By Kevin E. Noonan --
Judge Giles Sutherland Rich famously said that, in patent law, "the name of the game is the claim." One of the weaknesses in the "gene patenting" debate, as well as in the District Court's opinion in AMP v. USPTO (the Myriad case), is the lack of clarity about the scope and meaning of the claims of the challenged patents. The District Court did not bother to hold a Markman hearing to construe the claims, and thus neither that Court nor the Federal Circuit in reviewing Judge Sweet's decision brought any clarity to what the parties were arguing about, instead drawing generally inapt analogies to what other claims in other patents might claim. The type and number of claims at issue are sufficiently similar, having generally the same structure and using the same language that claim construction might help, and the language used is certainly no more arcane than in any other patent case.
Construing claim language to determine the scope and meaning of the claims is a matter of law for the district court in the first instance (Markman v. Westview Instruments, Inc., 52 F.3d 1321 (Fed. Cir. 1995) (en banc), aff'd 517 U.S. 370 (1996)). The Federal Circuit reviews claim construction de novo, giving no deference to either the district court's ultimate construction of claim terms or any factual determinations underlying the court's legal determination of claim scope (Cybor Corp. v. FAS Techs., Inc., 138 F.3d 1448 (Fed. Cir. 1998) (en banc)). The way courts are to perform claim construction was set forth (most recently) by the en banc Federal Circuit in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en banc). This procedure can be summarized in a few simple rules. First, claim terms are to be given their "ordinary and customary" meaning as understood by those with skill in the art. Id. at 1313. Next, the specification is consulted to determine whether the patent applicant has acted as his or her own lexicographer, a capacity recognized as black letter patent law (see Autogiro Co. Amer. v. U.S., 384 F.2d 391 (Ct. Claims 1967). This ability to define terms is limited only insofar as the applicant attempts to define a commonly accepted term in a way that is "repugnant" to its meaning in the art. In re Hill, 161 F.2d 367 (CCPA 1947). In the event that there is any ambiguity about the "ordinary and customary" meaning of a claim term, resort to the specification is preferred under Phillips as "the single best guide to the meaning of a disputed term." Id. at 1315 (citing Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed. Cir. 1996). Finally, the prosecution file history should be consulted to ensure that a patentee cannot advance a narrow construction of a term during patent prosecution, only to argue a broader construction in an infringement action. Phillips at 1317.
With these concepts and principles in mind, generally two types of claims were at issue in the Myriad case:
1. Isolated DNA claims (as well as claims to oligonucleotides); and
2. Diagnostic method claims.
(The one other claim at issue, claim 20 of U.S. Patent No. 5,474,282, is in some ways sui generis to the other claims, being directed to methods for screening chemical compounds for anti-cancer properties.) Claim 1 of U.S. Patent No. 5,837,492 is representative of the first type of claim:
An isolated DNA molecule coding for a BRCA2 polypeptide, said DNA molecule comprising a nucleic acid sequence encoding the amino acid sequence set forth in SEQ ID NO:2.
(Claims 1 and 7 of the '282 patent; claims 6 and 7 of the '492 patent; and claim 1 of U.S. Patent No. 5,693,473 are variants of this general claim type.) The italicized words and phrases in the exemplary claim need construing, the other terms being expressly defined with regard to other terms in the claims (e.g., the amino acid sequence of the BRCA2 polypeptide is defined as SEQ ID NO: 2; as a consequence, any ambiguity in the meaning of this term devolves to any ambiguity in the nature of the expressly recited sequence).
Of the italicized terms, the parties and the Federal Circuit focused most attention on the term "isolated." This term has a meaning accepted by the Patent Office (and in the biotechnology arts) with regard to nucleic acids, wherein the term means "isolated from its natural state and processed through purifying steps that separate the gene from other molecules naturally associated with it." Utility Examination Guidelines, 66 Fed. Reg. 1094 (2001). An associated definition is found in the specification of the patents in suit:
"Isolated" or "substantially pure". An "isolated" or "substantially pure" nucleic acid (e.g., an RNA, DNA or a mixed polymer) is one which is substantially separated from other cellular components which naturally accompany a native human sequence or protein, e.g., ribosomes, polymerases, many other human genome sequences and proteins. The term embraces a nucleic acid sequence or protein which has been removed from its naturally occurring environment, and includes recombinant or cloned DNA isolates and chemically synthesized analogs or analogs biologically synthesized by heterologous systems.
