DREAM study: Extended prenatal ART protects against mother-to-child transmission of HIV at low and high CD4 levels

This article is more than 15 years old. Click here for more recent articles on this topic

Antiretroviral therapy (ART) administered during pregnancy and breastfeeding reduces the risk of a woman transmitting HIV to her infant to just 2% by the time the child is six months of age – even without caesarean sections or formula feeding, according to a retrospective analysis of participants in the Drug Resource Enhancement against AIDS and Malnutrition (DREAM) programme in sub-Saharan Africa, presented this week at the International AIDS Society conference in Cape Town.

Without ART, even women with CD4 cell counts over 350 cells/mm3 were at high risk of transmitting the virus to their infants: “In our cohort, 37% of the transmissions that occurred were in this group with CD4 cell counts over 350,” said Professor Leonardo Palombi, of the University of Tor Vergata in Rome, who presented on behalf of the DREAM Programme.

But the benefit of ART was observed regardless of whether the mother had a low or high CD4 cell count when she went on treatment – and the longer that she was on treatment, the better. In fact, a shorter course of pre-delivery ART (less than 30 days) was strongly associated with HIV transmission and death at 1 month and between 1 and 6 months of age. Again, “this effect was also apparent in the women with higher CD4 cell counts,” said Prof. Palombi.




multivariate analysis

An extension of multivariable analysis that is used to model two or more outcomes at the same time.

retrospective study

A type of longitudinal study in which information is collected on what has previously happened to people - for example, by reviewing their medical notes or by interviewing them about past events. 

loss to follow up

In a research study, participants who drop out before the end of the study. In routine clinical care, patients who do not attend medical appointments and who cannot be contacted.


Short for logarithm, a scale of measurement often used when describing viral load. A one log change is a ten-fold change, such as from 100 to 10. A two-log change is a one hundred-fold change, such as from 1,000 to 10.

Going on ART during pregnancy was also associated with a dramatic reduction in adverse pregnancy outcomes, including a 70% reduction in the rates of spontaneous abortion and stillbirths, a 72% reduction in prematurity rates and a significant reduction in maternal mortality regardless of baseline CD4 cell counts. Toxicity was infrequent and mainly associated with AZT and d4T, even though there were concerns of nevirapine-related hypersensitivity in women with higher CD4 cell counts.


The DREAM programme supports HIV care and treatment for around 75,000 adults and children in 10 countries within sub-Saharan Africa. Their approach to the prevention of mother-to-child transmission (PMTCT) is much closer to the norm in industrialised countries than the typical African PMTCT regimen (a short course of AZT monotherapy during late pregnancy, followed by single-dose nevirapine [sdNVP] as the woman goes into labour).

“In DREAM, we start as soon as possible and continue for at least six months while women are breastfeeding,” said Prof. Palombi. Women with CD4 cell counts below 350 cells/mm3 are deemed to need ART for their own health, and given nevirapine with either AZT/3TC or d4T/3TC as early as the 14th week of gestation and continuing for life (or treatment switch). Those with CD4 cell counts over 350 cells/mm3 are given the same regimen, starting from the 25th week of pregnancy, and continuing during breastfeeding and stopping after weaning (about six months after birth).

Since nevirapine has a long half-life, the women then continue to take a AZT/3TC ‘tail’ for three weeks to reduce the risk of becoming resistant to nevirapine. All infants are also given sdNVP within 72 hours of birth.

Participants in the programme also receive more intensive monitoring than is routinely available in their countries (with routine viral load, CD4 cell counts and toxicity monitoring).

Impact on vertical transmission

The retrospective analysis presented by Prof. Palombi involved a record review for 3148 live births among 3273 HIV-positive women enrolled in the programme in Malawi and Mozambique from July 2005 to December 2008.

Of the live births, at one month, 93 infants were lost to follow-up and seven had died (for a neonatal mortality rate of 0.23%) but 2994 infants were tested by bDNA (and testing was still pending for 54 at the time of analysis).

Out of the 2994 infants tested at one month, 22 have been confirmed as being HIV-positive (and five with positive results await confirmation), for an HIV transmission rate of 0.7%.

At month 6, 2120 of the children who were HIV negative at month 1 were again tested (another 642 are waiting to be tested, while between month 1 and 6, 143 have been lost to follow-up and 41 have died (for a 1-6 month mortality rate of 1.9%). At this timepoint, 15 more children have been confirmed to be positive, and 5 are waiting to be confirmed as positive.

The cumulative transmission rate at 6 months is between 1.4 and 1.9%, the cumulative mortality rate at 6 months is 2.1% and the loss to follow-up rate at six months is 7.5%. These are rates as low as typically reported in well-resourced countries – but again, these are breastfeeding women who would be more likely to transmit HIV if not on treatment.

Benefit associated with time on ART

Because of factors such as late diagnosis and presentation, not every women started treatment as per protocol. This enabled the researchers to detect an association between outcomes and length of time on ART.

