In 2016, approximately 790,000 women were newly infected with HIV-1, contributing to the estimated 17.8 million women living with HIV-1 worldwide.1 Among other adverse perinatal outcomes, women living with HIV-1 have increased risk for delivering low birthweight (LBW) infants.2 LBW, defined as birthweight less than 2500 g, regardless of gestational age at delivery, is a composite indicator of a complex condition. LBW may be caused by preterm birth (PTB; birth before 37 completed weeks' gestation), intrauterine growth restriction, or congenital/genetic abnormalities.3,4 LBW infants are at increased risk for multiple health problems, including lung diseases, diabetes, high blood pressure, and diseases of the brain and nervous system.5,6 From a meta-analysis of 31 studies in 16 countries, the summary odds ratio of LBW related to maternal HIV infection was 2.1 (95% CI: 1.9 to 2.4).5 Thus, the HIV epidemic has both significantly increased the number of children at risk for LBW and dramatically increased the likelihood that children born to HIV-infected women will have childhood and adult morbidities.
In 2015, in an effort to prevent HIV-1 mother-to-child transmission (pMTCT) and improve maternal health, the World Health Organization (WHO) introduced Option B+, a treatment approach involving lifelong antiretroviral therapy (ART) for all pregnant and breastfeeding HIV-positive women, regardless of disease severity.6 This recommendation expanded on the previous treatment approach introduced in 2013 (option B), where treatment was only continued after pregnancy and breastfeeding if the woman was eligible for ART for her own health.4 An estimated 77% of all pregnant HIV-infected pregnant women received ART for pMTCT in 2015, compared with 53% coverage in 2009.7 This increase in coverage, largely attributed to option B/B+, has resulted in a substantial reduction in the number of HIV-infected infants.2
Although the massive ART rollout has decreased HIV-1 MTCT from 15% to 45% without ART to less than 5% with ART, a similar dramatic decrease in the prevalence of LBW and PTB among infants born to HIV-infected women has not been realized.8,9 For example, in the United States from 1989 to 2004, maternal ART use increased from 2% to 84%, MTCT decreased from ∼25% to less than 2%, but the prevalence of LBW only decreased from 35% to 21%; among women living with HIV, the prevalence of LBW was 3 times the LBW prevalence in the general US population.9 The reasons behind the still-elevated risk for adverse birth outcomes among infants of HIV-infected women, even HIV-1–exposed and HIV-1–uninfected infants, are not fully understood. Some research suggests that exposure to specific ART regimens in utero may be associated with adverse birth outcomes; specifically, in Europe, higher rates of PTB have been observed among women receiving ART.10–12 By contrast, studies conducted in the United States and Latin America have not observed an association between adverse birth outcomes and in utero ART exposure.13–15 Certain ART regimens may be more detrimental than others: relative to other ART regimens, protease inhibitor (PI)-based ART has been associated with an increased risk of PTB,16–18 particularly among those who initiate therapy earlier in pregnancy.16 Cautioning against the use of PI-based ART remains controversial; however, as this regimen was initially preferred for the treatment of women with more advanced disease, leading some to suggest that the association between PI-based ART and PTB could be the result of confounding by indication.19
Compared to receiving no therapy, ART use during pregnancy brings significant and undisputed benefits: it reduces the risk of HIV-1 MTCT, infant mortality, and fetal demise.20 However, it is critical to assess which ART regimens maximize maternal health and also minimize the risk of adverse birth outcomes, including HIV-1 MTCT, but also, whenever possible, PTB and LBW. Previous reviews on adverse birth outcomes and ARTs in pregnancy focus on studies from higher income countries18 or—if they include studies from lower income countries—do not contain studies published within the last 5 years.21 Thus, this systematic review focuses on pregnant women living with HIV-1 in low- and middle-income countries (LMICs). We sought to identify meaningful trends in associations between classes of ART regimens and risk of LBW or PTB.
