CAFC Applies Lilly to Invalidate Carnegie Mellon’s Plasmid Claims

Carnegie Mellon University v. Hoffman-La Roche (Fed. Cir. 2008)

Carnegie Mellon's patents cover recombinant plasmids used to enhance expression of an DNA polymerase. On appeal, the CAFC affirmed the lower court holding that the patent claims fail to meet the written description requirement under Lilly.

35 U.S.C. §112 requires that the patent document "contain a written description of the invention and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.

The written description requirement is used to make sure that the patentee only claims what has been invented. The public gets a "meaningful disclosure in exchange for being excluded from practicing the invention for a limited period of time." Quoting Enzo Biochem (Fed. Cir. 2002).

A common written description argument is that the claims have not been disclosed to their full scope. In the Lilly case, for instance, the CAFC found that a generic claim directed to "any vertebrate and mammalian cDNA" were not supported by a specification that only discussed one species – rat cDNA.

To be clear, written description is not about enablement. Patent claims may well be enabled based on PHOSITA's knowledge of the art, but still fail the written description requirement because the patentee did not disclose the entire scope of its invention.

A representative claim from Carnegie's patents reads as follows:

1. A recombinant plasmid containing a cloned complete structural gene coding region isolated from a bacterial source for the expression of DNA polymerase I, under operable control of a conditionally controllable foreign promoter functionally linked to said structural gene coding region, said foreign promoter being functional to express said DNA polymerase I in a suitable bacterial or yeast host system.

In reviewing the claim, the CAFC noted that the DNA coding sequence is "broadly defined … only by its function of encoding DNA polymerase I" and that the claims are not limited to any particular bacterial or yeast species.

The specification only discloses one operative gene – the E. coli polA gene. And at the time of the patent, only three polA genes had been cloned (out of thousands of bacteria strains). "[W]ith regard to the promoter, the patents fail to disclose the nucleotide sequence or other descriptive features for a polA gene (including the promoter sequence) from any bacterial source other than E. coli."

"To satisfy the written description requirement in the case of a chemical or biotechnological genus, more than a statement of the genus is normally required. One must show that one has possession, as described in the application, of sufficient species to show that he or she invented and disclosed the totality of the genus. . . . [W]e conclude that that requirement was not met here.

Gentry Gallery: The district court also applied Gentry Gallery to invalidate other claims. In Gentry, the CAFC found a patent claim invalid because the claim failed to recite an "essential element" of the invention. Here, Roche argued that the patents were directed to avoiding the problem of lethality to host cells, but that the claims did not include that limitation. As it has done repeatedly, the CAFC rejected the idea that Gentry Gallery created an essential element test. (The claims remain invalid under the Lilly analysis.).

Comments

[SD]

I printed out this case and got as far as reading that it is authored by Judge Lourie -- then tossed it into the pile of other WD/biotech cases that he wrote. We all know what it is going to say. Lourie has never met a biotech claim that he couldn't invalidate on WD grounds as being too broad -- continuing, as Judge Rader has complained, to perpetuation the conflation of WD and enablement.

My only surprise is that the Fed. Cir. continues to claim, with a straight face, that cases are randomly assigned. Is it truly a coincidence that Lourie always seems to get these cases involving allegedly over-broad biotech claims that don't meet his definition of the WD requirement?

[random person that posts here often]

It's a small world after all world SD.

Also, I wonder why they haven't expanded these requirements to other arts as well. Chem/bio aren't the only ones that may not be sufficienty disclosed to issue overbroad claims.

[somebody]

"Also, I wonder why they haven't expanded these requirements to other arts as well. Chem/bio aren't the only ones that may not be sufficienty disclosed to issue overbroad claims."

If the claims are "overbroad" why are they being issued? With super-vigilant examiners like you, 6k, isn't the problem of "overbroad" claims already on its way to being solved? I mean, I'm sure you've written tons and tons of rock solid WD rejections. Air tight cases, no doubt. Cases abandoned in droves in the face of your irrefutable logic.

LOL

[SF]

"Also, I wonder why they haven't expanded these requirements to other arts as well. Chem/bio aren't the only ones that may not be sufficienty disclosed to issue overbroad claims."

They actually have applied similar enablement requirements in both software and mechanical cases:

http://www.patentlyo.com/patent/2008/02/enablement-cont.html (Sitrick)

http://www.patentlyo.com/patent/2007/03/the_new_law_of_.html (Liebel-Flarsheim)

[TJ]

SD, the cases are randomly assigned for paneling. Judge Lourie does have a disproportionate number of patent cases assigned to him for opinion-writing, because (a) he is very senior and has the power to assign opinions to himself, and (b) he is one of the most experienced judges in patent law on the court, even relative to the other judges. Judge Newman writes many patent law opinions for the same reason.

