The n e w e ng l a n d j o u r na l of m e dic i n e

Original Article

Benefits and Risks of Antiretroviral Therapy
for Perinatal HIV Prevention
M.G. Fowler, M. Qin, S.A. Fiscus, J.S. Currier, P.M. Flynn, T. Chipato, J. McIntyre,
D. Gnanashanmugam, G.K. Siberry, A.S. Coletti, T.E. Taha, K.L. Klingman,
F.E. Martinson, M. Owor, A. Violari, D. Moodley, G.B. Theron, R. Bhosale,
R. Bobat, B.H. Chi, R. Strehlau, P. Mlay, A.J. Loftis, R. Browning, T. Fenton,
L. Purdue, M. Basar, D.E. Shapiro, and L.M. Mofenson,
for the IMPAACT 1077BF/1077FF PROMISE Study Team*​​

A BS T R AC T

BACKGROUND
The authors’ full names, academic degrees, Randomized-trial data on the risks and benefits of antiretroviral therapy (ART) as compared
and affiliations are listed in the Appendix. with zidovudine and single-dose nevirapine to prevent transmission of the human immuno-
Address reprint requests to Dr. Fowler at
the Department of Pathology, Johns deficiency virus (HIV) in HIV-infected pregnant women with high CD4 counts are lacking.
Hopkins University School of Medicine, METHODS
600 N. Wolfe St., 443 Carnegie, Baltimore,
MD 21287, or at ­mgfowler@​­mujhu​.­org. 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
* A complete list of members of the Inter-
national Maternal Pediatric Adolescent plus a 1-to-2-week postpartum “tail” of tenofovir and emtricitabine (zidovudine alone);
AIDS Clinical Trials Network (IMPAACT) zidovudine, lamivudine, and lopinavir–ritonavir (zidovudine-based ART); or tenofovir,
1077BF/1077FF Promoting Maternal and emtricitabine, and lopinavir–ritonavir (tenofovir-based ART). The primary outcomes
Infant Survival Everywhere (PROMISE)
Study Team is provided in the Supple- were HIV transmission at 1 week of age in the infant and maternal and infant safety.
mentary Appendix, available at NEJM.org.
RESULTS
N Engl J Med 2016;375:1726-37. The median CD4 count was 530 cells per cubic millimeter among 3490 primarily black
DOI: 10.1056/NEJMoa1511691
African HIV-infected women enrolled at a median of 26 weeks of gestation (interquartile
Copyright © 2016 Massachusetts Medical Society.
range, 21 to 30). The rate of transmission was significantly lower with ART than with zido­
vudine 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 zidovu-
dine 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 teno-
fovir-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.)

1726 n engl j med 375;18 nejm.org  November 3, 2016

The New England Journal of Medicine
Downloaded from nejm.org on January 8, 2017. For personal use only. No other uses without permission.
Copyright © 2016 Massachusetts Medical Society. All rights reserved.

no clinical or immune-related zidovudine plus intrapartum single-dose nevira­ indication for triple-drug ART. All pregnant women provided written in- nancy outcomes with maternal ART than with formed consent. Zambia. or tenofovir.5 g per deciliter. emtricitabine. Malawi. Uganda. neutrophil count of at least 750 cells per cubic lamivudine. zidovudine. 2011) and formula-feeding settings (enrollment vention of mother-to-child transmission for started in July 2011) had separate protocols women with a CD4 count of more than 350 cells (available at NEJM. studies have shown higher rates of adverse preg. intrapartum. partum. feeding settings (enrollment started in April When enrollment began in 2011. All rights reserved. n engl j med 375. No other uses without permission. birth until postpartum randomization. Receipt of one or two anti. or a fetus with is available at NEJM. South Africa. the prespecified maternal nevirapine resistance. the results. and the data and least 350 cells per cubic millimeter (or a country. and no serious preg. and lopinavir–ritonavir (zidovudine- millimeter.1. shows the antepartum regimens and dosing.org infant.18 nejm. Copyright © 2016 Massachusetts Medical Society.org on January 8. in the Supplementary Appendix. women were least 14 weeks and not in labor. Results evaluating early trans.6 early stopping guideline for efficacy for the ante- Eligibility criteria included a CD4 count of at partum component was met. ceipt of tuberculosis treatment within 30 days ful trial that used zidovudine single-drug prophy. with postpartum randomization (breast-feeding set- tings) at 6 to 14 days to maternal ART or infant Me thods prophylaxis during breast-feeding and “maternal Trial Sites and Participants health” randomization of women receiving ART The PROMISE trial was conducted at 14 sites in to continue or stop ART after breast-feeding ces- seven countries (India. and Zimbabwe). the full text of this article at NEJM.org). Antiretrovir al Ther apy for Perinatal HIV Prevention A  ntiretroviral regimens used for and for 30 days or fewer during the current the prevention of mother-to-child trans. an estimated creatinine clearance of more imens were continued through 6 to 14 days post than 60 ml per minute. child transmission during pregnancy among asymptomatic HIV-infected pregnant women with Trial Design and Treatment Regimens high CD4 counts. The PROMISE trial compared open-label strate- mission through the week 1 postpartum trial gies and included sequential randomizations: visit (in utero. Figure 1 retroviral agents for the prevention of mother. an absolute bine post partum (zidovudine alone). The trial was approved by local regimens containing fewer antiretroviral agents. Breast- Tanzania. A Quick Take antiretroviral therapy (ART) for the mother and tural or conduction heart defect. standard pre. and less than 2. 2016 1727 The New England Journal of Medicine Downloaded from nejm. 2017.5 times the upper limit of the normal lopinavir–ritonavir (tenofovir-based ART). antepartum randomization (at 14 weeks of ges- feeding transmission) are presented.org). an alanine aminotransferase level of based ART). Key mission of the human immunodeficiency exclusion criteria were active tuberculosis or re- virus (HIV) have evolved from the first success. no previous use of randomly assigned to one of three regimens: triple-drug ART. At the November a 1-to-2-week “tail” of two nucleosides to prevent 2014 planned interim analysis. to-child transmission in previous pregnancies Antiretroviral agents were donated by AbbVie. sation or birth (formula-feeding settings). a hemoglobin pine with 6 to 14 days of tenofovir and emtricita­ level of at least 7. or very early breast. . a struc. pregnancy before enrollment was permitted. For personal use only. the complete PROMISE per cubic millimeter in these countries was zidovu­ trial schema is shown in Figure S1A and S1B in dine with intrapartum single-dose nevirapine and the Supplementary Appendix. All reg- range.2 tion requiring HBV treatment (patients who did Although there are clear benefits of combination not require HBV treatment could enroll).3-5 and collaborating institutional review boards and The Promoting Maternal and Infant Survival reviewed every 6 months by an independent data Everywhere (PROMISE) trial compared the relative and safety monitoring board (members are listed efficacy and safety of various proven antiretro. safety monitoring board recommended releasing specific threshold for initiating triple-drug ART. these do not come without risks. tation or later) to one of three regimens. gestation of at In the antepartum component. available with viral strategies for the prevention of mother-to. some a serious congenital malformation.org  November 3. if that threshold was higher). hepatitis B virus (HBV) infec- laxis in 1994 to current triple-drug regimens. All infants received nevirapine from nancy complications. before trial entry.

