Compiled by Dr.R.R.Gangakhedkar, Assistant Director, National AIDS Research Institute, Pune
for Pediatric HIV/AIDS, First National Conference (Held on 3^rd & 4^th November, 2001 at Delhi)

Introduction

Spectacular advances have been made in the field of mother-to-child transmission (MTCT) of HIV since its initial description. UNAIDS estimates indicate that more than 33 million people are currently infected with HIV of which about 1.6 million are children, mostly from developing countries. More than 1.2 million children were estimated to be infected by HIV by 1998.The male-to-female ratio is fast approaching 1:1 level. Majority of the women are in the child bearing age group. HIV prevalence rates in pregnant women range from 0 ? 0.3% in North America, 1 ? 5% in South America, and 35 ? 45% in sub?Saharan Africa. Each day about 1600 infants are born with HIV infection worldover.

HIV infection is spreading rapidly in India. According to a report based on the sentinel surveillance data (1998), the HIV reactivity rates among pregnant women in five Indian states -Maharashtra, Karnataka, Andhra Pradesh, Manipur and Tamil Nadu have reached more than 1 %, a sign of well-established epidemic in the low risk population. Certain cities such as Mumbai, Pune and Pondicherry are reporting HIV reactivity rate among pregnant mothers up to 4%. The efficiency of male-to-female transmission is 2.3 folds higher than that of female-to-male transmission for various well-known anatomical, biological and sociological reasons. Thus, women and children are poised to occupy the centre-stage in the HIV epidemic in India in near future.

The median survival in pediatric AIDS patients is almost half of that seen among their adult counterparts or children who have acquired HIV postnatally. About 26 million deliveries take place in India, every year. Even if the HIV prevalence rate among antenatal women is presumed to be around 0.8%, pediatric AIDS is going to place a major burden on the health care delivery system in India in near future. Moreover, effective prophylactic regimes to reduce mother-to-child transmission (MTCT) of HIV-1 are now becoming available. Offer of some of the cost-effective regimes through public sector and prioritising research to identify newer ultra-short duration antiretroviral regimen/s to reduce MTCT will prevent it from becoming a public health problem in India. In order to understand various interventions, this article will attempt to review the existing knowledge about risk determinants in MTCT of HIV-1 before dwelling on various approaches related to reduction in MTCT.

Transmission Efficiency

In the initial phase of the epidemic, the reported efficiency of MTCT from African countries was high as women with high risk or clinical suspicion of HIV disease were enrolled in the studies resulting in a selection bias. The efficiency of MTCT of HIV-1 prior to PACTG 076 trial of zidovudine (ZDV) prophylaxis in HIV-infected pregnant mothers ranged from 7% to 13% in Europe about 25% in the United States, upto 40% in Africa. The efficiency of MTCT is reported to be 48% in India. There are major geographical differences in transmission rates between developed and developing countries. These differences may be mainly due to: i) differences in infant feeding practices (breast feeding and its duration); ii) differing study methodology especially with respect to definition of HIV infection in child; the duration of follow up; non-accounting of infant deaths as AIDS deaths using verbal autopsy; differences in sample sizes etc.; iii) differences in presence of other co-factors such as ; STDs, chorio-amnionitis etc.;iv) duration and severity of HIV disease in mother; and v) genotypic and phenotypic differences in HIV variants.

The efficiency of MTCT of HIV-2 was reported to be 1.2% when compared to 24.7 % of HIV?1 that is almost 20 times lesser than that of HIV-1. Though HIV-2 infection is present in India, HIV-1 is the predominant infection.

