Abstract
IMPAACT 2001 (NCT02651259) evaluated the safety and pharmacokinetics (PK) of once‐weekly isoniazid and rifapentine in pregnant and postpartum women for tuberculosis prevention. Rifapentine (RPT) and desacetyl rifapentine (desRPT) PK data were collected in maternal plasma, breast milk, infant cord blood, and infant plasma. Longitudinal maternal plasma and breast milk RPT and desRPT concentrations were characterized using a multicompartment PK model with biophase effect compartments (NONMEM v7.4). Rate parameters describing drug entry into breast milk from plasma and breast milk‐to‐plasma ratio parameters were estimated. For mother–infant pairs, maternal plasma PK was simulated for the last RPT dose pre‐delivery (with HIV and partum status covariates), and concentration ratios (observed infant cord blood or plasma relative to simulated maternal plasma) were calculated at the infant PK sample collection time. For RPT and desRPT, derived equilibration half‐lives were 0.568 hr and 0.439 hr, and estimated concentration ratios of breast milk to maternal plasma were 0.0236 and 0.0411, respectively, indicating rapid equilibration between maternal plasma and breast milk and a minimal extent of drug excretion into breast milk. Maximum relative infant dose was estimated at <1%. Within 72 hours of maternal dosing, median concentration ratios of observed infant cord blood and plasma to predicted maternal plasma were approximately 0.280 and 0.612 for RPT and 0.493 and 1.83 for desRPT, respectively. RPT crosses the placenta and is measurable in infant plasma at concentrations unlikely to be clinically significant. Breast milk concentrations are very low, supporting safe use of RPT in breastfeeding women.
Study Highlights.
WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?
3HP is endorsed by the WHO and CDC for TB prevention in individuals ≥2 years old with latent TB infection, including people living with HIV. IMPAACT 2001 (NCT02651259) provided the first clinical data evaluating 3HP in pregnant and postpartum women. A population PK model was built to support the recommended rifapentine dose in pregnancy that achieves concentrations comparable to nonpregnant adults. No dose adjustment was necessary given the clearances of rifapentine and isoniazid were not significantly impacted by pregnancy.
WHAT QUESTION DID THIS STUDY ADDRESS?
What is the PK of rifapentine and desacetyl rifapentine in breast milk relative to plasma in postpartum women and what is the infant drug exposure following 3HP dosing during pregnancy and breastfeeding?
WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?
It establishes how infant drug exposure compares to maternal drug exposure and shows that 3HP administration during pregnancy or lactation results in low exposure to infants, which is likely not clinically significant.
HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?
Data on secretion into breast milk and effects on breastfeeding infants and milk production are absent in the rifapentine prescribing information. Including pregnant women in clinical trials permits these data to be collected, and ethics demand that a pregnant woman benefiting from treatment provided in a clinical trial be permitted to continue therapy postpartum. Regulatory bodies need to know if the drug amount ingested by a breastfeeding infant is low enough to qualify as minimal risk. This analysis establishes that the drug quantity ingested by a newborn is low enough to qualify as minimal risk per US regulatory standards.
Pregnant women have an increased risk of new tuberculosis (TB) infection or reactivation of latent TB infection (LTBI) during pregnancy. 1 , 2 This heightened risk is caused by immunological changes like suppression of the T‐helper 1 proinflammatory response, which reverses during the postpartum period. 3 Co‐infection with human immunodeficiency virus (HIV) increases the probability of developing TB for the mother and HIV transmission to the infant. 4 Infant TB and HIV are associated with neonatal death, low birth weight, and prematurity. 3 , 5 , 6 , 7 , 8 , 9 Collectively, these health burdens underscore the importance of effective preventative interventions for pregnant women with LTBI to improve both maternal and child health outcomes.
