Breast Milk Transfer and Infant Exposures to Dolutegravir, Tenofovir, and Tenofovir Alafenamide: Results from IMPAACT 2010/VESTED

Abstract

Background:

Limited information is available on breastmilk transfer and subsequent infant systemic exposure to dolutegravir (DTG), tenofovir alafenamide (TAF), and tenofovir (TFV). We evaluated concentrations of DTG, TAF, and TFV in maternal and infant plasma and breastmilk from participants of IMPAACT 2010, a clinical trial that enrolled women initiating either a DTG-containing or an efavirenz-containing antiretroviral treatment (ART) regimen in pregnancy.

Methods:

Time-matched postpartum maternal and infant samples were collected at 6 weeks postpartum. Validated assays quantitated concentrations of DTG, TAF, and TFV. Relative infant dose (RID) was estimated using an average milk intake of 150 mL/kg/day and observed breastmilk concentrations.

Results:

Data were available from 192 postpartum lactating women and their breastfed infants. Median maternal plasma concentrations of DTG, TAF, and TFV were 2810, 0.0, and 96 ng/mL, respectively. For DTG, median (interquartile range; IQR) concentrations in breastmilk and infant plasma were 91 ng/mL (67–123) and 69 ng/mL (41–95), respectively. TAF was below the quantitative limit (BQL) in nearly all breastmilk and infant samples. For TFV, median (IQR) concentrations in breastmilk were 8 ng/mL (6–12), and nearly all infant samples were BQL. The median RIDs of DTG, TAF, and TFV from breastfeeding were 1.9%, 0.00%, and 0.03%, respectively.

Conclusions:

Breastmilk transfer of DTG, TAF, and TFV is low and results in minimal systemic exposure in breastfed infants. The clinical relevance of infant exposure to low concentrations of these antiretrovirals – particularly DTG – is unknown but should be considered in the context of the risk of drug resistance in infants who acquire HIV.

INTRODUCTION

In 2010, the World Health Organization (WHO) recommended the use of antiretroviral therapy (ART) to prevent HIV transmission from mother to infant during breastfeeding.¹ The guidelines have since been updated to encourage mothers living with HIV who are adherent on ART to breastfeed their infants for up to 2 years or longer, with exclusive breastfeeding for the first 6 months of life – particularly in low- and middle-income countries (LMIC) where breastfeeding is critical for infant survival.¹ In high-income countries where alternative infant feeding methods are accessible, guidelines have generally advised against breastfeeding to eliminate the risk of HIV transmission.² However, perinatal guidelines in the U.S., U.K., and other countries are evolving to support informed, shared decision-making, and acknowledging the potential psychosocial impact that breastfeeding restrictions may have on women.^2–4 These recommendations are informed by increasing evidence of low risk of HIV transmission (<1% but not zero) while breastfeeding on ART with sustained viral suppression and numerous breastfeeding-related benefits to both the infant and the mother.^1,5 Overall, these recommendations have evolved to support health equity for mothers living with HIV who wish to breastfeed and encourages well-informed, patient-centered discussions on infant feeding.

Approximately 1.3 million women with HIV become pregnant each year and the majority of these women breastfeed their infants while receiving ART.^6,7 The WHO guidelines recommend triple-drug regimens containing dolutegravir (DTG), an integrase strand inhibitor, as the first-line treatment for all adults, including pregnant women living with HIV.⁸ ART should be started immediately upon identification of HIV pre-conception or in pregnancy and continued lifelong to minimize the risk of transmission and maximize maternal and infant health outcomes.^6,9 Antiretroviral drug concentrations in breastmilk may potentially contribute to preventing breastmilk HIV transmission.¹⁰ Further, any drug excreted into human milk may be absorbed by the breastfeeding infant. Potential risks to the infant include antiretroviral-associated adverse events and, for infants who acquire HIV while breastfeeding, the potential for developing drug resistance due to subtherapeutic systemic drug concentrations.^11–13