Timing Of MTCT

HIV transmission can occur during intrauterine, intrapartum, and post-partum period. In order to assess the contribution of each of these time points to the overall transmission, a uniform definition to ascertain the timing of infection is essential. A positive HIV culture or PCR within 48 hours of birth connotes the timing of the infection in intrauterine period. HIV infection is considered to have been transmitted in the intrapartum period, when the PCR or virus culture done within first 48 hours after delivery is found to be negative , but the subsequent testing done within 7 to 90 days after birth reveals that the infant is HIV-infected. When an infant is found to be negative by PCR test between 2 to 6 months of age and later the infant is detected to have HIV infection, it is termed as late postnatal transmission most often occurring due to breast-feeding.

The estimates of timing of vertical transmission differ between breast-feeding and non-breast-feeding population. Various studies have shown that the estimates for intrauterine, intrapartum, and postpartum period as ranging between 23 to 30 %, 45 to 50%, and 30 to 35 % among breast-feeding population. Whereas for non-breast-fed population, these estimates for intrauterine period range between 22 to 38% and during intrapartum period from 62 to 78%.

Risk Determinants

The determinants of the risk of MTCT of HIV-1 can be broadly divided into four categories-viral, immune, maternal and foetal/placental risk factors. Factors such as viral load, virus phenotype - syncytium inducing and its tropism, and genotype can be categorised in the viral factors. CD4 counts, immune status of the mother such as neutralising antibody, antibody-dependent cellular cytotoxicity, gp120 V3 antibody cytotoxic T lymphocytes responses and CD8 suppression are some of the important immune factors. Maternal factors such as clinically advanced disease, primary HIV infection, type of delivery, premature rupture of membrane, presence of co-factors like chorioaminionitis, sexually transmitted diseases (STDs), cracked nipple, breast abscess, vitamin A deficiency, and genetic factors such as chemokine receptor expression are important determinants of MTCT. Breast-feeding, prematurity, epithelial integrity in the oral cavity and first born twins can be grouped under foetal factors. Some of these factors are being discussed as general maternal factors and some while discussing the timing of the transmission of HIV in this review.

HIV Subtypes

The wide geographical differences noted earlier may be mainly related to the genotype of HIV and breast-feeding practices. Most of the industrialised countries reporting lower rates of MTCT than that in the developing country have predominant HIV-1 subtype B. The reports are generally from formula-fed cohorts. A recent study from Thailand among non-breastfed cohort , found that the efficiency of MTCT was 24.2% for HIV -1 subtype E. Majority of the developing countries especially those in Africa have HIV-1 subtype C as the predominant HIV subtype and breast-feeding is followed as a universal practice. In a study conducted in Zimbabwe, the efficiency of MTCT was reported to be 17%. Thus it appears that the differences in transmission rates may not be related to subtypes but may be a reflection of prevalent infant feeding practices. Some of the recent papers also indicate that chemokine receptor profile such as CCR 5 or SDF 1 may not be facilitating transmission of HIV infection.

CD4 Counts

The CD4 count and plasma viral load estimations are a proxy of the stage of HIV disease. Several studies have found out a strong correlation of low CD4 counts along with low CD4 percentage. The European Collaborative Study reported that low maternal CD4 counts and presence of p24 antigenaemia increased the risk of transmission by two fold. A positive acid dissociated p24 antigenaemia test at labour, a cheaper, feasible test in resource-poor setting was associated with increased risk of transmission. However, certain studies could not find low CD4 levels as a significant risk factor.

Plasma Viral Load

High maternal plasma viral load (PVL) has been reported to be an important risk determinant by some studies. These studies found that PVL is significantly higher among mothers who are transmitters than that in non-transmitters. However, this was contradicted by a study conducted among non?breast-feeding mothers. High maternal PVL alone could not explain the MTCT for HIV-1. A Thai study found that two-thirds of the infections were attributable to PVL of > 10,000 copies/ ml. Another study found that for every ten fold increase in maternal PVL, the risk of transmission increased by 1.9 times. Similar incremental increase in the risk was reported by another study. The WITS study found out that the risk of transmission increased by 1.6 fold when maternal PVL were between 10,000 to 50,000 copies per ml and by 3.7 fold, when the levels exceeded 50,000 copies per ml. However, it was found that these levels do not predict the timing of MTCT. These data indicate that maternal PVL appears to be an important determinant of MTCT for HIV-1. However studies have not yet been able to conclusively identify a threshold level of viral load beyond which the transmission does take place for sure. Even cases of HIV transmission have been reported where the viral load was very low, though at a lower rate.