Some in the gene patenting debate have characterized this meaning as encompassing isolation of total genomic DNA as falling within the scope of the claim; indeed this interpretation is one of the bases for the argument that Myriad's claims may inhibit "whole genome sequencing" methods, as Chris Hansen of the ACLU argued during oral argument in Myriad (see "AMP v. USPTO: Oral Argument at the Federal Circuit"). This interpretation is only persuasive, however, if the phrase "many other human genome sequences" is eliminated from the list of cellular components from which the BRCA-encoding DNA is separated. Thus, the ordinary and customary meaning of the term "isolated" is reinforced by the express meaning of the term as recited in the specification. And as interpreted using this meaning, the distinctions drawn in the majority and concurring opinions in Myriad, between what is isolated and what is "merely" purified, become less relevant, since DNA is understood both to be "purified" from other cellular components and "isolated" from other human genome sequences.
In addition, since claims must be construed as a whole, giving each limitation meaning and avoiding surplusage, what is "isolated" is a nucleic acid encoding an amino acid sequence "set forth" in the specification as SEQ ID NO: 2. Thus, properly interpreted, the claim is limited to DNA molecules that encode the BRCA 2 polypeptide. The only other claim term that could expand the scope of these claims to extend to nucleic acids other than those encoding the BRCA proteins (and only those having the specifically recited amino acid sequences) is the word "comprising." However, using another well-recognized rubric of claim construction, this word cannot be used to expand the scope of the claim to extend to human genomic DNA per se: first, because "claims should be so construed, if possible, as to sustain their validity," Rhine v. Casio, Inc., 183 F.3d 1342, 1345 (Fed Cir. 1999); see also Klein v. Russell, 86 U.S. (19 Wall.) 433, 466, 22 L. Ed. 116 (1873), but also because claims are interpreted in light of the specification, United States v. Adams, 383 U.S. 39, 49, 15 L. Ed. 2d 572, 86 S. Ct. 708, 174 Ct. Cl. 1293 (1966). The specification of the Myriad patents claiming DNA encoding BRCA1 and BRCA2 disclose the patentees' efforts to identify, isolate, and purify BRCA-encoding DNA. Thus, the proper interpretation of these claims is directed to DNA (and in particular, cDNA) encoding the BRCA proteins.
Turning to the diagnostic method claims, claim 1 of U.S. Patent No. 5,753,441 is representative:
A method for screening germline of a human subject for an alteration of a BRCA1 gene which comprises comparing germline sequence of a BRCA1 gene or BRCA1 RNA from a tissue sample from said subject or a sequence of BRCA1 cDNA made from mRNA from said sample with germline sequences of wild-type BRCA1 gene, wild-type BRCA1 RNA or wild-type BRCA1 cDNA, wherein a difference in the sequence of the BRCA1 gene, BRCA1 RNA or BRCA1 cDNA of the subject from wild-type indicates an alteration in the BRCA1 gene in said subject.
(Claims 1 of U.S. Patent No. 5,709,999; claim 1 of U.S. Patent No. 5,710,001; and claims 1 and 2 of U.S. Patent No. 6,033,857 are variants of this general claim type.) Here, the claim term "comparing" has no specific definition in the specification; however, Example 9, entitled "Analysis of BRCA 1 Mutations" sets forth the general method for extracting DNA from blood or tumor samples from patients with breast or ovarian cancer to detect mutations. The relevant claim term to be construed is thus from a tissue sample from said subject, and whether this implies a necessary transformation step to ascertain the individual patient's BRCA gene sequence. Both the District Court and the Federal Circuit treated this claim as if it had been construed to encompass mere inspection of and comparison between two sequences (i.e., printouts of the germline and patient sample sequences) that did not require a patient sample to be assayed to obtain the sequence. While not a necessarily unreasonable interpretation (since these claims do not expressly recite limitations drawn to obtaining a sample and determining the BRCA gene sequence therefrom), it is not the only interpretation. What is missing from any judicial analysis of the diagnostic method claims at issue is whether the skilled worker would understand the claim not to require these steps, and whether the specification mandates interpretation of the claim to include these steps. (The prosecution file history may also be informative as to what the patentee understood the invention to be and, more importantly, represented that understanding to the patent examiner, with regard to the scope of these claims.)