the protocol was modified to allow Healthcare/GlaxoSmithKline. and ran. to women coinfected with HIV and HBV. women without and 12 and then every 8 weeks. of lopinavir and 100 mg of ritonavir. 200 mg at onset of labor 300 mg of zidovudine and 150 mg of 200 mg of emtricitabine and 300 mg Emtricitabine–tenofovir. All women and infants received prophylaxis with trimethoprim–sulfamethoxazole. (6 to 14 days). at a dose of Emtricitabine–tenofovir. 8. nal visits occurred at trial entry. from birth through 6 weeks of age. at a dose of 400 mg Lopinavir–ritonavir. under version 3. domization to tenofovir-based ART was limited and 12 and then every 4 weeks until delivery. because at labor and delivery. 8. None of the pharma- ceutical companies had any role in the trial de. All rights reserved. ber 2012 through October 1. at a dose of lamivudine. at a dose of Nevirapine. Infants received once-daily nevirapine in all trial groups. The n e w e ng l a n d j o u r na l of m e dic i n e Pregnant women with CD4 count ≥350 cells/mm3 or country-specific threshold for treatment Randomization Zidovudine alone: zidovudine plus Zidovudine-based ART: zidovudine. and HBV coinfection were assigned only to zidovu. Boehringer Ingelheim. with dosing based on birth weight. at postpartum week 1. at delivery. All women who met specific treat- ment guidelines and all infants who were confirmed to be infected with the human immunodeficiency virus (HIV) began ART immediately. Copyright © 2016 Massachusetts Medical Society. Trial Procedures sign. at a dose of 400 mg of tenofovir. with increased data on tenofovir chain-reaction assay in the ART groups and by 1728 n engl j med 375. to any of the three regimens. once daily 200 mg of emtricitabine and 300 mg Lopinavir–ritonavir.0 of the trial protocol (period 1). and lopinavir–ritonavir emtricitabine. Randomization Schema and Drug Regimens during the Antepartum Component of the Trial. Gilead Sciences.org on January 8.0 (period 2). For personal use only. only women who were positive for hepatitis B surface antigen could be randomly assigned to tenofovir-based antiretroviral therapy (ART). chemical values (alanine aminotransferase and signed to tenofovir-based ART were to be limited creatinine levels and calculated creatinine clear- to a nested HBV substudy. Under version 2. and at postpartum week 1 benefit was felt to outweigh risk in that group. plan to compare safety across all three groups tractor to the Eunice Kennedy Shriver National for participants enrolled during period 2 (Octo- Institute of Child Health and Human Develop. twice daily of tenofovir.org  November 3. 300 mg twice daily Zidovudine–lamivudine. and lopinavir–ritonavir of tenofovir and emtricitabine Zidovudine. or manuscript Screening included confirmation of maternal preparation. with the exception women regardless of HBV status to be assigned that some nevirapine was purchased from Boeh. single-dose nevirapine. ment for use in the trial. owing to limited safety data on tenofovir in pregnancy. During period 1 (April ance) were obtained at trial entry. data collection. 2011 through September 2012). . Mater- safety data on tenofovir in pregnancy. and ViiV in pregnancy. there were limited and HBV surface antigen (HBsAg) level. centration (quantified by real-time polymerase- in October 2012. The plasma HIV RNA con- dine alone or zidovudine-based ART. once daily for 6–14 days of lopinavir and 100 mg of ritonavir. No other uses without permission. at weeks 2. after delivery twice daily twice daily Lopinavir–ritonavir increased to 600 mg Lopinavir–ritonavir increased to 600 mg of lopinavir and 150 mg of ritonavir of lopinavir and 150 mg of ritonavir twice daily during third trimester twice daily during third trimester Figure 1. Tenofovir-based ART: tenofovir. data analyses. A complete blood count and blood comparative analyses including the group as. Infants of women coinfected with hepatitis B and HIV received hepatitis B immune globu- lin and hepatitis B vaccine at birth and the full hepatitis B vaccine series. all women could be assigned to any of the three regimens. HIV status and measurement of the CD4 count When the trial began. followed by "tail" lamivudine. at weeks 4.18 nejm. with an analytic ringer Ingelheim at a discounted rate by a con. 2017. However. 4. 2016 The New England Journal of Medicine Downloaded from nejm. 2014).

