Objective: Assess adverse perinatal outcomes associated with antenatal antiretroviral therapy (ART) regimens.

Design: Systematic review and network meta-analysis of randomized controlled trials (RCTS).

Methods: We conducted a systematic literature review by searching PubMed, CINAHL, Global Health, EMBASE, and the Cochrane Central Register of Controlled Trials and four clinical trial databases from 1 January 1980 to 28 April 2018. We included RCTs of antenatal ART regimens in HIV-positive pregnant women, which assessed preterm birth (PTB), spontaneous preterm birth (sPTB), very preterm birth (VPTB), low birthweight (LBW), very low birthweight (VLBW), small-for-gestational-age (SGA), neonatal death (NND), and mother-to-child-transmission. We used random-effects network meta-analysis models to calculate relative risks for treatment comparisons and the hierarchy of treatments.

Results: Of 83 260 citations identified, 10 manuscripts were included, assessing 6285 women. Compared with zidovudine (ZDV) monotherapy, we found a higher risk of LBW after exposure to zidovudine/lamivudine/efavirenz (ZDV/3TC/EFV; relative risk 1.61; 95% CI 1.03-2.51), tenofovir disoproxil fumarate/emtricitabine/ritonavir-boosted lopinavir (TDF/FTC/LPV/r; 1.64; 1.18-2.29), or zidovudine/lamivudine/ritonavir-boosted lopinavir (ZDV/3TC/LPV/r; 1.87; 1.58-2.20). TDF/FTC/LPV/r carried an increased risk of VLBW, compared with ZDV monotherapy (5.40; 1.08-27.08). ZDV/3TC/LPV/r posed a higher risk of PTB than ZDV monotherapy (1.43; 1.08-1.91) and a higher risk of sPTB than zidovudine/lamivudine/abacavir (ZDV/3TC/ABC) (1.81; 1.21-2.71). LPV/r-containing regimens also carried the highest risks of VPTB, SGA and NND, although the limited data showed no significant differences.

Conclusion: Of the ART regimens assessed in RCTs in pregnancy, LPV/r-containing regimens were associated with the highest risks of adverse perinatal outcomes.