Preventing mother-to-child transmission of HIV


Dr J McIntyre, Perinatal HIV Research Unit, University of the Witwatersrand, PO Box 114, Diepkloof, Johannesburg 1864, South Africa. Email


HIV transmission from mother-to-child remains a major cause of infant morbidity and mortality in resource-poor settings. There is consensus that women who need antiretroviral treatment should receive this during pregnancy and beyond, and that an appropriate antiretroviral prophylactic regimen should be given to those who do not yet need ongoing therapy. Infant feeding remains a major source of infection and new antiretroviral strategies, for mothers or children, are emerging with the potential to control this. Access to HIV testing and antiretroviral treatment or prophylaxis remain very limited in low resource settings and needs to be expanded.


A decade and a half has passed since the first successful trials of antiretroviral therapy to prevent mother-to-child transmission of HIV (MTCT). In this time, research has demonstrated that transmission can be almost eliminated by a combination of appropriate antiretroviral treatment and avoidance of breastfeeding, leading to transmission rates in well-resourced countries of 1–2%.1,2 HIV-related child morbidity and mortality has been reduced dramatically and some perinatally HIV-infected girls are themselves becoming pregnant in these settings.3

By contrast, more than 90% of infections from mother to child occur in low resource settings, where HIV seroprevalence rates are high, mothers are often not diagnosed until late in pregnancy, antiretroviral access is limited and replacement feeding of infants of HIV-infected mothers is uncommon. HIV causes substantial mortality in infected infants in low resource settings, and mortality in HIV-exposed but uninfected infants is also higher than in HIV-unexposed.4,5 HIV remains a major cause of maternal morbidity and mortality in these settings, where universal access to antiretroviral therapy is not available and remains a high priority for intervention.6 The opportunity to use prevention of mother-to-child transmission of HIV (PMTCT) programmes as an entry point for care, not only for a pregnant woman, but for the whole family has not yet been adequately utilised.7

Despite these challenges, many PMTCT programmes have been established in low resource settings and are showing excellent results.8,9 Added to this, there has been a major expansion of access to antiretroviral treatment for people with advanced HIV disease in the past 5 years in low and middle income countries, where the World Health Organization (WHO) estimates that about three million people (2 700 000–3 280 000) were receiving antiretroviral therapy at the end of 2007, nearly a million more than the year before, and continuing to increase.10 This represents a 7.5- to 10-fold increase during the past 4 to 5 years, but still probably only represents a third of those in need.

There has been international consensus on an approach to PMTCT in low resource settings, as set out in the WHO guidelines.11 This includes the provision of antiretroviral therapy for pregnant women who require treatment, provision of an appropriate antiretroviral prophylactic regimen for those who do not yet need ongoing treatment and appropriate adaptation of infant feeding practices, depending on the circumstances of the HIV-positive woman.

Antiretroviral therapy for PMTCT

In well-resourced settings, the use of combination antiretroviral therapy for PMTCT has become the standard of care. Where this is available, it has resulted in transmission rates of below 2%, almost eliminating vertical transmission of HIV.1,12 In these settings, the research focus has hence shifted to the safety of antiretrovirals for both the mother and the fetus, with preterm delivery, low birthweight, haematological and mitochondrial toxicities being the main concerns. Current data is reassuring, and the general consensus is that the benefits of using antiretrovirals in pregnancy outweigh the potential risks to the mother, fetus and the infant, although clinicians still need to be aware of the potential complications.13

For the majority of women in low-resource settings, combination antiretroviral therapy for PMTCT is not available; the focus has been on investigating more feasible and less expensive regimens. According to the World Health Organization (WHO) guidelines, pregnant women who do not meet the eligibility criteria for combination therapy for their own health should receive short-course antiretroviral regimens.11 The single-dose nevirapine regimen administered intrapartum to the mother and within 72 hours of delivery to the baby remains one of the recommended and used regimens for PMTCT. Its ease of use without the need for early antenatal care and additional laboratory monitoring, and absence of significant adverse effects makes it still the most feasible option in many resource-limited settings, although not the most efficacious.

