Benefits and Risks of Antiretroviral Therapy for Perinatal HIV Prevention

Abstract

Background: Randomized-trial data on the risks and benefits of antiretroviral therapy (ART) as compared with zidovudine and single-dose nevirapine to prevent transmission of the human immunodeficiency virus (HIV) in HIV-infected pregnant women with high CD4 counts are lacking.

Methods: We randomly assigned HIV-infected women at 14 or more weeks of gestation with CD4 counts of at least 350 cells per cubic millimeter to zidovudine and single-dose nevirapine plus a 1-to-2-week postpartum "tail" of tenofovir and emtricitabine (zidovudine alone); zidovudine, lamivudine, and lopinavir-ritonavir (zidovudine-based ART); or tenofovir, emtricitabine, and lopinavir-ritonavir (tenofovir-based ART). The primary outcomes were HIV transmission at 1 week of age in the infant and maternal and infant safety.

Results: The median CD4 count was 530 cells per cubic millimeter among 3490 primarily black African HIV-infected women enrolled at a median of 26 weeks of gestation (interquartile range, 21 to 30). The rate of transmission was significantly lower with ART than with zidovudine alone (0.5% in the combined ART groups vs. 1.8%; difference, -1.3 percentage points; repeated confidence interval, -2.1 to -0.4). However, the rate of maternal grade 2 to 4 adverse events was significantly higher with zidovudine-based ART than with zidovudine alone (21.1% vs. 17.3%, P=0.008), and the rate of grade 2 to 4 abnormal blood chemical values was higher with tenofovir-based ART than with zidovudine alone (2.9% vs. 0.8%, P=0.03). Adverse events did not differ significantly between the ART groups (P>0.99). A birth weight of less than 2500 g was more frequent with zidovudine-based ART than with zidovudine alone (23.0% vs. 12.0%, P<0.001) and was more frequent with tenofovir-based ART than with zidovudine alone (16.9% vs. 8.9%, P=0.004); preterm delivery before 37 weeks was more frequent with zidovudine-based ART than with zidovudine alone (20.5% vs. 13.1%, P<0.001). Tenofovir-based ART was associated with higher rates than zidovudine-based ART of very preterm delivery before 34 weeks (6.0% vs. 2.6%, P=0.04) and early infant death (4.4% vs. 0.6%, P=0.001), but there were no significant differences between tenofovir-based ART and zidovudine alone (P=0.10 and P=0.43). The rate of HIV-free survival was highest among infants whose mothers received zidovudine-based ART.

Conclusions: Antenatal ART resulted in significantly lower rates of early HIV transmission than zidovudine alone but a higher risk of adverse maternal and neonatal outcomes. (Funded by the National Institutes of Health; PROMISE ClinicalTrials.gov numbers, NCT01061151 and NCT01253538 .).

Figure 1. Randomization Schema and Drug Regimens during the Antepartum Component of the Trial

Under version 2.0 of the trial protocol (period 1), owing to limited safety data on tenofovir in pregnancy, only women who were positive for hepatitis B surface antigen could be randomly assigned to tenofovir-based antiretroviral therapy (ART); under version 3.0 (period 2), all women could be assigned to any of the three regimens. Infants received once-daily nevirapine in all trial groups, with dosing based on birth weight, from birth through 6 weeks of age. All women and infants received prophylaxis with trimethoprim–sulfamethoxazole. All women who met specific treatment guidelines and all infants who were confirmed to be infected with the human immunodeficiency virus (HIV) began ART immediately. Infants of women coinfected with hepatitis B and HIV received hepatitis B immune globulin and hepatitis B vaccine at birth and the full hepatitis B vaccine series.

Figure 2. Probability of Infant HIV Infection or Death through the Week 1 Postpartum Visit (6 to 14 Days after Delivery)

The insets show the same data on an expanded y axis. TDF denotes tenofovir, TDF-based ART tenofovir, emtricitabine, and lopinavir–ritonavir, ZDV zidovudine, ZDV alone zidovudine plus single-dose nevirapine, and ZDV-based ART zidovudine, lamivudine, and lopinavir–ritonavir.