Secondary Logo

Journal Logo

HIV Reports

Tenofovir Disoproxil Fumarate Use during Pregnancy and Infant Bone Health: the Tenofovir in Pregnancy Pilot Study

Kourtis, Athena P. MD, PhD, MPH*; Wiener, Jeffrey PhD*; Wang, Liming MD; Fan, Bo PhD; Shepherd, John A. PhD; Chen, Lili MD§; Liu, Wei MD; Shepard, Colin MD; Wang, Linhong MD; Wang, Ailin MD; Bulterys, Marc MD, PhD

Author Information
The Pediatric Infectious Disease Journal: November 2018 - Volume 37 - Issue 11 - p e264-e268
doi: 10.1097/INF.0000000000002152
  • Free


Tenofovir Disoproxil Fumarate (TDF), an orally bioavailable form of tenofovir, is approved for therapy of both HIV and Hepatitis B virus (HBV) infections, and is included in the recommended first line antiretroviral drug regimens for HIV-infected individuals, including pregnant women, due to its efficacy and safety.1 Even though TDF is a preferred agent for first-line therapy among pregnant women for prevention of HIV and HBV transmission to the infant, there are still concerns about its safety for the bone development of the fetus and infant.

The Tenofovir in Pregnancy (TiP) pilot study was designed to test the safety, during pregnancy, of an antiretroviral combination containing TDF in women co-infected with HIV and HBV in Guangxi, China. In this article, we report on the effects of TDF on bone mineral content (BMC) and density of the infants through the first year of life.


Study Design and Participants

The TiP Study was a pilot phase II randomized controlled trial (ClinicalTrials Registration number: NCT01125696) that recruited participants between 2012 and 2015 in Guangxi, a Southwestern province in China; participant follow-up was completed in 2017. Study sites were Guangxi Provincial Mother to Child Hospital in Nanning Prefecture and at Liuzhou Mother to Child Hospital in Liuzhou Prefecture. Target sample size for the study was 80 mothers (40 per group), consistent with phase II study considerations. However, due to slow study recruitment and many exclusions only 35 mothers were enrolled.

Eligible participants had to have serologically confirmed HIV and HBV infection; pregnancy between 14 and 28 weeks’ gestation; age 20 years or older; no previous or current use of antiretroviral agents with the exception of short prevention of mother to child transmission regimens in any previous pregnancy; hemoglobin over 8 g/dL at recruitment; and creatinine (Cr) clearance greater than or equal to 60 mL/min estimated by the Cockroft-Gault formula for women.2

The study was approved by the Institutional Review Boards at the U.S. Centers for Disease Control and Prevention and Guangxi Provincial Center for Disease Control and Prevention.

Study Procedures

Participants were randomly assigned (1:1) to start either tenofovir/lamivudine/lopinavir-ritonavir or zidovudine/lamivudine/lopinavir-ritonavir. Tenofovir was included in the Chinese national guideline-recommended treatment for HIV-infected pregnant women in 2015.

Regimen assignments were generated using permuted-block randomization method. Patients and study investigators were aware of regimen allocation after the randomization envelope was open. There were 10 visits in the study: screening and enrollment, 2 weeks and 8 weeks post randomization, 34 weeks of gestation, delivery, 6 weeks postpartum and at 3, 6, 9, and 12 months of life. All newborn infants were provided with 6 weeks of zidovudine. Three doses of HBV vaccine starting within 24 hours of birth and 100 mg HBIG were given to all infants, according to local practice.3 All infants were formula-fed.