The danger in ignoring and failing to perform the fundamental and threshold exercise of construing the claims at issue in patent litigation is clearly illustrated in the briefs (both from the parties and the amici) and the decisions of both the District Court and the Federal Circuit. The Courts' analyses are understandably adrift without the anchor of having decided what the claims mean, turning them into a "nose of wax" to be susceptible to whatever argument is being advanced. Although it is perhaps too late to do so now, if the Federal Circuit grants rehearing or en banc review, it would be well if the Court mandated that the parties provide the equivalent of their Markman briefs, to enable the Court as a threshold matter to determine what the claims mean as a matter of law. Without that, it is hard to see how the Federal Circuit, or the Supreme Court if it grants certiorari, will have any reasoned basis for coming to a decision as to whether these claims recite patent-eligible subject matter. That outcome may be good for arguments based on emotion and rhetoric, but it won't be good for deciding this fundamental question of patent law.
Kevin,
Your article is absolutely on point and correct. There are unfortunately many who derisively smirk that patent-eligibiltiy (or any other issue of patentability) shouldn't depend on how the claim is written. But that's completely wrong: it all depends (including patent-eligibility) on how the claim is written.
Also, you note correctly the fundamental "take away" from the Phillips case: claim terms should be construed according to their "ordinary and customary meaning" unless the specification clearly indicates that the patent applicant intends those terms to have a different meaning (the "applicant can be their own lexicographer" collary to the Phillips rule). The ACLU and others may like a "warmer" and "fuzzier" view on claim construction, but it is completely wrong (and definitely legally flawed), including when determining questions of patent-eligibility under 35 USC 101.
Posted by: EG | August 24, 2011 at 06:51 AM
The problem, EG, is that some CAFC judges now seem to think that the issue (for deciding patentable subject matter questions) is the nature of the "underlying invention" (see Cybersource). When even the Federal Circuit starts eschewing patent claim language to reach policy-driven decisions, it is hard to fault non-patent lawyers for making semantic arguments that favor their position.
Thanks for the comment.
Posted by: Kevin E. Noonan | August 24, 2011 at 07:35 AM
Kevin,
Very good stuff. Regarding the diagnostic method claims in AMP v. USPTO, the CAFC was correct in the result, but incorrect in the rationale. Construing the claim to require the physical steps of obtaining the sample, extracting genomic DNA, blah, blah, blah, shouldn't happen, and even if it did, it shouldn't affect the outcome. That's because the "invention" is the correlation, no matter how dressed up as a process. You know this. The Supremes will fix this all in Prometheus, and then we can move on with our lives.
Posted by: Gary Johnston | August 24, 2011 at 08:26 AM
It's a DEAD issue. The genome has been sequenced. As is always the case, the invention must be new and non-obvious. A sequence of amino acids or nucleotides will have to be novel, not merely a segment of native DNA (which was previously novel, but no longer is because it would be obvious based on the genomic DNA). Now the invention, the sequence of DNA or amino acids must be an entirely new chemical entity (novel sequence) or used in a function (tied to a novel method/therapy).
This argument is WAAAY late, not relevant and using a 2011 standard on 1995 techniques.
Posted by: Mike | August 24, 2011 at 08:53 AM
"some CAFC judges now seem to think that the issue (for deciding patentable subject matter questions) is the nature of the "underlying invention" (see Cybersource).
Kevin,
Amen to that. It also makes the case law precedent chaotic and almost impossible to apply objectively to future situations/cases.