Search Strategy and Study Selection
We conducted a systematic review of studies that reported the association between ART and either LBW or PTB, according to the guidelines developed by the PRISMA group.22 We searched electronic databases Medline, COCHRANE, Web of Science and SCOPUS, and examined the reference lists of identified articles. In addition, we reviewed Conference Proceedings Citation Index-Science (CPCI-S) to identify any relevant additional literature. Key search terms included “adverse birth outcomes,” “preterm birth,” “PTB,” “preterm delivery,” “PTD,” “low birthweight,” “LBW” and “antiretroviral therapy.” The database search and initial screening of title and abstracts was performed by a single investigator (J.L.S.). Any relevant article published on or before January 10, 2018, was included. Three authors (J.L.S., J.J.K., and C.M.) then independently read all full-text publications to determine whether they met eligibility criteria (see below). Two authors (J.L.S. and C.M.) extracted relevant data from selected articles using a standardized data collection form. In the event of a disagreement, a fourth author (A.N.T.) was consulted.
We included studies based on the following study characteristics: (1) participants: HIV-1–positive women in LMICs, as defined by the World Bank's classifications based on gross national income per capita, (2) intervention: ART initiated before conception and continuing throughout pregnancy, or ART initiated in pregnancy (with a significant majority of participants initiating during the first or second trimester of pregnancy). Studies in which most women received ART only in the third trimester or during labor were not included because it is unlikely that this therapy would have a meaningful effect on prematurity or intrauterine growth, and we wanted to ensure that there were relatively equivalent periods of exposure for different treatment groups, (3) comparison: ART regimens were compared against each other and to monotherapy. To reduce selection bias, regimens were only compared if they were initiated during the same period (ie, either preconception or postconception). Studies which did not specify time of treatment initiation were excluded. In addition, studies which only provided comparison with treatments that are no longer clinically relevant (eg, dual therapy or no therapy) were also excluded, (4) Outcomes: LBW, defined as infant weight less than 2500 g at birth, very LBW, defined as infant weight less than 1500 g at birth, PTB, defined as gestational age less than or equal to 37 completed weeks of gestation at birth, and very PTB (vPTB), defined as gestational age less than 32 completed weeks of gestation at birth, (5) study design: randomized [randomized controlled trials (RCTs) and quasi-RCTs] and observational studies, including cohort (retrospective and prospective), case-control, and registry studies were included. Ecological studies as well as case studies and case series were excluded. Included studies were further limited to articles written in English with human participants (ie, we did not include nonhuman primate studies). Nonenglish articles were excluded.
For each reviewed study, we extracted information on the study design, population, treatment regimen and dose, effect measures, and adjustment variables. Different ART regimens were grouped for comparison according to drug classification. We assessed study quality using the Newcastle Ottawa Scale for nonrandomized studies.23 Studies are assigned stars based on their selection, comparability, and exposure, with more stars reflecting higher quality. To assess the quality of randomized studies, we used the Cochrane Bias Assessment Tool, which reports the risk of several sources of bias, including selection bias, performance bias, detection bias, attrition bias, and reporting bias.24
Study Selection and Characteristics
Thirteen studies met the inclusion criteria and are included in the review (Fig. 1). Treatment regimens were classified into 4 broad categories: (1) monotherapy [which included monotherapy given together with single-dose nonnucleoside reverse transcriptase inhibitors (NNRTIs) in labor], (2) PI-based ART, (3) NNRTI-based ART [both efavirenz (EFV)-based and nevirapine (NVP)-based], and (4) NRTIs-based ART. Regimens were further delineated by drug bases and drug backbones. All the included studies took place in sub-Saharan Africa after 2005, with 1 study taking place in both India and sub-Saharan Africa.6 A majority (69%) were secondary analyses of existing medical record data (see Tables 1 and 2, Supplemental Digital Content, http://links.lww.com/QAI/B182 for descriptions and summary statistics of included studies).
