Single and multi-laboratory validation of a droplet digital PCR method

The authorisation of genetically modified food and feed in the EU is subject to the provision of evidence of safety and of the availability of reliable analytical methods. These methods represent an essential tool for official laboratories to enforce a harmonised market control. Here the validation of droplet digital PCR (dPCR) methods has been performed for studying if the performance and acceptance parameters set by EU and other international guidelines for the analysis of genetically modified organisms (GMO) in food and feed are suitable and achievable also with such methods. The single-laboratory validation study showed that performance requirements set for GMO analysis by real time PCR can also be used to assess dPCR-based methods. Moreover, trueness and precision were assessed for both simplex and duplex formats in a multi-laboratory validation study organised according to international standards. Overall, the data on trueness, repeatability and reproducibility precision resulting from the collaborative study are satisfying the acceptance criteria for the respective parameters as stipulated in the EU and other international guidance such as the Codex Committee on Methods of Analysis and Sampling (CCMAS). For instance, the duplex droplet dPCR method for MON810 showed relative repeatability standard deviations from 1.8% to 15.7%, while the relative reproducibility standard deviation was found to be between 2.1% and 16.5% over the dynamic range studied. Moreover, the relative bias of the dPCR methods was well below 25% across the entire dynamic range. In addition, other aspects supporting the application of digital PCR for the control of GMOs on the market were experimentally assessed such as the conversion of the measurement results from copy number ratio to mass fraction, the influence of the DNA extraction step and of the ingredient content. It was found that the DNA extraction step added only a limited contribution to the variability of the measurement results under the studied conditions. The decreasing amount of the target ingredient content may decrease the level of precision of the method, although within the acceptance range of GMO performance parameters.