**11. Drawbacks of using cytotoxicity to predict safety of herbal medicines in TM**

There are some shortcomings to extrapolating cytotoxicity studies to the safety of herbal medicine used in traditional medicine. Among this is the fact that tissue responses due to *in vivo* toxicity cannot be addressed by toxic responses in cells [37]. According to McGraw et al. [37], a critical factor in toxicology is metabolism *in vivo*, as some substances lacking toxicity initially may produce toxic metabolites after being exposed to liver enzymes, while other substances that are toxic in vitro may become detoxified. Other factors such as the capacity of the substance to penetrate the tissue, and clearance and excretion of the product cannot be accounted for using the cellular model. The time of exposure and the rate of change for these extracts are not the same in both *in vitro* and *in vivo* studies. Notwithstanding these limitations of cytotoxicity assay, it still needs to be an integral part of evaluating the safety of medicinal plants because they provide direct information at the cellular level which may be important in assessing the true toxicity of such plants.

**13. Toxicity studies on some African medicinal plants**

**13.2.** *Artemisia afra* **(Jacq. Ex. Willd), 'African wormwood'**

*C. sanguinolenta* (Apocynaceae) is a West African climbing shrub. The aqueous extract of the root has been used for centuries in African traditional medicine for the treatment of diseases including malaria, bacterial infections, hepatitis and rheumatism. It is also used as a spasmolytic and tonic [49]. In Ghana, several cryptolepis-based products are prescribed in herbal medicine clinics, sold in pharmacies, licensed chemical and herbal medicine shops for the

Toxicity and Safety Implications of Herbal Medicines Used in Africa

http://dx.doi.org/10.5772/intechopen.72437

73

Acute and sub-acute oral toxicity evaluation of the aqueous extract of the root suggested general safety at oral dosages below 500 mg/kg in Sprague Dawley rats. The extract did not exhibit either physiological or behavioral abnormality [50]. However, the ethanolic extract of the stem demonstrated localized systemic acute and sub-chronic toxicity by selectively stimulating the bone marrow leading to an increase in platelet counts in albino rats [48]. On the other hand, the aqueous extract of the root demonstrated genotoxicity against the Chinese hamster lung fibroblast (V79) cell line inducing mutagenicity at high concentrations and causing DNA damage [42, 47]. The ethanolic extract of the stem thus poses hematological challenges to white blood cells and platelets and showed localized systemic toxicity by selectively stimulating the bone marrow.

*A. afra* has been used for coughs, colic, fever, loss of appetite, earache, headache, malaria and intestinal worms [51]. Several studies have been conducted to substantiate the traditional use of this herb; it is also being investigated in diseases like diabetes, cancer and respiratory

In acute toxicity studies of aqueous extract of *A. afra* in mice administered doses (i.p., 1.5–5.5 g/kg) caused a regular dose-dependent increase in the death rate and also of general adverse behavior, but with single doses (2–24 g/kg) administered orally, the previous observed increases in the incidence of death rate and adverse general behavior that did not show were doseindependent. The route of administration, acute intraperitoneal and oral doses, showed LD50

In the chronic studies, rats administered *A. afra* aqueous extract (0.1 or 1 g/kg/day) survived the 3 months of daily dosing with LD50 greater than 1 g/kg. No significant changes were observed in the general behavior, hematological and biochemical parameters except for a transient decrease in aspartate aminotransaminase (AST) activity. No significant changes were observed in the organ weights and histopathological results showed no morphological alterations. High doses of the extract were also shown to be hepatoprotective. The aqueous extract of *A. afra* has been shown to be nontoxic in acute use and low chronic toxicity potential in rodent models [51].

**13.1.** *Cryptolepis sanguinolenta* **(Lindl.) Schltr**

treatment of malaria [21].

*13.1.1. Animal and cell toxicity*

diseases among others [51].

*13.2.1. Animal toxicity*

of 2.45 and 8.96 g/kg, respectively [51].

Other limitations in cytotoxicity studies with regard to safety prediction for herbal medicines is the use of organic solvent such as methanol, dichloromethane, petroleum ether, ethyl acetate, and so on extracts as against water decoctions/extracts. In situations like this, it is not reasonable to compare the cytotoxicity results of the organic solvent extract with what pertains in TM. Such studies are common because the focus of most cytotoxicity studies has not been the safety assessment of the plants as used in TM but to determine the fractions which contain the potentially safe and efficacious compounds. In some cases, while efficacy study was conducted for both organic solvent and aqueous extracts, cytotoxicity was determined for only the organic solvents. This makes it difficult to relate the toxicity of the plant to safety in traditional use. In most cases, the organic extracts tend to be more efficacious than the aqueous extracts. This could imply that the organic extracts are more cytotoxic than the aqueous extracts since, generally, they (organic extracts) tend to extract more active compounds, which are both efficacious and cytotoxic.
