Registration Dossier

Data platform availability banner - registered substances factsheets

Please be aware that this old REACH registration data factsheet is no longer maintained; it remains frozen as of 19th May 2023.

The new ECHA CHEM database has been released by ECHA, and it now contains all REACH registration data. There are more details on the transition of ECHA's published data to ECHA CHEM here.

Diss Factsheets

Toxicological information


Currently viewing:

Administrative data

Description of key information

According to the EU risk assessment report for hydrogen peroxide a number of animal studies on the carcinogenicity of hydrogen peroxide were carried out (European Commission 2003). In the majority of these studies the test material was administered via the oral route, however, some studies administered the test substance also via the dermal route.

Key value for chemical safety assessment

Justification for classification or non-classification

The following text is copied from the EU Risk Assessment Report (2003), pg. 130-131:

"A drinking water study in a catalase-deficient mice strain showed that hydrogen peroxide caused with a dose response duodenal hyperplasia (at a high frequency) and localised duodenal carcinomas (at a low frequency) (Ito et al., 1981a;b; Ito et al., 1982). A subsequent study with different strains of mice showed that there was a strong negative correlation between the indicence of duodenal tumours (hyperplasia or neoplasia) and catalase activities in duodenal mucosa, blood and the liver (Ito et al., 1984). In a comparable study with rats, drinking water administration of hydrogen peroxide seemed not to be associated with the occurrence of tumours, and there were no tumours in the gastrointestinal tract at all (Takayama, 1980). In carcinogenicity and tumour promotion studies, treatment with hydrogen peroxide demonstrated a promoting effect in rat intestinal carcinogenesis initiated by methylazoxymethanol acetate (Hirota and Yokoyama, 1981), in Syrian hamster buccal pouch carcinogenesis initiated by 9,10-dimethyl-1,2-bentzanthracene (DMBA) (Weitzman et al., 1986), and “extremely weakly” in the Sencar mouse skin after DMBA treatment (Klein-Szanto and Slaga, 1982). Other studies of similar design have not shown carcinogenic or promotion activity (Takahashi et al., 1986; Marshall et al., 1996; Bock et al., 1975; Kurokawa et al., 1984). However, 1% hydrogen peroxide in drinking water for 32 weeks induced sqamous cell papillomas of the forestomach in rats (Takahashi et al., 1986). While it is clear that hydrogen peroxide has the potential, even if weak, to induce local carcinogenic effect in the duodenum of a sensitive mouse strain, it is notable that the lesions showed a marked tendency of regression and even disappearance after the cessation of treatment. The mechanism of carcinogenic effect is unclear. Given that hydrogen peroxide causes DNA damage, a genotoxic mechanism cannot be excluded. Unfortunately, no cytogenetic studies have been available on the target tissue in various stages of histopathological injury. As regards tumour promotion, several mechanisms might be operative: direct genotoxicity, impairment of DNA repair, and chronic inflammation.

The special nature of the demonstrated carcinogenicity of hydrogen peroxide, and the overall evidence available at this time, cast some doubt on whether hydrogen peroxide should be regarded as a carcinogen of practical significance. The weak effect found in complete carcinogenesis studies in mice as well as in some promotion studies suggest promotion type of activity and possible underlying genotoxic mechanisms. Given the fact that mammalian cells have of necessity built defences against reactive oxygen species arising in endogenous metabolism, the injuries caused by hydrogen peroxide may well be non-stochastic, i.e. have a dose/dose rate threshold. This evidence however is not sufficient to trigger classification."

Additional information

The following text is copied from the EU risk assessment report (2003), pg. 126-127:

“Different studies with hydrogen peroxide exploring carcinogenic effects have been conducted. Hydrogen peroxide revealed a weak potential to induce local tumourgenic effects in the duodenum of a sensitive mouse strain (Ito; 1981, 1982), it is notable that the lesions showed a marked tendency of regression and even disappearance after the cessation of treatment (Ito, 1982). No similar effects were reported from rats. In tumour promotion experiments hydrogen peroxide showed tumour promoting effects in rats and Syrian hamsters and extremely weak in Sencar mouse skin (Takayama, 1980; Weitzman et al., 1986; Klein-Szanto & Slaga, 1982). Several further studies were negative concerning tumour promoting effects of hydrogen peroxide.”

The study by Ito et al. (1981) (reliability 2) showed that hydrogen peroxide in drinking water may induce tumours (duodenal carcinoma) in catalase deficient mice, which represents a very special strain. In Ito et al. (1982) it was demonstrated that strains with higher activities of catalase show lower incidence of duodenal tumours, and vice versa. The latter study is however not fully reliable because it is for example not clear whether tumour incidences are from females or from both sexes.

The study of Takayama (1980) shows that hydrogen peroxide is not clearly carcinogenic in rats up to dose levels of 0.6% in drinking water. Tumour incidences were high without any clear dose–relationship. Most tumours were Leydig cell tumours in males and pituitary and uterine tumours in females. Slightly increased incidences in thyroid and adrenal tumours in females were observed; incidences were 1, 3, 7 for thyroid tumours and 1, 2 and 4 for adrenal tumours at 0, 0.3 and 0.6% dose levels, respectively. Incidences of uterus carcinomas were 0, 1 and 2 at 0, 0.3 and 0.6% dose levels, respectively. It is not possible to compare the incidences to historical control values because they are not provided in the study.