
Originally Posted by
ucla report
According to the literature review by Brusick et al. (2008) on the genotoxicity of steviol and stevioside, two of 16 studies showed genotoxic activity for stevioside and four of 15 studies (Brusick et al. did not include Pezzuto et al., 1985, and TM677 results by Matsui et al., 1996a) showed genotoxic activity for steviol (see Tables 4 and 5, respectively).
Stevioside showed positive results in Salmonella typhimurium (S. typhimurium) strain TA98 at 50 mg/plate for 99% pure stevioside (Suttajit et al., 1993). The results showed a 4-fold increase in revertants without S9 extract and a 2-fold increase with S9. That study used stevioside pre-incubated with and without β- glucosidase. Those results demonstrate that at 50 mg/plate, stevioside (without β-glucosidase or S9), steviol (stevioside + β-glucosidase), stevioside metabolite(s) (stevioside +S9), and steviol metabolite(s) (stevioside + β- glucosidase + S9) are all mutagenic in TA98.
Using the comet assay, Nunes et al. (2007) reported DNA breakage in blood, spleen, liver, and brain cells in Wistar rats exposed to 400 mg/kg of stevioside in drinking water. The strongest effects of stevioside were found in the liver cells.
Metabolically-activated steviol was found to cause dose-related positive responses in several mutagenicity tests. These results indicate that a steviol derivative is likely responsible for its mutagenic activity, but the metabolite has not been identified (Brusick et al., 2008). The mutagenicity of steviol metabolites needs to be further investigated.
Matsui et al. (1989) showed positive results for steviol in a plasmid mutagenesis study.
Matsui et al. (1996a) found that steviol is mutagenic in S. typhimurium strain TM677, caused chromosome aberrations in cultured Chinese hamster lung (CHL) cells, and mutagenic in CHL cells in the presence of S9. In the same study, steviol produced a weak positive response with or without S9 in the umu test.
A forward mutation assay using S. typhimurium strain TM677 found mutagenicity using 100 μg/ml steviol when assayed with S9 extract (Pezzuto et al., 1985).
Studies in rats have failed to produce any evidence of carcinogenicity of stevioside, though rebaudioside A, the subject of the GRAS notification, itself has not been tested. It is important to note that FDA normally asks for tests in two rodent species, usually rats and mice, in a compound with such a high predicted exposure level. Also, all three of the aforementioned studies were done with stevioside, not rebaudioside A. It is possible that differences in metabolism and toxicokinetics would result in different risks of carcinogenicity using the two steviol glycosides.
Older studies reported anti-fertility effects, as well as decreases in the weights of the testes, seminal vesicle, and cauda epididymides and a reduction in spermatozoa concentration, in rats administered crude stevia extracts (Mazzei-Plana and Kuc 1968; Olivereira Filho et al., 1989, and Melis, 1999).
In both the rat and human metabolism studies, investigators demonstrate that rebaudioside A and stevioside have similar metabolic pathways within each species. However, rebaudioside A’s extra glucose moiety causes differences in the two compounds’ pharmacokinetic parameters (Tables 1, 2, and 3). Because of those differences, toxicity data for stevioside cannot be assumed to be an appropriate basis for assessing the safety of rebaudioside A. Separate toxicity studies on rebaudioside A itself are necessary to make definitive conclusions about its safety.
Since steviol glycoside metabolism in rats and humans is not identical, the rat may not be an ideal model for evaluating human toxicity. Both species hydrolyze the glycosides into steviol by the gut microflora, but after absorption the metabolic pathways differ.
Hutapea et al. (1997) reported a steviol-16,17-epoxide stevioside metabolite. Given the structures of stevioside and rebaudioside A, an epoxide is a likely metabolite. The possibility of a steviol glycoside forming an epoxide metabolite needs to be investigated carefully, because epoxides may react with DNA and cause mutations.
Genotoxicity studies published in the Food and Chemical Toxicology supplement, as well as other studies, raise significant concerns. Suttajit et al. (1993) reported positive results for reverse mutations in the S. typhimurium strain TA98 with and without S9 extract at a 50 mg/plate dose of stevioside. The ability of stevioside and rebaudioside A to cause reverse mutations as indicated by TA98 needs to be further investigated, because such mutations suggest the possibility of carcinogenesis. Stevioside also caused DNA breakage in blood, spleen, liver, and brain cells in rats (Nunes et al., 2007). The mutagenicity of this compound requires further, careful investigation.
Steviol was found positive in an umu test, mutagenic in a forward-mutation assay, and caused chromosome aberrations and gene mutations in mammalian cells (Matsui et al., 1996a) and plasmid mutagenesis (Matsui et al., 1989). Pezzuto et al. (1985) found that steviol is both toxic and mutagenic in the TM677 assay using S9 extract. Matsui’s studies were all conducted with S9. These results indicate that steviol has a mutagenic metabolite that has yet to be identified. These finding are very important because rebaudioside A is hydrolyzed into steviol before it is absorbed by the GI tract. Before rebaudioside A can be generally regarded as safe, the mutagenic steviol intermediate needs to be identified and further studied. Overall, because of the warning flags raised by several studies, it is critical that further genotoxicity testing be conducted to clarify the potential risks.
further studies on rebaudioside A, including a study on mice, are needed for several reasons:
• The rat is an imperfect model for evaluating steviol glycoside toxicity and carcinogenicity risks in humans because of the differences in metabolism in the two species.
• Several genotoxicity studies that found that stevioside and steviol cause mutations, chromosomal damage, and DNA breakage indicate the need for greater reassurance of noncarcinogenicity.
• The differences in pharmacokinetics between rebaudioside A and stevioside indicate the need to test rebaudioside A itself in two rodent species.
• Based on a maximum estimated intake level of steviol glycosides of 1.7 mg/kg bw/day (steviol equivalent), steviol glycosides should be considered a concern level III chemical, for which the FDA recommends carcinogenicity studies in two rodent species (usually mice and rats) (FDA Redbook, 2000).
The value of testing chemicals in two species is indicated by the fact that bioassays of chemicals with a variety of structures that did not find carcinogenicity in rats did find carcinogenicity in mice.
In sum, a lifetime carcinogenicity study in mice of rebaudioside A must be conducted before that substance (or other steviol glycosides) can be accepted as a GRAS ingredient that likely would be consumed by tens of millions of people.
In conclusion, the FDA should ensure that the genetic toxicity studies that produced either positive or conflicting results be repeated. Studies that look at potential DNA adducts related to the potential reactive metabolites (C-13 carbonium ion or the epoxide) of steviol would be a strong addition to the genotoxicity data. Finally, the FDA should require carcinogenicity4 and toxicology studies in rats and in mice before accepting rebaudioside A as a GRAS substance or approving it as a food additive. Ideally, all those studies would be conducted by an independent party, such as the National Toxicology Program of the National Institute of Environmental Health Sciences.