The use of single-dose nevirapine, at best reduces transmission by around 50% to a rate of 10–15% at 6 weeks, depending on early infant feeding options.14 The addition of antenatal zidovudine, ideally from 28 weeks, together with a 1 week postpartum course for the baby has been shown to be more effective than nevirapine alone with reported transmission rates of <5% in formula-fed groups.15 Achieving low transmission rates is possible in low-resource settings with relatively little increase in the complexity of the regimens. The WHO has made the zidovudine plus single dose nevirapine regimen the first line recommendation in settings where the drugs are available and the health system infrastructure can deliver these combination regimens.11

The development of resistant viral strains following the administration of single-dose nevirapine, either alone or in dual combination regimens remains an important concern. It is well documented in several studies in both mothers and babies exposed to single-dose nevirapine.16 The concern regarding nevirapine resistance is about the negative effect on future antiretroviral therapy, specifically virologic failure with non-nucleoside reverse transcriptase inhibitor-based (NNRTI-based) regimens if initiated within six months of single-dose nevirapine. Data from a study conducted in Botswana confirmed findings from other studies and found a much higher rate of virologic failure in both women and children exposed to single-dose nevirapine if initiated on HAART within 6 months of the exposure.17 The recommended 7-day postpartum course of zidovudine and lamivudine does decrease the risk of resistance, and is part of the WHO recommendations.18 Single-dose nevirapine should not be repeated in the same pregnancy as this may further increase the risk of resistance. One reassuring fact about the use of single-dose nevirapine is that previous exposure does not appear to lower the efficacy of this regimen in a subsequent pregnancy.19,20 This is important in low-resource settings where HIV-infected pregnant women may have repeated pregnancies without having access to more complex combination prophylactic regimens.

Given the concerns about single-dose nevirapine, either used alone or in combination with zidovudine, and its limited efficacy in decreasing transmission, especially with extended breastfeeding which is the norm in many resource-limited settings, there have been calls for increased efforts to make HAART available for PMTCT in these settings. The cost of the drugs; potential toxicities, concerns about adherence and addition monitoring required make this approach unfeasible at present for most women in poor countries, and this situation may not change for some time to come. Hence, the short course PMTCT regimens still have a role to play in low-resource settings and the use of a prophylactic ‘tail cover’ regimen to reduce the risk of NNRTI resistance remains important.

Women who need HAART for their own health

International guidelines recommend that combination antiretroviral therapy for HIV-infected pregnant women who need it for their own health be started as soon as possible in pregnancy, as these women are also the group most likely to transmit HIV to their infants.11,12,21 As there is general consensus about this fact, the debate has now shifted to determining the appropriate CD4 threshold to commence combination therapy in pregnancy, balancing the risks and benefits to the mother and the fetus. According to the latest WHO (2006) PMTCT guidelines, all women with CD4 counts <200 should receive combination therapy irrespective of WHO clinical stage, and those with CD4 counts <350 should be treated if WHO clinical stage 3 or 4. In the absence of CD4 testing, which is still a reality in many resource-limited settings, clinical staging alone is used as criteria to start therapy.11

Wherever possible, CD4 counts should be available although for all HIV-infected pregnant women so that they can access the appropriate treatment option. Evidence is mounting that there are better maternal and fetal outcomes if a CD4 threshold of 350/mm3 is used for starting combination therapy in pregnancy. The majority of postnatal complications and maternal deaths and most of the perinatal transmissions occur in the group of HIV-infected women with CD4 counts <350/mm3. The WHO reports that country level data show that only around 12% of HIV-infected pregnant women were assessed for their eligibility to receive antiretroviral therapy in 2007, either clinically through a symptom assessment or immunologically by CD4 count. There is an urgent need to make CD4 testing more available for pregnant women in these settings and to accelerate their access to appropriate care.10

The challenge in resource-limited countries, where there is a limited variety of antiretrovirals available, is balancing the benefits of starting antiretroviral therapy with higher CD4 counts and the risks of toxicity. The NNRTI drugs nevirapine and efavirenz form the backbone of most first-line regimens in resource-limited countries. Symptomatic hepatic toxicity, which may be fatal, is reported to be more common with the use of nevirapine in women with CD4 counts >250/mm3, and there is an advisory notice cautioning about its use.22 Efavirenz should be avoided in early pregnancy as animal studies suggest a possible risk of congenital abnormalities and it is rated as a category D drug.23 Triple nucleoside regimens may be less effective in reducing viral load, leaving only protease inhibitor containing regimens as the most common choice for pregnant women with CD4 counts over 250/mm3.12 Avoidance of NNRTI-based regimens in pregnant women with higher CD4 counts has cost implications as protease inhibitor-based regimens are considerably more expensive, and are often reserved for second line therapy in low resource settings. These are challenges, but do not mean that less well-resourced countries should not strive to make HAART available to pregnant women with higher CD4 counts.

