If one believes individual stakeholders, a feed additive simply needs to be safe – for the target animal, for consumers, for users and for the environment. They say its usefulness, its efficacy, would then be proved in practice. If the additive was no use, it would disappear again by itself. This view is contradicted by the basic Regulation on additives for use in animal nutrition (REG (EC) No. 1831/2003). In this, feed additives are determined via their function (Art. 2 (2)). These functions are described in Art.5 (3). According to these descriptions, the additive must satisfy at least one of the following requirements: it shall

a) favourably affect the characteristics of feed;

b) favourably affect the characteristics of animal products;

c) favourably affect the colour of ornamental fish and birds;

d) satisfy the nutritional needs of animals;

e) favourably affect the environmental consequences of animal production;

f) favourably affect animal production, performance or welfare, particularly by affecting the gastro-intestinal flora or digestibility of feedingstuffs, or

g) have a coccidiostatic or histomonostatic effect.

The founding act of EFSA (REG (EC) No. 178/2002, Chapter III) defines the task of the Authority as “providing scientific advice and scientific and technical support for the Community's legislation and policies in all fields which have a direct or indirect impact on food and feed safety.” This says nothing about assessing the efficacy, so that accordingly this does not form part of the core scope of Efsa’s tasks. The EU Commission deals with this by simply including the efficacy test in the mandate to evaluate applications for authorisation (the official request to EFSA to assess the dossier). Accordingly, the examination for efficacy of an additive has reached the FEEDAP Panel in the practice of additive assessment.

How many documents and test reports does a panel require in order to be able to draw inferences about efficacy? According to the principles “Once is as good as never” and “Twice is not much better than once”, the Commission and FEEDAP have agreed on submission of three proofs of efficacy in the target animals. There are a few exceptions here, for instance in the nutritional additives for which one trial is sufficient and this can even simply be carried out with laboratory animals, and the coccidiostats, for which six trials are required. Efficacy in technological additives is generally documented by in-vitro experiments. Efficacy that is established in the target animal must also be proven in these animals. The following remarks all concern in-vivo trials only.

What can be concluded from submission of three positive trials with a significant “improvement” of selected criteria? How many trials with a negative outcome are there which were not submitted? How representative were the trial conditions (race, keeping, feed, climate), and is this possible at all for 27 Member States? What producer of active ingredients would be satisfied with three trials to convince himself and then his potential customers? Shouldn’t there be an obligation on parties bringing a substance into circulation to present to the Authority the results of all trials they have had performed?

Is it not true that depending on the substance and animal species/category to be tested, the actually convincing effect can only be proven after a practical trial over a period of about two years?

Consequently even after submission of three trials to determine the effects that show a significantly positive outcome, the FEEDAP Panel therefore does not conclude that the substance is “effective”, but (only) that it is “potentially effective”.

The “three-trial-principle” is assessed particularly critically where applicants are submitting a feed additive that is to be used for different animal categories. Is it then necessary to submit three trials each for instance for fattening pigs and sows, or for beef cattle and dairy cows, or for broilers and layer hens? The FEEDAP Panel has so far assumed that efficacy proved in piglets by a 42-day trial cannot automatically be transferred to sows. The Panel will probably also adhere to this.

Is a significant difference (P<0.05, for ruminants P<0.1) in a performance parameter sufficient to certify potential efficacy? There are trials in which a reduction of feed outlay per kg weight increase in broilers from 1.63 to 1.60 (in other words by 0.03 or 30 g feed/kg weight increase) was shown as significant – statistically soundly. Anyone who has dealt with calculating this quotient themselves will understand our doubts.

For certain function groups of additives short-term trials are accepted as evidence of efficacy – for instance digestibility experiments for enzymes or comparative bioavailability studies for new types of trace element combinations. Here authorised additives are often used in a kind of positive control group and the “new” additive is characterised as better or of higher quality as a result of the studies. Since the start of additive assessment by FEEDAP, the Panel has “become accustomed” to not undertaking such classifications because the data basis for this is too narrow for a publicly accessible comparative assessment. The same applies for substances that have come off worse in a comparison. The standard formulations here are, the substance “is at least as effective as” or “has the potential to be effective”. “Uncertainties” in interpreting the data are increasingly being taken into account in assessment of the data.

One aspect which has currently not yet been resolved satisfactorily from the perspective of applicants and FEEDAP Panel members is the selection of parameters (endpoints) which are ultimately to be used to assess the efficacy. This is due on the one hand to the fact that the efficacy criteria of Article 5 of REG 1831/2003 have not been transferred consistently into the additive categories of Article 6 of the same Regulation. The effect of microorganisms (probiotics) is not derived from possible (favourable?) influencing of the intestinal flora, but instead can be considered as given if the performance of the target animal has been measurably influenced. By contrast with this, for enzymes substrate-specific digestibility experiments are increasingly being accepted as documentation of a probable influencing of performance (growth, feed consumption, feed outlay). Better and well-founded indications of the “mode of action” are admittedly helpful, but cannot solve the problem altogether. Immunological parameters, for instance in new function groups, will find it difficult to permanently and convincingly establish the relationship with influence on performance in healthy animals, however welcome they may be from the scientific standpoint. The Panel is currently engaged in a far-reaching revision of the corresponding guidance document on “Efficacy”. Interested stakeholders are involved within the context of a discussion group. The draft will be presented to the public for comments (probably in autumn 2017).
The Authors
The authors – (Gerhard Flachowsky and Jürgen Gropp) – are long-serving members of the EFSA Panel on Additives and Products or Substances in Animal Feed (FFEDAP). They report on the work of the FEEDAP Panel. The articles contain personal views and representations, not those of EFSA or the EU. Consequently they do not necessarily coincide with the viewpoints of EFSA or the EU. Consequently they do not necessarily coincide with the viewpoints of EFSA or the EU. As experts working on behalf of EFSA, the authors are subject to certain confidentiality obligations.