a preterm delivery analyses (including HIV-free survival through before 37 weeks of gestation (according to the week 1) were to include pairwise comparisons of Ballard examination.18 nejm. and 95% power to were composite. We calculated that Diseases. and delivery or postpartum week 1. Antiretrovir al Ther apy for Perinatal HIV Prevention batch testing of stored samples in the zidovu. was defined as a positive infant HIV of the data and analyses and for the fidelity of nucleic-acid-testing result on the birth or week 1 the trial to the protocol. allowing for 10% loss to follow-up and grade adverse events.similar. Copyright © 2016 Massachusetts Medical Society. The PROMISE trial was designed by the protocol week 4. ing the lead. For preg. 2016 1729 The New England Journal of Medicine Downloaded from nejm. Chan School of Public Health analyzed the Full details are provided in the protocol. stillbirth (≥20 weeks of gestation). the composite for safety outcomes. or grade 3 or higher signs and the individual participant (mother or infant) or symptoms through week 1 post partum. T.org on January 8. 2017. For infants.all three groups. safety weight of less than 2500 g. For mothers. when available. All rights reserved. . n engl j med 375. with secondary analysis of indi. grade 3 or on an intention-to-treat basis. data according to the agreed-on statistical plans and with input from the protocol team and the Efficacy and Safety Outcomes data and safety monitoring board. Infant visits. the primary efficacy thors vouch for the completeness and accuracy outcome. with a 5% (two-sided) type I Adult and Pediatric Adverse Events was used to error. No other uses without permission. Infant HIV-free survival. spontaneous abortion. delivery. The unit of analy- higher hematologic abnormalities or abnormal sis was the mother–infant set for efficacy analyses blood chemical values. For multiple-gestation births. 3400 from breast-feeding set- Program of the Division of AIDS (DAIDS) of the tings and 1000 from formula-feeding settings. National Institute of Allergy and Infectious to be combined for analysis. the composite out. Severe adverse pregnancy outcomes zidovudine-alone group or the group assigned included a very low birth weight of less than to zidovudine-based ART were restricted to 1500 g. confirmed by a second positive result on a separate specimen obtained on a different Statistical Analysis day (as soon as possible after the first positive Randomization was stratified according to HBV result). the composite outcome was a low birth expected to differ between ART groups. vudine-alone group versus 2% in the two ART The DAIDS Table for Grading the Severity of groups combined.H. cal estimate). The team. or obstetri. Pairwise safety comparisons of the Program of the Centers for Disease Control and group assigned to tenofovir-based ART with the Prevention). group and the group assigned to zidovudine- or a congenital anomaly (according to the criteria based ART were based on data from both period 1 of the Metropolitan Atlanta Congenital Defects and period 2. including the International Maternal Pediatric Adolescent HIV nucleic acid testing and measurement of AIDS Clinical Trials Network (IMPAACT) Statis- safety laboratory values. outcome was grade 2 or higher hematologic ab. or a major con. occurred at birth (to tical and Data Analysis Center at the Harvard 5 days of age) and week 1 (6 to 14 days of age). All the au- Early infant HIV infection. was defined as an infant being more than 90% power to detect a cumulative alive and not HIV-infected through the week 1 early transmission difference of 4% in the zido­ visit. tests were performed in laboratories that status and country. The target sample was 4400 were certified by the Virology Quality Assurance mother–infant sets. for safety analyses.when there was concurrent randomization to genital anomaly. because adverse events were nancy. specimen. with the first author and the cochairs tak- CD4 count was obtained at week 12.7 Primary safety analyses two interim efficacy analyses. stillbirth. spontaneous abortion (<20 weeks Safety comparisons of the zidovudine-alone of gestation).all three groups. Trial Roles dine-alone group) was measured at trial entry.detect differences as small as 25% versus 20% vidual components. a very preterm delivery before 34 weeks. However. All analyses were conducted come was death from any cause. For personal use only.org  November 3. and postpartum week 1. The a priori analysis plan required combining normalities or abnormal blood chemical values the results of the two ART groups for efficacy or grade 3 or higher signs or symptoms during analyses because efficacy was expected to be pregnancy through week 1 post partum. a secondary this sample size would provide the trial with safety outcome. The clinical sites gathered the data. data from participants enrolled during period 2.

9% vs.1% vs. we enrolled week 1 was significantly higher (i. repeated R e sult s confidence interval. If the repeated confidence interval ex.org  November 3. ing zidovudine alone during the current preg- tial repeated confidence interval for the difference nancy before enrollment (19% vs. and asymptomatic (97% had World Health type I error spending function (for this interim Organization [WHO] had clinical stage 1 HIV analysis.4. The numbers vations in the alanine aminotransferase level) and flow of participants in the trial are shown in (5. tenofovir-based ART (period 2) (Fig. infection). 2017. 17. 26 Lan–DeMets approach with an O’Brien–Fleming years).5% vs. 2A) or 3 were determined not to be pregnant at enroll. with interaction tests assessing hetero. 2 had molar pregnancies.3%.26). no significant Baseline Maternal Characteristics heterogeneity in subgroup analyses) (Table 2). and 3% of of 5%). for the Supplementary Appendix provides further infant safety analyses. P = 0. 2016 The New England Journal of Medicine Downloaded from nejm.008) and a higher rate of grade 2 or higher including 57 multiple births with at least 1 live. 0.18 nejm. the confidence coefficient was 96. Copyright © 2016 Massachusetts Medical Society. 2014 (the The rate of infant HIV-free survival through cutoff date for the interim analysis). Table S1 in HIV infection in any of the live-born siblings. All rights reserved. (through Week 1 after Delivery) formed.. and Table S2 Figure S2A (randomization to zidovudine alone or in the Supplementary Appendix). Baseline maternal characteristics were well but there was no significant difference between balanced between the zidovudine-alone group and the two ART groups (P = 0. Uganda) and in the percentage of women receiv- Interim efficacy analyses used a group-sequen. . There receiving zidovudine alone (21.1 to −0. Most in the early transmission rate with the use of the women were African. the rate of 3529 mother–infant sets (3245 in breast-feeding early transmission or death was significantly settings and 284 in formula-feeding settings). significantly higher rate of grade 2 or higher zidovudine-based ART. 2261 (65%) were enrolled during pe. 1. Africa and during period 2 in Zimbabwe and tical significance for all analyses except efficacy.5%.05 was considered to indicate statis. There were differ- Categorical variables were compared between ences between periods 1 and 2 in country-specif- groups with Fisher’s exact test.e. of any grade 2 or higher adverse event than those riod 1 and 1229 (35%) during period 2. P<0. At screening. ment. −1. During period zidovudine-based ART in periods 1 and 2) and 2. lower) with zidovudine-based ART than with A total of 39 women were excluded from analyses: zidovudine alone (periods 1 and 2) (Fig. lower in the combined maternal ART groups than in the zidovudine-alone group (0.8%. A two-sided P value ic enrollment (higher during period 1 in South of less than 0. One prespecified subgroup analysis Efficacy and HIV-free Survival Results and two post hoc subgroup analyses were per. Grade 2 or higher the group assigned to zidovudine-based ART elevations in the creatinine level occurred in less (periods 1 and 2) and between the group assigned than 1% of women in all groups. difference. the median CD4 count to preserve an experiment-wise type I error rate was 530 cells per cubic millimeter. this would indicate a significant difference. 1. From April 2011 through September 10. the women were HBsAg-positive. 2B). or tenofovir-based ART abnormal blood chemical values than those re- in period 2 only) in the Supplementary Appendix. −2.001) (Table 3. For personal use only.3%. abnormal blood chemical values (primarily ele- born infant (24 during period 2). women receiving zido­ Of the 3490 mother–infant sets included in vudine-based ART had a significantly higher rate analyses. The n e w e ng l a n d j o u r na l of m e dic i n e perinatal transmission was defined as confirmed two trial groups (period 2) (Table 1. women receiving tenofovir-based ART had a Figure S2B (randomization to zidovudine alone.8% vs. and 34 were coinfected with HIV and HBV and were randomly Safety Results in the Antepartum Component assigned to receive tenofovir-based ART during Maternal Adverse Events period 1.03). ceiving zidovudine alone (2. fication according to group). cluded zero. P = 0.3 percentage points. each live-born sibling was details on CD4 cell count and viral-load quanti- included separately. The rate of early transmission was significantly geneity of treatment effects. No other uses without permission. young (median age. 28%). were 3202 live births (1045 during period 2). During periods 1 and 2.8%.org on January 8. Table S3 in the to tenofovir-based ART and each of the other Supplementary Appendix provides additional de- 1730 n engl j med 375.

