Background:  Maternal HAART during pregnancy and breastfeeding reduces mother-to-child HIV transmission (MTCT), but may increase the risk for infant anemia.

Methods:  The Mashi and Mma Bana prevention of MTCT trials enrolled HIV-infected pregnant women from 4 sites in Botswana for randomized comparisons of MTCT prevention strategies. The incidence of first severe (grade 3 or 4, DAIDS 2004 toxicity tables) infant anemia from birth to 7 months of age was assessed among infants in these trials. Analyses were restricted to scheduled measurements at birth and at 1, 3 to 4, and 6 to 7 months of age among first-born infants who remained HIV-uninfected.

Results:  There were 1788 (1096 Mashi, 692 Mma Bana) evaluable HIV-uninfected infants. Of these, 743 infants were exposed to maternal 3-drug HAART (either zidovudine [ZDV]/lamivudine [3TC]/lopinavir [LPV]/ritonavir [r]; ZDV/3TC/abacavir [ABC]; or ZDV/3TC/nevirapine ]NVP]), 1 month of post-natal ZDV, and breastfeeding (HAART+BF); 517 infants were exposed to in utero ZDV, 6 months of post-natal infant ZDV, and breastfeeding (ZDV+BF); and 528 infants were exposed to in utero ZDV, 1 month of post-natal ZDV, and formula feeding (ZDV+FF). There were 126 infants with severe anemia by 7 months of age among all study participants, with a cumulative incidence of 12.6 % in HAART+BF, 5.4 % in ZDV+BF, and 2.3 % in ZDV+FF (see Figure 1). Severe anemia was more frequent among HAART+BF infants than either ZDV+BF infants (OR 2.5, 1.6 to 4.0), or ZDV+FF infants (OR 6.1, 3.2 to 11.6). In pooled stepwise logistic regression, severe anemia was significantly associated with HAART+BF (AOR 2.4, 1.5 to 3.8 and AOR 5.7, 3.0 to 10.7) for comparison with ZDV-BF and ZDV-FF, respectively; birth weight <2.5 kg (AOR 2.41, 1.5 to 3.9); and male sex (AOR 1.51, 1.0 to 2.2). Maternal HAART regimen was not associated with differences in infant anemia. Among infants in the HAART+BF group, either microcytosis or hypochromia was present in 39 of 89 (43.8%) infants with severe anemia. Severe anemia was treated with iron/multivitamin supplementation, and 9 infants (7.1%) were transfused. The median time to improvement in anemia to <grade 3 was 33 days (IQR 2 to 62). At the time of the last available follow-up, 7 infants (5.6%) remained severely anemic.

Conclusions:  Exposure to maternal HAART significantly increases severe but reversible infant anemia. The clinical implication of this finding requires further investigation to ensure that the established benefits of using HAART for MTCT prevention are maximized for all infants.