16 December 2016

CRISPR Patent Interference Hearing – How to Define an ‘Invention’ and Why It Matters

Gene editorOn 6 December 2016, and for the first time since the University of California (‘UC’) first sought to provoke interference proceedings over CRISPR/Cas9 gene editing technology in April 2015, the parties in dispute over just who has the rightful claim to inventorship appeared at an oral hearing before three judges of the US Patent and Trademark Office (USPTO) Patent Trial and Appeal Board (PTAB).  Remarkably, considering the complexity of the scientific, technical and legal issues involved, UC and opposing claimant the Broad Institute of MIT and Harvard (‘Broad’) each had just 20 minutes to present its arguments, plus an optional five minutes for rebuttal of the other side’s submissions.  According to a journalist present at the hearing, it was all over in 45 minutes total.

It is important to understand, however, that while the judges’ decision here (which is expected within about two-to-three months) could put an end to the interference proceedings at the PTAB, the hearing was not actually about the ultimate question of who was first to invent the CRISPR technology at issue.  It was more about what each party may, or may not, have actually ‘invented’, and whether there is even any interference to be decided.

For any readers coming across this story for the first time, you could go back and read all of my articles on the topic (and the many good sources of information linked therein).  However, the key points may be briefly summarised as follows:
  1. UC Berkeley Professor of Chemistry Jennifer Doudna and her collaborator Emmanuelle Charpentier were the lead researchers on a team that, in around 2012, developed CRISPR/Cas9 – a biotechnology system consisting of an engineered RNA and a ‘cutting’ protein that can be used for highly-targeted gene editing;
  2. a US patent application is pending (serial no. 13/842,859), which is based on the UC team’s work and a series of filings beginning in May 2012;
  3. the UC team’s original work related to the discovery of the CRISPR system in relatively simple bacterial cells (examples of ‘prokaryotes’, or simple cells without distinct nuclei), but in January 2013 they were successful in applying the technology to human cells (examples of ‘eukaryotes’, or more complex cells with, among other characteristics, distinct nuclei);
  4. roughly in parallel, a team led by Feng Zhang at the Broad reported similar success using CRISPR to edit human genes;
  5. a number of US patent applications were filed, beginning in December 2012, based upon the Broad team’s work (i.e. more than six months later than Doudna’s earliest filing);
  6. the Broad requested prioritised examination under the USPTO’s Track One program and, as a result, was awarded the first patent on the basic CRISPR technology – US Patent No. 8,697,359, ‘CRISPR-Cas systems and methods for altering expression of gene products’, issued on 15 April 2015.
The UC applications, and the earliest Broad applications, were filed under the ‘old’ US system, according to which the applicant entitled to receive a patent is the first one to invent the claimed subject matter.  This differs from the ‘first-to-file’ priority system that has long existed throughout the rest of the world, and in the US since 16 March 2013.  Under the first-to-invent system, the administrative procedure used to determine who should receive a patent when there are competing claims is called an ‘interference’.

Therefore, the UC parties suggested that the USPTO should institute interference proceedings, in order to determine who is really entitled to own the patent rights to the CRISPR/Cas9 technology, which it did on 11 January 2016.  However, the purpose of the recent hearing was not to determine who was the first inventor, but rather to decide exactly what is the invention at issue, and whether the interference should proceed any further at all.

Defining the Invention

It might seem odd that there is a dispute of what the invention actually is.  After all, the researchers on both sides are experts in the field and know exactly what they mean by CRISPR/Cas9 technology.  But that is not the way patent law works.  In order to secure a legal right over an invention, it is necessary to be quite explicit about just what it is that you claim to have invented.  For those less familiar with this concept, perhaps a simple hypothetical example may help.

Imagine that I invent a better mousetrap.  My specific design might employ some relatively complex arrangement of hinges, springs and latches, however I believe that my invention is, in fact, a more general mechanical principle of operation that this arrangement is designed to implement.  In common patent law parlance, I might refer to this principle as my invention, and to my specific design as an embodiment of my invention.

Now suppose that somebody else also develops a new mousetrap.  Her specific design is different from mine, and if you were to put the two products side-by-side you might say that they are plainly not the same.  You might even assume that this means that they are different inventions.  But then I might claim that, despite the difference in the two embodiments, both mousetraps are based upon the same underlying mechanical principle that I claim to have invented.

If I have obtained a patent on my mousetrap, and the issue is whether or not the other product infringes my patent, the question of whether or not it is legally the ‘same’ invention may be relatively straightforward.  My patent will include actual written claims, examined and approved by the patent office, defining the mechanical elements that are generally required to implement my principle of operation.  If the other product includes all of these elements, albeit that they may be differently embodied, then in the eyes of the patent law the mousetrap is the ‘same’ invention.

The patent interference situation is somewhat trickier, however.  Instead of a patent with approved claims in final form, and an allegedly infringing article, we have two (or more, but let’s keep it simple) patent specifications, containing somewhat different descriptions of different work by different inventors who may, or may not, have come up with the same invention at around the same time.  The claims in each case may be in a final form (as some of Broad’s are, since their earliest filings have issued as patents already), or they may be subject to further change and refinement (as UC’s are, because their application has yet to be approved).  Either way, however, the specific wording of the claims cannot entirely resolve who invented what, or when they invented it.  There may even be an invention in dispute that one of the parties had not previously claimed.

