Background:  There have been no clinical trials in resource-rich settings to address the questions of which HAART regimens are more effective for optimal viral response during pregnancy and when ART for prevention of mother-to-child transmission should be started in pregnancy in ART-naïve pregnant women.

Methods:  Data on 240 women starting HAART in pregnancy and enrolled in the prospective European Collaborative Study between 1997 and 2004 were analyzed. HAART was defined as a regimen of ≥3 ART drugs consisting of a nucleoside reverse transcriptase inhibitor (NRTI) backbone and including a non-nucleoside reverse transcriptase inhibitor (NNRTI), or a protease inhibitor (PI). An interval censored survival model using a Weibull distribution was used to assess whether factors including type of HAART regimen, race, and baseline immunological and virological status were associated with the time taken to suppress viral load below undetectable levels through to time of delivery.

Results:  Most women were black (59%) and were diagnosed with HIV antenatally (64%). Median CD4 cell count at initiation was 328 cells/mm³ (IQR 210 to 480) and median HIV RNA 4.16 log₁₀ copies/mL (IQR 3.62 to 4.58). Most (70%) women started HAART in the second trimester. PI-based HAART, mostly including nelfinavir (80%), was initiated in 156 (65%) women, with the remainder (84, 35%) receiving a nevirapine (NVP) -based regimen. Undetectable viral loads were achieved by 73% of all women by time of delivery. Relative hazards (RH) of achieving this were 1.54 (95% confidence interval, CI, 1.07 to 2.21) for NVP-based HAART vs PI-based regimens, 1.50 (95%CI 1.06 to 2.14) for black vs white women, 3.37 (95%CI 1.49 to 7.63) for baseline HIV RNA viral load <4 log₁₀ copies/mL vs viral load ≥5 log₁₀ copies/mL and 1.13 (0.57 to 2.22) for women initiating in the second trimester compared with those initiating earlier. Baseline CD4 cell counts were not associated with achievement of viral suppression by delivery in adjusted models.

Conclusions:  Women initiating with NVP-based HAART and those of black race experienced shorter times to reaching undetectable HIV RNA levels. However, reassuringly, there was no evidence to suggest that immunosuppressed women or those initiating in the second trimester were at a disadvantage compared with more immuno-competent women or those initiating earlier.