Are Functional Foods ‘Therapeutic’? Insights from the EPO’s Latest Ruling

Abstract: EPO decision T 1396/23 confirms that functional foods can support medical‑use claims when their therapeutic effects are credible, data‑backed, and tied to recognised clinical outcomes. This provides some comfort to innovators in this space, but many questions of implementation remain unanswered.

*              *              *

We have previously discussed the challenges in securing IP protection for functional foods. On paper, functional foods seem straightforward to patent, as the presence of a functional effect such as the ability to treat or prevent a disease could normally be relied on as an inventive effect. This is true either for entirely novel food compositions, which may be protected as such, or for new uses of known foods through the use of purpose-limited medical use claims (“composition X for use in treating or preventing disease Y”).

Both these practices are long-established for pharmaceuticals, but create several challenges for functional foods. The first is regulatory. Between challenging frameworks for “novel foods” (i.e. those comprising ingredients not previously recognised as safe), and prohibitions on explicitly marketing foods for the treatment or prevention diseases required to infringe a claim, a careful approach is needed to provide a valuable scope of protection.

However, a more fundamental challenge is whether or not medical use claims are even appropriate for foodstuffs, and when and whether therapeutic effects can be relied on for non-pharmaceutical interventions in the first place. At the EPO, medical use claims are only available under Article 53(c) EPC for “substances or compositions” for use in treatment by surgery or therapy. Historically, this has been interpreted narrowly, applying primarily to active pharmaceutical ingredients, and it is not clear whether foods which do not have a clearly defined, pharmaceutical-style “drug-target-effect” mechanism of action, would qualify for these formats.

Moreover, in order to proceed to grant, a medical use claim requires evidence that the disease or condition is actually treated or prevented. Often, this data is not available in full at the time of patent filing, and a degree of inference is required from in vitro, in vivo, or animal models. This results in a requirement that the claimed effect is “credible” (usually based on mechanistic principles), backed up by at least some supporting data in a valid model. Yet, for functional foods, both mechanisms of action and the correlation between ingredients and effects can seem unclear and even nebulous when compared to drugs. As the case law for assessing these requirements was developed for pharmaceuticals, it is unknown whether foods can meet them.

In recent decision T 1396/23, the EPO’s Boards of Appeal considered these questions, and provided some useful guidance.

An appeal from the opposition division, the claims of the patent at suit related to complexes of whey protein and polysaccharides such as alginate.

For example, claim 1 read:

1. Composition comprising polysaccharides and whey protein micelles for use in the treatment or prevention of a disorder linked to an increase in plasma postprandial insulin selected from the group consisting of diabetes, for example gestational diabetes; impairment of glucose metabolism; hyperinsulinemia or insulin resistance, in a subject; wherein the polysaccharides have a negative zeta potential at a pH value in the range 2.5 to 4.5 and are selected from the group consisting of alginate, xanthan, pectin, gum karaya, gum Arabic and carrageenan; the weight ratio of whey protein micelles to polysaccharide is between 30:1 and 0.8:1; the whey protein micelles are obtainable by adjusting the pH of a demineralized native whey protein aqueous solution to a value between 5.8 and 6.6 and subjecting the aqueous solution to a temperature between 80 and 98°C for a period of between 10 seconds and 2 hours, and the polysaccharides and whey protein micelles are in the form of polysaccharide-whey protein micelle complexes.

The patent as filed contained data in an animal model (healthy minipigs) which showed that WPM complexes prompted a reduced (yet still sufficient) insulin response compared to conventional whey isolate, the intake of which is a known measure for lowering insulin demand. This is relevant, and whey isolate is already used to formulate nutritional formulae for patients with diabetes and other glucose metabolism disorders.

The key question at stake here was whether or not it was credible that the claimed WPM complexes were capable of carrying out the claimed therapeutic treatment.

The Opponent/Appellant argued that, as the WPM complexes were nutrients and not drugs, cannot be absorbed into blood stream, and do not target a specific target for a specific disease, they were incapable of decreasing postprandial insulin release induced by a meal. Secondly, they put forward a robust critique of the data in the application, effectively arguing it was insufficiently convincing to show actual treatment or prevention of the specified disease. In particular, the use of a healthy animal model, and comparison of the slow-digested WPM complexes to fast digested whey isolate was questioned.

The Board was unconvinced by these arguments. They were clear that the claimed complexes had a relevant and beneficial effect and “it is immaterial what the mechanism of action is, whether there is a specific target, and also whether the complexes act on a specific target … In the context of the invention, the claimed compositions comprising WPM complexes can thus be regarded as “a substance or composition” within the meaning of Article 54(5) EPC that is used to carry out the claimed therapeutic method.”

Thus, when viewed as a whole, the application was seen to teach that whey is known as a nutrient, and benefits with reducing the risk of, and treating, metabolic diseases associated with high-fat diets and/or postprandial insulin levels. It was therefore reasonable to assume the WPM complex would have these effects. The application further made clear the intention to replace whey with the claimed WPM complexes, to provide a nutritional source that induces a reduced, but still adequate, insulin response, thereby preserving pancreatic function and improving insulin sensitivity. The minipig model was valid in this context.

The Board concluded that, when taken as a whole. it was credible that the treatment worked and, as a result, a medical use format was viewed as appropriate, and furthermore the claims were acknowledged as inventive.

Analysis

Functional foods often skirt a regulatory and legal grey area, but developments like this help to provide some welcome clarity.

Medical use claims remain a route to patentability for functional foods. By acknowledging that foods can qualify as a “substance or composition” within the meaning of Article 54(5) EPC, and therefore for medical use claims, the Board provides welcome precedent for pursuing these kinds of claims in the future.

The holistic principles for assessment of data in the application are also encouraging. The Board was willing to view the evidence through the lens of nutrition, not pharmaceutical science, and drew its conclusions based on data that showed effects that fell short of actual treatment or prevention of disease. In this, the Board is aligned with another recent decision T 2036/21, concerning a fatty-acid composition for use in the prevention of dementia. In that decision, the Board went even further, noting that “In proceedings before the EPO it is not a prerequisite to perform a statistical analysis of the results and to determine a specific confidence interval, as is most often required in biomedical research and by health authorities granting marketing authorisations for medicinal products”. In both these cases, strong mechanistic implications of data appear to be sufficient for providing the required support. This will be comforting for functional foods producers, hoping to avoid costly and extensive medical trials.

Key, however,  is the fact that improving insulin response is clearly known as a cornerstone of diabetes therapy. Would the Board have reached the same conclusion if the effect were a less well-defined disease, such as “gut health”? Our view is that it is not clear they would, so applicants should consider carefully how they tie their claimed effects to factual data points when drafting these kinds of claims.

Finally, it is instructive that the claims in this patent concerned treatment or prevention of disease. There is some question in Europe as to the value of second medical use claims for functional foods that treat or prevent disease as, in some territories, the underlying products would have to be offered for sale for this curative purpose in order to be infringed. This conflicts with prohibitions against labelling foods with assertions they treat or prevent diseases, for example EU Regulation 1169/2011. It would be interesting to see whether the principles established in this case would be applicable to medical use claims directed towards reduction of risk of disease, as these face a favourable regulatory environment. However, for now, these remain untested at the EPO.

This article was prepared by Patent Director Andrew Tindall