The 3HP regimen (900 mg isoniazid (INH) and 900 mg rifapentine (RPT) administered weekly for 3 months) is endorsed by the World Health Organization (WHO) 10 and recommended by the Centers for Disease Control (CDC) 11 for TB preventive treatment in individuals ≥2 years old with LTBI, including people living with HIV. RPT is hydrolyzed by an esterase enzyme to 25‐desacetyl rifapentine (desRPT), which is an active metabolite with lower clinical activity against M. tuberculosis than the parent drug. 2 RPT has a long elimination half‐life of approximately 16 hours making it suitable for less frequent (e.g., once‐weekly) dosing. 12 The maximum concentration is reached about 4–5 hours after administration. 13 RPT is an inducer of cytochrome P450 drug‐metabolizing enzymes and induces its own metabolism. Autoinduction is dependent on dosing frequency; drug accumulation is lower with once‐weekly dosing compared to daily dosing resulting in less autoinduction. 14
The 3HP regimen is not recommended during pregnancy due to insufficient safety data. The first study to test 3HP in pregnant women was IMPAACT 2001 (NCT02651259), a Phase I/II clinical study that evaluated the pharmacokinetics (PK) and safety of 3HP in pregnant women and established the RPT dose in pregnant and postpartum women with LTBI, with or without HIV. PK data for RPT were collected in maternal plasma during the second and third trimesters of pregnancy and postpartum. 15 A nonlinear mixed effects model was fit to maternal plasma PK data for RPT and its active metabolite desRPT, which was used to support the recommended RPT dose in pregnancy that achieves concentrations comparable to nonpregnant adults. 15
IMPAACT 2001 also collected PK data in breast milk from postpartum women and in cord blood and plasma from infants. Infant cord blood reflects how much drug crosses the placenta while infant plasma can reflect both drug from cord blood and from ingested breast milk, depending on the timing of the collection and the drug's properties. Small molecule drugs primarily enter breast milk by passive diffusion via capillaries in the mammary gland, but some can be actively secreted into breast milk via transport proteins that are expressed on mammary epithelial cells. 16 In utero, drugs can move from maternal to fetal blood by passing through the placenta using passive diffusion, facilitated diffusion, or active transport. 17 Transporters and drug‐metabolizing enzymes in the placenta can further affect drug transfer and exposure to parent drugs and metabolites. 18 For RPT, data on secretion into breast milk and effects on breastfeeding infants and milk production are absent in the prescribing information. 2 When a pregnant woman benefiting from treatment wishes to continue therapy postpartum, she and her healthcare provider can make informed decisions if the drug exposure a newborn is likely to experience through breastfeeding is known. Clinical pharmacology is well‐equipped with the knowledge and tools to provide this characterization when breast milk and infant concentration data collection are prioritized in early clinical trials.
Herein, we extend our quantitative framework 15 to (i) characterize the PK of RPT and desRPT in breast milk relative to plasma in postpartum women and (ii) incorporate cord blood and infant PK data by comparing the RPT and desRPT concentrations in infants relative to their mother's predicted exposure following her last pre‐delivery RPT dose administration. Together, these aims give insight into infant drug exposure following 3HP dosing during pregnancy and through breastfeeding, providing an understanding of infant exposure to TB preventative treatment.
METHODS
Study design and pharmacokinetic sampling
The full IMPAACT 2001 study design has been published. 15 Briefly, the PK, tolerability, and safety of 12 once‐weekly directly observed oral doses of 900 mg INH and 900 mg RPT were evaluated in pregnant and postpartum women (N = 50) living with HIV or without HIV, and with LTBI. Women living with HIV were taking an efavirenz‐based antiretroviral regimen. All women were enrolled into one of two cohorts based on gestational age1: second trimester (≥ 14 to <28 weeks) or2 third trimester (≥ 28 to ≤34 weeks). Maternal plasma PK sampling was performed at 0 (pre‐dose), 0.5, 1, 2, 4, 5, 8, 12, 24, 48, and 72 hours (intensive) after the first dose administration and 1, 4, 24, and 48 hours (sparse) after the last dose administration. Enrolled women were followed through pregnancy and up to 24 weeks postpartum. In women still taking RPT postpartum who agreed to provide breast milk samples (N = 22), breast milk and plasma PK sampling were performed during the first week after delivery at 3 hours post‐dose administration, during the second week postpartum at 6 hours post‐dose administration, and 24 hours post‐last dose administration. For infants (N = 13) born to women whose most recent RPT dose was administered within 72 hours of delivery, cord blood was collected during delivery at sites with capacity. Infants with birth weight greater than 1,000 g and without congenital malformations or medical complications had a single plasma PK sample collected within the first three days of life if the mother's most recent dose was within 72 hours of the blood draw. Infants were followed up to 24 weeks after birth. Figure 1 illustrates the PK study design. The trial was approved by local and collaborating institutional review boards, with interim reviews conducted by the IMPAACT Network.
Figure 1.
Schematic of IMPAACT 2001 study design. Women with LTBI and living with and without HIV were enrolled in their second trimester (Cohort 1) or third trimester (Cohort 2). Women in Cohort 1 were sampled intensively during the second trimester and sparsely during the third trimester. Women in Cohort 2 were sampled intensively during the third trimester and sparsely postpartum. In the postpartum period, breast milk was also sampled along with infant plasma and/or cord blood within 3 days of delivery. RPT, rifapentine; desRPT, desacetyl rifapentine; DOT, directly observed therapy; PO, by mouth; QWK, every week; PK, pharmacokinetics; HIV, human immunodeficiency virus.
