Question: Should HIV-positive mothers breastfeed in developing countries and in industrialized countries?
Breast milk transmission of HIV to an infant was first described in 1985, from a mother who acquired HIV from a post-natal blood transfusion.14 Since then, risk factors for breast milk transmission of HIV have been shown to include the duration of breastfeeding, the mother’s viral load and other maternal characteristics, characteristics of the breast milk and the type of breastfeeding.15
In an individual patient data meta-analysis of nine clinical trials in sub-Saharan Africa, the cumulative probability of transmission from HIV-infected women to their infants increased with increasing duration of breastfeeding (from 1.6% at age 3 months to 9.3% at age 18 months).16 For children whose timing of infection was known, up to 44% of transmission could be attributed to breastfeeding.
Maternal factors associated with transmission of HIV by breastfeeding include younger maternal age and higher parity, recent HIV infection associated with high viral load and breast abnormalities, such as mastitis or breast abscess. More advanced maternal disease, with low CD4+ cell counts and higher maternal peripheral blood or human milk viral loads, is a risk factor for late post-natal transmission of HIV.15
Oral candidiasis (thrush) has been described as an infant risk factor associated with increased MTCT of HIV by breast milk, but this may be merely an epiphenomenon indicative of greater immune dysfunction.
The main breast milk-associated risk factor for HIV transmission is the breast milk viral load (both cell-free and cell-associated). Another characteristic of breast milk which may be associated with a higher risk of transmission is reduced concentrations of antiviral substances, such as lactoferrin, lysozyme and epidermal growth factor.15
Studies have shown that mixed feeding with breast milk and formula or solids is associated with a higher risk of transmission of HIV than exclusive breastfeeding.16–20 It is thought that foods other than breast milk in early infancy may damage the gastrointestinal tract, causing increased permeability and allowing HIV in breast milk to penetrate.
Complete avoidance of breastfeeding may be a problem for women in some developing countries. Potential interventions to reduce or prevent transmission of HIV through maternal breast milk include decreasing the duration of breast milk exposure (either by complete avoidance of breastfeeding or by stopping breastfeeding early), reducing maternal infectivity (e.g. by using chemical or heat treatment of breast milk to lower breast milk HIV viral load or by giving the breastfeeding mother antiretroviral treatment), reducing factors which increase mother-to-child HIV transmission (e.g. mixed feeding, maternal breast problems, infant thrush).15
A Cochrane systematic review of the efficacy of interventions to decrease late post-natal MTCT of HIV identified six RCTs and one intervention cohort study.15 Two trials examined eliminating or shortening the duration of exposure to breast milk from HIV-positive mothers.21,22 In a trial of breastfeeding versus formula feeding, the cumulative probability of the infant developing HIV infection was 36.7% at 24 months in the breastfeeding arm and 20.5% in the formula arm (p = 0.001), but at this age the rates of mortality and malnutrition were not significantly different in the two groups.21 In a trial of early cessation of breastfeeding, HIV-free survival and mortality at 24 months was similar for children who ceased breastfeeding abruptly at around 4 months of age and children who continued breastfeeding for at least 6 months.22
A trial of maternal vitamin A supplementation surprisingly found more cases of HIV infection in children of mothers who were given vitamin A.23 Maternal multivitamin supplements had no significant effect on HIV transmission.23
One intervention cohort study involving 1276 infants evaluated the effect of infant feeding modality on MTCT.24 Breastfed children who also received solids during the first 6 months of life were significantly more likely than exclusively breastfed children to be infected with HIV (HR 10.9; 95% CI 1.5–78). The cumulative 3-month mortality in exclusively breastfed infants was 6.1% (4.7–7.9) compared with 15.1% (7.6–28.7) in infants given replacement feeds (HR 2.06; 1–4.27).24
Three trials evaluated antiretroviral prophylaxis to breastfeeding infants. In a trial in Botswana, breastfeeding plus zidovudine prophylaxis during the first 6 months of life (7-month HIV infection rate 9%) was less effective than formula feeding (7-month HIV infection rate 5.6%) in preventing post-natal HIV transmission.25 The breastfeeding plus zidovudine infants had a lower mortality rate at 7 months, but mortality and HIV-free survival at 18 months were comparable.25
Two trials studied infant nevirapine prophylaxis for breastfed infants.26,27 In one study, which combined data from trials in Ethiopia, India, and Uganda, a 6-week course of nevirapine resulted in a significantly lower risk of HIV transmission at 6 weeks, but not at 6 months of age.26 In a second study in Malawi of infants who were exclusively breastfed for 6 months, 14 weeks of either nevirapine alone (5.2%) or of nevirapine with zidovudine (6.4%) reduced MTCT at 9 months of age significantly compared to a control regimen of two-dose nevirapine prophylaxis (10.6%). HIV-free survival was significantly better at 9 months for both extended prophylaxis groups, and at 15 months in the extended nevirapine group.27
There are no randomized studies of heat-treating breast milk in developing countries. A small feasibility study of 20 mother–infant pairs in Zimbabwe showed that heat treatment of expressed breast milk is a possible option for feeding HIV-exposed, uninfected children after 6 months of age.28
Rarely, an HIV-positive mother in an industrialized country may want to breastfeed, and this can raise tricky child protection issues.29,30 In the past an HIV-positive woman who wanted to breastfeed triggered an immediate notification to child protection authorities. The risk of breastfeeding in a woman with undetectable virus load is low and needs to be weighed against the harms of trying to enforce no breastfeeding. Most authorities no longer feel an immediate notification is mandatory.
