About the Authors

  • The Authors and Contributors of "Patent Docs" are patent attorneys and agents, many of whom hold doctorates in a diverse array of disciplines.
2018 Juristant Badge - MBHB_165
Juristat #4 Overall Rank

E-mail Newsletter

  • Enter your e-mail address below to receive the "Patent Docs" e-mail newsletter.

Enter your email address:

Delivered by FeedBurner

Contact the Docs

Docs on Twitter


  • "Patent Docs" does not contain any legal advice whatsoever. This weblog is for informational purposes only, and its publication does not create an attorney-client relationship. In addition, nothing on "Patent Docs" constitutes a solicitation for business. This weblog is intended primarily for other attorneys. Moreover, "Patent Docs" is the personal weblog of the Authors; it is not edited by the Authors' employers or clients and, as such, no part of this weblog may be so attributed. All posts on "Patent Docs" should be double-checked for their accuracy and current applicability.
Juristat #8 Overall Rank


« PatentBuddy.com Examines Impact of Economic Recession on Patent Practice | Main | Manufacturing Alliance on Patent Policy: Apportionment of Damages Provision Will Have Adverse Effects »

January 14, 2009


*****Judge Rader then asserted that "Valiente had already produced p38," to which Ms. Rudolph reminded the Court that had been done for the protein and not for the cDNA. Specifically, she correctly asserted that "you need a starting cell library that will have a sufficient amount of that particular gene that will produce that particular protein." ****

This is all dust kicked up by Kubin. No serious biotech entity in the 21st century faces serious difficulties identifying a gene sequence for an isolated protein.

"GONGOLA: ... the Board has not identified any basis for identifying a specific binding region -- what the structure of that binding region is --"

That's inherent to the sequencing of the gene.

"...and it hasn't identified any basis for finding CD48-binding obvious."

Meaningless Rochester-style functional claiming that should be ignored. If CD48 binding was some awesome unexpected result (a la the alleged awesomeness of the isolated enantiomer in Sanofi v. Apotex) then it would matter. But not here. And particularly not when the claim covers a billion different sequences, none of which were investigated by Kubin.

What Kubin really needed to do was what Sanofi did with their entantiomer arguments: make a lot of noise about how terribly difficult it was to do ("millions of dollars, thousands of hours") and how nobody really wanted to do it until some poor guy stuck his neck sooooooooo far out to take it upon himself to do the oh-so-unlikely-to-succeed work of purifying the entantiomer according to methods that nobody ever thought would work because they had only been invented 125 years ago.

Dear Prediction:

You have hit on the problem we cited in our discussion of the oral argument: the assumption of something that isn't a fact. Valiante (USP 5,688,690) did not purify p38, and did not teach any way to do so. Valiante produced monoclonal antibodies the old-fashioned way - by immunizing an animal with human NK cells, producing hybridomas and then testing each hybridoma for NK cell-specific cross-reactivity. He chose the antibody against p38 because it was the only one that consistently reacted with human NK cells, and then probes a Western blot of NK cell proteins to identify the putative molecular weight of the protein (hence, "p38").

Before you argue that it would have been trivial to isolate the protein once you have an antibody (may be, or maybe not, depending on the amount of p38 produced by the cells, whether this cell-surface protein could be solubilized, and if so whether a reliable N-terminal sequence (which, of course, would lack the leader sequence) could be obtained), first let's just acknowledge that Valiante did NOT possess an isolated p38 protein in the art. As Judge Rader said, Valiante provided a probe, nothing more.

Thanks for the comment.

Kevin, thanks for the post. How disheartening for us prep/proc guys - even if you do everything right (hard enough in and of itself), it still may not matter b/c the judges who ultimately decide don't understand the technology. And I say this as someone who thinks that Judge Rader usually gets things right.

Dear Dan:

Barbara Rudolph pointed out what was missing in the prior art, and if the judges were listening (or if the review the argument, or if they have dedicated clerks), then there is a chance they will realize why the PTO Emperor has no clothes.