Population PK analysis for maternal plasma and breast Milk
Longitudinal maternal plasma and breast milk RPT and desRPT PK data were analyzed using nonlinear mixed effects modeling. A multicompartment PK model was built for maternal plasma with biophase effect compartments for maternal breast milk (Figure 2 ). The maternal plasma PK model and parameter estimates are described elsewhere. 15 In brief, a 1‐compartment model with oral absorption described using transit compartments was fit to maternal plasma concentrations of RPT. HIV status (living with vs. without HIV) and pregnancy status (antepartum vs. postpartum) were evaluated as covariates on apparent clearance (CL/F). Maternal plasma concentrations of desRPT were subsequently added, and both RPT and desRPT PK data were modeled simultaneously. In this analysis, a breast milk matrix for RPT and desRPT was added to the plasma matrix compartments for RPT and desRPT, respectively, using biophase effect compartments. 19 With maternal plasma PK parameters fixed, rate parameters characterizing the delay of drug entry into breast milk from plasma (kP‐BM and kP‐BMmet) and ratio parameters quantifying the concentration ratio of drug in breast milk relative to plasma (RP‐BM and RP‐BMmet) were estimated for RPT and desRPT, respectively. A complete description of the model development for RPT and desRPT in maternal plasma and breast milk is presented in the Supplementary Materials.
Figure 2.
Schematic of multicompartment PK model for RPT and desRPT in plasma and breast milk matrices. One compartment plasma PK model with oral absorption following RPT administration and subsequent conversion to desRPT (white squares) with unique apparent clearance (CL/F) parameter estimated for each HIV status (with and without HIV) and pregnancy status (prepartum, postpartum) combination (gray circles). Plasma PK is linked to breast milk matrix via effect compartments (black ovals). RPT, rifapentine; desRPT, desacetyl rifapentine; ktr, transit compartment rate constant; ka, absorption rate constant; CL, clearance of RPT; V, volume of distribution of RPT; CLmet, clearance of desRPT; Vmet, volume of distribution of desRPT; PM, plasma, Mom; BM, breast milk, Mom, RPT; BM met, breast milk, Mom, desRPT; kP‐BM and kP‐BMmet, delay of drug entry into breast milk from plasma for RPT and desRPT, respectively; RP‐BM and RP‐BMmet, concentration ratio of drug in breast milk relative to plasma for RPT and desRPT, respectively.
Simulation of pre‐delivery maternal plasma PK and comparison to observed infant cord blood and plasma PK
The time and date of a woman's last RPT dose prior to delivery was obtained from study sites since medication administration was provided by site personnel via directly observed therapy. For each mother–infant pair, the time of infant cord blood and plasma PK collection relative to this maternal dosing time was calculated (i.e., within 72 hours). Each mother's plasma PK was simulated (NONMEM V7.4) with the final maternal plasma PK model for RPT and desRPT using a single dose administration and with consideration for all significant covariates (i.e., HIV status and pregnancy status). Observed RPT and desRPT concentrations for infants from cord blood and plasma were superimposed onto observed and individual predicted maternal plasma concentration‐time profiles. Subsequently, the concentration ratio for each mother–infant pair (i.e., CPinfant/CPmother) in each infant matrix (cord blood vs. plasma) and for each analyte (i.e., RPT vs. desRPT) at the time of infant PK sample collection was calculated and summarized.
RESULTS
Infant drug exposure through breastfeeding: Quantitative characterization of maternal Rifapentine exposure in breast Milk
Maternal plasma and breast milk PK data were available for 50 and 22 participants, respectively (Table 1 ). The longitudinal time course of RPT and desRPT concentrations in plasma have been previously reported 15 ; breast milk PK profiles are shown in Figure 3 . The multicompartment PK model for plasma and breast milk matrices was fit to 1,346 and 1,344 observations for RPT and desRPT, respectively, from 49 participants; a single participant was classified as an outlier and removed from the modeling analysis. The final population parameter estimates are shown in Table 2 ; parameters were fixed for the plasma matrix and estimated for the breast milk matrix. Breast milk parameters, specifically the rate describing the delay of drug entry into breast milk from plasma (kP‐BM (parent; RPT) and kP‐BMmet (metabolite; desRPT)) and the concentration ratio of drug in breast milk relative to plasma (RP‐BM and RP‐BMmet), were well estimated with high precision (i.e., relative standard errors (RSE) < 30%). Breast milk concentrations of RPT and desRPT were low compared to maternal plasma. Both the rates and ratios for RPT and desRPT were similar in magnitude (1.22/h for kP‐BM vs. 1.58/h for kP‐BMmet and 0.0236 for RP‐BM vs. 0.0411 for RP‐BMmet). The visual predictive check (VPC) (Figure S1 ), which provides an internal validation or objective assessment of how well the model predicts the data on which the model is conditioned, illustrates good alignment between the observed data and the model predictions with no significant model misspecifications. Additionally, HIV status was tested as a covariate on breast milk parameters (concentration ratio) in the multicompartment PK model and not found to be significant.
