Following a request from the European Commission, the Panel on Food Additives and Nutrient Sources added to Food (ANS) was asked to deliver a scientific opinion on the safety of allyl isothiocyanate (AITC) when used as a food preservative, via addition to the packaging of certain foods. The volatilisation of the AITC from the packaging materials provides exposure of the food to AITC vapours for the purpose of protection from spoilage, resulting in a longer preservation of the food. The safety evaluation of the use of AITC in food contact materials is outside of the remit of the Panel.
AITC is an organic compound which occurs naturally in food, e.g. in horseradish and mustard. It is also used as a flavouring substance. For the proposed uses as a food additive, it is produced from seeds of brown mustard (Brassica juncea).
The use of AITC as a flavouring substance has been evaluated by the Joint FAO/WHO Expert Committee on Food Additives (JECFA) and by the EFSA Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food (AFC) which found “No safety concern at estimated levels of intake” based on the Maximised Survey-derived Daily Intake (MSDI) approach (EFSA Flavouring Group Evaluation 85).
Isothiocyanates including AITC are readily cleared from rat and mouse tissues so that within 24 hours after administration less than 5% of the total dose was retained in tissues. Generally, the highest concentration of AITC was observed in male rat urinary bladder tissue. Clearance was primarily via the urine. In the mouse, hydrolysis of AITC to thiocyanate was the major metabolic pathway, whereas in the rat glutathione conjugation followed by further metabolism to the mercapturic acid was the major route. In human studies, the mercapturic acid conjugate was found to be a major urinary metabolite. The major portion of this metabolite was excreted within 8 hours. Overall, the available studies indicate that isothiocyanates are readily conjugated to glutathione with the glutathione conjugates being further transformed successively into the corresponding cysteine conjugates and mercapturic acids. Since the mercapturic acid conjugate of AITC was the major urinary metabolite in humans and rats, the rat appears to resemble humans more than mice in AITC metabolism. However, rats have a much slower clearance of AITC than humans (biological half lives of 140 vs 2 hours).
AITC has been investigated in acute, short-term and subchronic toxicity studies as well as in carcinogenicity studies in rats and mice. The developmental toxicity has been investigated in rats, mice, hamsters and rabbits, however, a reproductive toxicity study (one or two generation) was not provided by the petitioner. Additionally, there are a number of in vitro and in vivo genotoxicity studies available.
The main effects observed in short-term toxicity studies in rats and mice, dosed by gavage, were a thickened mucosal surface of the stomach, adhesion of the stomach to the peritoneum and a thickened urinary bladder wall (the latter change being mainly observed in the males of both species). This indicates that AITC has irritant effects on these tissues. The No-Observed-Adverse-Effect-Levels (NOAELs) from short-term and subchronic toxicity studies in rats and mice which received AITC by gavage were in the range of 10 to 25 mg/kg bw/day. The NOAELs identified in subchronic toxicity studies were mainly based on the effects on kidney, stomach and urinary bladder observed at higher doses.
The Panel considered that AITC did not show any evidence of developmental toxicity in pregnant rats, hamsters and rabbits at oral doses up to 18.5, 23.8 and 12.3 mg/kg bw/day, respectively. AITC may be fetotoxic to mice at doses higher than 6.0 mg/kg bw/day, without exhibiting any teratogenic effects. The Panel noted that no reproductive toxicity study (one or two generation) was provided by the petitioner and that no information was identified in the published literature.
Both positive and negative results have been observed with AITC in genotoxicity tests in bacterial and mammalian cells in vitro in the absence of metabolic activation, with positive results occurring usually at or near cytotoxic concentrations. In the presence of metabolic activation, however, the results were generally negative. While mixed results were reported in non-standard genotoxicity assays in vivo, the results were consistently negative in assays for micronucleus formation in mice and rats and unscheduled DNA synthesis in mice after oral application. Additionally, AITC did not induce dominant lethal mutations in a germ cell test in mice after a single intraperitoneal injection. Hence, the Panel concluded that there is no concern with respect to genotoxicity.
AITC has been investigated for carcinogenicity in rats and mice in doses of 12 and 25 mg/kg bw/day administered by gavage at 5 days per week for 103 weeks. It was not carcinogenic in mice. In the rat carcinogenicity study, the incidence of transitional cell papillomas of the urinary bladder of males showed a significant positive trend. The Panel noted that transitional cell papillomas are benign lesions originating from epithelia of the urinary bladder and that AITC is not genotoxic in vivo in mice and rats. Hence, it could be assumed that there is a threshold mechanism underlying these effects on the urinary bladder. Although the incidences of papillomas and epithelial hyperplasias in the low dose group were not statistically significantly different from controls, the Panel considered it justified to conclude that the 12 mg/kg bw/day should be regarded as the Low-Observed-Adverse-Effect-Level (LOAEL) rather than the NOAEL based on the statistically significant positive trend observed in the incidence of these urinary bladder lesions in males and the well known very low incidence of these lesions in historical controls.
In studies conducted to investigate the goitrogenic effects of AITC in rats, increased thyroid weights and decreased iodine uptake were observed. However, these studies were considered to be of limited relevance for the human situation, keeping in mind the differences in the sensitivity between humans and rats.
The Panel noted that AITC may cause contact hypersensitivity (an immunologically-mediated adverse reaction which is mechanistically different from food allergy), however, it considered that it is extremely unlikely that AITC acts as a direct food allergen.
The Panel considered the long-term study in rats as the pivotal study and the dose of 12 mg/kg bw/day as the LOAEL (corresponding to a daily dose of 9 mg/kg bw/day when adjusted from the 5-day dosing week to a 7-day week). In order to cover uncertainties resulting from the extrapolation from this LOAEL to a NOAEL and uncertainties related to the absence of data on reproductive toxicity, the Panel applied an uncertainty factor of 5 in addition to that of 100 which would normally be applied. Thus, the Panel derived an ADI of 0.018 mg/kg bw/day which was rounded up to 0.02 mg/kg bw/day based on the LOAEL of 9 mg/kg bw/day established for transitional cell papillomas of the urinary bladder observed in the carcinogenicity study in male rats and applying an uncertainty factor of 500.
The Panel noted that intakes of AITC resulting only from application as an antispoilage agent could be estimated for children ranging from 0.3 up to 2.8 times the ADI and for adults ranging from 0.2 up to 1.6 times the ADI, based on a market share of 20 and 100%, respectively.
Furthermore, the Panel noted that the mean daily total exposure to AITC from all sources including natural occurrence in food, use as a flavouring substance and application as an antispoilage agent, based on the more refined model, results in a two to four-fold exceedance of the ADI in children (3-5 fold exceedance at the 95th percentile for the 20 and 100% market share, respectively), and up to eight-fold exceedance in the case of 95th percentile adult consumers.