[random person that posts here often]

Thanks for the cites SF, now I recall having read those same posts when they were posted. I'm still not quite sure how this can be implemented in the corps that well. I mean, the claims are hardly ever going to be fully enabled because they leave out buckets of things that are not necessarily "known" to a person of ordinary skill (such as the jacket/jacketless from the case you cited). Frankly I don't see how this is practical to have examiners making rejections like this, if we're going to make them fully enable the thing every time and put in the claim only that which is enabled then a huge amount of claims that I've ever read would be invalid and many of the applications would straight up have nothing that they would be able to claim.

I'd have to look into the cases a bit more but something just seems fishy here.

As for me somebody I don't believe that any 112 1sts that I have hypothetically written have come back properly argued and overcome. 112 1st is usually reserved for new matter and only very glaring cases of the claim being not enabled. A spaceship that can travel faster than light anyone?

[Leopold Bloom]

"A spaceship that can travel faster than light anyone?"

I was planning to respond to today's post last week, but my spaceship is in the shop.

[Freddie Mac]

Anyone who saw the BCP presentation today, and any admin law experts--how many times can an administrative agency change its "interpretation" of the statute and not be considered arbitrary and capricious, where no change in case law is driving the "reinterpretation"? I think this is not an issue that has been considered, but it came to mind in watching Mr. Weber's discussion of how written description and enablement are now being evaluated for sequence-based claims. I was grateful for the very articulate and thoughtful comments from the audience and am myself stymied as to why the previous apparent policy of the PTO for deeming allowable claims to say 90% sequence identity with specified function (and specified assay for said function in the specification) has now been changed to a "case by case basis" which, with all due respect to Mr. Weber, I am convinced is in practice as illusory and impossible to meet as the RCE standard under the limbo-ed rules would be. While I believe that Mr. Weber means well, I totally disagree with the change in the examination practice, and I believe that in the absence of a Federal Circuit case changing the application of 112, 1st to such claims, it could possibly be successfully argued that the PTO is doing something improper (either arbitrary and capricious, or substantive without proper notice and comment, or whatever the alternative might be for the PTO under admin law). This is why I love Jean Witz so much--even if I disagree with her, she always seems to go through the case law and statutes and regulations and actually *interpret* them, as opposed to going out on a limb to pursue some personal agenda as it sometimes seems with others (there being no apparent basis in case law, etc. for doing so). I really hope somebody takes this up, or that the Alonso case is decided in such a way as to make sense, because otherwise, as one of the excellent comments today noted, sequence-based specifications are just an "invitation to design around." Just my two cents. And thank you, Jean Witz; I hope you get to be the next Commissioner, if you want to be.

[Kevin E. Noonan]

Dear Freddie Mac:

Keep an eye on the Kubin case yet to be argued before the CAFC. The types of WD issues in the recent PTO guidelines changes are involved, and the Court may take the opportunity to straighten this out (although whether or not it is for the better may depend on the panel).

[Tony B.]

Does anyone know whether the Federal Circuit has held a claim invalid for obviousness-type double patenting over a commonly-owned patent with the same expiration date? I've been looking at the cases and so far they all involve different expiration dates (so that the invalidity prevents de facto extension of the patent term), but I'm wondering in view of the current PTO emphasis on terminal disclaimers whether there's case law (or dicta) to back that up.

[Malcolm Mooney]

"why the previous apparent policy of the PTO for deeming allowable claims to say 90% sequence identity with specified function (and specified assay for said function in the specification) has now been changed to a "case by case basis" "

The reason is that the previous policy was arbitrary, not grounded in any case law whatsoever (i.e, no case held that there is some "per se" percent identity that applicants should be entitled to) and therefore wrong.

Deal with it.

This Carnegie-Mellon case is a joke, by the way. Right result, ridiculous "reasoning" ... but that's usually the case any time that "written description" is used to tank a patent.

[Ed]

>>>>This Carnegie-Mellon case is a joke, by the way. Right >>>>result, ridiculous "reasoning" ...

Malcom,
How would you have done it?

[Malcolm Mooney]

"How would you have done it?"

I'd squeeze into non-existence between enablement and obviousness. For cripes sake, CMU was going after Taq polymerase with those claims. Did you catch the bizarre footnote at the end of the case by the way, regarding the Fed Cir's "understanding" of why CMU was flailing for so hard and so long? Truly sad.

[zebra]

I'm curious about the following "what could have been" scenario:
Suppose CMU filed an initial app with the genus claims, and further suppose that the PTO gave a written description rejection for the genus claims from the get go, but held the species claims as allowable. CMU reckons that the species is nice, but not truly valuable because there are potentially many species that provide design-arounds. Can CMU rush back to the lab and work up a second species, then file a CIP and go for the genus claims? If that is rejected, file a third species, and so forth. Could CMU eventually bootstrap their way to a valid genus claim?

[John prosecutor]

a year after the initial application is published it is prior art against the CIP's, right

[Malcolm Mooney]

"Could CMU eventually bootstrap their way to a valid genus claim?"

Depends on the genus.