Table 1.
Description of observed PK data in IMPAACT 2001
| Population | Mother | Mother | Infant | Infant |
|---|---|---|---|---|
| Matrix | Plasma | Breast milk | Cord blood | Plasma |
| N participants | 50 | 22 | 5 | 7 |
| Concentration range (μg/mL) | ||||
| Rifapentine | 0.039–49.4 | 0.011–1.18 | 0.841–4.63 | 0.972–4.37 |
| Desacetyl rifapentine | 0.039–23.3 | 0.007–0.573 | 0.941–5.30 | 3.62–8.97 |
Concentration range reflects concentration of all participants over all time that data were collected for each matrix listed.
μg, microgram; mL, milliliter; N participants, number of participants.
Figure 3.
Observed concentration‐time profiles of (a) RPT and (b) desRPT in breast milk stratified by HIV status. μg, microgram; mL, milliliter; HIV, human immunodeficiency virus; RPT, rifapentine; desRPT, desacetyl rifapentine.
Table 2.
Parameter estimates of multicompartment PK model for RPT and desRPT stratified by matrices (plasma and breast milk)
| Matrix | Drug | Parameter | Description | Unit | Population estimate (RSE, %) | Inter‐individual variability, CV% (RSE, %) |
|---|---|---|---|---|---|---|
| Plasma | RPT | |||||
| CL/F | Apparent clearance | – | – | 9.70† | ||
| [HIV‐uninfected, antepartum] | – | L/h | 1.20† | – | ||
| [HIV‐uninfected, postpartum] | – | L/h | 1.53† | – | ||
| [HIV‐infected, antepartum] | – | L/h | 1.56† | – | ||
| [HIV‐infected, postpartum] | – | L/h | 1.60† | – | ||
| V/F | Apparent volume | L | 29.4† | 16.6† | ||
| F | Bioavailability | – | 1† | 25.1† | ||
| MTT | Mean transit time | h | 2.80† | 46.0† | ||
| Additive error | Residual error | μg/mL | 0.326† | – | ||
| Proportional error | Residual error | % CV | 35.0† | – | ||
| Plasma | desRPT | |||||
| CLmet/(F*fm) | Apparent clearance | L/h | 2.75† | 22.9† | ||
| Vmet/(F*fm) | Apparent volume | L | 17.8† | – | ||
| Additive error | Residual error | μg/mL | 0.0677† | – | ||
| Proportional error | Residual error | % CV | 42.2† | – | ||
| Breast milk | RPT | |||||
| KP‐BM | Rate of drug entry into breast milk from plasma | 1/h | 1.22 (12) | ‐ | ||
| RP‐BM | Concentration ratio between breast milk and plasma | ‐ | 0.0236 (6) | 17.9 (5) | ||
| Additive error | Residual error | μg/mL | 0.138 (8) | ‐ | ||
| Proportional error | Residual error | % CV | ‐ | ‐ | ||
| Breast milk | desRPT | |||||
| KP‐BMmet | Rate of drug entry into breast milk from plasma | 1/h | 1.58 (4) | 122.9 (18) | ||
| RP‐BMmet | Concentration ratio between breast milk and plasma | – | 0.0411 (2) | – | ||
| Additive error | Residual error | μg/mL | 0.0049 (11) | – | ||
| Proportional error | Residual error | % CV | 29.9 (1) | – |
%CV, percent coefficient of variation; μg, microgram; †, fixed estimate; BM, breastmilk; desRPT, desacetyl rifapentine; F, bioavailability; fm, fraction metabolized; HIV, human immunodeficiency virus; h, hour; L, liter; met, metabolite; mL, milliliter; P, plasma; RPT, rifapentine; RSE, relative standard error; V, volume of distribution.