Intrauterine Transmission

Intrauterine transmission of HIV-1 is a scientifically well documented. Few studies have shown that it occurs early in the pregnancy, even as early as 8 weeks of gestation. Generally placenta acts as good barrier. However, HIV can infect the placental cells which express CD4 receptor on their cell walls. Contrary to the belief, placental damage such as placental tears, does not seem to be necessary for the transmission. However, the risk of transmission increases when there are placental tears. Transannexial or transplacental passage through a process of transcytosis and up-regulation of chemokine receptors on the syncytiotrophoblasts, Hofbauer cells and placental macrophages play a major role in intrauterine transmission. Intrauterine transmission is more likely to occur in the last trimester of pregnancy when materno-fetal leaks are common. Placental damage due to chorioaminionitis, cigarette smoking and substance abuse has been associated with increased risk of transmission. It is generally felt that not all the viral variants in maternal blood can be transmitted through placenta. Strains having macrophage tropism and those which are syncytium inducing were more likely to be selected during transmission.

Intrapartum Transmission

Majority of the transmission (50%) occurs during intrapartum period. Direct contact during the passage with the maternal blood and her genital secretions, which contain HIV as well as ascending infection from the vagina and cervix may cause this transmission. The newborn may engulf these contaminated secretions and it may result in the transmission of HIV. This risk increases in the presence of oral sores, hypochlorhydria in the infant; and cracked nipples and breast abscess in the mother. The first-born twin has more than two fold higher risk of HIV infection when compared to the second-born twin.

Premature Rupture of Membrane

Premature rupture of membrane of more than 4 hours and chorioamnionitis has been reported to enhance the risk of MTCT as it increases the exposure of the foetus to HIV. Presence of chorioaminionitis alone was found to increase the risk by 4.4 fold, but histologically confirmed diagnosis of chorioaminionitis, a strong predictor of transmission was difficult to be picked up clinically. Presence of sexually transmitted disease in such women has been associated with enhanced transmission. Premature delivery increased the exposure to HIV and increased the risk of acquiring intrapartum infection by 3.7 fold.

Mode of Delivery

The mode of delivery became a subject of interest when a few studies found that caesarean section reduced the risk of MTCT. A few subsequent studies distinguished it further and found that elective caesarean section was associated with this protective effect. However, this was contradicted by some studies probably due to the small sample size of these studies and inability to separate between elective and emergency caesarean section, retrospectively. The French Perinatal Cohort study came out with startling results where the elective caesarean section along with ZDV prophylaxis reduced the MTCT to 0.8% from that of 11.4% for emergency caesarean section. Additionally, a recently published meta-analysis of 15 prospective cohort studies has shown that the risk of MTCT was found to be 10.4% for women who undergo elective caesarean section to that of 19.0 %, if delivered by any other route, even in the absence of ZDV prophylaxis. Whereas the transmission efficiency was 2.0% for those who underwent elective caesarean section against 7.3% in women who were delivered by any other route, when they received ZDV prophylaxis as per PACTG 076 protocol. Clearly, elective caesarean section appears to have a protective effect on MTCT of HIV-1. The reasons for such a difference in MTCT between elective and emergency caesarean section are debatable. Elective caesarean section is more likely to reduce exposure to cervicovaginal secretions. Moreover, the micro-transfusions that occur once labour begins, are 5 fold lower than in emergency caesarean section. Though caesarean section among women without HIV infection have been shown to associated with higher maternal and neonatal morbidity than that observed in non-surgical deliveries, the elective caesarean section has been shown to have lower maternal morbidity than after emergency caesarean section.

