Divide and destroy: an overview of poisonous divisionals
A patent application can in effect destroy itself. It works like this: a priority filing might claim a kinase which has an acceptable homology with a particular sequence listing—call that kinase-SL. The filing which is eventually made claims the kinase as originally claimed and a more general description of it, such as any kinase which has general binding characteristics for a class of proteins, of which kinase-SL is a member—call that the class of kinases.
During the prosecution of the eventual filing a divisional application is made which makes a claim to kinase-SL and also to the class of kinases. Once granted the patent claims only the class of kinases. What is the status of the divisional application? The answer is that it can destroy the patent from which it was originally divided.
How does this work? The priority filing allows priority to be derived only as far as kinase-SL is concerned. As kinase-SL is not claimed in the eventual application, the granted patent is not entitled to priority and therefore its relevant date is its actual filing date. However, the divisional is entitled to priority in relation to kinase-SL. At this point one should recognise that the patent, despite claiming a generalised form of kinase, can be revoked for lack of novelty owing to the anticipation by a prior published specific example. In other words, if kinase-SL is published earlier then it can destroy the novelty of an invention claiming the class of kinases.
Can the divisional application do this? Yes. Why? Because the divisional has the priority date and the patent does not. Prior art includes not only material in the relevant technical area made available to be public before the priority date of the invention but also any patent application filing (whether made available to be public or not) which has a priority date earlier than that of the patent in issue.
What is the reason for this rule? To ensure that the integrity of the first-to-file system is kept intact. Imagine that the facts cited above are slightly different (but not materially so). Instead of the divisional application being made, imagine that a separate and independent application is made by a separate and independent inventor acting very shortly after the priority filing. Under a first-to-file system the first inventor to file should get the benefit of a granted patent. In the normal course, and provided that the patent is supported by the priority filing, the priority date is kept intact.
Ensuring priority
In order to ensure first filing priority, a rule should exist which says that any second filing should be anticipated by any first filing. Such a rule does exist and its essence of operation is cited above. Then imagine that the priority date is lost because of lack of support, ie, the patent claimed something which existed only in the eventual filing and not the priority document. Then the second application is the first to be filed. If the second application anticipates the patent then that is well and good—the first-to-file system is kept intact. The difficulty is when this all happens as a result of the folly of one entity and there is no danger of the first-to-file system resulting in the adverse consequences of the first-to-invent system.
Why do we have a first-to-file system? The answer is that, like democracy, it is a terrible system but it is better than all the others. The only real alternative to the first-to-file system is the first-to-invent regime. That is, in its purest form, the first person to invent an invention is entitled to get a patent relating to it no matter when he or she filed or files an application to obtain a patent in relation to the invention (if ever he or she does).
This can lead to some very costly and unnecessary court cases where the notion of who did what first is determined—in the US (under a slightly more restrictive set of rules) these are called interference proceedings. It should be added that the US has a hybrid between first to invent and file, and policy seems to suggest that the US is moving towards a first-to-file system.
“The divisional application shall be deemed to have been filed on the date of filing of the earlier application and shall enjoy any right of priority.”
The European system avoids the burden of interference proceedings by having a first-to-file system. It should be noted that there is still a type of interference system in place where there is a dispute about who the proper inventors of an invention are or who the proper applicants for a patent in relation to that invention are—these cases are often known as entitlement proceedings, although they are nowhere as costly or burdensome as their US interference counterparts. There are probably other reasons, but avoiding interference proceedings is, in my opinion, one of the better ones.
Are there any ways around the problem? There may be if, for instance, one was to make a class claim but specifically disclaimed the novelty-destroying prior art (the class of kinases but excluding kinase-SL), but this seems a bit unfair if there is no threat to the first-to-file system and the prior art in question remains unpublished. Additionally, novelty-saving carve outs are notorious for the complexities which they actually create.
In Disclaimer/PPG (G1/03) and Disclaimer/Genetic Systems (G2/03) in 2004, the Enlarged Board of Appeal (EBoA) held that disclaimers which act as carve outs (and can only act as carve outs) to avoid novelty objections must be narrow and act only to avoid the objection, and must be as clear and concise as claims are required to be—as they form part of the claim.
More recently, in Partial Priority/Infineum (T 0557/13) on July 17, 2015, the Board of Appeal in referring the question to the EBoA stated: “Although issues of priority, divisional applications and/or the purpose of article 54(3) EPC were addressed by the Enlarged Board of Appeal in … [previous cases], these … [cases] are silent on the specific question of whether a parent and any divisional stemming from it may be considered as state of the art under article 54(3) EPC in respect of each other.”
Divergent arguments
Consequently the matter was duly referred and a decision is pending. However, while we are waiting it might be instructive to look at the arguments presented by the parties in the Infineum appeal. The parties favouring the current ‘novelty-destroying’ approach stated that:
(1) There was no way out of the language of the EPC unless the convention’s words are construed in a manner inconsistent with the words which it uses;;
(2) Divisional requests, if not novelty-destroying, had the effect of extending the time during which advantage could be taken of the priority period; and
(3) Calling a divisional application a divisional application does not mean anything special—it is to be treated like any other application (as being independent of its parent application).
The parties favouring the approach of saying that the divisional application is not novelty-destroying stated that:
(1) The situation was clearly problematic and unfair and the legislator did not foresee it, and had he done so he would have expressly dealt with the problem;
(2) A number of experienced commentators supported the ‘novelty-maintaining’ view;
(3) The Paris Convention would be contradicted since the convention supports the right to “freedom of division” and causing a divisional to be ‘toxic’ would impair that freedom;
(4) A proper construction of article 54(3) of the EPC deemed a divisional not to be novelty-destroying provided that it was not published before the later filing date; and
(5) A divisional is entitled to the filing date of its parent and the priority that such a filing enjoys (to which there may be a corollary that if no priority date is enjoyed then the divisional does not enjoy it either).
On this last argument, this turns on the wording of article 76(1) of the EPC, which reads that “the divisional application shall be deemed to have been filed on the date of filing of the earlier application and shall enjoy any right of priority”.
What ought to be done? It is highly unlikely that the EBoA will construe the EPC in a manner which is inconsistent with the language it uses.That said, there is much attraction to the last argument of the anti-toxic brigade that a divisional enjoys (or doesn’t) the same status in terms of priority of their parent applications. So if the parent application loses priority because of added matter, for example, then the divisional—whether matter is added or not—does too.
The difficulty with this submission is that the word “enjoy” implies benefit. If a priority date is lost to the parent filing then it can hardly be said to be beneficial to the junior divisional for the priority date to be lost—there is nothing to be enjoyed.
Ashley Roughton is a consultant at Nabarro. He can be contacted at: a.roughton@nabarro.com
