Cancer risk assessment of ethyl carbamate in alcoholic beverages from Brazil with special consideration to the spirits cachaça and tiquira

BACKGROUND: Ethyl carbamate (EC) is a multi-site carcinogen in experimental animals and probably carcinogenic to humans (IARC group 2A). Traces of EC below health-relevant ranges naturally occur in several fermented foods and beverages, while higher concentrations above 1 mg/l are regularly detected in only certain spirits derived from cyanogenic plants. In Brazil this concerns the sugarcane spirit cachaça and the manioc (cassava) spirit tiquira, which both regularly exceed the national EC limit of 0.15 mg/l. This study aims to estimate human exposure in Brazil and provide a quantitative risk assessment. METHODS: The human dietary intake of EC via alcoholic beverages was estimated based on WHO alcohol consumption data in combination with own surveys and literature data. This data comprises the EC contents of the different beverage groups cachaça, tiquira, other spirits, beer, wine, and unrecorded alcohol (as defined by the WHO; including alcohol which is not captured in routine government statistics nor taxed). The risk assessment was conducted using the margin of exposure (MOE) approach with benchmark doses obtained from dose-response modelling of animal experiments. Lifetime cancer risk was calculated using the T25 dose descriptor. RESULTS: Considering differences between pot-still and column-still cachaça, its average EC content would be 0.38 mg/l. Tiquira contained a considerably higher average EC content of 2.34 mg/l. The whole population exposure from all alcoholic beverages was calculated to be around 100 to 200 ng/kg bw/day, with cachaça and unrecorded alcohol as the major contributing factors. The MOE was calculated to range between 400 and 2,466, with the lifetime cancer risk at approximately 3 cases in 10,000. An even higher risk may exist for binge-drinkers of cachaça and tiquira with MOEs of up to 80 and 15, respectively. CONCLUSIONS: According to our risk assessment, EC poses a significant cancer risk for the alcohol-drinking population in Brazil, in addition to that of alcohol alone. Model calculations show that the implementation of the 0.15 mg/l limit for cachaça would be beneficial, including an increase of the MOE by a factor between 3 to 6. The implementation of policy measures for tiquira and unrecorded alcohol also appears to be advisable.