Infant drug exposure following maternal dosing: Simulated maternal Rifapentine exposure in plasma relative to observed infant Rifapentine exposure in cord blood and plasma
Five and eight infants provided cord blood and plasma drug concentration levels, respectively, for RPT and desRPT (Table 1 ). PK was collected in only cord blood for one infant, only plasma for four infants, and in both cord blood and plasma for four infants. The corresponding infant RPT and desRPT concentration ranges are shown in Table 1 . A single mother–infant pair was removed from the analysis as the infant plasma RPT and desRPT concentrations were below the lower limit of quantification and the maternal dosing time was recorded to occur after (not prior to) delivery. For each mother–infant pair, Figure 4 compares the observed infant RPT and desRPT concentrations in plasma and/or cord blood relative to their mother's predicted exposure in plasma following her last dose administration pre‐delivery. Table 3 summarizes the median RPT and desRPT ratios of observed infant concentrations to predicted maternal plasma concentration (i.e., CPinfant/CPmother) at the time of infant PK sample collection stratified by infant matrix (cord blood or plasma). Within 72 hours of maternal dosing, median concentration ratios in infant cord blood and plasma were approximately 0.3–0.6‐fold lower for RPT and similar (cord blood) or higher (plasma) for desRPT compared to their mothers (Table 3 ).
Figure 4.
Predicted maternal plasma exposure following last RPT dose administration pre‐delivery relative to observed infant cord blood and plasma RPT and desRPT concentrations. Individual predicted (a) RPT (top row) and (b) desRPT (bottom row) plasma PK profiles of mothers (black solid line) overlaid with observed RPT and desRPT PK data from maternal plasma (black filled circles), infant cord blood (blue filled circles) and/or plasma (red filled circles), and maternal breast milk (purple filled circles). RPT, rifapentine; desRPT, desacetyl rifapentine; μg, microgram; mL, milliliter; PK, pharmacokinetics; LLOQ, lower limit of quantification. Vertical dashed lines demarcate 72 hours (gray) and time of delivery (orange) post‐last maternal RPT dose pre‐delivery.
Table 3.
Calculated RPT and desRPT median ratios of observed infant cord blood and plasma concentrations to predicted maternal plasma concentration
| Drug | Matrix | Concentration Ratio (CPinfant/CPmother) Median (Range) |
|---|---|---|
| Rifapentine | Cord blood | 0.280 (0.163–0.455) |
| Rifapentine | Plasma | 0.612 (0.343–0.915) |
| Desacetyl rifapentine | Cord blood | 0.493 (0.186–1.16) |
| Desacetyl rifapentine | Plasma | 1.83 (1.53–3.73) |
RPT, rifapentine; desRPT, desacetyl rifapentine.
DISCUSSION
Pregnant women are among the highest‐risk groups for development of TB disease from LTBI. Providing preventative therapy to this population during their pregnancy and then continuing therapy postpartum during lactation has proven challenging. There is a lack of robust information about drug concentrations in cord blood and breast milk (and, it follows, infant exposures) for the most oft‐used regimens in current clinical practice, including RPT. In fact, the RPT prescribing information remarks on the absence of parent or metabolite data on secretion into human or animal breast milk as well as effects on breastfeeding infants. 2 To help fill this knowledge gap, IMPAACT 2001 provided the first clinical data evaluating 3HP for TB prevention in pregnant women, with PK data collection for RPT and desRPT PK in plasma of pregnant and postpartum women, breast milk of postpartum women, and both plasma and cord blood of infants following delivery. Collectively, these data enable an understanding of infant exposure via in utero cord blood delivery and via breastfeeding among mothers taking TB preventative treatment. Using a modeling and simulations framework for PK analysis, an approach that is especially valuable for sparse and difficult‐to‐collect samples (such as breast milk or cord blood), we quantitatively characterized the PK of RPT and desRPT in breast milk relative to plasma in postpartum women, and established how RPT and desRPT concentrations in infants compare to their mother's exposure following her last pre‐delivery dose administration.
