Evaluating the Response to Two Antiretroviral Medication Regimens in HIV-Infected Pregnant Women, Who Begin Antiretroviral Therapy Between 20 and 36 Weeks of Pregnancy, for the Prevention of Mother-to-Child Transmission

A successful outcome was defined as maternal HIV-1 RNA plasma viral load less than the lower limit of quantification (LLQ) for the testing assay, which could vary. Most (99%) women had their viral load measured using an assay with LLQ equal to 40 or 20 copies/mL.

If the viral load at delivery was missing, the last observed viral load within 21 days prior to the delivery date was considered.

A successful viral load decrease was defined as follows: for women having HIV-1 RNA viral load greater than or equal to 10,000 copies/mL at entry, a viral load <200 copies/mL; for women with VL less than 10,000 copies/mL at entry, a Log10 viral load decrease of at least 2.0 from entry.

Change in viral load from entry (or screening, if there was no entry viral load) to each study week prior to delivery, calculated on the log10 scale as log10(week X RNA) - log10(baseline RNA).

For this analysis, HIV-1 RNA values that were censored below the lower limit of quantification (LLQ) were imputed to be equal to the LLQ - 1.

The Week 4 and 6 participant viral loads were the viral load results obtained closest to (within four days of) the target date for that visit from initiation of treatment (for Week 4, day 24-32 after first dose; for Week 6, day 38-46 after first dose).

The Week 4 and 6 participant viral loads were the viral load results obtained closest to (within four days of) the target date for that visit from initiation of treatment (for Week 4, day 24-32 after first dose; for Week 6, day 38-46 after first dose).

Vaginal swabs produce much less testable sample volume than blood plasma draws. Each vaginal swab specimen had to be diluted, and this dilution factor raised the lower limit of quantification (LLQ). The most commonly observed LLQs were 300 and 1200. For consistency, the higher LLQ was considered the threshold for this outcome measure.

The goal of this outcome measure was to address an objective relevant to protease inhibitors, one of which was originally included as a third arm in the Version 2.0 of the study. This outcome measure was included to assess how the infectivity of plasma changed over time among women receiving protease inhibitors, and whether this differed from other classes of antiretroviral drugs. However, the lopinavir/ritonavir arm was later dropped in Version 3.0, and only Version 2.0 women who received efavirenz or raltegravir were included in the study analyses. Therefore, because no women included in the study analyses received lopinavir/ritonavir, this outcome measure was not analyzed.

The unit of analysis was the mother-infant set. All sets where the woman received at least one dose of study treatment and remained on study through delivery were eligible. In the case of twins, the worst outcome (i.e. a stillbirth) was used.

The unit of analysis for this outcome measure was the mother-infant set. A mother-infant set was counted as having a premature delivery if any infant in the mother-infant set was delivered prior to 37 weeks gestation (i.e. in the case of twins, if either of the twins was delivered prior to 37 weeks gestation then this set would count as one premature delivery outcome).

All mother-infant sets that delivered at least one live birth on study were eligible for this outcome.

The unit of analysis for this outcome measure was the mother-infant set. A mother-infant set was counted as having an extremely premature delivery if any infant in the mother-infant set was delivered prior to 34 weeks gestation (i.e. in the case of twins, if either of the twins was delivered prior to 34 weeks gestation then this set would count as one extremely premature delivery outcome).

Only women who enrolled prior to 34 weeks gestation were included in this analysis. Those that enrolled from 34 to less than 37 weeks gestation were excluded because they were already past the gestational age where this outcome could have occurred at entry.

The unit of analysis was the mother-infant pair or set; in the case of multiple gestation, the worst outcome was considered in analysis (e.g. if two twins were delivered to one mother, one at 2,000 grams and one at 3,000 grams, this mother-infant set would count as one instance of low birth weight in analysis because at least one of the infants had the outcome).

The unit of analysis was the mother-infant pair or set; in the case of multiple gestation, the worst outcome was considered in analysis (e.g. if two twins were delivered to one mother, one at 2,000 grams and one at 1,000 grams, this mother-infant set would count as one instance of extremely low birth weight in analysis because at least one of the infants had the outcome).

Infants were considered infected if they had both a positive HIV nucleic acid test and a subsequent confirmatory test on a different sample. Uninfected infants were those that had no positive test results and negative test results obtained at two or more of the following visits: Week 6, Week 16, and/or Week 24 postpartum.

Genotypic resistance to each class of study drug (reverse transcriptase inhibitors and integrase inhibitors) was assessed separately among HIV infected infants.

Consensus sequencing was performed on a sample from screening. Women were evaluated for integrase and reverse transcriptase resistance mutations separately.

Additionally, consensus sequencing was performed among women who had an inadequate virologic response (defined in the protocol) on a sample taken at that time of inadequate virologic response.

Genotypic resistance among women who stopped antiretroviral therapy was not assessed. Because World Health Organization guidelines have been updated to indicate all people living with HIV should remain on antiretroviral therapy, even postpartum women, no women stopped antiretroviral therapy after delivery. Therefore, this aspect of the outcome measure is no longer relevant and was not assessed.