A new study of embryotoxicity of various Chinese herbal medicines with unreliable results

Axel Wiebrecht

L. Li and a team of scientists from Hong Kong and Germany have published a study of 10 Chinese herbal medicines, in which they compared (alleged) existing and/or non-existent embryotoxicity according to tests on animals, with results of new in vitro tests they carried out (1). The following herbal medicines with the designated characteristics were tested:

Tripterygium wilfordii (lei gong teng), a herbal medicine that is considered obsolete due to its 
   considerable toxicity
Hirudo (shui zhi), the leech, traditionally contraindicated during pregnancy
Trichosanthis Radix (tian hua fen), traditionally contraindicated during pregnancy
Coptidis Rhizoma (huang lian), according to modern research contraindicated during pregnancy (2)
Astragali Radix (huang qi) – no known risk during pregnancy
Paeoniae Radix alba (bai shao) – no known risk during pregnancy
Scutellariae Radix (huang qin) – no known risk during pregnancy
Notoginseng Radix (san qi) – traditionally to be used with caution during pregnancy
Carthami Flos (hong hua), according to modern research contraindicated during pregnancy(2) 
Salviae miltiorrhizae Radix (dan shen), traditionally contraindicated during pregnancy due to risk of miscarriage

First, these herbal medicines were categorised according to their level of embryotoxicity as known from the literature (strong – weak - non-existent). However, the classifications made are incorrect in several respects.

Tripterygium wilfordii (lei gong teng) and Hirudo (shui zhi) are accurately classified as strongly embryotoxic. For Trichosanthis Radix (tian hua fen), which is similarly classified as strongly embryotoxic, no usable data applicable to a decoction or granules exists. The study by Li et al. presents data on Trichonsanthin, a protein isolated from Trichosanthis Radix, which has a strong abortive effect and was also used clinically by injection for this purpose. Due to frequent allergic reactions, this application has been discontinued. In addition to an abortive effect, Trichosanthin displayed embryotoxic and teratogenic effects in animal tests. The protein is most likely denatured through the impact of heat and is thus rendered ineffective. Even if embryotoxic risk cannot be ruled out, these data cannot be applied to a Trichosanthis Radix whole extract derived from a decoction. Embryotoxicity of Trichosanthis Radix can neither be confirmed nor negated by the available data.

In the case of Astragali Radix (huang qi), the data is also inconclusive regarding possible embryotoxicity. Only animal testing-based studies exist relating to the substance Astragalosid IV. In rabbits and/or rats there was an increased number of stillbirths and in some cases developmental retardations, but no malformations (3-6). A statement regarding the whole extract of Astragali Radix is not possible based on these data.

Coptidis Rhizoma (huang lian) is described by Li et al. on the basis of literature data as „weakly embryotoxic“. A comparatively minor study by Chuang et al. (7) is quoted, which shows a slight, non-significant decrease in birth weight of newborns, yet, incomprehensibly, a crucial study by the same authors is ignored, which demonstrated a significant risk of malformation when mothers had taken this medicine during the first trimester of pregnancy (8). Coptidis Rhizoma must therefore be regarded as strongly embryotoxic, which is of particular relevance due to the existence of human data.

Paeoniae Radix alba (bai shao)is assigned a weak embryotoxic risk, however literature references are not provided. A decoction of the medicine showed no embryotoxic or teratogenic effects in mice. A test on stem cells revealed no significant cytotoxicity (9). 

A lack of embryotoxicity for Scutellariae Radix (huang qin), Notoginseng Radix (san qi)and Salviae miltiorrhizae Radix (dan shen)is consistent with the literature. On the other hand, the pronounced embryotoxicity of Carthami Flos (hong hua) was not acknowledged. A decoction or aqueous extract of this medicine showed, as well as a higher rate of miscarriage, an increased rate of stillbirths, delays in fetal development and deformities in rats (10, 11). In mice, there was an increased incidence of embryo death and likewise a higher rate of stillbirths and deformities (12, 13). 

Thus much of the information on allegedly identified embryotoxicity of the 10 herbal medicines represents either an under- or overestimate; in two cases an evaluation is not possible due to a lack of meaningful data. The study by Li et al. used various in vitro tests for determining the embryotoxicity of these medicines, and compared their results with the (allegedly) known embryotoxicity from in vivo tests. In vitro methods offer the advantage of sparing experimental animals. The question is, how well can they predict embryotoxicity risk and/or to what extent do their results correlate with those of in vivo tests. The study group applied three different testing systems: the embryonic stem cell test (EST), the micromass (MM) and the whole embryo culture (WEC) in two forms: WEC (PM1) and WEC (PM2). 

