Summary of Review Findings

This report updates a 2012 USPSTF review on prenatal screening for HIV infection.^1,2 Evidence reviewed for this update is summarized in Table 4. As in the 2012 USPSTF review, we found no direct evidence on effects of prenatal screening versus no screening on risk of mother-to-child HIV transmission or maternal or infant clinical outcomes. We also identified no studies on the yield of repeat prenatal screening versus one-time screening or different frequencies of screening for HIV in pregnancy. Although three studies conducted in the United States or the United Kingdom identified no cases of HIV infection among women who were rescreened for HIV during the third trimester of pregnancy,^115–117 results were difficult to interpret because the HIV risk of women who underwent rescreening was unclear, and not all women underwent rescreening. As discussed in the prior USPSTF report, the yield of repeat prenatal screening depends on HIV infection incidence during pregnancy. In addition, detecting HIV acquired during pregnancy may be important because some data suggest markedly higher risk of mother-to-child transmission compared with HIV acquired prior to pregnancy.¹²⁰

Table 4

Summary of Evidence.

New evidence identified for this update^51,60,61,74 confirm findings from the 2012 USPSTF review that full-course combination ART is highly effective at reducing the risk of mother-to-child transmission, with some cohort studies reporting rates of mother-to-child transmission of less than 1 percent when started early in pregnancy.^61,74 Cohort studies and RCTs also found that combination therapy started in the second or third trimester are effective at reducing the risk of mother-to-child transmission. Shorter courses of ART were not as effective as full-course regimens, but also reduce risk of mother-to-child transmission compared with no ART, supporting benefits of screening and initiation of therapy later in pregnancy.^44,80–82

New evidence on harms of ART was also largely consistent with the 2012 USPSTF review. Although some ART agents and regimens may be associated with increased risk of infant or maternal harms, such harms may be mitigated or reduced through appropriate selection of ART regimens. As in the prior USPSTF review, evidence from primarily observational studies found prenatal combination ART with a boosted protease inhibitor associated with increased risk of preterm delivery.^44,54,72,78 An African RCT found tenofovir-containing, lopinavir/ritonavir-based combination ART associated with greater risk of early infant death than zidovudine-containing, lopinavir/ritonavir-based combination ART.⁴⁴ However, there were methodological limitations with this trial, including two periods with different randomization protocols, and different rates of some adverse birth outcomes depending on period of randomization. The RCT found no difference in early infant death between tenofovir-based combination ART and treatment with zidovudine alone and no difference between treatments in risk of stillbirth. Tenofovir is a preferred nucleoside reverse transcriptase inhibitor for use in pregnancy in most, but not all, guidelines due to its demonstrated efficacy, acceptable toxicity, ease of use, and no established teratogenicity; however, the use of lopinavir, rather than a preferred protease inhibitor, makes the ART combination evaluated in this trial an alternative regimen (lopinavir is associated with more nausea than preferred protease inhibitors).⁴⁷ The increased risk of early infant death in the trial could be related to a higher risk of very preterm delivery associated with this tenofovir-containing ART regimen, which is associated with increased risk of infant mortality in low-income settings. Other African studies found no association between tenofovir-based ART and risk of stillbirth^62,78,79 or neonatal death.^65,78 A recent review of TDF in pregnancy and breastfeeding did not identify safety issues versus non-TDF regimens, but included HIV-negative women and unpublished studies.⁴³

For other birth outcomes (low birth weight, small for gestational age, stillbirth, and overall birth defects), results were mixed and depended on the specific antiretroviral drug or drug regimen given and timing of prenatal therapy. As in the prior USPSTF review, some evidence indicated that ART may be associated with cardiac findings such as ventricular septal defects^70,71 and echocardiographic changes,^58,70 although the clinical significance of findings is unclear. Evidence on congenital abnormalities was limited by small numbers of studies and imprecise estimates, although some studies found exposure to different drugs in the first trimester associated with increased risk of congenital abnormalities. Studies in older children exposed to ART in utero suggested no association with worse neurodevelopmental outcomes compare with unexposed children.^64,77

Evidence on long-term maternal harms associated with short-term exposure to ART during pregnancy or ART started during pregnancy and continued after pregnancy remains sparse, although one new study found evidence of increased short-term nonspecific adverse events.⁴⁴ Women found to be HIV-infected through prenatal screening would also benefit from standard HIV treatments following pregnancy, including long-term combination ART, prophylaxis for opportunistic infections, immunizations, and indicated screenings.^47,121

Limitations

We excluded non-English–language articles, which could result in language bias, although we identified no non-English–language studies that would have met inclusion criteria. We did not attempt to pool studies because of differences in study designs, populations, study setting, antiretroviral regimens evaluated, and outcomes assessed. Because we could not pool studies, we also could not formally assess for publication bias with graphical or statistical methods. We included observational studies, which are more susceptible to bias and confounding than well-conducted RCTs, although we restricted inclusion to observational studies that performed statistical adjustment for potential confounding. Another limitation is that RCTs of combination ART have only been conducted in Africa. The applicability of studies conducted in resource-poor and high-prevalence settings to U.S. practice is limited by differences in the antiretroviral drugs evaluated, evaluation of shorter regimens, inclusion of women who breastfeed, and other factors. Although we focused on new studies published since 2012, in most studies results were reported for individual ART agents and classes, rather than for currently recommended ART regimens, which could reduce applicability of findings to current U.S. practice. Restricting analyses to studies in which patients received treatment after 2006 (the earliest year a current preferred regimen was approved), or in whom results for currently recommended regimens could clearly be identified, did not appear to change conclusions, although formal stratified analyses were not possible.

Emerging Issues/Next Steps

ART regimens for use during pregnancy and indications for initiating long-term ART continue to evolve. Regularly updated guidelines on selection of ART in pregnant women are available.⁴⁷

Relevance for Priority Populations, Particularly Racial/Ethnic Minorities

HIV disproportionately affects women from racial/ethnic minority populations, who often have less access to prenatal care, leading to reduced opportunities for early screening and initiation of ART during pregnancy. Identification of HIV infection during pregnancy provides an opportunity to link affected women to long-term care and treatment. Therefore, improving access to prenatal care is an important challenge for reducing the effect of HIV infection in these populations.

Future Research

Although there are no studies comparing the effects of screening for HIV in pregnancy versus no screening, such studies may no longer be indicated given epidemiological evidence showing marked decreases in the number of children with perinatally acquired HIV infection in the United States and strong evidence on the effectiveness of ART on preventing mother-to-child transmission. Studies comparing one-time screening versus repeat screening or that perform rescreening in well-defined cohorts of women would be helpful for understanding the yield of rescreening and for understanding when rescreening is indicated. Future research is needed to further clarify the effectiveness and harms of currently recommended antiretroviral regimens, effects of in utero exposure to ART on pregnancy outcomes, and long-term harms in exposed children to optimize selection of ART regimens during pregnancy and to understand the effects of screening and treatment with ART in pregnant adolescents.

Conclusions

Combination ART is highly effective at reducing risk of mother-to-child HIV transmission. The USPSTF previously determined that avoidance of breastfeeding and Caesarean delivery in women with HIV RNA levels greater than 1,000 copies/mL near the time of delivery is also effective at reducing mother-to-child transmission, and that prenatal screening is accurate at diagnosing HIV infection. Use of certain ART regimens during pregnancy is associated with increased risk of preterm delivery and other adverse pregnancy outcomes. Although more evidence is required to better understand short- and long-term maternal and infant harms, selection of ART regimens may help mitigate or reduce harms.