AZT/3TC before and after labour reduces HIV transmission whether infants fed with formula or breastfed and weaned early

Adding 3TC to a short perinatal course of AZT (and single dose nevirapine (sd-NVP) during labour) and adding a seven day maternal course of AZT/3TC after labour, results in less mother to child transmission (MTCT) through 18 months than previously observed with a short course of AZT and sdNVP according to the most recent iteration of the ANRS 1201/1202 Ditrame Plus study in the Côte d’Ivoire, which was reported last week at the International AIDS Conference in Toronto.

Perhaps more importantly, the updated regimen was similarly effective whether infants were breastfed and weaned early (at around four months) or formula fed. The updated PMTCT-regimen achieved transmission rates as low as 6-7% — substantially lower than observed with a historical reference group of women who took short-course AZT but breastfed without early weaning.

According to an accompanying poster presentation, there was no difference in two year mortality between the two infant feeding interventions.

Ditrame studies in the Côte d’Ivoire

Since the mid-1990’s, French researchers have been tweaking regimens used to prevent MTCT (PMTCT) in Abidjan, Côte d’Ivoire in a project known as ANRS 049/Ditrame and ANRS 1201/1202 or Ditrame Plus. Although the studies are now open label, and have been conducted a couple of years apart from each other, the French researchers have attempted to extract as many lessons as possible from the PMTCT programme by comparing the outcomes in each cohort.

In the current analysis, three successive cohorts of live births from Ditrame/Ditrame Plus were compared, and a subset analysis looked at the two infant feeding options.

The original Ditrame study, which serves as a reference cohort (n=238) in this analysis, was conducted from 1995-1997. During this period, breastfeeding women received short-course AZT from week 36 of gestation to delivery.

In Ditrame Plus A, conducted from 2001-2002, pregnant HIV-infected women were given short course AZT from week 36 onward, then AZT and sdNVP during labour, and the infants (n=375) received sdNVP with seven days of AZT. Women then chose to either breastfeed and then abruptly wean their infant after four months (n= 169) or to formula feed (n=195).

In Ditrame Plus B, which ran between 2002-2003, the regimen was essentially the same as Ditrame Plus A, except that the short course regimen began earlier (from week 32 onward) and included 3TC (n=336). In addition, mothers received AZT/3TC for seven days after childbirth. Women in this study also chose between breastfeeding with early weaning (n=198) and formula feeding (n=126).

It should be stressed that women in Abidjan should in theory have access to clean drinking water (as opposed to many other, especially rural, settings in Africa). Formula and equipment for bottle feeding and sterilisation were freely provided to the study participants until the age of nine months and all women had access to nutritional counsellors. Paediatric HIV infection was defined as a positive HIV RNA PCR at any age, or if aged >18 months, a positive HIV serology. In the later cohorts, HIV screening was by PCR.

The cohorts were generally similar with the exception that viral loads and the percentage of women eligible for antiretroviral therapy (using WHO guidelines) was lower for women in the Ditrame reference cohort (~15.5% vs an average of 22% overall).

Results

Of the 926 infants for whom cumulative HIV results were available by 18 months, 107 infants were infected; 27 of whom were infected post-natally (with one infection in the formula feeding subset).

Rates of transmission in the different cohorts were:

• 22% in Ditrame
• 16% in Ditrame Plus A short term breastfeeding
• 9% in Ditrame Plus A formula feeding
• 7% in Ditrame Plus B short term breastfeeding, and
• 6% in Ditrame Plus B formula feeding.

“Cumulate rates of infection [for Ditrame Plus A and B combined] were lower for the formula fed versus the short term breast fed groups,” said Dr. Valériane Leroy of the Institut de Santé Publique, at the Université Victor Segalen Bordeaux who presented the findings. But she noted that the difference was not so marked in the Ditrame B group.

After adjusting for maternal eligibility for ART and low birthrate, all Ditrame Plus groups were superior to the reference cohort. After adjustments, only the type of regimen, its duration and the maternal viral load at delivery were predictive of infection at 18 months. The addition of 3TC, Dr. Leroy said, resulted in a 62% reduction in transmission.

She also concluded that both infant feeding options were significantly better than long term breastfeeding, although it is unclear how the analysis could divorce the impact of the the feeding intervention from the more aggressive ART regimens.

Patient characteristics

As women were identified and included in the programme earlier and earlier as the study went on, the length of perinatal antiretroviral exposure increased significantly from 20 days in Ditrame to 29 days in Ditrame Plus A up to a median of 50 days in Ditrame Plus B. The median duration of breastfeeding was longer in Ditrame (around 8 months), but similar (around 4 months) in Ditrame Plus A and B among the women who chose to breastfeed. Among the women who chose formula feeding, 15% failed to do this exclusively and breast fed at least once.

Concerns and cautions

But it could be a mistake to attribute all of the reduced transmission to the addition of 3TC to this regimen, since this was not the only variable changed in the Ditrame Plus B regimen. The relative contribution of seven days of post-natal AZT/3TC should not be underestimated (studies have shown that most post-natal transmission occurs within the first couple weeks of childbirth). A protective effect of postnatal AZT/3TC would go a long way towards explaining why there was so little difference between the short term breastfeeding and formula feeding arms in Ditrame Plus B while the difference between feeding options in Ditrame Plus A was much more marked.

And adding seven weeks or more of 3TC is worrisome for other reasons — namely, that there could a significant danger of 3TC resistance associated with the use of 3TC in a dual suboptimal combination for so long. Most studies suggest that resistance to 3TC typically occurs within three to six weeks in people taking AZT/3TC without nevirapine, efavirenz or a protease inhibitor. It is unclear what impact 3TC resistance could have upon subsequent responses to first-line ART in these women.

Finally, although there was little difference in mortality between formula feeding and short-term breastfeeding, it should be remembered that this trial was conducted in an urban setting with better access to clean drinking water and nutritional support than is found in most of Africa (or even in Côte d’Ivoire). Women in other settings have much more trouble opting for formula feeding because of cultural bias, lack of a consistent supply of infant feeding formula and safe drinking water.

Even in Botswana, which is one of the few middle income countries in Africa, data suggest that formula feeding could be associated with a higher mortality rate than breastfeeding.