NVP has several properties that make it an attractive candidate for antiretroviral therapy to interrupt HIV-1 transmission in the intrapartum and early postpartum period. The pharmacokinetic profile suggests that NVP would be rapidly absorbed by the mother and transferred to the infant in utero when given during labor and delivery. The HIV-1 antiviral activity is rapid with significant reduction in plasma virus occurring within a few days of drug administration. In addition, NVP has been shown to penetrate cell-free virions and inactivate virion-associated reverse transcriptase (RT) in situ. This property would be potentially useful in inactivating cell-free virions in the genital tract as well as in breast milk. These characteristics of NVP suggest that treatment of an HIV-infected pregnant woman in labor with an oral dose of NVP may provide a prophylactic level of NVP in the infant during the time of exposure to virus in the birth canal and/or in the maternal blood. In addition, NVP may inactivate the virion-associated RT present in cell-free virions in the genital tract or breast milk.

Mothers are randomized to receive either a single oral dose of NVP during labor or the corresponding NVP placebo. Randomization occurs at any time after the 28th week of gestation. To assure balance between the treatment groups, the randomization is stratified using 2 factors: (1) antiretroviral therapy during the current pregnancy (no antiretroviral therapy at all, monotherapy [with no multi-agent therapy] for any duration, or multi-agent therapy for any duration), and (2) CD4 cell count at the time of randomization (less than 200 cells, 200 to 399 cells, or 400 cells or greater). Mothers are followed on-study for 4 to 6 weeks postpartum. All mothers are required to incorporate zidovudine (ZDV) into their current treatment regimen and should continue ZDV during delivery and give ZDV to their infants as recommended. ZDV will not be provided as part of the study.

Infants receive a single oral dose of NVP (or the corresponding placebo) administered between 48 and 72 hours of life. The infant's study drug is the same as the mother's randomized treatment assignment. Infants are dosed with study drug according to their randomization group regardless of whether the mother received study drug or not. Infants are followed for 6 months of life and are tested for HIV at birth, 4 to 6 weeks of life, 3 months of life, and 6 months of life.