The International Maternal Pediatric Adolescent AIDS Clinical Trials (IMPAACT) 2010 trial was a multicenter, open-label, randomized study that assessed the efficacy and safety of three ART regimens started in pregnancy: DTG+emtricitabine (FTC)/tenofovir alafenamide fumarate (TAF); DTG+FTC/tenofovir disoproxil fumarate (TDF); and efavirenz (EFV)/FTC/TDF.¹⁴ The study found that when started in pregnancy, the TAF-containing regimen (DTG+FTC/TAF) had the fewest adverse pregnancy and neonatal outcomes compared to the TDF-containing regimens (DTG+FTC/TDF and EFV/FTC/TDF). Prior data on breast milk transfer of EFV and TDF exist, but there are limited breast milk data on DTG and TAF. ^15–17

In this analysis of mother-to-infant transfer of antiretrovirals during breastfeeding, we evaluated the concentrations of DTG, TAF, and tenofovir (TFV) from IMPAACT 2010 in time-matched mother-infant dyads (maternal plasma, breast milk, and infant plasma) at 6 weeks postpartum.

METHODS

Pregnant women (aged ≥ 18 years) with confirmed HIV in 9 countries were randomized 1:1:1 to start one of three open-label ART regimens at 14–28 weeks of gestation. Participants were ART-naïve with the following exceptions: up to 14 days of ART during current pregnancy; prior TDF or TDF/FTC pre-exposure prophylaxis; or ART during prior pregnancies (last dose ≥ 6 months previously). Eligible participants were randomly assigned to receive DTG 50 mg once daily with a fixed-dose combination of FTC 200 mcg/TAF 25 mg once daily, DTG 50 mg once daily with a fixed-dose combination of FTC 200 mg/TDF 300 mg once daily, or a fixed-dose combination of EFV 600 mg/FTC 200 mg/TDF 300 mg once daily (NCT 03048422).¹⁴ Infants received non-study antiretroviral prophylaxis consistent with local standards of care. Study ART was continued through 50 weeks postpartum and maternal plasma, breast milk, and infant plasma samples were prospectively collected at 6 weeks postpartum.

We aimed to include samples from 100 mother-infant dyad pairs each from the DTG+FTC/TAF arm and DTG+FTC/TDF arm (200 mother-infant dyads in total), from participants who enrolled in sites in Botswana, India, South Africa, Tanzania, Uganda, and Zimbabwe. This sample size was based upon non-probability sampling to obtain a representative number of participants given the various constraints of the study (including assay costs). Samples were selected by simple random sampling from among those available to arrive at the target sample.

DTG, TAF, and TFV concentrations in breast milk, infant plasma, and maternal plasma were measured at 6-weeks postpartum from mother-infant dyad pairs. One breast milk sample, one infant plasma sample, and one maternal plasma sample were collected at random at the same time between 0–24 hours after self-reported maternal dosing. DTG concentrations were measured in samples from the DTG+FTC/TAF and DTG+FTC/TDF arms, TAF concentrations were measured in samples from the DTG+FTC/TAF arm, and TFV concentrations were measured in samples from the DTG+FTC/TDF arm.

Drug concentration measurements were performed at the IMPAACT Pediatric Clinical Pharmacology Laboratory at the University of California, San Diego, USA, which participates in the DAIDS-supported external pharmacology quality control program. DTG, TAF, and TFV concentrations were determined in plasma and whole breastmilk by validated liquid chromatography tandem mass spectrometry (LC-MS/MS) assays. The lower limit of quantitation (LLOQ) for DTG, TAF, and TFV in breast milk was 7.8, 0.195, and 0.0977 ng/mL, respectively. The LLOQ for DTG, TAF, and TFV in plasma was 9.8, 3.9, and 0.977 ng/mL, respectively.

Maternal baseline characteristics were collected at enrollment and infant characteristics at birth. Mean and standard deviations were used to describe continuous variables and counts and percentages for categorical variables. For drug concentrations, samples below the LLOQ (BQL) were imputed to 0 for summary statistics. The milk-to-plasma ratio describes the amount of drug transferred into breast milk and was calculated using the average drug concentrations observed in breast milk and maternal plasma.

