2025 Volume 48 Issue 4 Pages 337-343
The potential risks to the fetus or infant should be evaluated before initiating pharmacotherapy during pregnancy and breastfeeding. However, safety information and experiences during pregnancy and breastfeeding are often lacking because these populations are generally excluded from clinical drug development studies. Perinatal mental health is important. Based on quantitative and toxicological analyses, we focused on medications used in psychiatry and neurology during the perinatal period. As the placenta serves as a temporary but crucial organ for ensuring successful pregnancy and appropriate fetal growth, we assessed the effects of antiepileptic drugs (AEDs) on placental functions such as transport mechanisms, nutrient transport, and trophoblast differentiation. Several AEDs have been suggested to be transported to the placenta via carrier-mediated pathways in a series of studies. Valproic acid, known to pose several risks to the fetus, affects the gene expression of nutrient transporters and trophoblast differentiation. Furthermore, we established several quantitative methods, such as those for antianxiety and hypnotic drugs, to evaluate the safety of pharmacotherapy during breastfeeding using liquid chromatography-tandem mass spectrometry. The validated methods were applied to clinical samples donated by lactating women. In a series of studies, the importance of choosing a suitable method for sample preparation for each biological matrix has been highlighted. The results obtained from the clinical samples suggest the possibility of differences in transfer properties among drugs categorized in the same class. Furthermore, this research emphasizes the critical need to assess breast milk transfer in human studies because species differences have been suggested in some cases.
The potential risks to the fetus or infant should be evaluated before initiating pharmacotherapy during pregnancy and breastfeeding. Medication use is common during these periods. In Japan, 78.4% of pregnant women use drugs or supplements before pregnancy, 57.1% during the first 12 weeks, and 68.8% after the first 12 weeks.1) Furthermore, a questionnaire survey conducted in Japan showed that many women have used prescription drugs while their children are in infancy.2) However, safety information and experiences during pregnancy and breastfeeding are often lacking because these populations are generally excluded from clinical studies on drug development. Assessing drug exposure levels and associated factors, along with toxicological impacts, is crucial for estimating potential risks to the fetus or infant. Several diseases require the continuation of medication during the perinatal period. In particular, we focused on medications for psychological and neurological disorders and investigated their effects during pregnancy and breastfeeding using quantitative and toxicological analyses.
The placenta is a transient yet vital organ that supports a healthy pregnancy and ensures proper fetal development. It performs multiple critical functions, including nutrient delivery to the fetus, metabolic processing, waste elimination, gas exchange, and the secretion of essential hormones. Placental trophoblasts originate from the trophectoderm and play important roles in placental function.3) After proper apposition to maternal tissues, cytotrophoblast (CT) cells differentiate and fuse to generate syncytiotrophoblasts (STs), which are important constituents of the blood–placental barrier.4)
Epilepsy is among the most prevalent neurological disorders worldwide.5) In recent years, the approval of various antiepileptic medications has led to significant advances in seizure management in patients with epilepsy. For most pregnant women with epilepsy, the continuation of antiepileptic treatment is essential because uncontrolled seizures can negatively affect both maternal and fetal health.6) Additionally, unintended pregnancies have been observed in women with epilepsy.7) Consequently, evaluating the toxicological effects of antiepileptic drugs on placental function is crucial for assessing the potential risks to fetal development. In this study, we multidirectionally assessed the effects of antiepileptic drugs (AEDs) on placental function, including (1) transport mechanisms, (2) nutrient transport, and (3) differentiation and hormone secretion.
