Poison and antidote

Poisonous or toxic divisionals have been a somewhat exotic but much commented upon topic, that has led to the revocation of a handful of patents in the past few years.

The excitement and dread generated in the European patent profession will culminate when the Enlarged Board of Appeal of the EPO, to which the topic has recently been referred, issues its decision G 1/15, probably within the next couple of years.

It looks like most, if not all, of the profession hopes that the Enlarged Board will put an end to a theory which is often viewed as nonsensical. I am not sure that I concur with this unanimity. But what everyone will agree on is that there is no predicting with absolute certainty what the Board will have to say on this subject.

So, since the world will not stop turning and patent applications will not stop being filed while we are all waiting for the verdict of the wise people in Munich, it may be worth looking for an antidote to the poison.

To put it more simply, today’s question is whether there is a way to immunize newly filed patent applications against a possible “poisonous divisional” objection further down the road, in case the Enlarged Board does not dispel this objection as being rubbish.

In order to answer this question, we need to look more closely at what the problem really is about. This is where the notion of “partial priority” comes into play.

Let’s assume a first application A, and a subsequent European application B claiming the priority of A. Let’s also assume that the invention claimed in B is broader than what was originally disclosed in A. For example A discloses a process with a temperature in the range of 10-100°C, whereas a broader range of 5-500°C is claimed in B. Or A discloses an apparatus comprising one part soldered to another part, whereas B claims the same apparatus wherein the two parts are attached to each other by any possible means.

In such scenarios, quite clearly, the B claims do not fully benefit from the priority of A, since B does not relate to the same “subject-matter” as A (in accordance with opinion G 2/98).  The tricky issue is whether the priority claim is fully invalid or whether it is only partially invalid.

If the priority is held fully invalid, and if A is a European patent application which gets to be published (in principle after the filing date of B), then A becomes prior art under art.54(3) EPC, i.e. prior art for the purposes of the assessment of novelty only (not inventive step). Therefore, if A discloses one or more embodiments falling within the scope of the B claims, these claims are invalid. We could call this a “toxic priority” situation.

Moreover, even if A itself is not prior art, for instance because it never gets published, or because it is not a European patent application, the existence of a divisional application C stemming from B can also be prejudicial to the novelty of B. Indeed, the part of C which does benefit from the priority of A can be opposed to B for the purposes of novelty only. In such as case, the divisional application is said to be toxic to the parent. But the reverse is also true, and the parent B could be toxic to its child C.

One solution which has been proposed to get out of this mess is the notion of partial priority.

If we accept that the claimed subject-matter in B can be divided up into two parts, a first part that benefits from the priority of A and a second part that does not benefit from the priority of A, then an argument can be made that:

  • said first part is immune to any toxicity because it benefits from the earliest filing date in the family; and
  • said second part is also immune, because it is not disclosed in A, so that there can be no anticipation by A (or any other application in the family validly claiming the priority of A).
Can a poison container safely be generalized to a beverage container without losing the priority for the poison embodiment?
Can a poison container safely be generalized to a beverage container without losing the priority for the poison embodiment?

Whether and to which extent partial priority can be conjured up is one of the main questions that the Enlarged Board of Appeal has to deal with in G 1/15 (see the referral decision T 557/13). In the founding opinion on priority G 2/98 it was mentioned that “the use of a generic term or formula in a claim for which multiple priorities are claimed […] is perfectly acceptable […] provided that it gives rise to the claiming of a limited number of clearly defined alternative subject-matters”. There has been a debate as to what this exactly means, and whether this implies restrictions on the conditions under which partial priority in a generic claim can be acknowledged.

With all this in mind, and without being able to guess at the further guidance which the Enlarged Board will offer, it is reasonable to state that there is a risk of invalidation when a European application is filed:

  • which claims a priority,
  • which comprises claims that have been generalized relative to the teaching of the priority document,
  • if the priority document is a European patent application which gets to be published, or if one or more divisional applications are filed at any point of time.

But, oddly enough, there are instances in which more poison could lead to a cure.

In fact, one solution when the above risk is identified at the time of drafting a priority-claiming application consists in adding a fallback position containing a feature which is not disclosed at all in the priority document. Indeed, if the priority claim becomes invalid because of an added feature (as opposed to just because of a generalization), then there can be no anticipation by the priority document or by any parent or divisional application – since the added feature restores novelty over those. Ideally, the added feature should be minimal and not significantly affect the scope of protection. Remember: it is only necessary to restore novelty over a potential art.54(3) EPC disclosure, so inventive step is not at stake.

If we go back to the above examples, if A discloses a process with a temperature in the range of 10-100°C, and if this range needs to be broadened to 5-500°C in B, why not also add in B another innocuous feature not at all disclosed in A, at least in a dependent claim? For instance, we could add that the process is performed within a time frame of 1 minute to 100 hours, if this is relevant.

Or, if A discloses an apparatus comprising one part soldered to another part, and if the claims in B need to be broadened to all possible modes of attachment between the parts, why not also add in a dependent claim in B reciting that the apparatus weighs from 100 g to 100 kg, assuming this is a relevant range.

Thus, even if partial priority is denied for the generalized claim, it is always possible to rely on the fallback and decide to completely forget about priority and at the same time make sure that there can be no anticipation by A or any application claiming the priority of A.

This suggestion comes with two caveats, though.

The first caveat is that, in case any other relevant disclosure took place during the priority year, the complete loss of the priority date may not be an option. In that case, the only solution seems to go back to what was originally disclosed in the priority document A and stick to that.

