Background:  HIV⁺ women receive HAART during pregnancy, either from the second trimester until delivery, or throughout the pregnancy if indicated. This study evaluated longitudinal peripheral blood mononuclear cells (PBMC) and placenta mtDNA levels.

Methods: In this prospective cohort study, blood was collected at 3 time periods in pregnancy (13 to 23 weeks of gestation, >23 to 30 weeks, >30 to 39 weeks), and for HIV pregnancies, at delivery and 6 weeks post partum. Placental tissue was also collected from fetal and maternal sides shortly after delivery. MtDNA content was measured in isolated PBMC, except for the infants for whom whole blood was used. Between groups comparisons were done using the Mann-Whitney test. For placenta fetal vs maternal side analyses, Wilcoxon signed-rank test and Pearson's correlation were used.

Results:  HIV⁺ women, including those on HAART since the second trimester (group A, PBMC n = 27, placenta n = 30), those on HAART since before conception (group B, PBMC n = 9, placenta n = 7), and HIV^– controls (group C, PBMC n = 24, placenta n = 23) were studied. Nucleosides were mainly zidovudine (AZT)/lamivudine (3TC) in group A and varied in group B. At 13 to 23 weeks of gestation, there was no difference in PBMC mtDNA levels between group A (pre-HAART) and group C (p = 0.25). PBMC mtDNA levels in group C increased near delivery. Such increase appeared delayed in the other groups, whereby group B did not increase throughout the antenatal period and a modest increase was seen in group A, which decreased post-partum (off HAART). At birth (0 to 3 days), infants born to group B mothers tended to have lower blood mtDNA than group A infants (p = 0.12). In group C, maternal and fetal side placenta mtDNA levels were very similar and very strongly correlated (R = 0.9, p <0.0001). In contrast, group A placenta mtDNA levels tended to be higher on the maternal side than the fetal side (p = 0.076) and were more weakly correlated (R = 0.48, p = 0.008). Compared with group C, group A placenta mtDNA levels tended to be higher on the maternal side (p = 0.077) but not on the fetal side (p = 0.53).

Conclusions:  A physiological increase in PBMC mtDNA is seen in normal pregnancy, possibly to meet high energy demands. HAART appears to delay this increase and alter placenta mtDNA, particularly on the maternal side of the organ. This likely affects metabolism and energy production within the placenta. There is concern that mtDNA toxicity related to nucleoside therapy in a perinatal setting may have amplified consequences. Further studies are required to identify antiretrovirals with the least toxicity in pregnancy.