2.1. Transport Mechanisms AEDs in Trophoblast ModelWe investigated the accumulation of several AEDs, such as gabapentin (GBP), lamotrigine (LTG), levetiracetam, and topiramate, in the trophoblastic models BeWo and JEG-3. The accumulation levels of AEDs were not consistent with predictions based on chemical properties.8) In particular, intracellular concentrations of GBP and LTG were higher than their extracellular concentrations.9,10) In BeWo cells, various transporters have been detected at the gene level,11) and l-type amino acid transporter 1 (LAT1/SLC7A5) is the main contributor to GBP transport.9) GBP transport is strongly inhibited by endogenous LAT1 substrates, including amino acids and thyroid hormones. Furthermore, our research suggests that GBP transport is facilitated by intracellular amino acids, such as histidine and methionine.12) This study has partially elucidated the transport properties of drugs, and further studies may lead to the prediction of fetal risks and the development of strategies to prevent fetal exposure.
2.2. Effects of AEDs on the Expression of Nutrient TransportersWe investigated the effect of 16 AEDs on folate transport, a process essential for fetal development. Short-term exposure to AEDs does not alter folate uptake by BeWo cells.13) In contrast, prolonged exposure to valproic acid (VPA) affected folate uptake by upregulating the expression of folate carriers (FRα/FOLR1 and PCFT/SLC46A1). Given that VPA inhibits histone deacetylase (HDAC) activity, we examined its role in altering gene expression. VPA treatment enhances acetyl-histone H3 (Lys9/Lys14) expression in BeWo cells.14) Our results indicate that pharmacological inhibition of class I HDACs influences folate transporter expression in BeWo cells. Exposure to certain AEDs has been linked to an increased risk of adverse fetal development. Notably, the use of VPA during the periconceptional period increases the likelihood of fetal malformations in a dose-dependent manner.6) Additionally, VPA use has been linked to a higher risk of cognitive impairment and autism spectrum disorders in the offspring.15,16) In our in vivo experiments, we examined the effects of oral administration of VPA on the placenta. Repeated VPA exposure decreases placental weight and downregulates Folr1 expression during late gestation.17) In this animal model, VPA affected various genes that contributed to the transport of nutrients and endogenous substrates, such as amino acids, carnitine, thyroid hormones, cyclic nucleotides, and prostanoids.18) Furthermore, our results suggest that the effects of VPA on the placenta differ following gestational development. The insights obtained in this study may contribute to the elucidation of the mechanisms underlying the onset of VPA toxicity in the placenta and fetus.
2.3. Effects of AEDs on Trophoblast Differentiation and Hormone SecretionThis study examined the effects of eight AEDs on ST. VPA exposure influenced the expression of differentiation-related genes in differentiated BeWo cells in a dose-dependent manner.19) In contrast, differentiation was unaffected by the clinical concentrations of the other AEDs. VPA also altered the expression of differentiation-related genes in human trophoblast stem-derived ST (ST-TSCT) (Fig. 1, left). Additionally, VPA exposure reduced the secretion of human chorionic gonadotropin (hCG) and cell fusion in BeWo and ST-TSCT. We further investigated the correlation between placental and neonatal parameters and the expression of syncytialization markers in human-term placentas. Notably, MFSD2A expression was positively associated with neonatal outcomes, including body weight, head circumference, chest circumference, and placental weight (Fig. 1, right). These results provide insights into the toxicological effects of VPA and the potential risks to placental and fetal development.
(Left) Effects of valproic acid (VPA) on cell fusion and gene expression of differentiation markers in human trophoblast stem (TS) cells. TSCT cells were first cultured for 3 d in TS-medium, followed by a 72-h exposure to differentiation medium (ST-medium) containing 1 mM VPA. The fusion index, which represents the percentage of nuclei in the syncytia, was calculated by counting the nuclei. **p < 0.01 compared with the TSCT. ††p < 0.01 compared with ST-TSCT. The mRNA expression levels of the differentiation markers were investigated using qPCR. *p < 0.05, **p < 0.01 compared with the control. (Right) Relationship between gene expression of differentiation markers and neonatal/placental parameters (body weight, stature, head circumference, chest circumference, and placental weight). The figures were partially reproduced using Toxicol. Appl. Pharmacol., 474, 116611 (2023) with permission.