One point of clarification on your comment about "it is hard to fault non-patent lawyers for making semantic arguments that favor their position." I frankly wish the "non-patent lawyers" would be honest and acknowledge their "technological ignorance" instead of "reveling" in it and being arrogant about that ignorance. Our Supreme Court is "Exhibit No. 1" of this unfortunate attitude.
Posted by: EG | August 24, 2011 at 08:54 AM
Dear Mike:
Agreed, but the mischief isn't over. This is a debate that should have happened 20-30 years ago, and the fact that most "gene patents" will die a natural death (20 years after filing) on or before about 2020 makes the issue much less important now than it would have been before the HGP.
Of course, it is historical revisionism to forget that one of the reasons we have a HGP is that Craig Venter energized the public sector by funding a private effort to sequence (and patent) human genes. There have even been statements by staunch gene patenting opponents that sequencing and publishing the sequences were strategies adopted to attempt to thwart gene patenting. So the effects of gene patenting certainly "promoted progress" in at least that way.
The risk now is having a court decision that adopts the rationale espoused by the government's amicus brief - which could render many biologic drugs patent-ineligible, or at a minimum disincentive commercialization. Judge Moore gets that - she asked the acting SG about the effects of his position on antibodies at oral argument. And in view of the tendency for black-and-white thinking when it comes to patent eligibility, there is a chance that we can get a decision that is meaningless for genes but devastating for the industry. After all, Myriad's patents expire in 3-5 years; does anyone really think they don't have a corporate strategy in place to deal with that?
Thanks for the comment.
Posted by: Kevin E. Noonan | August 24, 2011 at 10:38 AM
"It's a DEAD issue. The genome has been sequenced. As is always the case, the invention must be new and non-obvious."
I have to disagree on that point: If the human genome is prior art, a single isolated gene is a textbook example of a selection invention.
Posted by: James Demers | August 24, 2011 at 11:35 AM
My problem with claim 1 of the '441 patent is that it verges on nonsense. Filtering out the details, you can trim it down to
A method for screening for an alteration of a gene which comprises comparing the gene from a sample with the wild-type gene, wherein a difference from wild-type indicates an alteration in the gene.
Shorter version: Looking for an alteration by comparing two sequences, where a difference indicates an alteration. Good luck designing around this!
In the absence of limitations to the sequences or methods taught in the specification, how is this even remotely patentable?
Posted by: James Demers | August 24, 2011 at 11:53 AM
Quite happy with your discussion of the product claims. But not so sure about the process. I agree that construction of the claims is needed. But I'm not sure that it need be explicit. I suggest that the Federal Circuit construed the claim to be a one-step process - that of comparison of two sequences. They objected to that on the ground that it was not transformative. And, I suggest, properly. Otherwise, if the two sequences were printed side-by-side in a book, the claim would be infringed by readers. That's not reasonable. Of course, this objection could be overcome by drafting an explicit two-step claim including as a first step the sequencing of the sample. However, the court has to consider the actual claim, not what the patentee might have done 'if he had been well advised or bolder' (Nobel v. Anderson, per Romer J.).
Posted by: Tim Roberts | August 24, 2011 at 01:04 PM
Tim:
Exactly - if the claim is broad enough to encompass comparing two sequences written on a piece of paper it should be patent-ineligible. That is one way to interpret the claims, and perhaps the right way. My point is that you cannot assume that's what the claim means - you need to construe it using the rules set forth in Phillips. If that is done and that is the meaning a court gives the the claim, I think the outcome is correct.
But it shouldn't be an assumption.
Thanks for the comment.
Posted by: Kevin E. Noonan | August 24, 2011 at 03:16 PM
Dear James:
I'm not sure a claim needs to be capable of being designed around for it to be valid. But some alternative uses for BRCA gene mutations could be: correlations with other cancers; association of particular mutations with aggressiveness of a cancer; assays for protein expression that shows differences in the amount of BRCA protein or differences in affinity to anti-BRCA antibodies. All these would not fall within the scope of the claim and yet benefit from the disclosure of the BRCA gene by the inventors.