Alternatively, there is a possibility that the coyness of Janet Gongola's presentation with regard to the status of Deuel in view of KSR might lead to a very narrow holding even if the court affirms. While this will not be the right result for Amgen, the rest of us might be able to live with a decision that a cDNA sequence is obvious if the prior art provides: an isolated and thoroughly characterized protein; a probe antibody that is immunospecific for the protein; a cell or tissue source unambiguously taught in the art to express the protein; and a prophetic method that the art establishes will predictably produce the cDNA when followed exactly.

Thanks for the comment.

Thanks for the helpful analysis.

Bear in mind the analytical link between Bell/Deuel (obviousness) and Lilly (written description), which seem to create a parrellel universe of sorts for nucleic acid claims. Judge Rader's concerns with Lilly and its progeny are well-documented. His questioning in this case may reflect that he sees restrictions on Bell/Deuel (even subtle ones) as a necessary step in the narrowing/elimination of Lilly. In that light, one wonders were the "industry" really wants this to come out.


"You have hit on the problem we cited in our discussion of the oral argument: the assumption of something that isn't a fact. Valiante (USP 5,688,690) did not purify p38"

The issue is whether enough protein was identified to move forward with the mundane steps of cloning the gene. I'm still bothered by the haze around this issue, though. You yourself wrote in the original post: ******Judge Rader then asserted that "Valiente had already produced p38," to which Ms. Rudolph reminded the Court that had been done for the protein and not for the cDNA.*****

Do you see the problem? Judge Rader asserted that Valiente **had already produced the p38**

One would expect Ms. Rudolph to reply, "No, your honor, that is a falsehood that the PTO keeps repeating. It is not true. I can not emphasize this enough. That is NOT true. Valiente had NOT produced p38."

But instead (according to your summary), Ms. Rudolph admits that "it had been done for the protein ..."

So I may ask to be pardoned for misunderstanding some of the facts in this case. But Kubin and his reps seem to be trying to keep aspects of the record less than crystal clear. There's often a good reason to do that but it can lead to lengthy disputes ... and comment threads. ;)

Dear Prediction:

You raise a good point, although I'm not sure I think Kubin deserves the blame for covering anything up here. I hesitate to criticize counsel at oral argument, since they have to juggle what they want to say/emphasize and at the same time respond to the judges's questions. In this case, Judge Rader was coming at Kubin's counsel pretty hard for about 10 minutes with just a brief letup (so she could answer Judge Friedman's question), and I think it's understandable that you and I might be able to come up with a more succinct response in the comfort of our computer consoles.

I presume that this point was argued in the briefs. What I put in my comment and the post was based on my review of the Valiante reference.

There is an open question, however, about whether the combination of expression cloning methods + a specific antibody may make unnecessary the traditional protein purification/sequencing/degenerate probe screening methods of the past (while recognizing that the existence of databases of human genome sequence makes the older ways "easier" insofar as clear mistakes in protein sequencing should be more readily detected using in silico methods). But even if cloning has gotten more routine, in this case there was no known source of NAIL cDNA - based on the teaching away from human NK cells in Matthew (even in view of the failure of the art to recognize that 2B4 = NAIL), and the fact that if you did expression cloning with human NK cells just like Valiante said to do and the PTO maintains Matthew did with mouse cells, you would not be able to detect human NAIL cDNA-harboring clones because they were not made in "resting" human NK cells - and Matthew, Valiante and the rest of the art did not tell you that you needed to activate human NK cells (indeed, one of the characteristics of NK cells is that they are not MHC-restricted and thus are not "activated" in the same way as T cells). Despite Judge Rader's statements about how equivalent mice and humans are, they are not identical and there are many instances of surprising differences betwee the species.

Thanks for the comments.

The comments to this entry are closed.

August 2020

Sun Mon Tue Wed Thu Fri Sat
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31