To link maternal plasma PK with maternal breast milk PK, we extended the previously developed plasma PK model, 15 which contains distinct compartments for RPT and desRPT, to include biophase effect compartments for RPT and desRPT measured in breast milk. The effect compartment model structure is used in nonlinear mixed effects modeling to describe the movement (no information on directionality) and equilibration of drug levels between the plasma, where concentrations are typically (and easily) collected, and another site in the body (e.g., cellular or tissue matrix). Most drugs enter breast milk by passive diffusion and their breast milk concentration is directly proportional to their plasma concentration. 20 The first‐order effect compartment rate constant (ke0) characterizes the rate at which the effect compartment is equilibrating; a larger magnitude indicates that the concentration in the effect compartment closely matches the concentration in the plasma compartment. 21 The multicompartment PK model estimated kP‐BM and kP‐BMmet as 1.22 hour−1and 1.58 hour−1, respectively, indicating that the breast milk concentration closely follows the plasma concentration. Using the rate estimates, an equilibration half‐life (defined as ln2/ke0) 21 can be derived for RPT and desRPT as 0.568 hr and 0.439 hr, respectively, implying rapid equilibration of the effect compartment for both parent and metabolite. The milk‐to‐plasma ratio, estimated as the concentration ratio of drug in breast milk relative to plasma, describes the extent of drug excretion into breast milk. 20 The multicompartment PK model estimated RP‐BM and RP‐BMmet to be much lower than one at 0.0236 and 0.0411, respectively. The low magnitude indicates that breast milk concentrations of RPT and desRPT are several folds lower than the corresponding maternal plasma concentrations, making infant exposures low and likely not clinically significant. A similarly low milk‐to‐plasma ratio (0.16–0.23) has been observed for rifampicin, another member of the rifamycin family of anti‐TB antibiotics. 22
To understand infant drug exposure in utero following maternal dosing, we simulated the maternal plasma concentration‐time profiles for a mother's last RPT dose prior to delivery and overlaid her infant's PK data from cord blood and plasma matrices. The calculated median ratios of observed infant cord blood and plasma concentrations to predicted maternal plasma concentration (i.e., CPinfant/CPmother) at the time of infant PK sample collection showed approximately 0.3–0.6‐fold lower levels for RPT in infants compared to mothers for both matrices and similar (cord blood) or higher (plasma) levels for desRPT. With this level of drug exposure from their mothers, no infants had treatment‐related serious adverse events at birth. 15 Of infants born to mothers taking 3HP at the time of delivery, two (4%) experienced grade 2 or higher prothrombin time (PT) values that were deemed possibly related to the study drug; neither infant demonstrated clinical signs of bleeding and repeat tests revealed normal PT values. 15 Therefore, the measured infant concentrations in cord blood and plasma likely did not result in any clinically significant safety events.
Assuming the standardized milk consumption rate used by FDA of 150 mL/kg/day up to 200 mL/kg/day in early infancy and a typical infant weight range of 2.5–4 kg, along with the maximum observed RPT breast milk concentration in IMPAACT 2001 (1.18 μg/mL; Table 1 ), the maximum ingested amount of RPT in a day for an infant is 944 μg or approximately 1 mg (calculated using the maximum infant weight and maximum infant milk consumption rate). The relative infant dose (RID), which is the ratio of the infant weight‐adjusted dose received via breast milk (mg/kg/day) to the maternal weight‐adjusted dose (mg/kg/day), can also be calculated. A conservative estimate of the infant weight‐adjusted dose is 0.236 mg/kg/day obtained using the maximum observed RPT breast milk concentration in IMPAACT 2001 (1.18 μg/mL; Table 1 ) and the maximum standardized milk consumption rate (200 mL/kg/day). An average estimate of the maternal weight‐adjusted dose is approximately 15 mg/kg/day obtained using the maternal dose of 900 mg/day and the average maternal body weight in IMPAACT 2001 (61 kg). The ratio of 0.236 mg/kg/day to 15 mg/kg/day is 0.0157 or 1.57%, the upper bound of the RID, which is well below the 10% reference point generally considered safe for an infant. 23
While an RID of <10% is a standard value accepted by the breastfeeding medicine community representing minimal risk, it is also important to contextualize the RID within the existing knowledge of pediatric safety for the given drug. The current WHO‐recommended dose of RPT is 75 mg for children weighing 3–5.9 kg who are under 3 months of age and 105 mg for children weighing 3–5.9 kg who are over 3 months of age. These values are well above the estimated maximum ingested dose of RPT in a day for a breastfeeding infant (<1 mg). The current doses of RPT for children recommended in the WHO guidance, without need for routine laboratory safety monitoring, are based on the favorable safety, tolerability, and PK data from TBTC Study 35 (manuscript under review). 24 , 25 Therefore, RPT concentrations in breast milk appear to be far lower than the infant dose, with an RID significantly lower than 10%.