Infant feeding options

Infant feeding is one of the most difficult issues in HIV management in low-resource settings. Breast milk transmission contributes significantly to MTCT rates with about 150 000 new infant infections every year occurring through this route, mainly in low-resource settings.24 The WHO recommends that where replacement feeding is ‘acceptable, feasible, affordable, sustainable and safe’, (‘AFASS’ criteria), HIV-infected women should avoid breastfeeding.11 In resource-rich countries, where replacement feeding is a feasible option, complete avoidance of breastfeeding by HIV-infected women has eliminated postnatal transmission. Consequently, these countries have been able to achieve very low MTCT rates through replacement feeding being an integral part of their PMTCT strategies. However, in low-resource countries, selecting an appropriate infant feeding option remains a challenge, as extended breastfeeding remains the norm and replacement feeding is not feasible in many settings. This has resulted in postnatal transmission rates through breastfeeding remaining unacceptably high and accounting for up to half of perinatal transmissions. Several risk factors for transmission of HIV through breastfeeding exist and these include increased duration of breastfeeding; mixed feeding; breast abnormalities such as abscesses, mastitis and nipple lesions; and advanced maternal HIV disease.25,26 As the duration of breastfeeding increases, so does the cumulative risk of HIV transmission with a six-fold increase in risk if breastfeeding extends to 18 months of age.25 Very few mothers are able to sustain exclusive breastfeeding for 6 months and mixed feeding, which has been shown to more than double the risk of transmission, is a common practice.26 Sustaining exclusive breastfeeding requires a lot of support for the mother such as intensive counselling and home visits that may not be available in many settings. Advanced maternal HIV disease, which is associated with a low CD4 cell count and a high viral load, is another risk factor for postnatal transmission in women who breastfeed.25 Breast lesions and infections are associated with a higher breast milk viral load and, consequently, a higher transmission risk for the baby.27

Despite the knowledge about breast milk transmission, the reality is that breastfeeding remains the only option for many women in low-resource settings as the criteria for replacement feeding are hard to meet, and the use of formula carries a significant risk of morbidity and mortality to the baby in these settings, mainly through diarrhoeal and respiratory infections.28,29 Results from the Mashi study conducted in Botswana showed that formula feeding was more effective than breastfeeding with 6 months of zidovudine prophylaxis in preventing postnatal HIV transmission, but was associated with higher mortality at seven months, mainly due to diarrhoea and pneumonia.28 Even in settings where replacement feeding is a feasible option women may still elect to breastfeed due the HIV-associated stigma, which unfortunately remains prevalent.30 Healthcare providers may also be reluctant to promote replacement feeding because of their own views, lack of knowledge or stigma.31

Given that breastfeeding remains the only option for many women in resource-poor settings, it has become a priority for research to find ways of reducing breast milk transmission while maintaining the benefits of breastfeeding. Earlier recommendations were that if breastfeeding was the infant feeding option of choice, it should be exclusive for 6 months followed by rapid weaning, thereby potentially decreasing the risk of HIV transmission by shortening the period of breastfeeding and avoiding mixed feeding once complementary foods are introduced.26,32 However, studies have subsequently shown that early cessation of breastfeeding in resource-poor settings was associated with significant infant morbidity and mortality, and there was no significant difference in long-term survival between those who were weaned early, and those who had extended breastfeeding.33,34 The WHO now recommends that infants should be weaned at 6 months only if the AFASS criteria for replacement feeding are met at that time, otherwise breastfeeding should continue with additional complementary foods.11

Given that early weaning may be detrimental to the infant, one of the priority areas for research looking at approaches to reduce breast milk transmission has been the use of extended antiretroviral prophylaxis in HIV-exposed infants and the use of combination antiretroviral therapy in breastfeeding mothers. In the PEPI study in Malawi, an extended 14-week course of nevirapine, with or without added zidovudine, commenced at birth in HIV-exposed infants decreased the risk of HIV transmission by as much as 50% at 9 months compared with the standard regimen of single dose nevirapine and 1 week of zidovudine.35 The benefit was however lost beyond 12 months where breastfeeding was prolonged. In the SWEN study—trials conducted in Ethiopia, India and Uganda—there was a significant difference in HIV transmission rates at 6 weeks between HIV-exposed infants who received a single postpartum dose of nevirapine and those who received an extended regimen of daily nevirapine for 6 weeks.36 However, at 6 months, the difference was not significant. Other studies on infant prophylaxis, including the MITRA and SIMBA studies, also showed efficacy of extended antiretroviral regimens in decreasing the risk of breast milk transmission, but the duration of breastfeeding was short and the rate of exclusive breastfeeding high in both studies, which in reality is not the norm.37 Direct comparison of the studies is not easy because of the fact that maternal treatment, duration of breastfeeding, extent of exclusive breastfeeding, duration of and type of antiretroviral drugs used for infant prophylaxis were different.38