Risk of Bias of Included Studies
We assessed the potential for bias for the 9 nonrandomized studies using the Newcastle Ottawa Scale tool and found low to moderate bias in most studies (see Table 1, Supplemental Digital Content, http://links.lww.com/QAI/B182). For many of the nonrandomized studies included within this analysis, the selection of control groups or nonexposed cohorts could also have induced bias by indication: women who received monotherapy later in pregnancy often made up the control groups/nonexposed cohorts of our included studies, and these women may have had less severe HIV-1 disease or may have presented later for antenatal care for unknown, potentially confounding, reasons. However, the majority (56%) of nonrandomized studies controlled for maternal HIV-1 disease severity (ie, CD4+ T-cell count or WHO disease staging), which we hypothesized a priori could induce bias by indication. Six of the 9 nonrandomized studies also controlled for other potential confounders in their analyses, such as obstetric history, education, maternal age, and body mass index (BMI) (see Table 2, Supplemental Digital Content, http://links.lww.com/QAI/B182). For the 4 randomized trials, we found low risk of bias using the Cochrane Bias Assessment Tool (see Table 1, Supplemental Digital Content, http://links.lww.com/QAI/B182).
Synthesis of Results: PTB
Compared to NRTI [abacavir (ABC)]-based ART and zidovudine (ZDV) monotherapy, PI [lopinavir/ritonavir (LPV/r)]-based ART with a ZDV + lamivudine (3TC) backbone was associated with increased PTB.17,25 This association remained after adjusting for CD4+ T-cell count or WHO disease staging (Table 1) and other confounders. By contrast, there was no consistent association, unadjusted or adjusted, between PTB and other PI-based ART regimens when compared with EFV-based ART or monotherapy (Fig. 2 and Table 1).25–27
When compared to all other regimens, NNRTI (EFV)-based ART either exhibited null effects or was protective against PTB. One large observational study conducted in Botswana (N = 5780) found that EFV-based ART with a tenofovir disoproxil fumarate (TDF) + emtricitabine (FTC) backbone was protective against PTB when compared with NVP-based ART with a ZDV + 3TC backbone, and when compared with PI (LPV/r)-based ART with a ZDV + 3TC backbone.28 Regarding NNRTI (non-EFV)–based ART, a large study conducted in Botswana (N = 5726) found that NVP-based ART with a ZDV + 3TC backbone was associated with increased risk of PTB when compared with ZDV monotherapy,29 although a smaller study (N = 2280) conducted in Tanzania did not find significant association when comparing the 2 regimens.30 Both studies controlled for maternal disease severity.
Three of the 14 studies included in this analysis also reported on vPTB. Associations with vPTB generally reflected those of PTB.17,28,30 Similar to their findings for PTB, Li et al30 observed no association between PI (LPV/r)-based ART with a ZDV + 3TC backbone and vPTB when compared with NRTI (ABC)-based ART. Relative to their findings for all PTB, Zash et al28 found a greater effect of NVP-based ART (with a ZDV + 3TC backbone) and LPV/r-based ART (with a TDF + 3TC backbone) on vPTB when these regimens were compared with EFV-based ART (with a TDF + FTC backbone) [ORadj (95% CI): 1.44 (1.07 to 1.95); and ORadj (95% CI): 2.21 (1.29 to 3.79), respectively]. Conversely, Powis et al17 did not find a significant association between vPTB and LPV/r-based ART when compared with ABC-based ART, in contrast to the significant positive association they found for all PTB (Table 2).
Synthesis of Results: LBW
When compared to ZDV monotherapy, both PI-based and non-PI–based ART regimens had a consistent, harmful association with LBW (Fig. 3),25,31,32 with one exception for ART (combined LPV/r- and ABC-based) with a ZDV + 3TC backbone, which did not show a significant association.33 We observed mixed results regarding the effects of EFV-based ART vs. NVP-based ART on LBW, with 1 study reporting that EFV-based ART increased the risk of LBW,27 and another study reporting no significant differences in LBW risk between the 2 regimens.31 Overall, results from unadjusted analyses agreed with results from adjusted analysis among the studies that controlled for maternal disease severity and other confounders (Table 3). No studies provided adequate information to compare the effect of ART regimen on very LBW.
This review summarized the wide array of ART regimens used by pregnant women living with HIV in LMICs and their association with LBW or PTB. From our assessment of the available evidence, we conclude the following: (1) when compared with monotherapy, both PI-based and non-PI–based ART increase the odds of LBW; (2) there is mixed evidence suggesting both potential harm and potential benefit for most other regimens on risk of LBW and PTB; and (3) the harmful or protective effects of certain regimens varies depending on the drug backbone, with PI (LPV/r)-based ART exhibiting harmful effects on PTB when paired with a ZDV + 3TC backbone, and EFV-based ART exhibiting protective effects against LBW when paired with a TDF + FTC backbone.