The quantitative framework used in this analysis is informed by the data from IMPAACT 2001. Even with sparse infant PK sample collection (one or two samples only per participant), we initially extended the multicompartment PK model to include effect compartments for infant cord blood and plasma RPT and desRPT; however, estimating model parameters with high certainty was challenging. As more infant PK data are collected, 25 the mother–infant modeling framework can be revisited to improve our understanding about infant exposure (i.e., rate of equilibration between maternal plasma and infant plasma or cord blood compartments as well as mother‐to‐infant concentration ratios) following maternal dosing and breastfeeding. It may also provide an understanding of whether infant desRPT concentration levels are a reflection of maternal levels or infant metabolism, helping to explain why infant desRPT in utero exposure compared to maternal levels is similar in cord blood and higher in plasma. As more infant safety data are collected and if a favorable safety profile continues to be observed, the model may also be used to simulate infant dosing regimens that confer protection from TB infection.
In conclusion, the IMPAACT 2001 study, which employed a longitudinal design with matching maternal–infant pairs, provided the first PK data for RPT and desRPT in maternal breast milk and in infants born to mothers living with and without HIV and taking the once‐weekly 3HP TB prevention regimen. Using modeling and simulations, we estimated maternal milk‐to‐plasma ratios for RPT and desRPT that suggest low potential for infant exposure following breastfeeding. Infant exposure in utero following maternal dosing is also ~50% lower than maternal levels. Taken together, the data suggest that 3HP administration during pregnancy or lactation does not result in clinically significant infant exposure to RPT or desRPT. Given a favorable infant safety profile following maternal dosing and breastfeeding, and the risk for poor maternal and child health outcomes from untreated TB, the benefit‐to‐risk ratio for RPT use in pregnant and postpartum women remains high. Data generated from larger studies 26 are needed to confirm the safety of 3HP for pregnancy and infant outcomes, and a quantitative mother–infant modeling framework may continue to be used to further our understanding of infant exposure from maternal LTBI treatment.
FUNDING
This work by the International Maternal Pediatric Adolescent AIDS Clinical Trials Network (IMPAACT) was supported by the National Institute of Allergy and Infectious Diseases (NIAID) with co‐funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) and the National Institute of Mental Health (NIMH), all components of the National Institutes of Health (NIH), under Award Numbers UM1AI068632–15 (IMPAACT LOC), UM1AI068616–15 (IMPAACT SDMC) and UM1AI106716–15 (IMPAACT LC), and by NICHD contract number HHSN275201800001I. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. PJ was supported by the National Institute of General Medical Sciences (NIGMS) [grant T32 GM007546]. KED is supported by NIAID grant 5R37AI184117.
CONFLICT OF INTEREST
PJ is currently an employee of Regeneron Pharmaceuticals, Inc. and owns stock or stock options in the company. All other authors declare no competing interests for this work.
AUTHOR CONTRIBUTIONS
PJ, KED, and JSM wrote the manuscript; RS, JSM, Sa P, St P, AG, KED, ET, PB, GM, SB, LW, JN, TC, VR, DL, and AC designed the research; PJ, RS, KED, and JSM performed the research; PJ, NZ, RS, KED, and JSM analyzed the data.
Supporting information
Figure S1.
ACKNOWLEDGMENTS
The authors would like to thank the participants of the IMPAACT 2001 study and the contributions of the site investigators and staff who conducted the study.
Previous Presentation: Measuring TB drug levels in breast milk: results of IMPAACT 2001. 55th Union World Conference on Lung Health, Bali, 2024. Int J Tuberc Lung Dis 2024: 28 (11 Suppl 1): S1 – S814.
References
- 1. Mathad, J.S. , Yadav, S. , Vaidyanathan, A. , Gupta, A. & LaCourse, S.M. Tuberculosis infection in pregnant people: current practices and research priorities. Pathogens. 11, 1–18 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Sanofi Aventis . Priftin (Rifapentine) [Package Insert] (US Food and Drug Administration, Morristown, NJ, 2014; ) <https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/021024s017s018lbl.pdf>. [Google Scholar]