There are certainly claims in hte patents-in-suit that are directed to detecting specific mutations in the BRCA genes. Since the claim of the '441 patent seems to encompass all differences between germline and BRCA gene sequences there may be a lack of utility or enablement argument that could be made to invalidate the claims (since it seems to include differences in nucleotide sequence that are silent on an amino acid level).
I think we can agree that there may be several avenues for attacking the patentability of various claims in the Myriad patents. But first we need to construe the claims to know what they mean. That hasn't been done, and until it is we are grasping around in the dark.
Thanks for the comment.
Posted by: Kevin E. Noonan | August 24, 2011 at 03:27 PM
Dear James:
I'm not sure a claim to any particular gene is a selection invention, since the sequence of the gene (i.e., the specific chemical structure) is expressly disclosed in the prior art and thus a claim to the DNA is anticipated (under In re Hall, since the gene sequences are annotated and public). Since many times the function of the protein encoded by these genes is unknown there may be an opportunity for patenting the protein, methods and cells for making the protein, antibodies, etc.
Posted by: Kevin E. Noonan | August 24, 2011 at 03:31 PM
I have to agree with Kev agreeing with my position that claim construction must always take place and must always come first. I've been saying this for awhile now.
I make an exception for instances where it is discovered that the claim construction was blatantly wrong at the get go and allow for it to be redone, at least at the office anyway.
Other than that, do ur claim constructions. Do em first. Stop the madness of everyone arguing past each other.
I also note that if the Fed simply construed the claim then they might not even have to reach any really interesting issues of lawl because of the claim construction.
Posted by: 6 | August 24, 2011 at 04:05 PM
"I have to agree with Kev agreeing with my position that claim construction must always take place and must always come first."
Hey, 6, we've reached what I call "common ground" (i.e., we agree on something). It does happen.
Posted by: EG | August 25, 2011 at 12:29 PM
Kevin,
Very nice analysis.
I quite agree it's too bad there was not Markman hearing to clarify how the claims would be interpreted. Would have helped a lot, and made this much more useful in setting case law precedent.
Three quick points:
1. I think it's possible the claims do mean both purified from other stuff and isolated from other "genes" or DNA. But how much other DNA? It would have been pretty straightforward to specify, but the plain language of the claims does not do this, nor the definition in the specification; so I don't think your interpretation is the only interpretation, even among the cognoscenti. Your interpretation has the word "isolated" doing two things, but isn't it just as reasonable to assume it means just one--but which one. Lourie says it was not the "purified" meaning; the other two judges are not so specific, and seem to use sometimes one meaning and sometimes the other. Would be really useful to have some case law here.
2. It would be nice if the oligomer claims hanging off the independent cDNA claims really did mean "actually derived from DNA encoding the protein," but again it would have been straightforward to so narrow the claims, and yet the claims did not do that. And it does not work to indicate the claimed sequences are derived from cDNA, because no one makes a cDNA in doing diagnostics, so either (1) the claims are useless for diagnostics (but then why have them--the purpose was clearly to capture oligos that would go on microarrays, PCR fragments, or hybridization probes?); or (2) they're entirely useless because they would claim any subfragment of any therapeutic molecule or construct--extremely broad, so much so that they'd be hard to defend.
So even beyond the problems of "metastasis" to other genes that raise problems of prior art, I think these claims are deliberate efforts to get short DNA molecules useful in diagnostics that were never expected to have a cDNA intermediary. They were deliberately worded to make them broad. I think that's fatal; I gather you don't think so.
3. I think your analysis of the method claims does raise a lingering uncertainty. The language from Judge Lourie (who is the only one who went deeply into the method claims because the opinion on that point was unanimous) implies that if they'd merely specified that they were extracting and then sequencing before comparing they could have passed muster.
Your argument that it was implied might be right, in which case they are valid? Geez. They're still amazingly broad, and I presume that whole-genome sequencing that identifies BRCA alternations would infringe. Or not?
Whole-genome sequencing would not infringe if there were the "isolation" language, but there is no such language in the broad method claims. If I follow your logic on these claims it must also somehow be implied that the methods pertain only to BRCA DNA because they are linked to the specification. I quite agree that's possible. But I also think other interpretations are equally plausible that are not so narrow. Through your analysis, can you answer Judge Bryson's first question to Castanias: does whole genome infringe these patents or not? If not, why not?