Vitamin A

Vitamin A deficiency in HIV-infected pregnant women has been associated with higher risk of transmission which has been postulated to affect the T-cell and B-cell functions, decreased transfer of maternal antibodies to the foetus, decreased integrity of placenta, increasing susceptibility of trauma to the birth canal, and increasing viral load. Studies have shown that the lower vitamin A concentration was associated with a significant increase in the risk of MTCT of HIV-1.

Immune Response

The plasma viral load is dependent on the host immune responses such as CD8 cell-mediated immune response. A study revealed that the CD8 cells of 81% non-transmitting and 41% transmitting women were able to suppress HIV replication in vitro, in the third trimester. Women who could suppress < 50% of HIV replication were 3.4 fold more likely to transmit HIV to their infants independent of ZDV use. Lack of neutralising antibodies against mother's viruses may increase the risk of transmission.

Transmission Through Breast Milk

Breast milk is considered as an ideal food for infants among mammals due to its age appropriate nutritive values and the presence of certain protective immunological factors. However, it is known to transmit viruses such as cytomegalovirus, and retroviruses like HTLV-1 & HIV in humans and certain viruses in animals. HIV has been isolated from the breast milk. It is known to exist in both, cell-free and cell-associated forms. Studies in animal models have shown that certain retroviruses can cause infection through cell-free viruses. About 58% samples of breast milk revealed presence of HIV-1 DNA with the proportion of HIV infected cells ranging between 1:10,000 to 1:3 cells, more so in women having severe vitamin A deficiency (20 folds higher).

Colostrum contains higher concentration of cells, IgG, vitamin A, lactoferrin and antimicrobial neutralising activity than mature milk., It also contains secretory leucocytic protease inhibitor, the role of which is still not clearly elucidated. A study reported a lack of association between the history of intake of colostrum and transmission of HIV, probably due to its lower consumption by a neonate compared to mature milk. Paradoxically, samples from mature milk were more likely to show presence of cell-free HIV than colostrum.

The presence of neonatal hypochlorohydria, oral sores and maternal conditions like cracked nipple and breast abscess facilitate HIV transmission through breast milk. Additionally, specialised epithelial immune cells like M cells may also be involved in presenting HIV to the macrophages despite intact intestinal barrier. A novel in vitro model demonstrated that HIV can penetrate an intact epithelial barrier through a process named transcytosis. The HIV-infected cells probably stimulate the enterocytes to engulf HIV particles. This process can be blocked by the secretary IgA and IgM antibodies against envelope proteins of HIV through induction of mucosal immunity. The late postnatal transmission increases from a recently seroconverted mother who breastfeeds her infant. Presence of HIV?1?infected cells and a lack of IgM and IgA antibody in breast milk at 15 days after birth were strongly associated with transmission in breastfed children suggesting a protective role of mucosal immunity against late postnatal transmission. Glycosaminoglycan, a factor that inhibits the binding of CD4 to HIV envelope glycoproteins, was found in the breast milk of HIV-infected as well as uninfected women. The presence of lipases in breast milk are alleged to inactivate the virus on standing.