At one month, there was a 0.9% transmission rate for infants whose mothers had received at least one dose of ART before delivery (these women had a viral load of 3.55 log) but a 5.1% transmission rate among infants whose mothers waited until delivery to begin ART (with a viral load of 4.51 log, p = 0.001). The one-month HIV-free survival rate for infants was 97.6%.

When infant outcomes were stratified according to maternal baseline CD4 level, pre-delivery ART use appeared to have an impact regardless of whether mothers had low or high CD4 cell counts. Among 1388 children of women whose baseline CD4 levels were below 350 cells/mm3, there was a 1.3% rate of HIV transmission or death with 30-plus days of maternal ART use (16 of 1231 infants) versus a 3.8% rate when maternal ART was used for 30 days or less (six of 157 infants) (odds ratio [OR] 0.33, confidence interval [CI] 0.12-0.86). Among 1533 children of women whose baseline CD4 levels were 350 cells/mm3 or higher, there was a 0.7% rate of either HIV transmission or death with 30-plus days of maternal ART use and a 2.0% rate with a shorter duration of maternal ART use (OR 0.33, CI 0.10-1.09).

In multivariate analysis, 30 days or less of pre-delivery ART was found to be associated with either infant death or HIV transmission at one month postpartum after adjusting for baseline viral load, CD4 cell count, haemoglobin and body mass index.

During the period of breastfeeding – between 1 to 6 months – the association between the length of time on ART pre-delivery was significantly associated with both HIV transmission on its own and the combined endpoint of transmission and death – again, this was apparent in both CD4 cell strata, but most pronounced in the women with lower CD4 cell counts.

Among the 990 children of women whose baseline CD4 levels were below 350 cells/mm3 who were tested for HIV at month 6, there was a 3.1% rate of HIV transmission or death with 30-plus days of maternal ART use (27 of 876 infants) but an 8.8% rate when maternal ART was used for 30 days or less (10 of 114 infants) (OR 0.33, CI 0.16-0.70).

Among 1105 children of women whose baseline CD4 levels were 350 cells/mm3 or higher, there was a 1.8% rate of either HIV transmission or death with 30-plus days of maternal ART use and a 2.1% rate with a shorter duration of maternal ART use (OR 0.81, CI 0.24-2.83).

In a multivariate analysis, both baseline maternal viral load and receiving 30 days or less of pre-delivery ART were found to be associated with either infant death or HIV transmission at 1-6 months postpartum after adjustments.

The six-month estimated HIV-free survival for the children of women who received less than 30 days of ART was 90.9%, while it was 96% in those who received extended therapy.

Finally, “the infant mortality rate and other parameters of infant health (such as birth weight, weight for age) were also positively affected by the approach,” said Prof. Palombi. “Breastfeeding under ART is a safe, feasible and acceptable choice in resource-limited settings where the health risks associated with no breastfeeding are extremely high.”

ART results in favourable pregnancy outcomes

The DREAM researchers also looked pregnancy outcomes (such as spontaneous abortion, stillbirth and prematurity), maternal mortality, toxicity and resistance to ARVs to determine how the approach impacted on the mother’s health.

Out of the 3273 pregnancies in the cohort, 42 died, for a maternal mortality rate of 1.2%. 3.1% of the pregnancies ended in stillbirth, 2.1% were spontaneous abortions (9 of these coincided with the mother’s death), and 19.1% of the infants were premature.

Of the entire cohort, 68 women received no ART, and 365 received less than 30 days of ART – and again, there were clear differences in outcomes. While only 5.2% of those on ART for more than 90 days had stillbirths, the rate was 26.5% in women who received no ART, and 7.1% in those who received ART for less than 30 days. The rate of stillbirth and abortion in those who received less than 30 days of ART was higher regardless of the CD4 cell count but those with CD4 cell counts below 200 cells/mm3 were most likely to have poor outcomes. Likewise, prematurity was strongly associated with shorter duration of ART – regardless of CD4 cell strata. And although the numbers were rather small, maternal mortality followed the same pattern.

ART-related adverse reactions were rather uncommon. Grade 3/4 anaemia was reported in 8.6% of subjects. Likewise, d4T-related peripheral neuropathy was observed in 6.9% of subjects – mostly after pregnancy. Nevirapine-related toxicity was less common. Grade 3/4 liver toxicity was observed in only 2.2% (one patient's LFT increased more than 5 times the upper limit of normal within six weeks of ART). There were 39 cases of Stevens-Johnson Syndrome (1.2%), while grade 3/4 rash was reported in 2.4% of the subjects.

Finally, no viral resistance has been detected among a substudy of 26 women who discontinued ART after weaning.


Prof Palombi conceded that as a retrospective study, there is a potential for selection bias. Some in the audience pointed out that the better outcomes seen in the women who accessed ART earlier might simply be due to better access to care. However, Prof. Palombi doesn’t think this is likely – the various benefits are too consistently associated with longer ART across CD4 strata.