- 3. Mathad, J.S. & Gupta, A. Tuberculosis in pregnant and postpartum women: epidemiology, management, and research gaps. Clin. Infect. Dis. 55, 1532–1549 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Gupta, A. et al. Maternal tuberculosis: a risk factor for mother‐to‐child transmission of human immunodeficiency virus. J Infect Dis 203, 358–363 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Gupta, A. et al. Postpartum tuberculosis incidence and mortality among HIV‐infected women and their infants in Pune, India, 2002‐2005. Clin. Infect. Dis. 45, 241–249 (2007). [DOI] [PubMed] [Google Scholar]
- 6. Cantwell, M.F. et al. Brief report: congenital tuberculosis. N. Engl. J. Med. 330, 1051–1054 (1994). [DOI] [PubMed] [Google Scholar]
- 7. Jana, N. , Vasishta, K. , Jindal, S.K. , Khunnu, B. & Ghosh, K. Perinatal outcome in pregnancies complicated by pulmonary tuberculosis. Int. J. Gynaecol. Obstet. 44, 119–124 (1994). [DOI] [PubMed] [Google Scholar]
- 8. Jana, N. , Vasishta, K. , Saha, S.C. & Ghosh, K. Obstetrical outcomes among women with extrapulmonary tuberculosis. N. Engl. J. Med. 341, 645–649 (1999). [DOI] [PubMed] [Google Scholar]
- 9. Lin, H.C. , Lin, H.C. & Chen, S.F. Increased risk of low birthweight and small for gestational age infants among women with tuberculosis. BJOG 117, 585–590 (2010). [DOI] [PubMed] [Google Scholar]
- 10. WHO Consolidated Guidelines on Tuberculosis (World Health Organization, Geneva, Switzerland, 2024) <https://www.who.int/publications/i/item/9789240096196>. [Google Scholar]
- 11. Sterling, T.R. et al. Guidelines for the treatment of latent tuberculosis infection: recommendations from the National Tuberculosis Controllers Association and CDC. Morb. Mortal. Wkly. Rep. Cent. Dis. Control Prev. 69, 1–11 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Weiner, M. et al. Pharmacokinetics of rifapentine at 600, 900, and 1,200 mg during once‐weekly tuberculosis therapy. Am. J. Respir. Crit. Care Med. 169, 1191–1197 (2004). [DOI] [PubMed] [Google Scholar]
- 13. Dooley, K.E. et al. Safety and pharmacokinetics of escalating daily doses of the antituberculosis drug rifapentine in healthy volunteers. Clin. Pharmacol. Ther. 91, 881–888 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hibma, J.E. et al. Rifapentine population pharmacokinetics and dosing recommendations for latent tuberculosis infection. Am. J. Respir. Crit. Care Med. 202, 866–877 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Mathad, J.S. et al. Pharmacokinetics and safety of 3 months of weekly Rifapentine and isoniazid for tuberculosis prevention in pregnant women. Clin. Infect. Dis. 74, 1604–1613 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Beers, J.L. , Hebert, M.F. & Wang, J. Transporters and drug secretion into human breast milk. Expert Opin. Drug Metab. Toxicol. 21, 409–428 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Hudson, R.E. et al. Drug exposure during pregnancy: current understanding and approaches to measure maternal‐fetal drug exposure. Front. Pharmacol. 14, 1111601 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Dallmann, A. , Liu, X.I. , Burckart, G.J. & van den Anker, J. Drug transporters expressed in the human placenta and models for studying maternal‐fetal drug transfer. J. Clin. Pharmacol. 59(Suppl 1), S70–S81 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Sheiner, L.B. , Stanski, D.R. , Vozeh, S. , Miller, R.D. & Ham, J. Simultaneous modeling of pharmacokinetics and pharmacodynamics: application to d‐tubocurarine. Clin. Pharmacol. Ther. 25, 358–371 (1979). [DOI] [PubMed] [Google Scholar]
- 20. Loebstein, R. , Lalkin, A. & Koren, G. Pharmacokinetic changes during pregnancy and their clinical relevance. Clin. Pharmacokinet. 33, 328–343 (1997). [DOI] [PubMed] [Google Scholar]
- 21. Upton, R.N. & Mould, D.R. Basic concepts in population modeling, simulation, and model‐based drug development: part 3‐introduction to pharmacodynamic modeling methods. CPT Pharmacometrics Syst. Pharmacol. 3, e88 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lenzi, E.S.S. Preliminary observations on the use of a new semisynthetic rifamycin derivative in gynecology and obstetrics. Atti Accad. Lancisiana Roma 13, 87–94 (1969). [Google Scholar]
- 23. Wang, J. et al. Evaluation of the safety of drugs and biological products used during lactation: workshop summary. Clin. Pharmacol. Ther. 101, 736–744 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. World Health Organization . TB Knowledge Sharing Platform Operational Handbook (World Health Organization, Geneva, Switzerland, 2024) <https://tbksp.who.int/en/node/1271>. [Google Scholar]
- 25. Salazar‐Austin, N. et al. DOLPHIN‐KIDS: Phase I/II Trial of Dolutegravir With 3HP for TB Prevention in Children With HIV (ABSTRACT 150) CROI; February 22‐25, 2026; Denver, CO.
- 26. Mathad, J.S. et al. DOLPHIN Moms: Pharmacokinetics and HIV Suppression with Dolutegravir and 1HP or 3HP in Pregnancy CROI; March 9‐12, 2025; San Francisco, CA.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1.