In this situation the question arises: exactly what ‘invention’ are the parties fighting over?  Is there just one invention that both may have a claim over, in which case the purpose of the interference proceeding is to decide who was the true first inventor?  Or are there two (or more) distinct inventions, such that each party might instead be entitled to their own independent patent(s)?

They Like to Count... (‘AH AH AH AH AH!’)

The solution to the invention identification problem under the US interference system is a concept known as a ‘count’.  One way to think of a count is that it is a type of notional patent claim.  It may be the same as, or similar to, actual claims appearing in the interfering patent specifications, or it may be somewhat different.  The point is, however, that if there is an interference, then there must be at least one count that defines distinct inventive subject matter that both parties would be entitled to claim as their invention, whether or not they have actually done so.

The corollary to this is that if it is not possible to devise a suitable count, then there is no interference, and both parties can just go their own way with their own separate inventions.

The PTAB has settled upon a count that it regards as suitable, which defines the invention in dispute as the application of CRISPR/Cas9 in eukaryotes (i.e. the cells of higher organisms) using any suitable targeting RNA.  UC disputes that this count is appropriate, and is pushing to substitute its own preferred definition which is not limited to eukaryotes (i.e. it covers both prokaryotes and eukaryotes) but which stipulates the use of single-molecule targeting RNA.

Remember, the UC team was first to get CRISPR/Cas9 gene editing to work.  They did it in prokaryotes (i.e. bacterial cells), and they did it using a single-molecule DNA-targeting RNA.  UC contends that once CRISPR/Cas9 had been demonstrated in prokaryotes, it was obvious (in the sense required by the patent law) that it could be made to work also in eukaryotes, and thus the subsequent success by Broad (and by UC, and by about half a dozen other teams) in doing so cannot constitute a distinct and separate invention.  (To be an ‘invention’, as required by the patent law, it is not enough for something merely to be new, it must also be nonobvious.)  On this basis, UC also contends that it is entitled to claim as its invention CRISPR/Cas9 gene editing technology generally (i.e. not limited to prokaryotes, just like I hypothetically claimed the broader mechanical principles underlying my mousetrap), using its method of employing single-molecule RNA.

Broad disagrees.  It says that getting CRISPR/Cas9 to work in eukaryotes is a separate and distinct invention, and was not obvious based upon the UC team’s work with prokaryotes.  Furthermore, the Broad team’s initial successful experiments employed two-molecule guide RNA.  Broad thus contends that what it is entitled to claim as its invention (and UC is not) is the more general use of guide RNA (whether or not a single molecule) to achieve CRISPR/Cas9 gene editing in eukaryotes.

Conclusion – It’s All About Obviousness

On the basis of the arguments made by UC and Broad, whether or not there is an interference to be decided comes down to a question of obviousness.  It perhaps does not matter so much how the count is defined (although if UC is correct, then its preferred count may be more appropriate) – what is important is whether or not the Broad team actually invented something when they got CRISPR/Cas9 to work with eukaryotes, or whether this was merely an uninventive and inevitable development once the UC team’s success in prokaryotes became known.

It has been reported that the PTAB judges seemed sceptical of UC’s submission that the extension to eukaryotes was obvious, and involved ‘no special sauce’.  However, it is not surprising that they would have particularly probed UC’s lawyers on this point, considering its centrality to the decision before them.  It is common that parties arguing that an alleged invention is obvious, in a range of contexts, are pressed harder for evidence supporting their case.  The fact is that following publication of the UC team’s results with prokaryotes, a number of other researchers succeeded in getting CRISPR/Cas9 to work with eukaryotic cells.  That it took a few months, and some trial and error, is not necessarily evidence of nonobviousness.  If the path to success is evident, such that there is a ‘reasonable expectation’ that it will be achieved, then neither the length of that journey, nor the number of potholes encountered, will necessarily negate the obviousness of the route.

Nonetheless, if Broad’s argument prevails, then there is no interference.  This would mean that each party is entitled to a patent on its own invention – UC for the use of single-molecule guide RNA with prokaryotes, and Broad for the use of any guide RNA with eukaryotes.  Commercially, this would be a win for Broad, because it is very clear that editing of the genes of higher organisms, including humans, is where the money lies.

On the other hand, if the UC argument prevails at this stage, then it would still remain to be determined who has priority as the first inventor.  That might place UC back in the box seat, but would still leave the interference proceedings very far from a conclusion.

The other thing to keep in mind is that decisions of the PTAB are appealable to the US Court of Appeals for the Federal Circuit.  Given what is at stake in the CRISPR priority dispute, and barring any settlement between the parties, it is highly unlikely that any PTAB ruling will be the last word on the matter.


Post a Comment

Copyright © 2014
Creative Commons License
The Patentology Blog by Dr Mark A Summerfield is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Australia License.