with Stevens-Johnson syndrome [2]. Side effects associated with specific antiretroviral agents among pregnant and breastfeeding females are largely the same as those experienced in the general population with HIV. (See "Overview of antiretroviral agents used to treat HIV" and "Safety and dosing of antiretroviral medications in pregnancy".)
Fetal and infant toxicity — Adverse birth outcomes, such as low birth weight, stillbirth, prematurity, and small for gestational age have been an issue of concern with the use of antiretroviral medications during pregnancy [83]. (See "Prevention of vertical HIV transmission in resource-limited settings", section on 'Adverse effects/safety'.)
Some, but not all, studies have reported elevated rates of hematologic toxicity among breastfeeding infants of mothers who were receiving zidovudine-based combination antiretroviral regimens during pregnancy and/or breastfeeding [99-101]. However, studies that have evaluated the impact of breast milk exposure to tenofovir-lamivudine-efavirenz (the recommended antiretroviral regimen) on infant laboratory values, growth, and markers of bone growth or resorption have not shown evidence of toxicity in the breastfeeding infant [102-104].
Infant administration of zidovudine, nevirapine, and lamivudine have all been studied for extended infant HIV prophylaxis. Postnatal infant zidovudine is associated with hematologic toxicity and does not appear to provide any additive benefit in preventing transmission over nevirapine alone [85]. Infant prophylaxis with lamivudine was well tolerated in one open-label prospective study and non-inferior to lopinavir-ritonavir prophylaxis [89,105]. In the BAN trial, a small percentage of infants receiving daily nevirapine (<2 percent) developed a hypersensitivity reaction, which resolved in all cases with discontinuation of nevirapine and substitution of lamivudine as the infant prophylaxis regimen. In the PROMISE study, in which breastfeeding infants received 18 months of daily nevirapine, toxicity rates were low and similar to breastfeeding infants receiving only six weeks of daily nevirapine (with exposure to maternal lopinavir-ritonavir-based antiretroviral therapy) [3].(See "Overview of antiretroviral agents used to treat HIV", section on 'Non-nucleoside reverse transcriptase inhibitors (NNRTIs)'.)
EFFICACY OF ALTERNATIVE FEEDING STRATEGIES — Replacement feeding, shortened breastfeeding, and mixed feeding have been assessed to determine if alternative strategies in infant nutrition may limit the risk of HIV, in the absence of antiretroviral drug interventions. As discussed below, all of these approaches have their limitations.
Replacement feeding — Avoidance of breastfeeding is a certain way to prevent HIV transmission via breast milk during the postnatal period [106]. In fact, following the first report of HIV transmission through breast milk in the United States, the Centers for Disease Control issued guidelines recommending replacement feeding for infants without HIV born to mothers with HIV [107]. However, in resource-limited countries, replacement feeding is associated with other significant risks, such as increased infant morbidity and mortality from diarrheal disease, pneumonia, and other infectious diseases, thus limiting the utility of this option [108-113]. Decreased rates of overall morbidity and mortality in infants who breastfeed have been attributed to protective antibodies passed from mother to infant in breast milk and decreased exposure to infectious pathogens via unsafe water [45].
In a few areas with access to clean water supplies (eg, Nairobi), replacement feeding has been associated with significantly lower rates of HIV transmission compared with breastfeeding [114]. However, this trial preceded antiretroviral prophylaxis studies that demonstrated marked efficacy of decreasing transmission of HIV in both replacement feeding and breastfeeding females [115-117]. Thus, in the context of antiretroviral medications, the benefits of replacement feeding (eg, improving HIV-free survival) are no longer retained in resource-limited settings [118].
Shortened breastfeeding — Some studies have suggested that a shortened course of breastfeeding may be somewhat protective against HIV infant transmission [119]. In order to limit the risk of HIV transmission and to facilitate the benefits of passive maternal antibody transfer, prior WHO guidelines recommended breastfeeding for a limited period of time (eg, six months). However, subsequent studies noted increased infant morbidity, mortality, and growth compromise with early breastfeeding cessation compared with longer periods of breastfeeding [120-123]. Furthermore, in a study of Zambian Exclusive Breastfeeding, HIV-free survival was found to be similar among infants who abruptly stopped breastfeeding at four months and those who continued for longer periods of time [124]. (See 'Timing of HIV transmission through breast milk' above.)