A meta-analysis estimating the transmission efficiency of HIV through breast milk based on reported it to be ranging between 14% to 29%. It is almost 2 times higher in incident HIV infection than that in prevalent HIV infection. A meta-analysis of four studies each from industrialised and developing countries showed that the overall estimated risk of late postnatal transmission of HIV-1 through breast milk was 3.2 per 100 child years of breast-feeding follow up and found that no transmission will occur if breast-feeding is stopped at four months. Certain other studies have found this transmission rate to be ranging between 4 to 12% per breastfed year. Recently a few prospective studies have been published on MTCT through breast milk. A multi-centric study among four sites each in Africa and Europe was found that the overall incidence of HIV in infants was 3.2 per 100 person-years of breast-feeding. An observational study from Malawi has reported that the HIV incidence amongst infants who are breastfed between 1-5 months is higher (0.7%/month), when compared to that between 6-11 months (0.6% /month) and between 12-17 months (0.3%/month). The higher transmission rate between 1-5 months may be due to immaturity of the immune system and high cellular content in early milk. Longer duration of breast-feeding is reported to enhance the risk of transmission. However, early weaning has been associated with adverse impact on growth, morbidity, and mortality. Therefore, early weaning may not be an appropriate strategy. But, the issue of quality of protective immune factors in breast milk of an HIV-infected woman is still unsettled. Interestingly, a study found no additional protection against common childhood illnesses, failure to thrive and rate of disease progression among exclusively breastfed children of HIV-infected women. A study based on statistical modelling to assess the impact on survival of children suggested that alternatives to breast-feeding and wet nursing should not be selected in countries until antenatal HIV prevalence reaches beyond 40%. Wet-nursing from an HIV-infected woman and ingestion of pooled breast milk has been shown to be associated with MTCT of HIV-1. About 51% of the infants are exclusively breastfed up to 3 months. However, supplementary feeding is a costly option. According to the Breast-feeding Promotion Network of India, the average cost of artificial feeding (powdered milk) for a child above 6 months age is about Rs.1100/ month. This monthly cost increases from Rs.782 in the first month to Rs.1284 in the sixth month. If all the women having infants were to artificially feed their children, the total national cost for animal milk or tinned milk will be Rs. 59160 millions and Rs. 118320 millions, respectively in India, which is equal to its central plan outlay for the department of industry during 1998-99 in India. Thus the old dilemma about breast-feeding by HIV infected women continues to exist.

Intervention Trials

Though exact mechanisms involved in MTCT of HIV-1 are still not clear, the viral load in maternal blood and genital secretions, maternal HIV disease stage, infant feeding practices and maternal immune response appear to be critical in transmission. Four major approaches to reduce MTCT emerge as important. The first approach aims at reduction of the maternal viral load by using antiretroviral drugs. The second focuses towards reducing the exposure of the foetus during intrapartum period to HIV through various approaches. The third attempts to develop safer infant feeding practices and the last approach works over promotion of protective immune responses and also primary prevention of HIV infection through behavioural strategies.

Antiretroviral drugs are known to reduce the plasma viral load that is important in MTCT. Hence a trial to reduce MTCT of HIV was undertaken in the United States. The PACTG 076 clinical trial, based on administration of zidovudine to HIV infected pregnant mothers was a ground-breaking study as the first successful approach to prevent HIV transmission from mother- to-child. zidovudine - 100 mg five times a day, was administered orally between 14-34 weeks of gestation to HIV- infected pregnant women till labour ensues. During labour, these women received intravenous zidovudine in the dosages of 2 mg per kg over one hour that was followed by its continuous hourly infusion of 1 mg/kg. Subsequently, the newborn received zidovudine orally for 6 weeks ( 2 mg/kg 6 hourly), postnatally. The MTCT of HIV-1 was reduced by two thirds from 27.7% in placebo group to 7.9 % among zidovudine group. This trial was prematurely stopped due to the significant reduction in transmission efficiency considering its immediate public health and ethical implications and the protocol was approved for use in the US as public health policy. These women were zidovudine-naive and had CD4 counts of more than 200/cumm. There were no congenital abnormalities reported due to zidovudine administration in the study. Though ZDV in animal studies has been shown to have tumorogenicity, no such effect is observed among humans.

However, the possibility of poor adherence to such protocols led to a scare of increasing the zidovudine-resistant strains in the mother. Though transmission and development ZDV resistance has been reported, the proportions developing it, is small. PACTG 076 protocol became a mainstay of management of pregnant women in the developed countries. The proportion of infections due to intrauterine transmission was found to be similar among women who received zidovudine prophylaxis and those who did not. This may be an indicator that zidovudine prophylaxis probably reduces intrauterine or postnatal transmission of HIV infection.