The relative infant dose (RID) for DTG, TAF, and TFV were estimated by using a standard feeding milk volume of 150 mL/kg/day, observed whole breast milk concentrations, and the daily maternal dose, as follows:

$$Dailyinfantdosage\left(\frac{mg}{kg}\right)=Medianbreastmilkconcentration\left(\frac{mg}{mL}\right)\times 150\frac{mL}{\frac{kg}{day}}$$

$$RID=\frac{\mathit{Daily}\mathit{infant}\mathit{dosage}(\frac{mg}{kg})}{\mathit{Daily}\mathit{maternal}\mathit{dosage}(\frac{mg}{kg})}\times 100$$

The study was approved by Institutional Review Boards at each site. All maternal participants provided written informed consent for their own and their infant’s participation. The study was monitored by an independent Data and Safety Monitoring Board.

RESULTS

Data were available from 192 postpartum lactating women and their 192 breastfed infants (out of target 200 mother-infant dyads). The mean age (SD) of mothers at enrollment was 26 (6) years, and most of the participants lived in Zimbabwe (59%), Uganda (21%), or South Africa (9%). The median (interquartile range; IQR) sampling time was 6.0 (5.9–6.3) weeks postpartum. Overall 55% of infants were female at birth, with a mean gestational age (standard deviation [SD]) of 40 (2) weeks. Infants at birth had a mean weight (SD) of 3089 (490) g and a length of 50 (3) cm. Most infants received prophylaxis with nevirapine and small proportions received zidovudine and/or lamivudine, generally in combination with nevirapine.

Samples collected between 0–24 hours post self-reported dose were used for analysis, with the the majority of time-matched samples collected between 12–20 hours post-dose. Maternal plasma concentrations of DTG, TAF, and TFV are displayed in Figure 1. Concentrations were generally within the range of steady-state exposures observed in non-pregnant adults.^18–21 The median (IQR) maternal DTG and TFV plasma concentrations were 2805 ng/mL (2108–3678) and 123 ng/mL (96–353), respectively. About 24% of DTG maternal plasma concentrations were above the 95^th percentile seen in nonpregnant adults. All TAF concentrations sampled after 2.5 hours post-dose were BQL (0.977 ng/mL).

Figure 1. Maternal Plasma Concentrations at 6 Weeks Postpartum.

Shaded regions display the exposures (5^th – 95^th percentile) of each respective drug in nonpregnant adults. (A) DTG concentrations (n=180). Six concentrations sampled after 24 hours are not shown. Six concentrations below the quantitative limit of the assay are plotted as half of the lower limit of quantitation (½ LLOQ = 0.0049 mcg/mL). (B) TAF concentrations (n=98). Four concentrations sampled after 24 hours are not shown. Ninety-one concentrations below the quantitative limit of the assay are plotted as half of the LLOQ (½ LLOQ = 1.95 ng/mL). (C) TFV concentrations (n=89). Five concentrations sampled after 24 hours are not shown. Two concentrations below the quantitative limit of the assay are plotted as half of the LLOQ (1/2 LLOQ = 0.4885 ng/mL).

DTG, TAF, and TFV concentrations in breast milk and infant plasma are shown in Figure 2. For DTG, the median (IQR) concentrations in breast milk and infant plasma samples were 91 ng/mL (67–123) and 69 ng/mL (41–95), respectively. The maximum DTG concentration in breast milk was 233 ng/mL. Using paired data, the median (IQR) milk-to-plasma ratio was 0.032 (0.28–0.38). TAF was BQL in nearly all breast milk and infant plasma samples. Median (IQR) TFV concentrations in breast milk were 8 ng/mL (6–12), and nearly all infant plasma were BQL for TFV. In breast milk, 9 (5%), 96 (98%) and 3 (3%) samples were BQL for DTG, TAF, and TFV, respectively. In infant plasma, 11 (6%), 98 (100%), and 93 (99%) samples were BQL for DTG, TAF, and TFV, respectively.