Finally, with expanded antiretroviral drug availability in resource-limited settings, the risk of breast milk-associated HIV transmission becomes low, making early breastfeeding cessation relatively more risky than longer periods of breastfeeding, in terms of nutrition and other benefits.
Mixed feeding — Although it might be anticipated that mixed feeding (eg, breastfeeding combined with other types of liquids or solids) might be associated with a lower risk of HIV transmission compared with exclusive breastfeeding due to less exposure to HIV-infected breast milk, studies have not supported this hypothesis.
Several studies from South Africa, Zambia, West Africa, and Zimbabwe have demonstrated that mixed feeding during the first three to six months of life is associated with an increased risk of infant HIV transmission compared with breastfeeding alone in infants of mothers not receiving antiretroviral therapy [48-50,125]. The reason for this observation is not well understood; some have suggested that infants may have greater exposure to infectious agents through liquids or solids, which compromise gut integrity, allowing for HIV transmission [126]. In the context of maternal antiretroviral use and infant prophylaxis, the relative influence of mixed feeding on breast milk HIV transmission may be attenuated, however, exclusive breastfeeding confers many benefits independent of HIV and should be supported in all females regardless of HIV infection.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: HIV treatment in pregnant women" and "Society guideline links: Breastfeeding and infant nutrition".)
SUMMARY AND RECOMMENDATIONS
●Vertical HIV transmission occurs in utero, peripartum, and postnatally via breastfeeding; the risk of HIV transmission to the infant is significantly reduced with antiretroviral medications. (See 'Introduction' above.)
●The risk of HIV transmission through breast milk is greatest in the first several months of life; however, a lower but constant risk persists throughout the entire breastfeeding period. The risk also increases with increased levels of maternal HIV RNA in plasma or breast milk. (See 'Timing of HIV transmission through breast milk' above and 'Risk factors for HIV transmission through breast milk' above.)
●Replacement feeding is recommended for infants born to mothers with HIV in the United States, which is the standard of care in resource-rich settings. However, in resource-limited settings, replacement feeding is associated with greater infant morbidity and mortality from diarrheal disease, pneumonia, and other infectious diseases. Therefore, the World Health Organization (WHO) recommends breastfeeding with antiretroviral interventions (maternal antiretroviral therapy [ART] and short-term infant antiretroviral prophylaxis) in such settings to prevent transmission. (See 'Options for infant feeding' above.)
●In such settings, exclusive breastfeeding in combination with ART, is recommended for the first six months of life. Subsequently, breastfeeding, along with ART, adherence support, and appropriate complementary feeding should continue up to 24 months or longer, similar to the general population. (See 'Exclusive breastfeeding protocol' above.)
●The WHO recommends that ART be initiated in all individuals with HIV, including pregnant and breastfeeding females, regardless of CD4 cell count or clinical stage, and continued lifelong. A once-daily, fixed-dose combination of tenofovir, lamivudine (or emtricitabine), and dolutegravir is the preferred first-line regimen. (See 'Maternal antiretroviral use' above.)
●Antiretroviral medications significantly decrease the risk of vertical transmission of HIV during the antepartum, intrapartum, and early postpartum periods. However, risk of HIV transmission recurs if antiretrovirals are discontinued during the breastfeeding period. Although maternal and infant antiretroviral prophylaxis strategies during the breastfeeding period are comparably effective in reducing the rate of transmission, maternal ART is the preferred strategy, as it is recommended for all individuals with HIV and is essential for maternal health. (See 'Efficacy of antiretroviral drugs to prevent transmission during breastfeeding' above.)
●Although there is strong evidence that HIV is not transmitted sexually when the viral load is undetectable, breast milk transmission may still occur despite plasma viral suppression. (See 'Transmission risk despite viral suppression' above.)
●Infant antiretroviral use remains important as postexposure prophylaxis after delivery and in settings in which maternal antiretroviral use is delayed or interrupted during breastfeeding (table 1 and algorithm 1). (See 'Infant antiretroviral use' above.)