However, due to unaffordability and inaccessibility of zidovudine to most patients, the prenatal administration of zidovudine did not become popular in India. Additionally, despite such a large reduction in transmission of HIV from mother to child, the use of intravenous zidovudine during labour precluded the feasibility of implementation of such an approach to women in rural areas of India. Additionally intravenous administration of zidovudine requires the presence of a qualified practitioner. However according to WHO, 43% of deliveries take place without any attendant.

Therefore, scientists started looking for shorter duration intervention regimes. A retrospective analysis of experience with PACTG 076 protocol in US showed that though a mother may not have received zidovudine during pregnancy, initiation of oral ZDV in syrup formulation within 48 hours after birth has been shown to confer a significant protective effect for MTCT that is almost equivalent to that given in intrapartum period. An observational study conducted by Wade et al in New York State on abbreviated regimen of zidovudine showed that the rate of transmission was the lowest when ZDV was initiated during the antenatal period (6%). Administration of zidovudine during intrapartum period (10%) alone and that within 48 hours of birth (9.3%) to the newborn were equally successful in reducing the transmission efficiency. These findings indicated that even if a mother in labour and a neonate within 48 hours receive zidovudine prophylaxis, a significant proportion of MTCT can be reduced. This indicates that the short-duration regimes should cover the intrapartum and immediate neonatal period.

Simultaneously, encouraging reports from a clinical trial using of shorter duration zidovudine regimen to prevent mother-to-child transmission of HIV from Thailand were published. In this study, zidovudine was administered in the dose of 300 mg twice a day from 36 weeks onwards and 300 mg every 3 hours orally during labour. Zidovudine was not offered to the newborn. The transmission efficiency of HIV infection was reduced by about 50% ( from 18% to 9%). This was a placebo-controlled trial. The women enrolled in the study were not permitted to breast feed their children. Therefore, its applicability for developing countries where mothers cannot afford artificial feeds and those who are unable to maintain good hygiene became suspect.

However, results of four studies to reduce MTCT of HIV-1 where prophylaxis was offered orally to pregnant women who were permitted to breast feed their infants are encouraging. In RETRO-CI, placebo controlled study using zidovudine as in Thai-CDC regime among pregnant women who were permitted to breast-feed, a 37% reduction in MTCT of HIV-1 at 3 months age of the infant was reported. DITRAME study where pregnant women were administered zidovudine 300 mg bid from 36 to 38 weeks, 600 mg orally at the onset of labour and 300 mg bid for one week, found that the efficacy of this regime was 38% at 6 months. Interim analysis of another study having a similar regimen as DITRAME study except that the woman was offered zidovudine postnatally for 8 days, found the relative efficacy to be 31% at 6 months. Since maternal plasma viral load has been strongly associated with risk of MTCT of HIV-1 and a combination of antiretroviral drugs can reduce the viral load more efficiently in a short duration, PETRA, a placebo-controlled trial was designed using zidovudine and 3 TC. It reported a reduction in MTCT by almost 50% between PETRA-A regimen and placebo and 37% reduction between PETRA-B regimen and placebo at 6 weeks follow up of the infant. In PETRA-A regimen, 300 mg zidovudine and 150 mg lamivudine were given twice daily from 38 weeks of gestation until delivery along with additional 200 mg of zidovudine during labour. ZDV along with 3 TC were continued in the mother and also administered to the newborn (ZDV 5 mg/kg and 3TC 2 mg/kg every 12 hours) postnatally for a week. And in PETRA-B, the pregnant women were offered ZDV and 3 TC as in PETRA-A regimen except that they were not offered drugs during antepartum period. Thus it appears that even if the infants are breastfed, the reduction in MTCT of HIV-1 is around 31% to 51%. However, the duration of follow up in the interim analysis of PETRA-A is only upto 6 weeks whereas for other studies it is between 3 to 6 months. Therefore, conservatively one may infer that a third of the transmissions can be reduced using any of these regimes to reduce MTCT of HIV-1 even among breastfed infants, who are weaned off from breast milk before 6 months.