Figure 2. DTG, TAF, and TFV Concentrations in Breast Milk and Infant Plasma.

BM, breastmilk; IP, infant plasma. Solid horizontal lines represent median, and error bars represent interquartile range. Samples below the quantitation limit (BQL) are shown in the box. In breast milk, 9 (5%), 96 (98%), and 3 (3%) samples were BQL for DTG, TAF, and TFV, respectively. In infant plasma, 11 (6%), 98 (100%), and 93 (99%) samples were BQL for DTG, TAF, and TFV, respectively.

Based on a 150 mL/kg/day milk intake, the estimated median (range) RID from breastfeeding for DTG was 1.92% (0.00–4.89). The estimated RID for TAF and TFV were 0.00% (0.00–0.03) and 0.03% (0.00–0.11), respectively.

DISCUSSION

Following the first report of HIV transmission through breast milk in 1985, mothers living with HIV were initially advised to avoid breastfeeding and to use alterative feeding methods (i.e. infant formula, donor human milk). Breastfeeding restrictions are particularly challenging for mothers living in low-resource settings where replacement feeding is not often socially acceptable, feasible, affordable, sustainable, and safe (AFASS). Artificial feeding may expose women to stigma or discrimination and has greater risks related to unsafe water for formulation preparation, lack of access to formula (high cost, low quality products), reduced mother-infant bonding, and more.^22–24 Research has emerged to demonstrate low risk of breastfeeding transmission of HIV with sustained viral suppression on ART. Consequently, the WHO updated its guidelines for mothers living with HIV in LMIC who are receiving ART to exclusively breastfeed for the first 6 months and continue breastfeeding for up to 2 years or longer.^25,26 In high-income countries where mothers living with HIV were traditionally advised to avoid breastfeeding, guidelines are evolving to support informed, shared decision making and to offer clinical support for mothers who choose to breastfeed.²

Breastfeeding has numerous health benefits for infants and mothers. Breast milk is an ideal source of nutrition for children and aids in healthy growth and development.²⁶ It contains important antibodies that provide protection against childhood illnesses (i.e., asthma, diabetes) and reduces the risk of sudden infant death syndrome.^26,27 Breastfeeding can reduce the risk of infant morbidity and mortality from diarrhea and malnutrition, especially in countries with limited access to safe water and food products. Mothers who breastfeed for longer duration are at a lower risk for certain cancers (e.g., breast and ovarian), diabetes, and cardiovascular disease.^26,27 Additionally, breastfeeding allows for convenient feeding at any time without having to mix formula or prepare bottles and promotes the mother-infant bonding experience through skin-to-skin contact.²⁸

In 2019, the WHO recommended DTG as the preferred first- and second-line treatment for HIV in all populations, including pregnant women.⁸ Newer antiretroviral drugs such as DTG have limited breast milk transfer and infant exposure information. To our knowledge, the DolPHIN-1 trial, which included 17 mother-infant pairs, is the only study that has assessed breast milk transfer of DTG.²⁹ In DolPHIN-1, antiretroviral therapy was initiated between 28–36 weeks of gestation, and pharmacokinetic sampling of DTG was performed within 1–3 weeks after delivery. The study observed maximum concentration of DTG in breast milk at 171.2 ng/mL and a milk-to-plasma ratio of 0.03.

