Quantitative risk assessment of aflatoxicosis associated with milk consumption in Italy (2000-2004)

Currently available information has been examined to assess the risk of aflatoxin contamination in milk and give the public managers tools for the evaluation of mitigation strategies that have already been imposed or could be imposed to reduce risk. The assessment concerns milk for direct consumption produced in Italy, from 2001 to 2004, and also provides an estimate uncertainty associated with available data and determines whether available data have been critical and have driven the overall risk assessment. Data relative to milk contamination were generated by two independent control systems, namely a monitoring system regularly maintained by a nationally relevant private industry during the period January 2001-July 2004 and an extensive surveillance system implemented by Public Veterinary Services during the period September 2003- July 2004 as a consequence of critical environmental conditions leading to an \u201caflatoxin crisis in milk\u201d. Both concerned raw milk produced in the North and Centre of Italy during the last five years and consist of more than 9000 samples of milk which were analysed in two laboratories using validated ELISA methods. The mean estimated level of aflatoxin M1 in milk observed during the crisis period was approximately 0.035 \ub5g/kg in both data sets (Industry and Public sampling plans) and 95th percentile values were 0.073 and 0.080 \ub5g/kg respectively. These values were relatively higher than those relative to data recorded before September 2003 (mean = 0.027 \ub5g/kg; 95th percentile = 0.080\ub5g/kg). The data have been evaluated for their capability of representing the overall variability of aflatoxin level in milk produced in the region and the associated degree of uncertainty (accuracy and bias) of the analytical methods. The relative amount of milk presenting different levels of aflatoxin contamination has also been considered because the monitoring data provided by Industry were correlated with records of milk weight. Any processing steps, apart from mixing, are capable of producing change in aflatoxin level in the production of pasteurised or UHT milk. A mixing model considering the capacity of storage tanks at the processing plant and the relative weight of bulk milk supplied with different aflatoxin levels has produced probability distributions for aflatoxin in milk. An exposure assessment concerning aflatoxin M1 in milk has been carried out by using the contamination data in combination with milk consumption patterns observed in the Italian population. To this aim data of the Italian Institute of Nutrition relative to milk consumption have been used to produce second order parametric probability distributions for children (1-9 y) adolescents (10-17 y) adults (18-64 y) and the elderly (>= 65 y). The aflatoxin M1 hazard was characterized by using the carcinogen potency estimates outlined by the WHO panel of experts. This panel has recently analysed all available toxicological studies concerning aflatoxin M1 and has used the comparative experimental studies with aflatoxin B1 to estimate its relative carcinogen potency. The carcinogen potency ranges should encompass the different sensitivities to aflatoxin observed in the human population. The genotypic and phenotypic variability is at the basis of these differences, but a mechanistic modelling of cancer initiation and progression, which could directly account for them, cannot be accomplished at present, because probabilistic quantitative data are inadequate for risk assessment. Carriers of the Hepatitis B virus are significantly more susceptible to aflatoxin carcinogenicity than healthy individuals and their prevalence in the population has been accounted for in the risk assessment. However, many factors, including body weight, milk consumption and prevalence of Hepatitis B carriers, which are dependent on the age of exposed individuals are correlated and have to be consequently analysed. In order to assess the risk for the different age classes, w..