Studies that examine the effects of treatments initiated within pregnancy could be subject to selection bias because women who experience PTB before initiating therapy are systematically excluded from the analysis. Stoner et al34 contend that the association between PTB preconception initiation of ART, noted in previous research,29,35,36 is the result of this selection bias. In the absence of larger randomized trials, they used a simulated cohort of 1000 HIV-infected women to demonstrate that preconception ART was not associated with PTB in an intent-to-treat analysis. In recognition of the risk of selection bias, we limited our comparisons with treatments initiated contemporaneously (ie, preconception or postconception). Nevertheless, certain regimens may be initiated later in pregnancy relative to other regimens, potentially attenuating their effect on PTB. In general, ZDV monotherapy was initiated later in pregnancy compared with ART, which may explain, at least in part, the lower risk of PTB among women receiving monotherapy.
Simultaneous minimization of HIV-1 MTCT, PTB, and LBW outcomes is challenging. For example, in the PROMISE RCT, pregnant HIV-infected women who received zidovudine plus single-dose NVP had the highest HIV-1 MTCT prevalence (2%) and the lowest LBW prevalence (9%), whereas women who received zidovudine-based ART (zidovudine, lamivudine, and lopinavir/ritonavir) had the lowest HIV-1 MTCT prevalence (0.5%) and the highest LBW prevalence (20%).25 Although effective at reducing HIV-1 MTCT, compared with monotherapy, non-PI–based ART is consistently associated with increased LBW and PTB; the pooled odds ratio for nonrandomized studies within this review for non-PI–based ART verses monotherapy was 1.88 (95% CI: 1.34 to 2.63) for LBW, and 1.50 (95% CI: 1.40 to 1.6) for PTB. Although highly consistent across studies, interpretation of this finding is challenging because it compares the new standard of care (non-PI–based ART) to a regimen, which is no longer routinely offered (eg, monotherapy). This result clearly does not advocate for a return to a less-effective pMTCT regimen; rather, it suggests that current regimens should be optimized to maximize both prevention of MTCT and positive birth outcomes.
Seven of the 9 nonrandomized studies controlled for confounders in multivariable models. Most of the studies took place before 2013, when many LMICs reserved ART treatment for women with CD4+ T-cell counts <350 cells/µL. Therefore, we might reasonably expect to see more adverse birth outcomes among women on ART compared with those not using ART because these women by definition had severe enough HIV-1 disease to warrant treatment. Only 3 studies25,27,28 included data collected after 2013, and these studies varied in their regimen comparisons, and thus it is difficult to comment on any observed trend in the odds of LBW and PTB for studies conducted before vs. after 2013 (see Table 1, Supplemental Digital Content, http://links.lww.com/QAI/B182). In addition, only 2 studies included covariates that are prevalent in lower income settings (ie, anemia and opportunistic infections). Despite this, we did not note any meaningful differences in the results between studies that adequately controlled for these confounders and those which did not.
Our review has several limitations. Because of geographic variations in access to both antenatal care and specific ART medications, as well as local variation in the implementation of therapy guidelines, we observed great diversity across the literature on ART and pregnancy outcomes. This diversity presented a significant challenge for our original goal of direct comparison between regimens. Furthermore, we did not include studies that reported neonatal anthropometric measures only (weight-for-age, length-for-age, and weight-for-length), or small for gestational age as additional adverse birth outcomes of interest, because these measures are inconsistently reported in LMICs. However, investigations of associations between ART and these outcomes—when validly captured—could provide insight into intrauterine growth restriction, which is a hypothesized cause of LBW. Finally, although its use among pregnant women is increasing worldwide,19,20 there are currently no published studies that examine the relationship between dolutegravir-based ART and adverse birth outcomes.