Breastfeeding offers significant advantages to both mothers and infants. In infants, it enhances immune function and reduces the risk of infection, whereas in mothers, it lowers the likelihood of developing breast and ovarian cancer.20) Despite these benefits, breastfeeding is often avoided when the mother is taking medications because of insufficient data on drug transfer to breast milk and its impact on nursing infants.21) To estimate risks during breastfeeding, evaluating infant drug exposure is essential for estimating risks during breastfeeding. The milk/plasma (M/P) ratio and the relative infant dose (RID) are commonly used risk assessment metrics for nursing infants. The M/P ratio measures the extent of drug transfer to breast milk, and the RID represents the drug exposure level in infants.22)
Liquid chromatography-tandem mass spectrometry (LC/MS/MS) is a highly sensitive and specific technique for measuring drug concentrations in biological matrices. Breast milk has a diverse composition of carbohydrates, proteins, lipids, and essential vitamins. Lipids comprise approximately 3–5% of breast milk, with individual variability in content.23,24) Additionally, breast milk contains phospholipids, which can contribute to ion suppression during analysis.25) To address the matrix effects, a consistent and reproducible sample preparation protocol is essential. The development of robust methods for drug quantification in both plasma and breast milk is critical for generating reliable safety data.
Therefore, we developed quantification methods using LC/MS/MS and investigated the concentrations of drugs in breast milk, such as antiepileptic drugs, antipyretic and analgesic drugs, and antianxiety and hypnotic drugs.26–30) In this review, I present research on antianxiety and hypnotic drugs, such as benzodiazepines (BZDs) and orexin receptor antagonists (ORAs).
3.1. BZDsPregnant and postpartum women frequently experience anxiety and insomnia. Adverse consequences may arise if maternal anxiety disorders remain untreated or if psychotropic medications are abruptly discontinued.31) Thus, understanding the transfer of psychoactive medications to breast milk and their potential effects on nursing infants is essential for the well-being of both mothers and children. BZDs are prescribed to manage insomnia during the perinatal period. They are generally characterized by a low molecular weight, high lipophilicity, and strong plasma protein binding. However, these properties vary among the BZDs. Comprehensive data regarding the transfer rates of BZDs in breast milk are limited. To address this, we developed a validated LC/MS/MS-based method for quantifying BZDs in human plasma and breast milk and analyzed clinical samples. Sample preparation involved a straightforward liquid–liquid extraction (LLE) process using ethyl acetate. In this study, ion suppression was detected in biological samples, but the use of a deuterium-labeled internal standard (IS) successfully addressed the variability caused by matrix effects.26) The developed method was validated through a validation assessment.
Using the established method, we quantified the concentrations of drugs in breast milk and plasma, which were provided by lactating women undergoing treatment with BZDs, including alprazolam, bromazepam, brotizolam, clonazepam, clotiazepam, etizolam, flunitrazepam, lorazepam, and CM7116 (a metabolite of ethyl loflazepate). The M/P ratios estimated using quantification data obtained several days after delivery are shown in Fig. 2. This study included 19 lactating women. Some lactating women use multiple BZDs. Plasma and breast milk were obtained at the trough and peak (corresponding to the estimated maximum plasma concentration) following oral administration of BZDs. The M/P ratios at peak and trough levels were not significantly different. Previous studies have shown that the concentration–time profiles of alprazolam in breast milk and plasma follow similar patterns.32) Our study also suggests that the plasma and breast milk profiles of other BZDs are similar. The M/P ratios for all the BZDs investigated in this study were below 1, suggesting that the drugs were not concentrated in breast milk. However, the M/P ratios were not the same for all the BZDs. Typically, compounds characterized by low ionization, low molecular weight, minimal plasma protein binding, and high lipophilicity exhibit a greater propensity for transfer into breast milk.22) In our preliminary analysis, higher protein binding in the plasma was associated with a lower transfer of BZDs to breast milk. Future studies are needed to investigate the factors contributing to the transfer of BZDs to breast milk. In many BZD cases, the RID remained below 10%27) and was within the clinically acceptable threshold.22) In contrast, the RIDs for patients taking several long-acting BZDs, such as ethyl loflazepate, exceeded these criteria. Further data are required to assess the safety of breastfeeding.
