Table 9 Phase 2 evaluation summary on the adequacy testing of existing guidance documents for toxicology. Per default the existing guidance refers to the case study Table 3, and if needed an additional guidance is mentioned

Case	Specific evaluations of the case, including intended and unintended effects	Conclusions on the adequacy of existing guidance	Updates recommended for future guidance
1	The minimised genome B. subtilis strain derives from a QPS parental strain. The production of surfactin has been depressed and the genetic modification is not expected to raise additional concerns.	Adequate	No update needed
2 and 3	The toxicology of the presence of XNA and xenoproteins need to be assessed. It is considered that toxicity of xenocompounds can be detected as for other (chemical) compounds or proteins.	Adequate	No update needed
4	DNA of the Aspergillus oryzae remains in the product as may be the secondary metabolites produced by the production organisms.	Adequate	No update needed
5	The production strain is enlisted as QPS and autotrophy is not expected to change this QPS status.	Adequate	No update needed
6	Some cyanobacteria may produce cyanotoxins. In principle cyanotoxins and other microbial toxins (e.g. secondary metabolites including mycotoxins) can be detected using the standard toxicological approaches (genotoxicity and rodent studies).	Adequate	No update needed
7–9	The species is QPS granted. Therefore, no toxins or (geno)toxic metabolites are expected, and in vitro genotoxicity testing can be waived.	Adequate	No update needed
10	The product consists of bacteriophages.	The BIOHAZ guidance is general and not detailed, but there are two examples that applied an approach for non‐GM bacteriophages that are adequate	Update of the general guidances on the basis of gained experience is needed
11	Some cyanobacteria may produce cyanotoxins. In principle cyanotoxins and other microbial toxins (e.g. secondary metabolites including mycotoxins) can be detected using the standard toxicological approaches.	Adequate	No update needed
12	The parental species/strain is QPS. Therefore, no toxins or toxic (genotoxic) metabolites are expected. However, this product is designed to express viral and Salmonella antigens. Therefore, interactions with immune function should be checked.	Adequate	No update needed
13	The engineered strains [Pseudomonas simiae, P. putida and Ralstonia (Pseudomonas)] are non‐QPS organisms and toxicological studies may be necessary.	Adequate	No update needed
14	Regarding the oral uptake by humans and non‐insect animals, dietary non‐coding RNAs are generally rapidly degraded in the gastrointestinal tract and barriers (e.g. mucus, cellular membranes) limiting their cellular uptake in the gastrointestinal cells. Hence, the amount of RNAs taken up and absorbed after oral ingestion is considered negligible (EFSA, 2014; EFSA GMO Panel 2018) and therefore the dsRNA producing CTV viruses are not considered as hazardous. Oral ingestion by insects could trigger an off‐target RNAi effect.	Adequate for toxicological assessment of human or non‐insect animals. Guidance for testing off‐target RNAi effects in insects is provided in EU Regulation No 283/2013 and is adequate	No update needed No update needed
15	The SynBioM Snodgrassella alvi (producing dsRNA) is expected to be present in bee products (such as honey) similarly to the natural strain. Regarding the oral uptake by humans, dietary non‐coding RNAs are generally rapidly degraded in the gastrointestinal tract and barriers (e.g. mucus, cellular membranes) limiting their cellular uptake in the gastrointestinal cells. Hence, the amount of RNAs taken up and absorbed after oral ingestion is considered negligible (EFSA, 2014; EFSA GMO Panel 2018) and therefore the dsRNA producing bacterium is not considered as hazardous. The potential adverse effect of the feed additive on honeybee should be assessed. Specific guidance for testing adverse effect of agents on insects do exists in other areas (e.g. PPP and GMO), however they have different aims in line with the purpose of the product.	Adequate. Specific indications for tolerance and efficacy in insects as target species of feed additives are not available in the FEEDAP Guidances 2018 and 2017.	No update needed Update needed