Infant weight, length, and head circumference were measured at birth, 6 months, and 12 months of life by trained personnel using standard methods previously described.4,5

Dual-energy Radiograph Absorptiometry

Lumbar spine and whole body scans were performed in infants at delivery (median, 5 days of life; IQR, 4–32 days) and at 6 months postpartum/postnatal, using a Hologic scanner (Discovery W, Marlborough, MA) and scanning protocol according to the University of San Francisco Densitometry Center’s procedures. All scans were analyzed by an International Society for Clinical Densitometry–certified bone densitometry technologist at the University of California San Francisco Densitometry Center using Hologic APEX software version 3.4. The lumbar spine and hip scans were analyzed according to manufacturer’s instructions. Infant whole body scans were analyzed by placing 6 sub-regions: 1 for head and trunk, respectively, and 4 for each of the limbs. Due to frequent motion during infant scans, 3 attempts were allowed for each of the anatomic sites. The results were imputed by either combining multiple scans or replacing motion limb results with results from the non-motion side. Infants were swaddled and not sedated.

Statistical Analysis

The primary objective was to compare the mean BMC and bone mineral density (BMD) of tenofovir exposed and unexposed infants. The primary BMC outcome was whole body BMC (with head), but whole body BMC less head and BMC per body weight were also evaluated. Unadjusted mean differences in outcomes between the groups were calculated with 95% confidence intervals using the test separately for each study visit. Multivariable linear mixed-effects models were used to evaluate differences in BMC and BMC per body weight between treatment groups accounting for repeated measurements and adjusted for infant sex, body length, gestational age, maternal age, maternal body mass index (BMI), and maternal CD4+ T lymphocyte count and HIV viral load at baseline. All analyses followed the intention-to-treat principle and were conducted using SAS version 9.3 (Cary, NC).


Characteristics of Mothers and Infants

Seventy-seven HIV/HBV coinfected pregnant women were screened for study eligibility; 42 (54.5%) did not meet inclusion criteria: 20 were already on highly active antiretroviral therapy (HAART) when the pregnancy was identified, 8 presented for antenatal care (ANC) after 28 weeks of gestation, 10 decided to terminate their pregnancy, and 4 were not willing to participate in the study (Fig. 1). Thirty-five eligible women were enrolled and randomized in the study. One woman decided to terminate her pregnancy at 8 weeks after randomization due to complications resulting from advanced HIV disease. Four women on the ZDV/3TC/LPV-r arm were prenatally switched to the TDF/3TC/LPV-r arm due to anemia or liver function abnormalities at a median of 30 weeks gestation; one was switched from TDF/3TC/LPV-r to ZDV/3TC/LPV-r due to creatinine elevations after delivery.

Enrollment and disposition of study participants, Tenofovir in Pregnancy (TIP) Study

Baseline characteristics and laboratory measures were similar in the treatment arms (Table 1). Ninety-seven percentage (33/34) of enrolled women reported no missed drug doses before delivery. Only 1 woman on the TDF study arm reported missing a dose of LPV-r on the day before delivery.

Characteristics of Mothers and Infants by Study Arm, TiP Study

BMC of Infants by Study Arm at Birth and at 6 Months of Age

The mean (standard deviation [SD]) whole body BMC of tenofovir-exposed infants was 55.5 g (18.2) compared with 61.9 g (15.8) for unexposed infants (mean difference = -6.4 g; 95% CI = -19.9, 7.1, Table 2). At age 6 months, the corresponding mean (SD) BMC were 133.4 g (27.9) and 139.1 g (21.2), respectively (mean difference = -5.7 g; 95% CI = -26.0, 14.7, Table 2). These differences represent an unadjusted 10.3% difference in BMC (0.4 SD) between exposure groups at birth and an unadjusted 4.1% difference (0.2 SD) at 6 months of age, respectively, and were not statistically significant. For whole body BMC less head, the mean (SD) was 33.3 g (8.9) in the tenofovir-exposed group compared with 35.5 g (8.6) in the unexposed group at birth (mean difference = -2.2 g, 95% CI = -9.0, 4.7, Table 2), representing a nonstatistically significant 6.2% difference (0.3 SD) between exposure groups. At age 6 months, the corresponding mean BMC less head were 70.6 g (15.2) and 73.9 g (11.6), respectively (mean difference = -3.2 g, 95% CI = -14.3, 7.9, Table 2). Table 2 also shows lumbar spine BMC measurements; there was no difference in the 2 groups of infants.

