P1-71 Development and Inter-laboratory Validation of Aflatoxin M1 Analysis in Bovine Milk

Monday, August 4, 2014
Exhibit Hall D (Indiana Convention Center)
Yang Chen, U.S. Food and Drug Administration, Bedford Park, IL
Ramesh Yettella, Institute for Food Safety and Health, Bedford Park, IL
Salvador Lopez, Institute for Food Safety and Health, Bedford Park, IL
Ravinder Reddy, U.S. Food and Drug Administration, Bedford Park, IL
Introduction: Aflatoxins are of major concern to the dairy industry as ingestion of aflatoxin B1 contaminated feed by dairy cattle can result in milk containing aflatoxin M1. Validated method for quantitation of aflatoxin M1 in milk is important.

Purpose: To develop a simple sample preparation protocol together with a LC-MS/MS method for quantitation of aflatoxin M1 in bovine milk to meet both EU (50 ppt) and U.S. (500 ppt) action levels and to evaluate the performance through an inter-laboratory study.

Methods: A simple extraction procedure based on a QuEChERS-like protocol followed by protein precipitation and ultrasonication prior to LC-MS/MS analysis was developed, validated and recommended to use for a proficiency testing program. A total of eight blind coded bovine milk samples, six spiked with aflatoxin M1 in the range of 200 to 750 ppt and two unspiked samples were shipped to 37 Food Emergency Response Network (FERN) laboratories. The results from ten laboratories that used the developed method were statistically analyzed according to the internationally harmonized protocol ISO 13528:2005. The mean values and reproducibility for each sample obtained by the ten laboratories were compared to the corresponding consensus values obtained from all participating laboratories reported quantitative results.

Results: The mean concentrations of aflatoxin M1 obtained using the developed method (n = 10) were 212 ± 85, 299 ± 55, 412 ± 135, 509 ± 139, 754 ± 131ppt compared to the consensus values (n = 25) of 216 ± 84, 302 ± 84, 415 ± 183, 503 ± 175 and 739 ± 174 ppt for the spiking level of 200, 300, 400, 500 and 750 ppt, respectively. The aflatoxin M1 concentrations obtained using the developed method were not significantly different from the corresponding consensus values (P > 0.05).

Significance: These data showed that the developed method could be used for routine and accurate monitoring of aflatoxin M1 in bovine milk.