Over the last year or so, the concept colloquially known as “poisonous divisionals” or “poisonous priority” has been a hot topic for analysis and debate. It is a large topic that includes a number of pieces of legislation, opinions, and arguments.
This article aims to provide a broad overview of the current position at the EPO and in the UK courts.
For those who are not familiar with the terms “poisonous divisionals” and “poisonous priority”, they are used to describe the stance that claims of an application (or patent) that are not fully entitled to priority can lack novelty over the disclosure of a later-filed and later published divisional application or priority application in the same patent family, respectively, under Art. 54(3) EPC or S.2(3) of the UK Patents Act.
As an example, consider a first (parent) application EPA claiming priority from an earlier national application in which the claims of EPA are only partially entitled to the claimed priority date, because the claim of EP recites a numerical range 1-10 and the priority application has the range 4.2-8.3. Each of EPA and the priority application has an example using a value 5.5.
Later, a divisional application EPD is filed, which is given the filing and priority dates of EPA and has identical description to EPA. EPD has an identical description to EPA.
The “poisonous divisional” theory says that on publication, EPD can be used as novelty-only prior art against EPA for any subject-matter for which the claims of EPA are not entitled to priority. In this scenario, the example using a value of 5.5 renders the non-entitled numerical range 1-10 not novel.
An argument against the “poisonous divisional” theory says that if the claims of the parent application are only partially entitled to priority, so that there is subject-matter in the claim that is entitled to priority and subject-matter that is entitled only to the filing date, the divisional application can only be entitled to the same date as the parent application for each particular subject-matter (remember that the disclosure of the divisional cannot extend beyond that of the parent application).
Using the above example, the argument could be made that EPA is entitled to priority for the range 4.2-8.3, though not for 1-4.1 or 8.4- 10.
In this case, EPA does not lack novelty over EPD because the value 5.5 is inside the scope of the priority-entitled part of the claim.
Similarly, it can be argued that the priority document does not disclose the range 1-4.1 or 8.4-10, so that the range 4.2-8.3 and the example cannot be used to attack the novelty of the claim of EPA because EPA is entitled to priority for these subject-matters.
Commentators and practitioners have been awaiting a decision that confirms whether or not the EPO or the UK courts would accept that a divisional or priority application could be used as prior art against a member of its own patent family using the “poisonous” theory.
There are also broader questions about whether the fate of an application should rightly rest on whether or not it is divided, or whether the priority application is allowed to publish after the filing of the application.
This year, we have seen that at least some members of the EPO and in the UK courts consider that a divisional application and a priority application, respectively, can be “poisonous”.
In late 2013, the Board in T1496/11 held that claim 1 of the patent lacked novelty over its divisional application. Correspondingly, in Nestec SA & Ors v Dualit Ltd & Ors  EWHC 923 (Pat), the UK High Court concluded that the claims of a patent were not entitled to priority. The priority application had been allowed to publish, and therefore the claims of the patent in suit were found not novel over the published priority document.
The application of the “poisonous” approach depends to a large extent on whether or not the claims of the patent application (or patent) are entitled to priority; that is, whether the claims of the application (or patent) relate to “the same invention” as that of the priority document. G2/98 is a decision of the EPO’s Enlarged Board of Appeal concerning the requirement for the “same invention” in a priority document and a later application for a valid priority claim to be accorded. [This decision was also cited in the ruling of the UK court in Nestec v Dualit.]
In G2/98, the Enlarged Board said that multiple priorities could be allowed for a definition in a claim using a generic term/formula, provided that it “gives rise to the claiming of a limited number of clearly defined alternative subject-matters”. The meaning of “a limited number of clearly defined alternative subject-matters” has been the subject of several discussions at the EPO for determining whether or not a claim is entitled to priority. The Boards, so far, appear to have taken rather divergent approaches to the question of “the same invention”, as detailed below.
A strict approach was taken in T 1127/00. In that case, claim 1 had a broader, more general structural formula than the priority application that the Board considered covered a large number of structural and functional embodiments. The board said there was no “limited number of clearly defined alternatives…the fact that they might be intellectually envisaged to fall within the scope of the claim does not make up for a clear and unambiguous presence of these alternatives, individualized as such, in the claim”.
In T 1877/08, the claim contained a broader numerical range than the numerical range disclosed in the priority application. The Board said that priority could only be accorded for those elements of the claim included in the priority application. They interpreted “elements” as meaning “relating to separable alternative embodiments” (G2/98). In this case, the Board said that the numerical ranges represented a continuum of values and therefore were not distinct, separable alternatives.
In Nestec v Dualit, the court held that the broadening of the claim in relation to a housing meant that the claim could not be split into a limited number of clearly defined alternatives because two versions of the housing were not clearly defined as alternatives in the application. In each case, partial priority was not accorded.
A less strict approach to the question of priority entitlement was taken in T 1222/11, in which the EPO Board explicitly disagreed with T 1127/00.
In T 1222/11, the Board said that it was sufficient that it was possible to conceptually identify a limited number of clear alternative subject- matters for partial priority to be accorded.
The Board in T15/01 also considered the issues surrounding multiple and partial priority (reasons 33, 36 and 38), though not specifically in the context of “the same invention”. They opined that the Paris Convention “guarantees the right to divide patent applications while presenting the benefit of the right of priority also for the divisional application”, coming to the general conclusion that the EPC system encourages combining originally disclosed subject-matter with further improvements and additional embodiments developed in the priority period within one application. It is difficult to envisage how this Board could reconcile such a view with “poisonous divisionals” or “poisonous priority”.
To date, the case law appears to be focussed on having broader claims in the parent application compared to the priority application. It should be borne in mind that similar “poisonous” arguments might prevail for claims that are narrower in the parent application than in the priority application if they result in a loss of priority, for example using the rules relating to selection inventions to determine whether a narrower range is novel over a broader range.
So, what can applicants and patentees do to avoid these sorts of difficulties?
Of course, no issue arises if the EP application is identical to the priority application. Therefore it is always advisable to retain all of the disclosure from the priority application in the later application, including all explicitly-recited ranges and sub-ranges. This at least allows the possibility of later amendment to priority-entitled subject- matter.
In some cases, it might be possible, before filing the application, to abandon the priority filing and re-file a new priority case to include the new subject-matter (taking into account any risks relating to intervening art).
Alternatively, it might be useful to write claims explicitly in two parts in the manner “subject-matter fully entitled to priority OR subject-matter not fully entitled to priority”, to try to persuade a Board that there are a “limited number of clearly defined alternative subject-matters” in the claim.
In view of the different approaches taken as regards partial priority, the question of whether a divisional or priority application will be permitted to form part of the state of the art is presently rather difficult to predict. Such objections appear to be arising during the examination procedure as well as in oppositions before the EPO, and have also been raised in post-grant UK revocation proceedings.
Perhaps it is not unreasonable to speculate that a referral to the Enlarged Board of Appeal on this topic might (and probably should) arise in the near future.