Unadjusted mean infant BMC measures by study arm and age, TiP Study

BMD of Infants by Study Arm at Birth and at 6 Months of Age

The mean (SD) whole body BMD at birth of tenofovir-exposed infants was 0.15 g/cm2 (0.02) compared with 0.16 g/cm2 (0.02) for unexposed infants (mean difference = -0.01 g/cm2; 95% CI = -0.02, 0.01, Table 2). At age 6 months, the corresponding mean (SD) BMD were 0.20 g/cm2 (0.02) and 0.21 g/cm2 (0.02), respectively (mean difference = -0.01 g/cm2; 95% CI = -0.02, 0.01, Table 2). These differences represent an unadjusted 5.7% difference (0.5 SD) in BMD between exposure groups at birth and an unadjusted 2.9% difference (0.4 SD) at 6 months of age, and were not statistically significant. There were no significant differences in lumbar spine BMD at birth or at 6 months of age in the 2 groups of infants (Table 2).

Adjusted Difference in Whole Body BMC of Infants by Study Arm

When evaluating BMC over time using a multivariable linear mixed-effects model, the estimated unadjusted mean BMC was 5.9 g lower in tenofovir-exposed infants (95% CI = -18.0, 6.1, Table 3) compared with unexposed infants. After adjustment for relevant covariates including infant sex, body length, gestational age, maternal age, maternal baseline BMI, CD4+ T lymphocyte count, and HIV viral load, the estimated mean BMC was 6.6 g lower in tenofovir-exposed infants (95% CI = -16.6, 3.5). In a similar model evaluating BMC per body weight, the estimated unadjusted mean BMC per weight was 0.7 g/kg lower (95% CI = -2.5, 1.2) and the adjusted mean BMC per weight was 0.6 g/kg lower (95% CI = -2.9, 1.7) in tenofovir-exposed infants. None of the differences were statistically significant.

Unadjusted and Adjusted Mean Differences in Infant Whole Body BMC and BMC per Body Weight, TiP Study


Evidence on the effects of TDF use during pregnancy on infant bone health is very limited. In a study evaluating whole body dual-energy radiograph absorptiometry scans within 4 weeks of birth among 74 infants exposed to more than 8 weeks of TDF in utero and 69 infants with no TDF exposure, the infant BMC was significantly lower by 6.3 g (P = 0.004), or about 12% in the TDF-exposed infants.6,7 In contrast, another study did not note differences in quantitative bone ultrasound and parameters of bone metabolism in 33 children with in utero exposure to TDF compared with 35 unexposed children.6,8–10 No adverse effect of maternal TDF use on infant BMC was seen in the randomized PROMISE trial in HIV/HBV coinfected women (9) or on infant BMD in the randomized iTAP study in HBV-mono-infected women (10), both reported as conference abstracts.

The estimated decreases in infant BMC and BMD among in utero TDF-exposed and unexposed infants in our study were 10% and 4% at birth to 6% and 3%, respectively, at 6 months of age. These differences were small and not statistically significant, and were not attenuated by the possible confounders in the multivariable model. Even though their clinical significance is unclear, these results are consistent with previously reported findings.9,10 Our findings are overall reassuring and provide support for the current World Health Organization (WHO) and National Institutes of Health (NIH) recommendations that include TDF in the preferred regimens for HIV-infected and HIV/HBV coinfected pregnant women.

This study’s main limitation is it was a pilot phase II trial with a small sample size and not powered to detect statistical differences between the 2 arms. While the target sample size of n = 80 would have had 80% power to detect a difference of 18.5 g in BMC between study arms given the observed data, the actual reduced sample size increased this detectable difference to 28.4 g. A few women switched treatment arms before delivery due to laboratory abnormalities. Our analysis followed the intention-to-treat principle; however, a sensitivity analysis adjusting for the actual duration of time on each treatment arm did not yield different results. Results from this study may not be applicable to other racial/ethnic groups. Notwithstanding the small sample size, recently reported initial findings from larger trials in HIV/HBV coinfected (14) and HBV-monoinfected (15) women seem to confirm our findings.


