Twists and turns in CRISPR battles at the EPO

In 2012, the world was introduced to the concept of using CRISPR systems for genome editing.  Since then, the technique has revolutionised biotechnology, becoming widely utilised and allowing the ability to precisely target and edit specific sequences in genomes.  Due to its popularity, simplicity, and promise, there is significant interest in understanding who owns the patent rights for this technology.  However, the patent situation is complex, because different parties filed competing patent applications in 2012, and many users of the technology have been awaiting the outcome of ongoing inter-party actions at patent offices to understand to whom licence fees should be paid. It seemed that some clarity was appearing at the EPO, but there have been a series of surprising decisions in 2024 that have extended the period of uncertainty.

Bacterial defence to gene editing

CRISPR systems evolved to provide microbes, including bacteria and archae, with a defense mechanism against invading bacteriophages.  Before 2012, biotechnological interest in CRISPR systems was mainly due to its use in protecting yoghurt and other dairy cultures.  However, in 2012 numerous parties filed patent applications directed at the use of modified CRISPR systems (initially using the Cas9 version of the system) to specifically target and/or edit non-bacteriophage DNA, particularly genomic DNA. This flurry of filings has led to contentious proceedings in numerous jurisdictions.  These proceedings should establish which parties own the earliest, and broadest, patent rights, and the following will focus on the developments this year in Europe.

Nobel prize winners revoke their own patents

The University of California, the University of Vienna, and Emmanuelle Charpentier (“CVC”) were early filers, as co-applicants, in 2012. The 2020 Nobel prize in Chemistry was awarded to inventors Jennifer Doudna and Emmanuelle Charpentier of CVC for their development of CRISPR-Cas9, recognising their significant contribution. Their patent position in Europe has, therefore, seemed quite strong.  Their patent EP3241902, directed at the use of CRISPR-Cas9 for targeting heterologous activity to a specific sequence, was revoked due to insufficient teaching in the earliest application about how to generate a Cas9 with reduced nuclease activity. But their patents EP2800811 and EP3401400, directed at modifying DNA by cleaving it, were maintained by the opposition division with only minimal amendments and seemed to be in quite a strong position. However, appeals were filed against the decisions of the opposition division, and in July this year the board of appeal issued preliminary opinions that the earliest date for the patents is invalid.

In order to protect unintended cleavage of a bacteria’s own DNA, it is now known that there is a “PAM” sequence that must be present in the target (bacteriophage) DNA, in order for cleavage to occur.  Because the earliest CVC application did not disclose this PAM sequence, and because the skilled person at that time did not otherwise know they had to take account of the PAM sequence, the board’s preliminary opinion was that the earliest application did not provide enough information to enable the claimed inventions. Since the board was of the opinion that CVC could not rely on their earliest application date, the invention was considered to lack novelty over the inventors’ own article published in Science soon after.

In response to the preliminary opinions, CVC themselves revoked EP2800811 and EP3401400 in September by withdrawing their approval of the granted text, albeit at the same time they filed extensive arguments as to why they thought the board’s opinion was incorrect. Whilst such an action might seem surprising, by revoking their own patents before a final decision could be taken by the board, CVC have avoided receiving any negative final decision that could be detrimental to their ongoing prosecution and defence of other members of their patent family.

EPO U-turn reinvigorates patent portfolio

The Broad Institute, Harvard College and the Massachusetts Institute of Technology (“Broad”) are often seen as the main group, as co-applicants, on the other side to CVC. They filed their first CRISPR application comparatively late in 2012, on 12 December. However, Broad followed an accelerated patent strategy in Europe, so that they were the first to have their patents granted and opposed.

Broad’s patent EP2771468 was revoked in 2020 after opposition and appeal.  The board decided (T 844/18) that the priority claims to the earliest filing dates were invalid, because the rights in the earliest application were not all properly transferred to the applicants of the application from which the patent was derived.  Because Broad could not rely on their earliest application dates, it was decided the claimed subject-matter lacked novelty over intervening publications.

The board’s decision seemed to mean that Broad would not be able to get broad protection for CRISPR genome editing in Europe. This position seemed to be finally decided.

However, in October last year the enlarged board of appeal at the EPO issued decision G1/22 (consolidated with G2/22).  Our report of this decision can be found here. Essentially, though, with this decision the board changed how entitlement to priority should be assessed at the EPO, stating that there is a “rebuttable presumption” that a priority claim is valid, raising the hurdle that must be cleared by a third party in order to challenge entitlement to priority.  In paragraph 128 of the reasons for decision G1/22, the enlarged board referenced T 844/18, i.e. the decision to revoke EP2771468, suggesting that even in the circumstances underlying T 844/18, there might still be entitlement to priority.

The board of appeal followed G1/22 when they issued a decision for EP2784162, a divisional patent of EP2771468. The opposition division had previously issued a decision revoking EP2784162 in 2019, for reasons consistent with those the board gave for the parent case in T 844/18 .  Earlier this year, though, the board decided (T2360/19) that the opponents for EP2784162 had not rebutted the presumption of priority, so that the entitlement to priority is valid.  EP2784162 has been returned to the opposition division for further prosecution, now that the priority entitlement is to be considered valid.

Thus, the decision in EP2784162 is the complete opposite of that in EP2771468, despite the underlying facts of the cases being the same, and Broad can once again seek to defend broad protection for CRISPR-Cas9 methods.  Of course, it will need to be decided whether the broader protection they seek to defend is novel and inventive, particularly since even their earliest priority claims are quite late in 2012.

Looking to the future

There is less than ten years left of potential patent protection in Europe for the earliest CRISPR-Cas9 patents. It might have been hoped, and expected, that we would have clearer understanding by now of the ownership and validity of the broadest patent rights.  However, the ongoing uncertainty continues at the EPO; all of the parties who filed CRISPR applications in 2012 have had difficulties at the EPO, but all parties still have applications pending, and many final decisions have been avoided or delayed.

For those seeking to license and use this exciting technology, for example to bring new treatments to the clinic, the ongoing uncertainty means it is difficult to plan for the licenses they might require. Hopefully, the EPO’s efforts in recent years to increase the speed of opposition and appeal proceedings will mean that a clearer picture emerges soon.

In a former role, Emma wrote the arguments for Opponent 4 that were endorsed by the board in their preliminary opinion for EP2800811.