Milk/plasma (M/P) ratios calculated from the quantification results were plotted. “–” indicates the mean value of the data. LLOQ, lower limit of quantification. Figures were reproduced in part from J. Pharm. Biomed. Anal., 168, 83–93 (2019), Breastfeed. Med., 16: 424–431 (2021), and unpublished data.
Recently, new classes of hypnotics, including melatonin receptor agonists and ORAs, have been introduced into clinical practice. The U.S. Food and Drug Administration (FDA) approved suvorexant as the first ORA in 2014. Lemborexant is an ORA that was approved by the FDA in 2019. Additionally, newer ORAs, such as daridorexant and vornorexant, have been developed and are undergoing review for marketing authorization in Japan. Between 2010 and 2019, the proportion of new prescriptions for insomnia treatment in Japan shifted, with a decrease in BZDs and an increase in ORAs from 0 to 20.2%.33) Therefore, the use of ORAs is expected to increase among women of childbearing age; however, studies addressing their perinatal applications remain limited.
In this study, we developed and validated a method for quantifying ORAs (suvorexant and lemborexant) in human breast milk and plasma using LC/MS/MS. In the preliminary analysis, the matrix factor (MF) was calculated to assess the matrix effects. Plasma samples were pretreated with LLE using ethyl acetate, as matrix effects were minimal in the plasma (Fig. 3, left). However, predominant ion suppression was observed in breast milk. Although the matrix effect for suvorexant could be corrected using suvorexant-d6 as an IS, lemborexant showed limited correction. Attempts to use alternative extraction solvents with varying properties did not improve MF. Consequently, the breast milk sample pretreatment was switched to solid-phase extraction (SPE). Phospholipids and proteins, which are known contributors to ion suppression,25) are abundant in breast milk. Therefore, phospholipid removal columns (InertSep® Phospholipid Remover and Oasis® PRiME HLB) were utilized, resulting in improved MF values (Fig. 3, right). These findings suggest that phospholipids play a key role in suppressing ORAs in breast milk. This study highlights the importance of selecting a suitable sample preparation method for each biological matrix.
MF for each analyte and internal standard (IS: suvorexant-d6) was calculated by measuring the peak area in the presence of the matrix, and the peak area in the absence of the matrix. Figures were reproduced in part from J. Pharm. Biomed. Anal., 251, 116432 (2024) with permission.
The developed method was used to analyze clinical samples, allowing us to estimate the M/P ratio and the RID of the ORAs (Table 1). For lemborexant, the M/P ratio was below 1.0, suggesting that the drug did not accumulate in breast milk. The RID of lemborexant was calculated at 1.05%. Another group reported that the RID of lemborexant was 1.96% in a study of eight women.34)
| Case no. | Drug | Maternal dose (mg/d) |
Days after delivery |
Time after dose |
Concentration (ng/mL) | M/P ratio |
RID (%) | |
|---|---|---|---|---|---|---|---|---|
| Plasma | Breast milk | |||||||
| 1 | Lemborexant | 5 | A few days | 70 min | 9.97 | 4.69 | 0.47 | — |
| 565 min | 7.99 | 5.12 | 0.64 | 1.05 | ||||
| 2 | Suvorexant | 20 | A few days | 0 min | 67.3 | 1.36 | 0.02 | — |
| 70 min | 133 | 2.85 | 0.02 | 0.11 | ||||
| 3 | Suvorexant | 20 | A few days | Next day | 37.4 | 3.47 | 0.09 | 0.17 |
| One month | Next day | 80.3 | 4.09 | 0.05 | 0.20 | |||
The M/P ratio was <0.1 for suvorexant, indicating minimal drug transfer into breast milk. The concentration of suvorexant in the breast milk of rats is 9 times higher than that in the plasma of the dam.35) These findings suggest significant species differences in the transfer of suvorexant into breast milk between rodents and humans. A previous study documented the differences in the transferability of various drugs to breast milk between humans and mice.36) This study emphasizes the critical need to assess breast milk transfer in humans.
