Pilot plant investigations on cleaning efficiencies to reduce hazelnut cross-contamination in industrial manufacture of cookies.

Shared equipment in industrial food manufacture has repeatedly been described as a potential source of unlabeled food allergens, i.e., hidden allergens. However, the impact of shared equipment on allergen cross-contamination is basically unknown. Therefore, we sought to investigate systematically the extent of hazelnut cross-contamination in fine bakery wares as a model. A product change from cookies with 10% hazelnut to cookies without hazelnuts was simulated on pilot plant equipment. The extent of hazelnut cross-contamination (HNCC) was analyzed by enzyme-linked immunosorbent assay (ELISA) for each production device (kneaders, rotary molder, wire cutting machine, and steel band oven) and various cleaning procedures used between products. The experiments were performed repeatedly with finely ground hazelnuts and with roughly chopped hazelnut kernels. Cross-contamination from chopped kernels was distributed statistically but not homogeneously, and sampling and analysis with the ELISA was therefore not reproducible. Further analysis concentrated on homogenously distributed HNCC from ground hazelnut. Apart from product changes without intermediate cleaning, the highest HNCC was found after mechanical scraping: Up to 100 mg/kg hazelnut protein was found in the follow-up product after processing by one machine. After additional cleaning with hot water, the HNCC decreased regardless of the processing device to levels at or below 1 mg/kg hazelnut protein. In our pilot plant study, the application of an appropriate wet cleaning procedure in combination with quantitative monitoring of the cleaning efficiency reduced the hazelnut protein cross-contamination to a level at which severe hazelnut-related allergic reactions are unlikely to occur.