and Tanzania. For personal use only.8 3.3–4. Malawi. and lopinavir–ritonavir./total no.2–4.5) 3/3084 (<0.5) * Between-group differences were not assessed for statistical significance. (%)‡ Gestational age — wk Median 26 25 26 26 26 26 26 Antiretrovir al Ther apy for Perinatal HIV Prevention IQR 21–30 21–30 21–30 21–31 21–31 22–31 21–31 Copyright © 2016 Massachusetts Medical Society.* Characteristic Periods 1 and 2 Period 2 Only ZDV Alone ZDV-Based ART Total ZDV Alone ZDV-Based ART TDF-Based ART Total (N = 1543) (N = 1541) (N = 3084) (N = 413) (N = 410) (N = 406) (N = 1229) Age — yr Median 26 26 26 25 26 26 26 IQR 22–30 23–30 22–30 22–29 23–30 22–30 22–30 Race or ethnic group — no. lamivudine.5) 0 2 (<0.5 3. ART denotes antiretroviral therapy.org  November 3. emtricita­bine. IQR interquartile range.5) 0 0 0 0 Weight — kg Median 64 65 64 63 64 64 64 IQR 57–74 58–74 58–74 57–71 58–73 58–76 58–73 CD4 count — cells/mm3 Median 533 526 530 530 540 537 536 IQR 433–678 439–650 436–663 431–687 451–657 432–691 436–680 Viral load at enrollment — log10 copies/ml Median 3.4 3.2–4. and ZDV-based ART zidovudine. Region or country — no.9 IQR 3. (%)† Black African 1495/1543 (97) 1494/1541 (97) 2989/3084 (97) 412/412 (100) 409/409 (100) 405/405 (100) 1226/1226 (100) Indian 46/1543 (3) 46/1541 (3) 92/3084 (3) 0 0 0 0 Other 2/1543 (<0.4 3. No other uses without permission. All rights reserved. Southern Africa 267 (17) 267 (17) 534 (17) 111 (27) 111 (27) 110 (27) 332 (27) India 46 (3) 46 (3) 92 (3) 1 (<0. ZDV zidovudine. TDF tenofovir. (%)§ East Africa 718 (47) 719 (47) 1437 (47) 232 (56) 231 (56) 229 (56) 692 (56) South Africa 512 (33) 509 (33) 1021 (33) 69 (17) 67 (16) 67 (17) 203 (17) Downloaded from nejm.8 3./total no. ZDV alone zidovudine plus single-dose nevirapine. 2017. § East Africa includes Uganda. Baseline Maternal Characteristics at Entry to the Antepartum Component of the Trial.5) 1 (<0.4 The New England Journal of Medicine n engl j med 375.9 3. TDF-based ART tenofovir.9 3.2–4.3–4. and lopinavir–ritonavir.4 3.org on January 8. † Race or ethnic group was self-reported.2–4.8 3.18 nejm. Table 1. ‡ Persons with World Health Organization (WHO) clinical stage 1 human immunodeficiency virus (HIV) infection are asymptomatic.9 3.2–4. 1731 .4 3.5) 1/1541 (<0.5 3. and Southern Africa indcludes Zimbabwe and Zambia. 2016 WHO clinical stage 1 — 1483/1535 (97) 1498/1534 (98) 2981/3069 (97) 394/408 (97) 396/407 (97) 394/401 (98) 1184/1216 (97) no.

During period 2.4) Maternal gestational age at trial entry† 0. P = 0. no significant differences between teno­ vudine-based ART had significantly higher rates fovir-based ART and zidovudine alone were ob- than women receiving zidovudine alone of ad. No other uses without permission. served for composite or individual outcomes. How- verse pregnancy outcomes (40.1%.04) and a low birth weight of less alone zidovudine plus single-dose nevirapine. and HIV human immunodeficiency virus. composite or individual outcomes between women receiving zidovudine-based ART and those receiv- Adverse Pregnancy Outcomes ing zidovudine alone (periods 1 and 2).0% vs. Infant HIV Infection through Week 1 (Periods 1 and 2 Combined) in All Mother–Infant Sets and According to Subgroup. tails on signs.org  November 3. and ZDV- than 2500 g (16. a low birth weight of less than 2500 g significantly higher rates than those receiving (23.7% vs. The n e w e ng l a n d j o u r na l of m e dic i n e Table 2.04). 27. 8. dix provides additional details on the combined P = 0. TDF-based ART tenofovir.7 (−2. Probability of Infant HIV Infection weight and very preterm delivery).9%.68 <34 wk 16/1229 (1.5) −0.004).7) −1. but there based ART zidovudine. P = 0. preterm delivery before 34 weeks (6. emtric- itabine.4) 1/189 (0.70 350–499 cells/mm3 16/577 (2. † Data on maternal gestational age at trial entry were missing for one woman in the group assigned to ZDV-based ART. CI denotes confidence interval. No significant were no maternal deaths.8) 4/592 (0. zidovudine alone of adverse outcomes (34. Table S2 in the Supplementary Appen.2% vs. the other two subgroup analyses were post hoc analyses.1) 3/793 (0.001).3) 6/1230 (0. in the group assigned to zidovudine- outcome of low birth weight and preterm deliv- ery and the combined outcome of very low birth Figure 2 (facing page). 4% the three groups (Table S4 in the Supplementary of the women (120 of 3248) discontinued their Appendix lists congenital anomalies). and diagnoses. P<0.1) ≥34 wk 9/157 (5.5) ≥500 cells/mm3 9/809 (1. (%) percentage points (repeated CI) All mother–infant sets 25/1386 (1.2) Maternal viral load at trial entry 0.7) −2.9 to −0. and lopinavir–ritonavir.0) ≥1000 copies/ml 25/1083 (2.0) −4. ZDV zidovudine. 2017. of mother–infant sets/total no. and lopinavir–ritonavir.5) 2/268 (0. 2016 The New England Journal of Medicine Downloaded from nejm. ZDV Alone Interaction no. ZDV 27. P = 0.7) 1/136 (0.3) 6/1129 (0. symptoms. assigned antiretroviral regimen prematurely.6) 1/51 (2. women receiving zido­ period 2.6) −1. lamivudine.8) 7/1385 (0.0% vs.8 to −0. All rights reserved.8 (−8.7) Missing data 4 3 0 * The analysis of infant HIV infection according to maternal gestational age at trial entry was a prespecified analysis. or Death through the Week 1 Postpartum Visit (6 to 14 women receiving tenofovir-based ART had sig.0%.6) Maternal CD4 count at trial entry 0.7 to −0.3%.5) 1/274 (0. For personal use only.02) and very (Table 3. and With respect to severe pregnancy outcomes.4) 0/57 0.0% vs. was no significant difference for any outcome 1732 n engl j med 375.6%.18 nejm. P<0.1 to −0.22 <1000 copies/ml 0/299 1/253 (0.7) 1/154 (0. ZDV-Based ART and P Value for Subgroup Alone ART ART TDF-Based ART vs.9% vs. 0. . There between the two ART groups.2%.6 to −0. 12. women receiving tenofovir-based ART had P<0. TDF denotes tenofovir.1 (−3. 4.5%. Copyright © 2016 Massachusetts Medical Society.7% had a discontinuation that was assessed as no significant differences were observed for the being possibly related to a trial drug.001). 2. The rate of trial-drug differences in stillbirth or spontaneous abortion discontinuation was low (2 to 5%) and did not and congenital anomalies were observed among differ significantly among the three groups.4 to 1.5% vs.org on January 8. During During periods 1 and 2.4) −0.001) cy outcomes (9.5) 2/325 (0. nificantly higher rates than women receiving The insets show the same data on an expanded y axis. Days after Delivery). 13. ever.3 (−0.7 (−1.6 to 0. and preterm delivery zidovudine-based ART of severe adverse pregnan- before 37 weeks (20.* ZDV ZDV-Based TDF-Based Difference.3 (−2.8 (−1.