The RIDs for suvorexant and lemborexant were <10% and remained within the clinically accepted threshold. Although additional studies are necessary, these findings indicate the potential of ORAs to be used during breastfeeding under certain conditions. Suvorexant exhibited a lower M/P ratio than lemborexant, highlighting the possible differences in the M/P ratio across ORA drugs. Further studies are needed to investigate the mechanisms and factors contributing most to breast milk transfer.
This review presents studies on assessing medications during pregnancy and breastfeeding based on quantitative and toxicological analyses. Regarding pharmacotherapy during pregnancy, we conducted multidirectional assessments of the effects of AEDs on placental function. However, the extrapolation of data obtained from basic assessments directly into clinical settings should be performed with caution. Our study used cell lines (choriocarcinoma or placental stem), pregnant rats, and human-term placentas obtained from donors who gave birth via cesarean section. Each model has advantages and disadvantages. For example, choriocarcinoma cell lines such as BeWo can proliferate and differentiate, making them easy to manipulate. However, these cells are cancerous and may differ from normal cells in their toxicological mechanisms. Although animal models can be used to study the effects of drugs in vivo, there may be species differences in the placental structure and transporter expression. Indeed, we observed that these tendencies were inconsistent with the findings of in vitro studies using a human choriocarcinoma cell line and in vivo studies using pregnant rats. In analyzing transporter and syncytialization markers in the human placenta, we could not precisely unify patient characteristics. In addition, we investigated only term placentas; transporter expression levels must be considered to vary with gestational age. Regarding pharmacotherapy during breastfeeding, we revealed the properties of milk transfer of prescribed drugs, such as antianxiety and hypnotic drugs, using validated quantitative methods. Our study highlighted the importance of selecting a suitable sample preparation method for each biological matrix. The results obtained from the clinical samples suggest the possibility of differences in the transfer properties of the same categorized drugs. Furthermore, this study emphasizes the critical need to assess breast milk transfer in human studies because species differences have been suggested. This study has several limitations. In our study, plasma and breast milk samples were collected at a few points due to patient burden. Additionally, the detailed and long-term effects on breastfed infants and their plasma concentrations have not been evaluated. However, more evidence is needed to establish the safety of these medications during breastfeeding. Further studies using various approaches are required to improve the health of women of reproductive age.
I thank Dr. Masaki Kobayashi, Professor, Laboratory of Clinical Pharmaceutics & Therapeutics, Division of Pharmaceuticals, Faculty of Pharmaceutical Sciences, Hokkaido University, Japan, and Dr. Ken Iseki, Professor at Hokkaido University, for their support. I also thank all staff and students at the Laboratory of Clinical Pharmaceutics & Therapeutics. I am also grateful to my collaborators, Ms. Ayako Nishimura, Pharmacist, Department of Pharmacy, Hokkaido University Hospital, and Dr. Takeshi Umazume, Department of Obstetrics, Hokkaido University Hospital. The studies described in this review were supported in part by Grants from the Japanese Society for the Promotion of Science (JSPS) (Grant Numbers: 16K18929, 18K1497208, 20K07149, and 23K1941703), Japan Spina Bifida & Hydrocephalus Research Foundation, OTC Self-Medication Promotion Foundation, YOKOYAMA Foundation for Clinical Pharmacology, Mishima Kaiun Memorial Foundation, Kieikai Research Foundation, and Urakami Foundation for Food and Food Culture Promotion.
The author is currently associated with a laboratory funded by Sato Pharmaceutical Co., Ltd. to support its research activities.
This review of the authorʼs work was written by the author upon receiving the 2024 Pharmaceutical Society of Japan Incentive Award for Women Scientists.