This clinical trial was supported by the U.S. Centers’ for Disease Control and Prevention, the Guangxi Provincial Center for Disease Control and Prevention, and Gilead Sciences, Inc, through an award to the CDC Foundation. The authors especially thank the study team at the 2 study sites: Guangxi Provincial Mother to Child Hospital in Nanning Prefecture and Liuzhou Mother to Child Hospital in Liuzhou Prefecture for all their support and implementation of the study. The late Dr. Zhirong Tang cared tirelessly for the HIV/HBV co-infected patients in this study. Steven Schindler at the U.S. Division of Reproductive Health and Alison Kelly and Nannan Zhang at the US CDC Global AIDS Program Office in Beijing provided critical management and logistics support. Drs. Yu Wang, Zunyou Wu, Yiyun Hu, and Jiangping Sun of the China CDC and Dr. Minghui Ren at the Chinese Health and Family Planning Commission provided leadership support and guidance to the investigators. The authors also thank the members of the Data Monitoring Committee, and most of all, the women and infants who participated in the study.


1. P. Guidelines for the Use of Antiretroviral Agents in HIV-1-infected Adults and Adolescents. [Feb 18, 2016].Department of Health and Human Services;
2. Cockcroft DW, Gault MHPrediction of creatinine clearance from serum creatinine. Nephron. 1976;16:3141.
3. Liang X, Bi S, Yang W, et al.Epidemiological serosurvey of hepatitis B in China—declining HBV prevalence due to hepatitis B vaccination. Vaccine. 2009;27:65506557.
4. Jacobson DL, Patel K, Williams PL, et alPediatric HIVAIDS Cohort Study. Growth at 2 years of age in HIV-exposed uninfected children in the United States by trimester of maternal antiretroviral initiation. Pediatr Infect Dis J. 2017;36:189197.
5. Wang L, Wiener J, Bulterys M, et al.Hepatitis B Virus (HBV) load response to 2 antiviral regimens, tenofovir/lamivudine and lamivudine, in HIV/HBV-coinfected pregnant women in Guangxi, China: the Tenofovir in Pregnancy (TiP) study. J Infect Dis. 2016;214:16951699.
6. Siberry GK, Jacobson DL, Kalkwarf HJ, et alPediatric HIVACS. Lower newborn bone mineral content associated with maternal use of tenofovir disoproxil fumarate during pregnancy. Clin Infect Dis. 2015;61:9961003. doi: 10.1093/cid/civ437. PubMed PMID: 26060285; PMCID: PMC4551007.
7. Siberry GK, Jacobson DL, Kalkwarf HJ, et alPediatric HIVACS. Lower newborn bone mineral content associated with maternal use of tenofovir disoproxil fumarate during pregnancy. Clin Infect Dis. 2015. doi: 10.1093/cid/civ437. PubMed PMID: 26060285.
8. Viganò A, Mora S, Giacomet V, et al.In utero exposure to tenofovir disoproxil fumarate does not impair growth and bone health in HIV-uninfected children born to HIV-infected mothers. Antivir Ther. 2011;16:12591266.
9. Siberry GK, Tierney C, Strainx-Chibanda L, et al.Impact of maternal tenofovir use on HIV-exposed newborn bone mineral. 23rd Conference on Retroviruses and Opportunistic Infections (CROI), 2016, Boston, MA, #36, Abstract.
10. Jourdain G, Ngo-Giang-Huong N, Harrison L, et alTDF to Prevent Perinatal Hepatitis B Virus Transmission: A Randomized Trial (ITAP). Seattle, WA: CROI; 584LB 2017.

tenofovir desoproxil fumarate; HIV; in utero; infant; bone

Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.