ART use during pregnancy is an essential public health intervention that promotes maternal health and prevents HIV-1 MTCT. From this analysis, some ART regimens seem to have a harmful association with PTB and LBW. Because of the fact that LBW infants have increased risks of mortality,37 neurodevelopmental impairment (including cerebral palsy),38,39 impaired lung function and respiratory morbidities,40 and adult onset diseases such as type II diabetes mellitus, hypertension, and cardiovascular disease,41 reducing risk of adverse birth outcomes among children born to women living with HIV-1 should also be prioritized. The mechanisms linking HIV-1 to PTB and LBW remain unknown, but additional mechanistic insights could help to identify an optimal HIV-treatment modality for pregnant women living with HIV: one which prevents HIV-1 MTCT, promotes maternal health, and also minimizes adverse birth outcomes.
The authors thank Dr. Marcel Yotebieng and Dr. Maria Gallo for their guidance.
2. Evans C, Jones CE, Prendergast AJ. HIV
-exposed, uninfected infants: new global challenges in the era of paediatric HIV
elimination. Lancet Infect Dis. 2016;16:e92–e107.
3. Usha K, Sarita B. Placental insufficiency and fetal growth restriction. J Obstet Gynecol India. 2011;61:505.
5. Brocklehurst P, French R. The association between maternal HIV
infection and perinatal outcome: a systematic review of the literature and meta-analysis. Br J Obstet Gynaecol. 1998;105:836–848.
8. Neri D, Somarriba GA, Schaefer NN, et al. Growth and body composition of uninfected children exposed to human immunodeficiency virus: comparison with a contemporary cohort and United States national standards. J Pediatr. 2013;163:249–254.e2.
9. Schulte J, Dominguez K, Sukalac T, et al; Pediatric Spectrum of HIV
Disease Consortium. Declines in low birth weight and preterm birth among infants who were born to HIV
-infected women during an era of increased use of maternal antiretroviral drugs: pediatric spectrum of HIV
disease, 1989-2004. Pediatrics. 2007;119:e900–e906.
10. Thorne C, Townsend CL. A new piece in the puzzle of antiretroviral therapy in pregnancy and preterm delivery risk. Clin Infect Dis. 2012;54:1361–1363.
11. Townsend C, Schulte J, Thorne C, et al. Antiretroviral therapy and preterm delivery-a pooled analysis of data from the United States and Europe. BJOG. 2010;117:1399–1410.
12. Townsend CL, Cortina-Borja M, Peckham CS, et al. Antiretroviral therapy and premature delivery in diagnosed HIV
-infected women in the United Kingdom and Ireland. AIDS. 2007;21:1019–1026.
13. Cooper ER, Charurat M, Mofenson L, et al; Women, Infants' Transmission Study Group. Combination antiretroviral strategies for the treatment of pregnant HIV
-1-infected women and prevention of perinatal HIV
-1 transmission. J Acquir Immune Defic Syndr. 2002;29:484–494.
14. Szyld EG, Warley EM, Freimanis L, et al. Maternal antiretroviral drugs during pregnancy and infant low birth weight and preterm birth. AIDS. 2006;20:2345–2353.
15. Tuomala RE, Shapiro DE, Mofenson LM, et al. Antiretroviral therapy during pregnancy and the risk of an adverse outcome. N Engl J Med. 2002;346:1863–1870.
16. Cotter A, Garcia A, Duthely MÂ, et al. Is antiretroviral therapy during pregnancy associated with an increased risk of preterm delivery, low birth weight, or stillbirth? J Infect Dis. 2006;193:1195–1201.
17. Powis KM, Kitch D, Ogwu A, et al. Increased risk of preterm delivery among HIV
-infected women randomized to protease versus nucleoside reverse transcriptase inhibitor-based HAART during pregnancy. J Infect Dis. 2011;204:506–514.
18. Kourtis AP, Schmid CH, Jamieson DJ, et al. Use of antiretroviral therapy in pregnant HIV
-infected women and the risk of premature delivery: a meta-analysis (provisional abstract). AIDS. 2007;21:607–615.
19. Chougrani I, Luton D, Matheron S, et al. Safety of protease inhibitors in HIV
-infected pregnant women. HIV
AIDS (Auckl). 2013;5:253–262.
20. Marazzi MC, Palombi L, Nielsen-Saines K, et al. Extended antenatal use of triple antiretroviral therapy for prevention of mother-to-child transmission of HIV
-1 correlates with favorable pregnancy outcomes. AIDS. 2011;25:1611–1618.