of Mother– Death or HIV No Event Infant Sets Infection (censored) 1. TDF-based ART: P=0.0 ZDV Alone 341 15 326 ZDV-Based ART 346 3 343 TDF-Based ART 335 16 319 0.2 0.08 0.3 0.01 0.8 0. No other uses without permission.07 ZDV alone vs.5 ZDV alone 0.06 P=0.9 0.8 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days after Delivery No.07 0.org  November 3.2 0. All rights reserved.00 0.0 ZDV Alone 1411 52 1359 ZDV-Based ART 1406 23 1383 0. Antiretrovir al Ther apy for Perinatal HIV Prevention A Periods 1 and 2 No.09 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Days after Delivery No.001 by log-rank test 0.10 Estimated Probability of HIV Infection or Death 0.7 0.05 ZDV alone 0.84 by log-rank test 0.18 nejm.4 0.03 ZDV-based ART 0.05 0.04 0.1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0.10 Estimated Probability of HIV Infection or Death 0. of Mother– Death or HIV No Event Infant Sets Infection (censored) 1.6 ZDV-based ART vs.6 0. 2017. of Mother–Infant Sets at Risk ZDV alone 1411 1400 1386 1372 1369 1368 1366 1365 1295 1141 964 730 548 436 387 ZDV-based ART 1406 1401 1399 1396 1392 1390 1386 1383 1322 1187 1018 778 606 469 423 B Period 2 No.9 0.06 0.08 TDF-based ART 0.5 0. of Mother–Infant Sets at Risk ZDV alone 341 336 334 332 331 330 330 329 322 293 259 201 146 111 78 ZDV-based ART 346 346 346 345 344 344 344 342 331 318 285 224 191 139 103 TDF-based ART 335 333 327 325 323 323 323 322 313 292 257 205 155 117 82 n engl j med 375.1 0.01 0.09 0.00 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0.002 by log-rank test 0.02 ZDV-based ART 0. 2016 1733 The New England Journal of Medicine Downloaded from nejm.7 0.org on January 8. For personal use only.04 0. Copyright © 2016 Massachusetts Medical Society. .03 0.4 0.02 0.3 0. TDF-based ART: P=0.

30 Preterm delivery: <37 wk 46/341 (13.2) 2/346 (0.2) 9/346 (2.1) — 0.0) 61/385 (15.3) 29/314 (9.org on January 8.001 * For more details on maternal and infant safety. § Severe adverse pregnancy outcomes included a very low birth weight of less than 1500 g.03 0. a very preterm delivery before 34 weeks.09 0. ZDV-based ART Any severe adverse outcome§ 83/1399 (5. ZDV Alone vs. TDF- Alone ART ART ART ART Based ART number/total number (percent) Maternal adverse events Periods 1 and 2: ZDV alone vs. a preterm delivery before 37 weeks of gestation (according to the Ballard examination.13 Period 2 only: all three groups Any severe adverse outcome§ 22/329 (6.5) 563/1407 (40. spontane- ous abortion.25 0. 2016 The New England Journal of Medicine Downloaded from nejm.43 0. 2017.8) 0. ZDV-Based ZDV ZDV-Based TDF-Based ZDV-Based TDF-Based ART vs. ‡ The composite outcome was a low birth weight of less than 2500 g.26 chemical value Adverse pregnancy outcomes Periods 1 and 2: ZDV alone vs.001 chemical value Period 2 only: all three groups Any grade ≥2 adverse event† 59/393 (15.6) 20/335 (6.7) 11/380 (2.004 0.02 Very preterm delivery: <34 wk 11/341 (3. stillbirth.8) 60/380 (15.0) 0.6) 44/1406 (3.9) 65/319 (20.3) 318/1505 (21.7) 0. 80% of very preterm deliveries occurred Infant Safety Outcomes during period 1. † The composite outcome was grade 2 or higher hematologic abnormalities or abnormal blood chemical values or grade 3 or higher signs or symptoms during pregnancy through week 1 post partum.2) 0. .1) — 0.4) 0. or a congenital anomaly (according to the criteria of the Metropolitan Atlanta Congenital Defects Program of the Centers for Disease Control and Prevention). when available.10 0. with only 20% during period 2. ZDV-based ART Any grade ≥2 adverse event† 261/1510 (17.001 Preterm delivery: <37 wk 185/1411 (13.43 Infant deaths through wk 1 28/1432 (2.4) 51/301 (16. stillbirth (≥20 weeks of gestation).7) 62/335 (18. or obstetrical estimate).2) — 0. The n e w e ng l a n d j o u r na l of m e dic i n e Table 3. For personal use only.7) 14/322 (4. see Tables S2 through S7 in the Supplementary Appendix.22 Very preterm delivery: <34 wk 37/1411 (2.1) 288/1406 (20.77 >0.001 Low birth weight: <2500 g 161/1347 (12.9) 99/1385 (7.99 Grade ≥2 abnormal blood 3/392 (0. All rights reserved.8) — <0. through Week 1 Post Partum.0) — <0.5) — <0.6) 15/341 (4.04 0.3) 88/1505 (5.5) 111/320 (34. spontaneous abortion (<20 weeks of gestation).001 Period 2: all three groups Any adverse outcome‡ 91/334 (27.8) 18/385 (4.46 Low birth weight: <2500 g 28/315 (8.04 Infant deaths through wk 1 11/349 (3.9) 0.org  November 3.18 nejm.1) — 0. Maternal Safety and Pregnancy Outcomes.77 Severe adverse pregnancy outcomes Periods 1 and 2: ZDV alone vs.0) 306/1332 (23. No other uses without permission.* Outcome Antepartum Randomization Group P Value ZDV Alone vs. including Infant Deaths. Copyright © 2016 Massachusetts Medical Society. Among 3202 live births.5) 0.2) 123/328 (37. ZDV-based ART Any adverse outcome‡ 389/1414 (27.5) 68/346 (19. or a major congenital anomaly. based ART.0) 17/1419 (1.9) 0.008 Grade ≥2 abnormal blood 19/1510 (1.0) — <0. no significant differ- ences were observed in pairwise comparisons of the groups for composite or individual infant 1734 n engl j med 375.