Posted by: Robert Cook-Deegan | August 26, 2011 at 06:40 PM
Dear Bob:
1. I think the claims distinguish away from other DNA. They are to "an isolated DNA comprising a nucleic acid that encodes the polypeptide of SED ID NO: 2." I think that the claim as a whole, in light of the specification, says that the DNA claimed is isolated away from all the other stuff in cells (proteins, lipids, etc.) and in addition, is isolated from the other DNA because it is directed to the specific DNA that encodes the BRCA protein. The only word in the claim that extends past that could be "comprising," but I think the rule that you don't construe a claim to be invalid (if you can and if there is a basis for it) comes into play to prevent that interpretation. Keeping in mind that these sorts of claims were developed to claim DNA that could be used to produce the protein, such a construction is consistent with accepted practice. Any patentee who attempted to sue someone practicing WGS would lose on summary judgment of no infringement or invalidity, so as a policy matter this is not a problem in my view.
I think Judge Lourie did the conservative thing and said "Isolated" as in "cleaving covalent bonds" is enough. The real question (and perhaps the saving grace) is that genomic DNA in a cloning vector (which is how they are typically made) is certainly man-made and thus gets out of most of this. Of course, we can "win" this aspect and cripple the biologic drug industry if the Supreme Court were to accept the DOJ's childish dichotomy.
2. Similarly, if the dependent claim is limited to the DNA encoding SEQ ID NO. 2, which I think it must be, then oligos from the introns (for example) are not within the scope of the claim. Which pretty much vitiates the WGS straw man. I also think the patentee's intent is irrelevant; their claim language in my view limits the scope of the claims to oligos found within the cDNA sequence, and frankly I think your prior art analysis pretty much kills those claims (for reasons we have previously discussed).
3. I think Judge Lourie did imply that claims that say "comprising the steps of a) obtaining a patient sample; b) determining the nucleotide sequence of at least a portion of the BRCA genes in the sample; and c) diagnosing the patient as having an enhanced risk of developing breast or ovarian cancer when a mutant sequence according to Table 3 is detected" could pass muster under the MOT test. I think he is on solid ground in view of current Supreme Court precedent; clearly, how the Court decides Prometheus will be determinative.
Now, in this case WGS would infringe, and this is the question that has been present all along (and neglected to a certain extent by the plaintiffs' strategy). Patented genes are not the problem (especially since all those patents expire in about 10 years), and for most of them we don't know enough for them to commercially useful (which is not to say that they don't have patentable utility, just not a utility that will make anyone invest in them yet). Which is why the method claims are important - there is the possibility for patent thickets and the like. But as I've said on the blog and elsewhere, the alternative is worse.
Remind me to discuss with you some time how a ban on patenting diagnostic methods will encourage a return to the medieval guild system.
Thanks for the comment.
Posted by: Kevin E. Noonan | August 28, 2011 at 01:34 PM
It certainly goes without saying (though it bears repeating) that claim drafting and construction often determine patentability vs. patent-ineligibility. Proper claim construction at some point in this very important case seems, therefore, a no-brainer.
http://www.GeneralPatent.com/blog/
Posted by: patent litigation | August 28, 2011 at 10:33 PM
"Remind me to discuss with you some time how a ban on patenting diagnostic methods will encourage a return to the medieval guild system."
Dude that would be soooo awesome.
Kev, if you start your own guild can I be in it? Maybe an officer?
"Any patentee who attempted to sue someone practicing WGS would lose on summary judgment of no infringement or invalidity, so as a policy matter this is not a problem in my view."
Come on Kev, what if it was someone attempting to sue someone NOT practicing WGS with that claim? Do they lose or not? We still have to make a claim construction, and if that claim construction would lead one to the conclusion of invalidity in the instance that WGS was being used then it would also invalidate in the instance where WGS was not being used. However, in the instance where WGS was not being used, they don't have "no infringement" to fall back on.
So you kind of have a problem.
Posted by: 6 | August 29, 2011 at 03:11 PM