The second caveat is that features to be added in the second filing B should be as innocuous as possible (in order not to significant restrict the scope of protection); but not completely meaningless. Indeed, meaningless limitations would not be enough to distinguish over an art.54(3) disclosure, bearing in mind that the disclosure has to be read and assessed by the skilled person. For instance, if A discloses an invention on a car, and if we decide to add in B that the car has four wheels, it is very possible that the skilled person will be deemed to understand the car of A as necessarily comprising four wheels, in which case we would be back to square one.


CASE REFERENCE: Board of Appeal 3.3.06, T 557/13, Infineum USA LP v. Clariant Produkte (Deutschland) GmbH, July 17, 2015

Revocation out of thin air

Patent eligibility has been quite a hot topic in the U.S. since the Alice, Myriad and Mayo Supreme Court decisions. Whether a patent simply attempts to preempt an abstract idea or a law of nature or whether it is truly directed to an application of those – worthy of patent monopoly – is, as far as I can understand, the million-dollar question that everyone now has to struggle with on the other side of the Atlantic, and will have to struggle with in the foreseeable future.

Over here, on the Old Continent, things look pretty different, on the face of it. The EPO has set a low threshold for the so-called “technical character” requirement, actually such a low threshold that it is barely a threshold at all. Of course, this is not the end of the story, because “non-technical features as such, do not provide a technical contribution to the prior art and are thus ignored in assessing inventive step” as the November 2014, soon-to-be-revised Guidelines for examination nicely put it. So, inventive step can be a real challenge, but patent eligibility is at least one concern that applicants can forget about, right?

Well, we may have to think again, in view of a judgment from the Paris Tribunal de Grande Instance (TGI) of July 3, 2014.

This case is an action for revocation brought by a company named Evinerude against French patent No. FR 2822237 assigned to Aair Lichens. According to the first and only claim of the patent, the invention relates to the use of lichens exposed to sources of emission of chlorinated compounds, and used in the form of transplants or cultivars, for making quantitative measurements of polychlorodibenzodioxine (PCDD) or polychlorodibenzofurane (PCDF) compounds and evaluating their impact on the environment.

Indeed, it turns out that lichens are good indicators of the presence of PCDD and PCDF air pollutants. Thus, by measuring the amount of these chemicals in lichen samples, it is possible to evaluate the level of atmospheric pollution that they have been exposed to.

An elegant breathing device which could become handy in case of a polychlorodibenzodioxine or polychlorodibenzofurane leak.
An elegant breathing device which could become handy in case of a polychlorodibenzodioxine or polychlorodibenzofurane leak.

It looks like the PCDD and PCDF-loaded Parisian air was not favorable to Atlantic coast-based Aair Lichens, since the TGI revoked the patent based on Article L.611-10 of the Code de la Propriété Intellectuelle, which mirrors Article 52 of the European Patent Convention and provides a list of things which are not inventions, discoveries being the first item in the list.

The court started by recalling the long-standing principle that

A mere discovery cannot be patentable. Actually, the discovery preexists man’s intervention, whereas the invention is the result thereof. A discovery thus does not bring any novelty to the state of the art since it is at the stage of sheer knowledge. However, although the subject-matter of a discovery is not patentable, a practical application can give rise to the grant of a patent.

In this case, the invention relied on the discovery of a natural phenomenon, namely the absorption of PCDD and PCDF compounds present in the ambient air by lichens. The patent proprietor’s position was however that the patent was not directed to this discovery per se but rather to process steps of measuring the amounts of these chlorinated compounds in lichens and then evaluating their impact on the environment. But the court was not convinced, and replied as follows:

As it is drafted, the only claim of the patent does not protect the process steps but merely the statement that measurements can be made to evaluate the impact on the environment, which is not a process invention. The technical means for carrying out this statement could be protectable if they were described and claimed as such.

The patent was thus revoked.

To me, this looks more like a US-type approach than an EPO-type approach. After all, it seems that the court’s reasoning could be applied in a similar manner to many use claims which are routinely granted by the EPO.

An important caveat is that the patent at stake was somewhat light, with four pages of description and a single claim. It appears from the file wrapper that the application was drafted and prosecuted by the inventor himself, which is always risky. In fact, it is not every day that one comes across a claim starting with the words “the invention relates to”. The text does contain an example of measures performed in samples taken from various geographical areas, but there are no detailed explanations in the text on a possible process to be implemented. Also, the court noted that the skilled person, at the filing date, already knew that the pollution of a site by PCDDs and PCDFs results in atmospheric contamination, and that these compounds are easily absorbed by plants such as lichens [sic, I am not sure lichens can be considered as plants] so that these are good pollution indicators.

In other terms, one could infer from the court’s findings that the patent could also have been revoked for insufficiency of disclosure or lack of novelty or inventive step.

It is difficult to assess to what extent these flaws played a part in the TGI’s decision. Would the situation have been different if the patent had contained more technical information or if the prior art had been less relevant? It is certainly possible. But on the other hand, the way the claim was drafted did seem to play a critical part in the decision.

Therefore, when an invention relates to an application of a natural phenomenon, applicants would be well advised to draft at least some claims reciting the invention in terms of practical process steps and not just use claims broadly covering the basic concept of the invention.


CASE REFERENCE: Tribunal de Grande Instance de Paris, 3ème chambre, 1ère section, Evinerude SARL v. Philippe Giraudeau & SARL Aair Lichens, July 3, 2014, RG No. 10/14406.