Line or threshold? Evolving toxicological views on estragole

Introduction In toxicological assessments of trace constituents of natural products with potential toxicological relevance, one question is always key: Can data from studies be linearly transferred to the actual uptake with these products, which is several orders of magnitude lower? Or is ADME of high doses different, so that doses below a certain threshold are of less toxicological concern? This question is presently under discussion for estragole, a widely spread constituent of medicinal and food plants like fennel or basil [1].

Objective and Methods As a contribution to this discussion, relevant in vitro and in vivo data from the literature have been retrieved, presented and assessed.

Results Two recent in vitro studies are key for the understanding of the dose dependency of potential genotoxic effects of estragole [2] [3]. The first is describing the accumulation of DNA adducts in the human liver cell line HepaRG cell after repeated incubation with 50 µM estragole, for modelling repeated dietary exposure. According to PBK modelling, this concentration is reached in venous liver blood at a dose level of 36 mg/kg b.w. in humans, being substantially higher than normal human dietary intake estimated to values of 0.07 and 0.01 mg/kg b.w./day, respectively. In a different study [3] in cultured rat hepatocytes, the focus was on a concentration of estragole of 0−1 μM, i.e. concentrations corresponding to more realistic oral dose levels of 0−1.5 mg/kg b.w. In this concentration range, the E-3′-N2-dG DNA adduct formation increased in a sublinear way with an apparent threshold at 0.5 μM.

Regarding in vivo data, a study in female mice on the genotoxic carcinogenicity is of relevance, with a TD₅₀ of 50–100 mg/kg b. w. [4]. In addition, there is an NTP 3-month rat study with a wide dose-response range, not with a classical carcinogenicity endpoint, but with focus on acidophilic, eosinophilic, mixed-type and clear-cell foci in the livers [5]. Given the morphological continuity between preneoplastic foci and hepatocellular tumors [6] and considering that preneoplastic liver foci occur at lower dose levels and at much earlier timepoints than liver tumors [7], showing an about 10fold higher sensitivity for the development of preneoplasia, this 3-month study [5] seems to be a favorable starting point for setting a limit.

Conclusions The available toxicological data on estragole are quite diverse, some pointing to a linear dose effect relationship at high concentrations, some to a potential threshold in low concentrations, so that the different approaches to deduct limits from these are being discussed intensively. Given that also in the field of food regulation, a collection and assessment of data is ongoing, it seems to be justified to carefully reconsider the implementation of regulatory measures with potentially detrimental effects on certain treatment options in phytotherapy, while the scientific base of limits for estragole in the herbal medicinal products field is being further consolidated.

AcknowledgementMany thanks for fruitful discussions to D. Schrenk and J. Fahrer, University of Kaiserslautern

Publication History

Article published online:
14 June 2023

• References

• 1 EMA Committee on Herbal Medicinal Products (HMPC) Public statement on the use of herbal medicinal products containing estragole. EMA/HMPC/137212/2005 Rev 1
• 2 Yang S. et al. Toxicol Lett 2021; 337: 1-6
• 3 Schulte-Hubbert R. et al. Toxicology 2020; 444: 152566
• 4 Miller EC. et al. Cancer Res 1983; 43: 1124-1134
• 5 Bristol DW.. Toxicity Report Series 2011; 82: 1-111
• 6 Marsman DS, Popp JA.. Carcinogenesis 1994; 15: 111-117
• 7 Kunz HW. et al. Environ Health Perspect 1983; 50: 113-122