The current study provides additional information on the breast milk transfer of DTG in a larger sample size of 192 mother-infant pairs. The maximum DTG concentration in breast milk and milk-to-plasma ratio in the current study are consistent with the DolPHIN-1 study. The median DTG breast milk and infant concentrations were 91 ng/mL and 69 ng/mL, respectively, and the median relative infant dose of DTG from breastfeeding was 1.9%. The transfer of DTG in breast milk resulted in low but detectable systemic concentrations in breastfed infants. Notably, the breastfeeding child may be exposed to DTG levels higher than the in vitro 90% protein-adjusted minimal inhibitory concentration (IC90) of 64 ng/mL.³⁰ On the other hand, consistent low exposure to DTG through breastmilk could potentially predispose infants who acquire HIV to develop DTG resistance. In IMPAACT 2010, vertical transmission of HIV through the first 50 weeks after birth was rare (occurred in 4 of 617 live-born infants) and no integrase mutations conferring dolutegravir resistance were detected.³¹ Preventing perinatal transmission is critical, and clinical care providers should provide the necessary support to ensure maternal ART adherence while breastfeeding.

We also show that the breastmilk transfer of TAF and TFV is low, consistent with other studies.^13,16–17 TFV breastmilk concentrations are likely low in part because TFV is charged at physiological pH with poor membrane permeability. This also contributes to low oral bioavailability in infants (TFV is administered as a prodrug to improve absorption). Further, TAF has a short plasma half-life (25–30 minutes) which may contribute to limited drug transferred into breast milk.

DTG bioavailability via breastmilk appears to be higher than well-characterized commercial formulations such as dispersible tablets.³² Comparing infant DTG plasma concentrations with the breast milk “dose” (derived from the calculated RID) to infant exposures with the DTG dispersible tablet, concentrations of DTG in infant plasma from the current study were approximately 5 times higher than predicted.

Maternal postpartum DTG plasma concentrations were generally higher than those observed in non-pregnant adults. Increased exposure in the postpartum period has been previously reported for DTG and other antiretrovirals.^33–36 The majority of maternal TAF concentrations were BQL (0.977 ng/mL) which is expected given TAF has a short plasma half-life (approximately 30 minutes) and most samples were collected between 12–20 hours post-dose.²⁰ Postpartum TFV concentrations were mostly within the range of steady-state exposures observed in non-pregnant adults.²¹

This study has a few limitations. Only a single milk sample per participant was available and we did not collect the entire milk volume from both breasts over a dosing interval. Next, maternal dosing was self-reported and adherence information prior to sampling was not available. Therefore, missed doses may have resulted in a participant not being at steady state. However, maternal ARV concentrations were generally within the range of steady-state exposures observed in non-pregnant adults. we do not have detailed information on the timing of infant feeding, including the last feeding prior to sampling

In conclusion, breast milk transfer of DTG, TAF, and TFV is low and results in minimal systemic exposure in breastfed infants. The clinical relevance of infant exposure to low concentrations of these antiviral agents – particularly DTG – is unknown, but should be considered in the context of the risk of drug resistance in the small number of infants who acquire HIV.

IMPAACT 2010/VESTED Team Members

Renee Browning, DAIDS/NIH; Lynette Purdue, DAIDS/NIH; Nahida Chakhtoura, NICHD/NIH; Katie McCarthy, FHI 360; Sean Brummel, Center for Biostatistics in AIDS Research, Harvard T.H. Chan School of Public Health; Chelsea Krotje, Frontier Science Foundation; Nagawa Jaliiah, Mulago Hospital; Rivet Amico, University of Michigan; Judith Currier, David Geffen School of Medicine, UCLA; Lee Fairlie, Wits Reproductive Health and HIV Institute; Lisa Frenkel, Seattle Children’s Research Institute; Risa Hoffman, David Geffen School of Medicine, UCLA; Lewis Holmes, MassGeneral Hospital for Children; Gaerolwe Masheto, Botswana Harvard AIDS Institute Partnership; Mark Mirochnick, Boston University School of Medicine/Boston Medical Center; Paul Sax, Brigham and Women’s Hospital; Roger Shapiro, Harvard T.H. Chan School of Public Health; Lynda Stranix-Chibanda, University of Zimbabwe; Jeffrey Stringer, University of North Carolina at Chapel Hil; Kyle Whitson, Frontier Science Foundation; Frances Whalen, UCLA; Sikhulile Moyo, Botswana Harvard AIDS Institute Partnership.