This suggests there could be some ART (period 2) than with zidovudine alone. Table S7 in the Supplemen. ART groups is unclear. 1. P = 0. However. 0. of 17 vudine-based ART (periods 1 and 2) or the group deaths in the group assigned to zidovudine-based assigned to tenofovir-based ART (period 2). All rights reserved. resulted in a significantly higher rate of HIV-free survival among infants in the group assigned to zidovudine-based ART than among those in the Discussion group assigned to tenofovir-based ART during The PROMISE trial showed the efficacy of triple.18 nejm. A post significantly between the two ART groups during hoc analysis suggested that mortality by week 1 period 2 (2. The reason for differences in very preterm mission rate of less than 2%. but no significant Tables S5 and S6 in the Supplementary Appendix). No signifi. no significant difference between the two very preterm delivery and infant death in the ART regimens was observed. Specifically. for CD4 count and showed higher rates of pre- cant difference in infant deaths was observed term delivery and low birth weight among women between the group assigned to zidovudine-based initiating ART (primarily nevirapine-based) dur- ART and the zidovudine-alone group (periods 1 ing pregnancy than among those starting zido­ and 2) (1.04).6% vs.5% vs. Reassuringly. respectively. P<0.2%. How- ART.4% and 3. 88% (15 deaths) occurred during period 1 ever. 4. For personal use only.org  November 3. . the rate of very preterm delivery based ART had significantly more infant deaths did not differ significantly between the zidovu- than the group assigned to zidovudine-based dine-alone group and the group assigned to zido­ ART (4. 40 of 101 very preterm deliveries (40%) with higher rates of adverse events than zidovu. pharmacokinetic interaction between lopinavir– ods 1 and 2) and tenofovir-based ART (period 2) ritonavir and tenofovir. zidovudine-based ART. but the group assigned to tenofovir. as did with lower mortality rates among those born at early infant mortality (correlated with very pre- a gestational age of 34 to less than 37 weeks term delivery) (0. Antiretrovir al Ther apy for Perinatal HIV Prevention adverse outcomes other than death (Table 3. P<0.6% with zidovudine-based ART vs. Copyright © 2016 Massachusetts Medical Society. mortality.8-11 P = 0. period 2 (but no significant difference between drug ART for preventing transmission among infants in the group assigned to tenofovir-based HIV-infected pregnant women with high CD4 ART and those in the zidovudine-alone group).4%.8%). However. ART (primarily with nevirapine) showed no sig- ation due to toxic effects were low. This (3. 2017. vudine plus single-dose nevirapine. and there were no based ART (with efavirenz) or zidovudine-based maternal deaths. nificant differences in preterm delivery or still- Higher rates of adverse pregnancy outcomes birth.9% of live births) observed. Comparisons are limited sion than zidovudine alone (0.0%.6%. liveries and 88% of infant deaths occurred dur- based ART (periods 1 and 2) and tenofovir-based ing period 1.5%). ity and increased newborn or later postnatal dine-alone group (4.001). Our results are consistent with those by week 1 (Table 3. ing period 2. All three regimens studied had a trans. decreased renal clearance n engl j med 375. counts. but the ART regi. was highest among very preterm infants (27. P = 0.5. unknown confounder resulting in lower rates of ever. and 28 of 60 infant deaths (47%) occurred dur- dine alone. there group assigned to zidovudine-based ART during were no increases in hematologic toxic effects period 2. 2016 1735 The New England Journal of Medicine Downloaded from nejm.org on January 8. the rate of very preterm delivery differed and only 12% (2 deaths) during period 2. to concurrent randomization during period 2. However.2% and 2. 6.3%) and those born at term (0. 80% of very preterm de- malities were more common with zidovudine.13). both ART regimens were associated Overall. difference between the two ART regimens was There were 60 infant deaths (1. and than with zidovudine alone. No other uses without permission. of observational studies in Africa that controlled tary Appendix lists reported causes). A study in Botswana that compared with zidovudine-based ART or renal toxic effects birth outcomes in women receiving tenofovir- with tenofovir-based ART. delivery and infant mortality between the two mens had significantly lower rates of transmis. preterm delivery and low birth infant deaths was observed between the group weight can be associated with increased morbid- assigned to tenofovir-based ART and the zidovu.6 In resource- During period 2.12 A potential biologic cause may be a were observed with zidovudine-based ART (peri.0% with tenofovir-based ART.43). respectively.4% vs. Rates of trial-drug discontinu.0%). in the group assigned to Maternal grade 2 or higher laboratory abnor. How.001). no significant difference in limited settings.

