CORESTA Meeting, Smoke Science/Product Technology, 2011, Graz, ST 15

Development and validation of a UPLC-MS/MS method for quantification of mycotoxins in tobacco and smokeless tobacco products

Swedish Match Smokefree Products, R&D, Chemical Analysis, Stockholm, Sweden

Mycotoxins are a group of naturally occurring toxic/carcinogenic substances that could be found in a wide range of food commodities, but can also be found in tobacco and tobacco products. Many countries have adopted regulations to limit exposure from food. In Sweden, Aflatoxins are regulated in food and smokeless tobacco products (STPs), and Aflatoxin B1 is also included in the "Draft Proposed Initial List of Harmful/Potentially Harmful Constituents in Tobacco Products" presented by the FDA. Today there is no recommended method for the simultaneous analysis of Mycotoxins in tobacco and STPs. Therefore, a quantitative multi-method for determination of four Aflatoxins and Ochratoxin A has been developed and validated.

The samples are extracted with a methanol/water mixture and diluted with PBS buffer after the addition of deuterium-labelled internal standards. The aqueous extract is then transferred to a Gilson ASPEC™ XL4 robot for automated clean-up using immunoaffinity columns. By using UPLC-MS/MS for quantification, instead of traditional HPLC-FLD, the analysis time is typically reduced by 75% to below a 4 min cycle time. The considerable reduction in analysis time by using this robot and UPLC-MS/MS system offers significant benefits to laboratories undertaking routine analysis of Mycotoxins in tobacco and tobacco products as well as in food products.

During the initial method development, one main problem was to achieve a homogeneous tobacco sample. Therefore, a procedure for the homogenization of the tobacco was developed in order to obtain accurate and reproducible results. This method showed high sensitivity with low quantifications limits of 0.1 µg/kg and 0.5 µg/kg for Aflatoxins and Ochratoxin A, respectively. The recovery rates were between 87 and 106% and the precision (CV) below 20% for all analytes. When participating in several proficiency tests for food items organised by FAPAS using this method our laboratories have obtained very good Z-scores.