A major breakthrough was reported by the HIVNET 012 trial where an ultra-short regime of nevirapine single dose administration to the pregnant woman in labour 200 mg orally and to the breastfed neonate within 72 hours at 2 mg/kg dose, was reported to have reduced the MTCT of HIV-1 by 47% at 14-16 weeks when compared to zidovudine 600 mg at the onset of labour followed by 300 mg 3 hourly to the mother and oral zidovudine 4 mg/kg twice a day for 7 days after birth to the baby. Nevirapine has a long half-life in pregnant mothers and neonates. Hence its single dose regime appears to be efficacious in reducing MTCT. Of all the available regimes to reduce MTCT, this is the most cost-effective regime.

The PACTG 185 trial using HIV hyperimmune immunoglobulin (HIVIG) during pregnancy to reduce the viral load and thus the MTCT was designed to compare its efficacy with zidovudine prophylaxis (as in PACTG 076) had to be terminated prematurely due to its lower efficacy. Therefore, of the strategies to reduce MTCT by reduction in plasma viral load, the ultra-short regime with nevirapine appears to be the most feasible and fairly effective. However, the control group for HIVNET 012 trial was not administered a regimen of known efficacy in reducing MTCT of HIV-1. Despite this, HIVNET 012 regimen will be extremely useful in countries like India, where women tend to seek antenatal consultations late and some times may directly come in labour.

The reduction in exposure to HIV during intrapartum period can best be achieved through caesarian section. A recently published analysis of French Perinatal Cohort using zidovudine (ACTG 076 regimen) revealed that use of zidovudine and elective caesarian section reduced the risk of HIV transmission to almost negligible level - 0.8% as against 6.6% in children born after vaginal delivery. However, the feasibility of undertaking C-section in all HIV-infected women appears to be low for various well-known reasons. Use of vitamin A to reduce MTCT by reducing the viral load in the genital secretions and other risk determinants has not been successful. Vaginal douching with chlorhexidine has also been found to be not useful. Prevention and control of STDS during pregnancy can also reduce the transmission and hence, at least one evaluation for presence of STDs must be done in pregnancy.

The third strategy is to reduce the transmission occurring through breastfeeding. A significant proportion of MTCT occurs through breastfeeding, which probably increases with the duration though the serial point risk estimate decreases in its later part. Options such as artificial feeding, not offering colostrum and early weaning may be associated with increased morbidity and mortality during infancy. Therefore, a joint statement of UNICEF and UNAIDS suggested that the choice of breast-feeding should be offered to the mother after explaining the risks and benefits of breast-feeding in resource poor settings. This recommendation received a boost when a study from South Africa reported that the risk of MTCT of HIV-1 was the lowest for children exclusively breastfed for 3 months (14.6%) as against those who were partially breastfed (24.1%) and never breastfed (18.8%). This was attributed to the protective immune factors such as secretary leucocyte protease inhibitors, lactoferrin, complement, glycosaminoglycan and growth factors such as; epidermal growth factor and transforming growth factor that are useful in maintaining the epithelial integrity of the gut. Additionally, the higher risk of necrotising enterocolitis among non-breastfed children was also alleged to play an important role. However, research for strategies to inactivate breast milk such pasteurisation or simple heating and standing still needs to be pursued aggressively.

Raising the level of mucosal immunity through use of vaccines is being pursued in research. A successful vaccine to induce high level of protective mucosal immunity will be extremely useful adjunct in preventing MTCT through breast milk.

Apart from all these strategies to reduce MTCT, the most important is that which envisages primary prevention of HIV among women in child bearing age group. Voluntary counselling and testing, intensification of efforts to increase level of awareness about STDs/HIV and bring about sustained behaviour change, research on vaginal microbicide, and issues related to women empowerment are critical from public health point of view.