21. Alemu FM, Yalew AW, Fantahun M, et al. Antiretroviral therapy and pregnancy outcomes in developing countries: a systematic review. Int J MCH AIDS. 2015;3:31–43.
22. Moher D, Liberati A, Tetzlaff J, et al; PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6:e1000097.
24. Higgins JPT, Altman DG, Gøtzsche PC, et al. The cochrane collaboration's tool for assessing risk of bias in randomised trials. Br Med J. 2011;343:889–893.
25. Fowler MG, Qin M, Fiscus SA, et al. Benefits and risks of antiretroviral therapy for perinatal HIV
prevention. N Engl J Med. 2016;375:1726–1737.
26. van der Merwe K, Hoffman R, Black V, et al. Birth outcomes in South African women receiving highly active antiretroviral therapy: a retrospective observational study. J Int AIDS Soc. 2011;14:42.
27. Zash R, Souda S, Chen JY, et al. Reassuring birth outcomes with tenofovir/emtricitabine/efavirenz used for prevention of mother-to-child transmission of HIV
in Botswana. J Acquir Immune Defic Syndr. 2016;71:428–436.
28. Zash R, Jacobson DL, Diseko M, et al. Comparative safety of antiretroviral treatment regimens in pregnancy. JAMA Pediatr. 2017;171:e172222.
29. Chen JY, Ribaudo HJ, Souda S, et al. Highly active antiretroviral therapy and adverse birth outcomes
-infected women in Botswana. J Infect Dis. 2012;206:1695–1705.
30. Li N, Sando MM, Spiegelman D, et al. Antiretroviral therapy in relation to birth outcomes among HIV
-infected women: a cohort study. J Infect Dis. 2015;213:1057–1064.
31. Njom Nlend AE, Nga Motazé A, Moyo Tetang S, et al. Preterm birth and low birth weight after in utero exposure to antiretrovirals initiated during pregnancy in Yaoundé Cameroon. PLoS One. 2016;11:e0150565.
32. Ekouevi DK, Coffie PA, Becquet R, et al. Antiretroviral therapy in pregnant women with advanced HIV
disease and pregnancy outcomes in Abidjan, Cote d'Ivoire. AIDS. 2008;22:1815–1820.
33. Powis KM, Smeaton L, Hughes MD, et al. In-utero triple antiretroviral exposure associated with decreased growth among HIV
-exposed uninfected infants in Botswana. AIDS. 2016;30:211–220.
34. Stoner MCD, Cole SR, Price J, et al. Timing of initiation of antiretroviral therapy and risk of preterm birth in studies of HIV
-infected pregnant women. Epidemiology. 2017;29:224–229.
35. Nachega JB, Nachega JB, Uthman OA, et al. Timing of initiation of antiretroviral therapy and adverse pregnancy outcomes: a systematic review and meta-analysis. Lancet HIV
36. Machado ES, Hofer CB, Costa TT, et al. Pregnancy outcome in women infected with HIV
-1 receiving combination antiretroviral therapy before versus after conception. Sex Transm Infect. 2009;85:82–87.
37. McCormick MC. The contribution of low birth weight to infant mortality and childhood morbidity. N Engl J Med. 1985;312:82–90.
38. Hack M, Klein NK, Taylor HG. Long-term developmental outcomes of low birth weight infants birth weight infants. Future Child. 1995;5:176–196.
39. Lampi KM, Lehtonen L, Tran PL, et al. Risk of autism spectrum disorders in low birth weight and small for gestational age infants. J Pediatr. 2012;161:830–836.
40. Pike K, Jane Pillow J, Lucas JS. Long term respiratory consequences of intrauterine growth restriction. Semin Fetal Neonatal Med. 2012;17:92–98.
41. Longo S, Bollani L, Decembrino L, et al. Short-term and long-term sequelae in intrauterine growth retardation (IUGR). J Matern Fetal Neonatal Med. 2013;26:222–225.
HIV; pMTCT; adverse birth outcomes; ART; LBW; PTB
Supplemental Digital Content
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.