D. M. and Mrs. Patricia M.N. Renate Strehlau.D.... Susan A.. Harvard T. the statistical staff at the regimens in the PROMISE trial.. For personal use only. George K. Disclosure forms provided by the authors are available with whether the current WHO-recommended ART the full text of this article at NEJM. main to be defined.L. Anne S. Klingman. as compared tions to the PROMISE 1077 trial and HIV–AIDS research. Ph. M. in grateful memory of their many contribu- perior efficacy of triple-drug ART. Harvard T. for lular levels. Pendo Mlay. of AIDS.K.. Avy Violari. Boston (M.22-25 However.. The authors’ affiliations are as follows: the Department of Pathology.D..C.E.).. M.. Stephen Lagakos (Center for Biostatistics in nancy outcomes than zidovudine-based ART..P.P. tor–based ART and preterm delivery. Pharm.13-15 In in HIV-infected women with high CD4 cell counts the PROMISE trial. M. Los Angeles (J. the staff at the IMPAACT central laboratory at the Univer- birth weight or preterm delivery than the ART sity of North Carolina at Chapel Hill. M. the other available nonnucleoside need for continued research to assess ART in reverse-transcriptase inhibitor (NNRTI).. and we cannot comment on to this article was reported. Judith S.D. TN (P. Ed.).S.H. other tious Diseases of the NIH under award numbers UM1AI068632 (IMPAACT LOC)..G. Edward Handelsman (Division not associated with higher rates of adverse preg. Johns Hopkins University Bloomberg School of Public Health (T. M. M.. Martinson.D..D.D. search Foundation). Taha.). M.org  November 3.. College of Health Sciences.E.S. Currier.).18-21 Some studies have shown that protease inhibitors re..S.G. 2016 The New England Journal of Medicine Downloaded from nejm.).D.D.H.D. the WHO recom- pharmacokinetic studies showed decreased lopi. Dhayendre Moodley. National Institutes of Health. Chan School of Public Health..B. Ph.H. and the Division of AIDS. Chi.. Ch. Ch.. Harare 1736 n engl j med 375. the lopinavir–ritonavir dose but also showed higher rates of adverse events.S. K. St. Durham (A.D. Chan Botswana data are reassuring in that the combina. Fowler. Copyright © 2016 Massachusetts Medical Society.H. Linda Millar (Frontier Science and Technology Re- In conclusion. as well as the PROMISE staff at the clinical research sites. University of North Carolina School of Medicine. Dr.B. the Center for Biostatistics in AIDS Research. Fiscus. B. M. all paid to compare protease inhibitor–based ART with to her institution. Shapiro.D.D. the Center for Biostatistics in AIDS Research. 2017.B. and the data-management staff at the Frontier Science and Technology Research Foundation... M.M. Baltimore.26. regimen of tenofovir. .C. Ger- hard B..B. Ch. M.C. Ramesh Bhosale.3.). tion of tenofovir.D....D. the prevention of mother-to-child transmission ported with concomitant administration. UM1AI068616 (IMPAACT SDMC). M.B.O. Loftis. M. Harvard T..P. and Lynne M.. Ph.P. There are infected women and healthier outcomes for their inconsistent findings regarding protease inhibi.D. Chapel Hill (S. and there are clear benefits of pregnancy.F.L. A. M.17 ART for the prevention of mother-to-child trans- The ART regimens in the PROMISE trial were mission and maternal health. David E. and studies have shown that nevirapine-based ART is UM1AI106716 (IMPAACT LC). D.org. and FHI 360. B. Devasena Gnanashanmugam. M. Taha E.E. have been re.. and efavirenz We thank the mothers and their infants who participated in would be associated with a lower risk of low the trial.). particularly in women.. Ch. the UCLA Center for Clinical AIDS Research and Education. National Institute of Allergy and Infectious Diseases (D. emtricitabine. R.B. M. Francis E.5.. Tsungai Chipato. and efavirenz was We dedicate this article to Dr.) — both in North Carolina.12 AIDS Research.. Johns Hopkins University School of Medicine (M.B. Violari reports receiving consulting fees from Janssen and rapine.)..C. the Retrovirology Core Labora- tory. Ph. Bethesda — all in Maryland. it is protease inhibitor–based because nevirapine-based also clear that the most efficacious and safest ART was contraindicated in women with a CD4 triple-drug ART regimens during pregnancy re- count of more than 250 cells per cubic millimeter. with cofunding from the Eunice also associated with a higher rate of preterm Kennedy Shriver National Institute of Child Health and Human delivery than zidovudine plus single-dose nevi­ Development and the National Institute of Mental Health.B.D. Siberry.6 The PROMISE trial was not designed grant support from Gilead Sciences and ViiV Healthcare. was increased during the third trimester because On the basis of recent trials. Terence Fenton.. and the Department of Epidemiology. James McIntyre..D. emtricitabine.H.Q.F.S. M. The n e w e ng l a n d j o u r na l of m e dic i n e of tenofovir and increased plasma and intracel.G. No other potential conflict of interest relevant NNRTI-based ART. Ph. Ph.D.N. University of Zimbabwe.S.. Browning. M. No other uses without permission. Our findings emphasize the and efavirenz..D.S.T. Benjamin H. L... Flynn. The views expressed in this article are solely those of the au- thors and do not necessarily represent the official views of the duce progesterone production in the trophoblast National Institutes of Health (NIH).A. However. the PROMISE trial showed su. Theron.R.D. M. Coletti.. F.B. R. All rights reserved...A. Lynette Purdue.F. M.18 nejm. Chan School of Public Health)..F. M.K.B.H. Jude Children’s Research Hospital. the Department of Obstetrics and Gynecology. Dr. Ch. Renee Browning. Min Qin. School of Public Health. Maxensia Owor. NIH). Karin L. Memphis. uninfected infants. M. M. pregnancy to ensure safer pregnancies for HIV- traindicated in pregnancy at that time. mends ART for HIV-infected persons regardless navir–ritonavir levels with standard doses in late of CD4 cell count.org on January 8. Appendix The authors’ full names and academic degrees are as follows: Mary G. Supported by the National Institute of Allergy and Infec- sociated with preterm delivery.16. M. M..D.27 However. Raziya Bobat. Mofenson. Michael Basar. T. in vitro and that low progesterone levels are as.. Amy J. with zidovudine plus single-dose nevirapine. was con. and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (G.J..