An overview of the alleged and the actual known in vivo embryotoxicity, as well as the embryotoxicity prediction from the various in vitro test procedures is shown in the table. For the medicines Trichosanthis Radix (tian hua fen) and Astragali Radix (huang qi), no information regarding known embryotoxicity can be given, as no valid data exist. It is evident that many of the in vitro results differ from previously known risk data. The strong embryotoxicity of Hirudo is only identified as such in 2 out of 4 tests, that of Coptidis Rhizoma in only 1 out of 4; the strong, well-proven embryotoxicity of Carthami Flos is not recognized as such at all. On the other hand, results of several in vitro tests indicate a weak embryotoxicity for medicines to which none has previously been attributed.


The study by Li et al. does not accurately reflect previously known embryotoxicity from in vivo tests. The pronounced embryotoxicity of Coptidis Rhizoma and Carthami Flos is underestimated or negated, an alleged weak embryotoxicity of Paeoniae Radix alba is not substantiated and contradicts data from the literature. (Numerous other errors in citations and numerical data quoted, and units which differ from the literature, will not be addressed in detail here). The in vitro test procedures presented are an interesting contribution; however, their results for 8 herbal medicines bear limited correlation to previously known results from the literature. To what extent in vitro tests for reproductive toxicity of Chinese herbal medicines may replace in vivo studies, is a question which requires further research.

Axel Wiebrecht


1.Li L, Yin Tang L, Liang B, et al.Evaluation of in vitro embryotoxicity tests for Chinese herbal medicines. Reprod Toxicol. 2019;89:45-53.
2.Wiebrecht A.Sicherheit in der Chinesischen Arzneitherapie. In: Focks C, ed. Leitfaden Chinesische Medizin. 7thed. München: Elsevier; 2017:978-1028,1160-1164.
3.Zhu JB, Zhu YP, Zhang TG.[Evaluation of toxicity of astragaloside IV in rats and rabbits] (Chinese). Du Lixue Zazhi [Journal of Toxicology]. 2007;21(4):317-318.
4.Zhu JB, Wan XY, Zhu YP, et al.Effect of astragaloside IV on the embryo-fetal development of Sprague-Dawley rats and New Zealand White rabbits. J Appl Toxicol. 2009;29(5):381-385.
5.Zhu YP, Zhang TB, Wan XY, et al.[Study on the teratogenicity of astragaloside IV in SD rats] (Chinese). Zhong Chengyao [Chinese Traditional Patent Medicine]. 2010;32(10):1783-1785.
6.Wan XY, Zhu JB, Zhu YP, et al.Effect of astragaloside IV on the general and peripartum reproductive toxicity in Sprague-Dawley rats. Int J Toxicol. 2010;29(5):505-516.
7.Chuang CH, Lai JN, Wang JD, et al.Use of Coptidis Rhizoma and foetal growth: a follow-up study of 9,895 pregnancies. Pharmacoepidemiol Drug Saf. 2006;15(3):185-192.
8.Chuang CH, Doyle P, Wang JD, et al.Herbal medicines used during the first trimester and major congenital malformations: an analysis of data from a pregnancy cohort study. Drug Saf. 2006;29(6):537-548.
9.Xu W, Xu L, Deng B, et al.The potential impact of Radix Paeoniae alba in embryonic development of mice. Phytother Res. 2017;31(9):1376-1383.
10.Lin BH, Yan DP, Zhou LR, et al.[Toxicity and effects of Flos Carthami on rat pregnancy and embryonic development] (Chinese). Anhui Zhongyi Xueyuan Xuebao [Journal of Anhui University of Chinese Medicine]. 1998;17(4):50-52.
11.Zhao YL, Feng B, Zhou Y, Zhang J.[Study on selective expression of toxicity of Flos Carthami in pregnant rats] (Chinese). Liaoning Zhongyi Zazhi [Liaoning Journal of TCM]. 2011;38(11):2274-2276.
12.Nobakht M, Fattahi M, Hoormand M, et al.A study on the teratogenic and cytotoxic effects of safflower extract. J Ethnopharmacol. 2000;73(3):453-459.
13.Song XQ, Li JH, Wei HP, et al.[Effects of Flos Carthami on estrogen/progesterone and their receptors in early pregnancy of mice] (Chinese). Zhongguo Xiandai Yixue Zazhi [China Journal of Modern Medicine]. 2014;24(32):5-8.