Despite the fact that these studies show that infant prophylaxis can decrease breast milk transmission, it is difficult to translate these research findings into policy for resource-limited countries. No consensus has been reached yet about the duration of prophylaxis and the antiretroviral drugs to use. There are also several concerns about infant prophylaxis in these settings: availability of the drugs; adherence to treatment; risk of development of resistant HIV strains; implications for future treatment if the infant becomes infected with HIV despite the prophylaxis; and the risk of toxicities associated with these drugs.

Observational studies investigating the use of combination antiretroviral therapy in breastfeeding women who do not yet require treatment for their own health suggest that this approach could decrease breast milk transmission.39–41 Similar concerns raised with infant prophylaxis have also been raised with maternal prophylaxis. There is also the additional concern about the impact of prolonged use of combination antiretroviral therapy for prophylaxis and subsequent treatment interruption after cessation of breastfeeding. Treatment interruption in individuals with high CD4 cell counts appears to be associated with an increased risk of HIV disease progression and death.42 Recommendations about the prophylactic use of combination antiretroviral therapy during breastfeeding are yet to be made.

Infant feeding counselling in the context of HIV, in low-resource settings, remains complex.43 There is a need to minimise HIV transmission while maximising infant survival and the availability of resources to achieve this remains a challenge.

Turning knowledge into results

Despite the advances in knowledge of how to prevent mother-to-child transmission of HIV, access to services remains unacceptably low in many low-resource settings. The WHO reports that the global coverage of PMTCT services has improved in recent years but remains suboptimal.10 Only about 18% of all pregnant women in low- and middle-income countries (20.6 million of 115 million pregnant women) received an HIV test in 2007, although this has increased from estimates of 16% in 2006 and 10% in 2004 and 2005. This is despite 87 of 109 low- and middle-income countries reporting the implementation of provider-initiated testing and counselling in all or in some antenatal service sites. One of the major barriers to access of PMTCT services remains initial access to HIV testing for pregnant women. Although stigma and inadequate counselling do play a party in this, failure to be tested is most usually because the test is simply not offered to pregnant women. Several programmes have demonstrated that provider initiated testing can dramatically increase the uptake of antenatal HIV testing. In Botswana, antenatal HIV testing increased from 75% to 95% within 6 months of the introduction of routine provider initiated testing, and similar results have been reported from Malawi and Zimbabwe.44–46

Coverage of antiretroviral prophylaxis for HIV-infected pregnant women has also increased—from around 10% in 2004 to around 33% in 2007,10 but two-thirds of women in need are still receiving nothing to reduce the risk of transmission of HIV to their children. Although the more efficacious regimen of zidovudine with peripartum single dose nevirapine has been recommended by the WHO since 2006, only around a quarter of HIV infected women in sub-Saharan Africa are reported to have received this in 2007, with 50% receiving single dose nevirapine only, and <10% receiving a triple combination antiretroviral regimen.10

There are some consistent factors for success and expansion of PMTCT programmes. These include the four prongs of the WHO’s PMTCT approach—the prevention of primary infections in mothers, the provision of appropriate family planning, the prevention of transmission during pregnancy and the provision of care and treatment.47 Integrating PMTCT services into existing services and the strengthening of these services by innovative human resource utilisation and the use of lay workers adds to their sustainability and success.48The further involvement of men and communities in general in spreading awareness of and support for PMTCT services is equally important.

The prevention of transmission of HIV from other to child remains a global obstetric priority. Many children’s lives can be saved by the implementation of a combination of effective antenatal and postnatal PMTCT strategies, and obstetric services need to accept this challenge and act more effectively to do this.

Disclosure of interests

J.A.Mc.I has received research funding and speaker’s honoraria from Glaxo Smith Kline, Boehringer Ingelheim and Abbott Pharmaceuticals.

Contribution to authorship

C.N.M. and J.A.Mc.I wrote and reviewed the content of this manuscript.

Details of ethics approval

Not applicable.


Not applicable.