Feasibility Of Interventions

The recent successes in intervention research in reducing MTCT of HIV-1 world over have resulted in making these interventions as the most successful prevention interventions for HIV infection. The reduction of MTCT is a cost-effective approach over management of pediatric AIDS cases as well. Therefore, there is an urgent need to conduct operational research on how to deliver prophylaxis to reduce MTCT of HIV-1 in India. The feasibility of successful offer of this intervention depends on establishment of voluntary counselling and HIV testing (VCT) facilities for pregnant women, acceptance of this services and, cost-effectiveness.

Not all women who are offered VCT are likely to accept the test. Surprisingly, the acceptance of rates ranged between 3.4 to 51% in inner London and 30 to 41% in New York state, where zidovudine is routinely offered to HIV infected women. However, higher acceptance rates up to 92% have also been reported in certain parts of California. A recent study that collated data on acceptance of VCT from 13 different sites in developing countries showed that the median acceptance rates were as high as 92%. Though some women accept the offer for VCT, the proportion failing to return to collect the test report is found to be significant. Only 45 to 66% women returned back in a few sites in developed countries when compared to the median rate of 82% (range 33-100%) in developing countries. The overall acceptability of the offer was found to range between 33 to 95% with a median of 69%. No such studies are available from India. When the factors associated with the non-acceptance of the VCT offer were studied, it was found that fear about not being able to cope up with the diagnosis, fear of separation or domestic violence, educational status, fear of social discrimination, counselling done by midwife, and patient's knowledge about interventions were important. An offer of rapid HIV testing may be able to circumvent these problems.

However, an offer of test and its accessibility is only a part of the package, counselling in crowded antenatal settings is a major challenge, especially in our set up. One of the hospitals in Mumbai has innovatively replaced pre-test counselling by a pretest group discussion. However, such approaches will need mechanisms to conduct ongoing assessment of the comprehension. Post-test counselling presents a unique challenge. Conducting one-to-one counselling ot HIV-infected women alone may lead to indirect disclosure of the HIV test result by default. Moreover, one-to-one counselling will be crucial to those women who are not infected by HIV as an effort towards primary HIV infection. The issue of maintenance of confidentiality needs to be handled by training the staff and changing the record-keeping practices. Offer of VCT at the antenatal clinic alone can be a double-edged weapon. Since more number of women will be found to HIV-infected compared to men, this might expose them to discrimination and accusations of infidelity. Hence partner notification and testing need to be strengthened in our setting. Though the interventions to reduce MTCT through prophylaxis are cost-effective, the major cost of these interventions lies in voluntary testing and counselling. This needs to be reduced using innovative approaches. Issues like regularity of drug supply and monitoring drug adherence are important while offering antiretrovirals for prophylaxis. Integration of this programme in the existent maternal and child health services is critical for its acceptability and sustainability. Operational research on these issues must be accorded primacy in India.

Conclusions

Spectacular advances have been made in the field of mother-to-child transmission of HIV infection in the recent past. Though a lot has been learnt about the risk determinants facilitating the transmission, understanding its exact mechanism still eludes scientists. Exciting reports of short-duration antiretrovirals like zidovudine and nevirapine have raised hopes about their offer even in the developing countries. The dilemma about breast-feeding continues to hold.

There is a rapid rise in HIV reactivity rate amongst pregnant women in some of the high HIV prevalence states in India. Operational research on universal offer of HIV test and counselling in antenatal setting needs to be undertaken urgently. Research on role of breast milk in MTCT of HIV, its risk determinants and to development of appropriate risk-reduction approaches should be accorded priority. alternative approaches is crucial for India. Offer of chemoprophylaxis to reduce MTCT will greatly assist in minimising the potential adverse impact on the infant and under-five mortality rate, life expectancy and social structure of the country. In addition to these approaches, primary prevention for control of HIV infection needs to be strengthened further more.