Recommendation: Complete avoidance of breastfeeding reduces mother-to-child HIV transmission, but is associated with significant morbidity (e.g. gastroenteritis) in many developing countries.15 Two alternative strategies that reduce transmission for breastfed infants are, firstly, exclusive breastfeeding during the first few months of life as opposed to mixed feeding and secondly, extended use of antiretrovirals (nevirapine alone, or nevirapine with zidovudine).
In industrialized countries, where formula feeding is effectively as safe as breastfeeding, all HIV-positive mothers are counselled to use formula feeds and to avoid breastfeeding completely.
19.4.1.2 Human T-cell lymphotropic virus type 1
Human T-cell lymphotropic virus type I (HTLV-I) is a retrovirus, most prevalent in Japan, Africa, the Caribbean and South America, which can cause adult T-cell leukaemia, lymphoma and tropical spastic paraparesis.31 Known modes of transmission include vertical transmission (predominantly through breastfeeding), transverse transmission (sexual intercourse), transfusion of infected blood products and sharing of needles and syringes.31 Infected mothers and blood donors can be identified by screening.
The main route of MTCT is through breastfeeding. Although HTLV-1 proviral DNA can be detected in cord blood samples, HTLV-I proviruses in cord blood may be defective, and intrauterine transmission is rare. The placenta can be infected by HTLV-I, but infection does not reach the foetus. HTLV-I proviral DNA and antibodies against HTLV-I can also be detected in saliva, but there is no direct evidence to show that HTLV-I transmission via saliva occurs.32
Avoiding breastfeeding can prevent MTCT of HTLV-1 but, as with HIV, the risks of formula feeding in a resource-poor setting may outweigh the benefits.33
19.4.1.3 Cytomegalovirus
Like all herpesviruses, CMV persists after acquisition and can reactivate with or without causing symptoms. Using the sensitive technique of DNA amplification by polymerase chain reaction (PCR) it can be shown that CMV can be detected intermittently in breast milk from the majority of CMV-seropositive mothers.34–38
While post-natally acquired neonatal CMV infection is usually mild, the same is not true for extremely low birth-weight infants, who can develop severe, sometimes fatal disease.34–39 A sepsis-like syndrome, thrombocytopenia, neutropenia, hepatitis and pneumonitis have all been described (see Chapter 17).
A literature search on breast milk transmission of CMV found three non-systematic reviews34–36 and one non-Cochrane systematic review.38 The non-Cochrane systematic review analysed studies from 1966 to 2008 on transmission of CMV via breast milk to pre-term infants and their outcome.38 The reviewers included 26 studies. The proportion of CMV IgG seropositive mothers ranged from 52% to 100% (median 82%), and CMV IgG was detected in 67–97.2% (median 80%) of breast milk samples. The proportion of infants who became seropositive varied widely from 5.7% to 58.6%. Similarly, a wide range of infants from none to 34.5% (median 3.7%) developed symptomatic CMV disease and 0–13.8% (median 0.7%) developed severe sepsis-like syndrome. Pre-term infants with acquired symptomatic CMV infection had a low risk of mild neurologic and cognitive sequelae, without hearing impairment.38 Finding an association between the presence of CMV in breast milk and the infant becoming infected with CMV does not prove that milk is the vehicle of infection, because CMV is also found in respiratory secretions and many other body fluids (see Section 17.1).
Question: Should pre-term babies be fed breast milk if their mothers are CMV seropositive?
Breast milk transmission of CMV can be prevented by using formula feeds instead of breast milk, but in extremely low birth-weight infants the benefits of breast milk in preventing NEC and bacterial infections far outweigh the benefits of preventing post-natal CMV acquisition. Possible interventions to reduce CMV transmission are to freeze the milk, which has been shown to reduce transmission and reduce the severity of CMV infection but not eliminate it totally, or to pasteurize the breast milk, which will destroy proteins and is likely to reduce many of the benefits of breast milk. The American Academy of Pediatrics suggests that short-term pasteurization may be less harmful to the beneficial constituents of human milk, but sits firmly on the fence by saying that decisions about breastfeeding of infants by CMV-seropositive mothers should “include consideration of the potential benefits of human milk and the risk of CMV transmission.”40
Recommendation: The data do not support avoidance or pasteurization of breast milk for high-risk pre-term infants of CMV-positive mothers.
19.4.1.4 Varicella-zoster virus
Although VZV can be detected in the breast milk of mothers with chickenpox41 and zoster,42 the major risk to the infant comes from transplacentally acquired VZV in the absence of maternal IgG. Post-natally acquired VZV, even in the absence of IgG antibodies, is almost always benign, presumably because the viral load is far lower from post-natal than from transplacental exposure (see Chapter 17).
Babies whose mothers develop chickenpox around delivery are given varicella-zoster immunoglobulin and/or antivirals and the risk of severe neonatal chickenpox in a baby who has received suitable prophylaxis is negligible (see Chapter 17). There is no rational reason to separate a mother with chickenpox or zoster from her baby and no reason to stop her from breastfeeding.
VZV vaccine is a live attenuated vaccine. The vaccine strain has not been detected in breast milk, but it is possible that it could be transmitted in milk, although the risk to the infant is likely to be extremely low.40,43 If a VZV-seronegative woman is exposed to VZV post-natally, we recommend that she should be immunized with VZV vaccine and be counselled that it is probably safe to continue breastfeeding.
19.4.1.5 Herpes simplex virus
Only gold members can continue reading. Log In or Register to continue
