EPO revokes patent for low-calorie confectionery
Board of Appeal decision (T1384/16) highlights the importance of data in food and drink patents, as well as the pitfalls of opposition and appeal proceedings.
As consumers become increasingly health- and environmentally-conscious, food and drink companies are continually driven to innovate in order to meet complex demands, deliver new experiences and align with current trends. In today's competitive markets, patents can provide valuable protection for innovative compositions, helping companies to ring-fence their market niche. But how easy is it to get, and defend, a patent in this sector?
More often than not, new food and drink compositions are based on a novel combination of existing ingredients, making it more difficult to demonstrate the "inventive step" needed to obtain a patent. However, if a particular combination of known ingredients produces an unexpected effect, then an inventive step may be established. Patent applications for food and drink compositions are examined in much the same way as applications for other chemical inventions. Therefore, patent offices will typically want to see evidence, usually in the form of experimental data, of the unexpected effect that arises from the new combination. Robust data is important not only for obtaining a granted patent, but also for defending it against challenge by a third party. Cargill's case emphasises why:
How Cargill lost its cool
The patent (EP2136652) related to compositions containing erythritol, a sugar substitute used to reduce the calorie content of foods. The problem is that erythritol produces a cooling sensation which is undesirable in confectionery. Cargill claimed to have solved this problem by combining sugar esters and/or certain dietary fibres with erythritol at a particular weight ratio. The patent contained data showing that a selection of the sugar esters and fibres claimed were effective at reducing the cooling effect of erythritol.
The patent was challenged in opposition proceedings before the European Patent Office. The opponent argued that the patent was insufficient since the data didn't make it credible that a reduction in the cooling effect of erythritol could be achieved by all of the sugar esters and fibres claimed. However, in the absence of any supporting evidence, this attack failed to establish serious doubts that the effect could be achieved across the whole scope claimed. Nevertheless, the patent was revoked on other grounds.
In the subsequent appeal proceedings (T1384/16), the opponent reiterated that the examples in the patent could not be generalised to all of the fibres and sugar esters claimed. Surprisingly, despite the opponent failing again to present evidence to substantiate its attack, Cargill chose to submit its own additional data to defend the patent. This proved to be a costly mistake, as the new data revealed that a reduction of erythritol-induced cooling was not achieved by inulin, one of the fibres falling within the patent claims. Since there was no evidence that all of the other fibres claimed had a common structure or property not shared by inulin, the Board no longer considered it credible that all of the claimed fibres would be effective in reducing the cooling effect of erythritol. As such, the patent claims as granted were found invalid.
Cargill further came unstuck when the Board refused to admit all but one of their 16 auxiliary requests (amended claim sets) into the proceedings for being filed late and without adequate justification. The single auxiliary request admitted was found to contain an unallowable amendment. As a result, Cargill lost the appeal, and their patent.
Lessons for food and drink manufacturers
While this outcome may have left a bitter taste for Cargill, it should not deter companies from seeking patents for new food compositions. On the contrary, the fact that Cargill's patent was opposed and appealed indicates the value of IP in the food and drink sector.
As Cargill found, a single contradictory result can undermine a patent which claims a technical effect across a group of substances. Innovators can help to mitigate this risk by carrying out experiments to determine which ingredients and quantities deliver the effect, and which do not, before filing a patent application. When applications must be filed with incomplete data, as is often the case, suitable fall-back positions should be included so that non-working compositions can easily be excluded.
There are no concrete rules on the type or quantity of data needed to support a patent application; this will depend on the nature of the invention. If, for example, it is claimed that a beneficial effect is provided by a class of ingredients, ideally the patent application will contain evidence that this benefit is achieved for at least one representative member of the class. If the class covers a diverse range of ingredients, multiple data points may be required.
It is helpful to include comparative data where possible. Cargill's patent contained experimental results showing the temperature drop of water following addition of erythritol combined with various sugar esters or fibres, versus erythritol alone. However, data does not necessarily need to be quantitative; in some cases qualitative data, such as tasting panel results, may be sufficient.
Cargill's case also highlights the importance of strategy in contentious proceedings. The patent might have been saved, albeit in a restricted form, had Cargill's amended claims been considered. Even though the case was heard under the old Rules of Procedure of the Boards of Appeal (RPBA 2007), the Board appeared to apply a particularly strict interpretation – perhaps a sign of things to come under the new, more stringent RPBA introduced in January 2020. Patentees should therefore consider filing auxiliary requests at an earlier stage of proceedings, and ensure that adequate explanations are given as to why the amendments address the objections.
Patents can provide valuable protection for innovative food products in competitive markets. Any new food or beverage composition which delivers a non-obvious benefit may potentially be patentable, including animal food. However, without the right data and strategy, the pursuit of IP may ultimately be unpalatable.
Originally published 02 July, 2020
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.