A combination of AZT, lamivudine, and nevirapine (AZT/3TC/NVP) taken from week 28 of pregnancy and up to one month after delivery resulted in significantly reduced HIV levels in breast milk as compared to the HIV levels found in the breast milk of women who were not treated prophylactically.
The DREAM programme (Drug Resource Enhancement Against AIDS and Malnutrition) in Mozambique conducted this pilot study to examine the effect of using triple antiretroviral (ARV) therapy to reduce mother to child HIV transmission while breastfeeding. Participants in the study were 40 women from the DREAM programme (group A) who receive comprehensive HIV prevention and care services and 40 HIV-positive women from an antenatal clinic (group B) where HIV testing is not done and ARV prophylaxis is not offered.
Pregnant women in group A were enrolled into the study and began antiretroviral therapy at week 28 of pregnancy or as soon as possible after the first trimester. (If haemoglobin was < 8 g/dL, stavudine was substituted for AZT in order to limit the risk of AZT-associated anaemia.) In group B, women were tested for HIV at delivery, and if HIV-positive, were asked to participate. All infants received a single dose of NVP within 72 hours of birth. All mothers agreed to refrain from breastfeeding and free formula feeding was provided.
Breast milk was expressed manually five times a day and samples were analysed at delivery and seven days post partum. Plasma HIV-1 RNA measurements had a lower detection limit of 50 copies/mL; DNA measurements expressed as the number of HIV copies/106 had a limit of detection of 10 copies/106 cells.
The median age of the patients was 25 in group A and 26 in group B (overall range 15-39). In each group, 38 women had WHO stage 1 HIV disease and two had stage 2 disease. In group A, the pre-HAART median CD4 cell count was 538 cells/mm3 and women were on therapy for a median of 85 days (range 4-165). At delivery, the women receiving ART had a median CD4 count of 551 cells/mm3 (range 183-1291) versus a median of 347 cells/mm3 (range 28-1091) in group B (p = <0.001). Statistical analysis determined correlations between plasma and breast milk HIV RNA levels and between drug levels and plasma/breast milk HIV RNA.
At delivery, the median plasma viral load in group A was 2.2 log versus 4.8 log in group B. In group A, the median viral load level in breast milk was higher than in plasma (2.3 vs. 2.2 respectively) and in group B, the opposite was true (3.4 vs. 4.8). At day seven the median HIV RNA level in breast milk in both groups was lower than in plasma and the median plasma viral load had increased by 0.1 log.
In group A, the median HIV RNA level in breast milk was 1.9 vs. 2.3 logs in plasma. In group B, the median HIV RNA level in breast milk was 3.6 log vs. 4.9 in plasma. The proportion of women with HIV RNA levels <400 copies/mL in both plasma and breast milk in group A was significantly lower at delivery (almost five times more likely) and at day 7 (12 times) than that found in group B, with an odds ratio of 4.8.
Looking only at women who had detectable viral load at delivery and day 7, the median concentrations of NVP, 3TC, and AZT were 0.6, 1.8, and 1.1 times higher in breast milk than in plasma. Individual drug level concentration of each of the three drugs was similar at both points in time.
Only in the median plasma NVP level at day seven was there an inverse correlation to HIV viral load and the authors suggest this might be attributable to the small sample size. At delivery, there was no significant correlation between NVP concentration and plasma/breast milk viral load and similarly, there was a lack of correlation between concentrations of AZT and 3TC and plasma/milk viral loads.
In 10% of cases, NVP drug concentration was detectable in breast milk, but not in plasma, indicating that NVP might be eliminated from breast milk at a slower rate than in plasma. HIV viral load in breast milk was not significantly different among women who were partially or fully compliant to HAART.
Notably, this study found that in untreated women, the concentration of HIV RNA in breast milk was higher than was the case in other studies. The authors suggest that this could be a result of greater sensitivity in the processing of specimens, without ruling out the possibility that results might also be affected by variation in CD4 cell values in different populations. They did not find a higher than expected NVP plasma level in this group of women as was the case in a previous Malawian study.
Both HAART administration and CD4 cell count were independently associated with levels <50 copies/mL in breast milk at day 7, and in untreated women, there was a higher rate of detectable HIV DNA in breast milk cells (55% versus 32.5% in group A, p=0.07). In contrast to other studies though, the effect of HAART on DNA levels was lower than its effect on HIV RNA levels.
These results suggest that this particular ARV regimen, given during and after pregnancy, is able to significantly reduce HIV RNA viral load in both plasma and breast milk and suggest there may be a role for maternal HAART prophylaxis as a means to reduce breastfeeding-associated transmission.
No mention of hypersensitivity reactions was made although some of the women given NVP had CD4 cell counts above 250 cells/mm3, implying a higher risk of hypersensitivity reaction. The small sample size of this pilot study may have limited results in this area. In regard to future virologic studies of breast milk, a recommendation was made to use whole milk, as it showed an equal sensitivity to skim milk, a greater sensitivity than the lipid layer, and processing is not required. The authors also point out a need for longer-term postpartum ARV pharmacokinetic studies and further data on the infant safety of maternal HAART.
Source: Bio-Bio Technology