Thom 7. fumarate on coadministration with lopi. N Engl analysis for 2010 with time trends since J Infect Dis 2011. version 16. and the Centre for the AIDS Programme of Research in South Africa–Umlazi Clinical Research Site (D. Washington.​207:​612. Powis KM.​331:​1173-80. Marchant T. Papp E. Stek AM.who​ Pilot pharmacokinetic study of human progesterone the missing piece to the . et al. Re­ nancy. Spiegelman D. HIV protease inhibitor use during 1051-4. Sun J. 21. assuring birth outcomes with tenofovir/ 22. PLoS Med preterm birth. J Acquir Immune Defic laxis for HIV.E. Kacanek D. regional. Kearney BP.​211:​4-7. and national causes of use during pregnancy and risk of preterm mission of human immunodeficiency virus child mortality: an updated systematic delivery: more questions than answers. Recommendations for tional-age infants in low-income and apy during pregnancy: role of the ritona- use of antiretroviral drugs in pregnant middle-income countries: a pooled coun.int/​hiv/​pub/​guidelines/​earlyrelease-arv/​ weight for gestational age: individual par. 206:​1695-705. Clin Pharmacol Ther 2008.​213:​ emtricitabine/efavirenz used for preven. Cousens S. 2016 (http://aidsinfo​ mortality levels and trends by HIV status HIV-infected women randomized to pro- . Kampala. AIDS 2006. J Acquir Immune Defic 40. and the Department of Obstetrics and Gynaecology. et al. Kumwenda J. 2016 1737 The New England Journal of Medicine Downloaded from nejm. 10-8. Geneva: World Health Or- Neonatal mortality risk associated with Syndr 2008. . Pune.) and Durban Paediatric HIV (R.nih​. tion and spontaneous preterm birth. 26. Amherst.​20:​1931-9. 2010 (http://whqlibdoc​.S. in infants: recommendations for a public Syndr 2006.J. hibitors and adverse birth outcomes: is Organization. Cheng AK. et al. Malawi (F. DC (L. J Infect Dis 2016. J Infect Dis 2013. Willey B.pdf?ua=1). Venkataramanan R.​ The effect of lopinavir/ritonavir on the re. . growth restriction. Kilimanjaro Christian Medical Center–Duke University Collaboration.C. et al. Pruvost A. Powis KM.e1-6. Papp E. Uganda (M.​213:​1057. E. Stellenbosch University (G. Copyright © 2016 Massachusetts Medical Society. Relationship between 17-alpha Services. J Acquir tease versus nucleoside reverse transcrip- GL​. Connor EM. Rock. the Department of Obstetrics and Gynecology. Kitch D.com/ et al. et al.tech-res. University of Cape Town (J. et al.). Kiser JJ. University of the Witwatersrand. Liu L. Faculty of Health Sciences (A. Antiretroviral drugs for treating preg. Lilongwe.org  November 3. Negredo E. Souda S. Medical College. Cape Town. 2012.).​ Obstet Gynecol 2014. and Empilweni Services and Research Unit (R.M.​204:​506-14. J Infect Dis 2016. Moshi. kinetics and safety of tenofovir disoproxil inhibitor-based combination antiretrovi- 6.M. Ribaudo HJ. Mor. and adverse pregnancy outcome: the ele. the University of North Carolina Project. J Infect Dis 2015. Kamuzu Central Hospital–Tidziwe Center. et al.​ 83:​ Low prolactin and high 20-α-hydroxy­ Antiretroviral therapy in relation to birth 265-72. Aquilante CL. MD:​AIDSinfo. Chen JY. Mittan A. Geneva: World Health 15. Li N. Antimicrob Agents Chemother 2009. Katz J. Bobat).org on January 8.​210(2):​128. Chen JY. Best BM.C. 2. Protease in- health approach. grading_severity_of_adult_pediatric_ 17. Rahman 21. Mirochnick M. Man- dela School of Medicine. Immune Defic Syndr 2012. tase inhibitor-based HAART during preg- 3. No other uses without permission. Tanzania (P. tality risk in preterm and small-for-gesta. Sibiude J. Ebrahimi R. Katz J. Combination antiretroviral use and en/). et al. phant in the room? J Infect Dis 2016. Stek AM. during pregnancy. tial mechanism contributing to fetal verse birth outcomes among HIV-infected 13. National Institutes of Health. Lopinavir exposure with an increased dose viral therapy and on pre-exposure prophy- 8. of AIDS table for grading the severity of 53:​1937-43. et al.). Kourtis AP. Warszawski J. or lopinavir-ritona.​54:​1348- HIV-1-infected women for maternal health try analysis. Fowler MG.​211:​ women in Botswana.​204:​493-4.). et al.​43:​278-83. Antiretroviral therapy 12.B. cellular interactions between TDF and hydroxyprogesterone caproate concentra- National Institute of Allergy and Infec. adult and pediatric adverse events. B. the Centre for Infectious Disease Research in Zambia. tion of mother-to-child transmission of pregnancy is associated with decreased 4. and interventions to reduce perinatal HIV 11. patients. 2017. Antiretroviral et al. et al.). Capparelli E. Dadabhai SS.).V. Lusaka.​379:​2151-61. Mohammadi H. Johannesburg. Am J tious Diseases. Ogwu A. HIV in Botswana. August 2009 (http://rsc. For personal use only.​71:​428-36. Guideline on when to start antiretro- adverse_events.18 nejm. 2000. Department of Health and Human (TDF): investigation of systemic and intra. pregnant women exposed to protease 64. Mathias A. India (R.gov/​contentfiles/​lvguidelines/​Perinatal in Blantyre. Lancet 2012. Souda S. 24. Division vir. N Engl J Med docs/default-source/safety/table_for_ pregnancy. . Sperling RS. Global.​9(8):​e1001292. J Infect Dis 2012. Kozuki N. J Med 1994.int/​publications/​2010/​9789241599818_eng​ immunodeficiency virus-infected patients puzzle? J Infect Dis 2015.H. abacavir. The INSIGHT START Study Group. Sando MM. and the Elizabeth Glaser Pediatric AIDS Foundation. Carten ML. 19.pdf). vir boost? Clin Infect Dis 2012.0.​382:​417-25.). NY (M. MH.). 2015. Johnson HL. et al. 60. Zambia (B. J Acquir Immune Defic progesterone levels. References 1. Nelson R.). Gelber R.​61:​226-34. School of Public Health and Family Medicine. receiving tenofovir disoproxil fumarate 25.M.). Antiretrovir al Ther apy for Perinatal HIV Prevention (T. Premature delivery in HIV-infected Pregnant Women and Prevention of Peri. lamivudine. J Infect Dis 2011. Frontier Science and Technology Research Foundation.T. ganization.O. Mofenson LM. Panel on Treatment of HIV-Infected 10. 20. Caritis SN.​213:​1532- nant women and preventing HIV infection navir/ritonavir. Reduced lopinavir exposure during asymptomatic HIV infection. Zash R. Anova Health Institute (J. Williams PL. adjusted by 18. Tubiana R. Mirochnick M. Initiation of antiretroviral therapy in early 1. Perinatal HIV Research Unit. Lancet 2013. Division of AIDS. Loutfy MR. Makerere University–Johns Hopkins University Research Collaboration. Malawi: 1989-2009. nal clearance of tenofovir in HIV-infected 23. ticipant level meta-analysis. 27. Banko N.who​. lower progesterone levels in HIV-infected a cohort study. Watts DH. Bhosale).B. et al. Child Increased risk of preterm delivery among ville. women starting protease inhibitor ther­ natal Transmission. Durban — all in South Africa. Sayre J. et al. 5.). All rights reserved. suggesting a poten- Highly active antiretroviral therapy and ad. Pharmaco.pdf). et al. steroid dehydrogenase levels contribute to outcomes among HIV-infected women: 14. ral therapy.M.M. University of KwaZulu–Natal. Copyright © 2016 Massachusetts Medical Society. Taha TE. 9. Balogun K.M. Syndr 2016.​373:​795-807. type 1 with zidovudine treatment. Kumwenda N.). Reduction of maternal–infant trans.M. September 2015 (http://www​ preterm birth in East Africa. n engl j med 375. transmission in the United States.​49:​485-91. Shapiro RL. Lee AC. Théodoro F.