Treatment of mild mental disorders in pregnancy: how safe are phytomedicines?

Pregnancy and associated physiological changes, which begin in the first trimester and are most pronounced in the third trimester, alter the pharmacokinetics of drugs. In addition, most drugs enter the fetal-placental unit, which pose several challenges to medical care, as it affects both the unborn child as well as the expectant mother. This also applies to the treatment of non-psychotic mental disorders (NMDs) such as depression, restlessness, sleep disorders, and anxiety since the use of synthetic, central nervous system (CNS)-active drugs for treatment should be carefully considered in a risk-benefit ratio. Nevertheless, pregnancy is a vulnerable period for such disorders and preexisting mental illnesses may even worsen or relapse. Alternatively, there are some herbal medicines, so called phytopharmaceuticals, which possess antidepressant, sedative, or anxiolytic properties. St. John’s wort, for example, is used to treat mild to moderate depression, whereas California poppy, valerian, and hops are mainly known for their sleep-inducing effects. Finally, lavender essential oil represents an option for the treatment of restlessness and anxiety due to its calming, sedative, and anxiolytic effects. Most products based on these phytopharmaceuticals are available over the counter (OTC) and are generally recommended by a considerable proportion of health care professionals who deal with pregnant women, such as midwives and obstetricians. Despite the limited experimental and clinical evidence, and especially lack of safety data, pregnant women often resort to herbal medicines and perceive them as safe.
We designed a cross-sectional survey with which we determined the use of most common herbal medicines in pregnancy by obstetrics patients. Furthermore, some questions targeted the perceived effectiveness and tolerability of some of these plants. The survey revealed that a large majority of pregnant women make use of herbal medicines, and lavender (22%), valerian (4.7%), St. John’s wort (3.0%), and hops (1.7%) were among the most mentioned pharmaceutical herbal products for the treatment of mental disorders and/or symptoms (MDS). Although 52.0% of all participants suffered from MDS during pregnancy, only 1.3% reported taking (synthetic) psychoactive medications. The fact that the prevalence of MDS was higher in pharmaceutical herbal products users than in non-users suggests that pregnant women rely on herbal medicines to treat mild MDS. St. John’s wort, valerian, hops, and lavender were used to reduce mood and sleep disorders, restlessness, and stress with perceived good to very good effectiveness and tolerability; no participant reported the use of California poppy. 
To verify whether various herbal extracts such as St. John's wort, California poppy, valerian, hops, and lavender can be used for the treatment of NMD in pregnancy, their safety must be thoroughly investigated with a palette of in vitro assays. Those extracts and some of their active marker compounds were therefore evaluated for their cytotoxic and genotoxic potential and for their effects on metabolism and cell differentiation. The in-depth in vitro safety assessment indicates that extracts of St. John’s wort, California poppy, valerian, hops, and lavender – at concentrations up to 30 µg/mL – have no cytotoxic or genotoxic potential and do not compromise the viability, metabolic activity, and differentiation of placental cells. Moreover, none of the five extracts was able to significantly alter protein expression of BeWo b30 cells. For active marker substances, protopine (found in California poppy), valerenic acid (in valerian), and linalool (in lavender) showed no adverse effects in all experiments performed. The following phytochemicals might conceivably cause safety issues: hyperforin (in St. John’s wort) induced cell apoptosis (≥ 3 µM) and inhibited BeWo b30 cell differentiation (≥ 1 µM). Hypericin (in St. John's wort) decreased cell viability (≥ 1 µM) and induced cell apoptosis (30 µM). Valtrate (in valerian) decreased cell viability (≥ 10 µM), induced cell apoptosis (≥ 10 µM), and decreased the metabolic cell activity by reducing glucose consumption and concomitant lactate production. However, none of the tested phytochemicals resulted in genotoxic effects and thus are not DNA damaging. In summary, most of the phytochemicals were not of concern, but the attainment of high plasma concentrations of a few relevant phytochemicals – hyperforin, hypericin, and valtrate – deserves special attention.
In addition to investigating effects on placental cells, the question of transplacental permeation of certain phytochemicals needs to be addressed in order to better assess the risk of fetal exposure. For this purpose, an ex vivo placental perfusion model was optimized and modified that it can be used for the application of phytochemicals. Furthermore, an in vitro Transwell model with placental cells was established which served to complement the placental perfusion model. We successfully used the human ex vivo placental perfusion model for the first time for transport experiments with phytochemicals, after a thorough validation with various compounds, such as antipyrine, citalopram, and diazepam.
Only a small fraction of the initially present hyperforin (< 10%) reached the fetal circuit after 4 h, whereas hypericin did not cross the placental barrier and therefore remained in the maternal circuit. In contrast, protopine was transferred from the maternal to the fetal circuit, reaching a steady state after 90 min with no further changes in concentration. Also, valerenic acid was transferred gradually over a 4-hour period and reached an equilibrium with the maternal concentration. None of the phytochemicals affected placental viability or functionality, and histopathological evaluation of all placental specimens revealed no pathologic findings. In addition, in vitro translocation studies have confirmed that valerenic acid, from valerian, cannot cross the placental cell layer within 60 min, which might indicate that this phytochemical is not accessible to the unborn child in early pregnancy.
We were able to gain more detailed knowledge into the safety of St. John’s wort, California poppy, valerian, hops, and lavender extracts and a few of their active phytochemicals through in-depth in vitro assessments and an ex vivo model. In addition, we know that phytomedicines of these extracts, except California poopy, are already used in pregnancy in Switzerland with well-perceived effectiveness and tolerability. We hope that this thesis contributes to a rational basis for future decisions on the treatment of NMDs during pregnancy.