Abbvie v. Janssen: Another nail in the coffin of functional claiming

08.03.14 Posted in Federal Circuit Opinions by

Contact the author: Andrew Berks

AbbVie Deutschland GmbH v. Janssen Biotech, Inc., No. 2013-1338 (Fed. Cir. 7/1/2014)

This is a significant decision that extends the Federal Circuit’s holdings on written description from Ariad Pharms., Inc. v. Eli Lilly Co., 598 F.3d 1336 (Fed. Cir. 2010) (en banc), and has important lessons for patent applicants seeking to obtain functional claims, and for patent challengers facing blocking functional claims.

Abbvie (an Abbott spinoff starting in 2013) owns US Patents 6914128 and 7504485. These patents are directed to humanized antibodies that bind to and neutralize the activity of human interleukin 12 (IL-12). The over-production of IL-12 is a cause of psoriasis and rheumatoid arthritis. Slip op. at 5–6. Abbvie’s antibodies are based on a lead called “Joe-9,” that was developed through phage display technology. Its not mentioned in the opinion if Abbvie has an approved product based on the patented technology.

Centocor developed an anti-IL-12 antibody called “Stelara®” using transgenic mouse technology, approved by the FDA in 2009. Centocor filed a patent application, 10/912994 on its product, which apparently has not issued.  The key point with Stelara is that it is structurally substantially different than the Joe-9 derived antibodies from Abbvie.

Five claims were at issue, which each define the claimed antibodies by their function rather than their structure. The opinion cited claim 29 of the ‘128 patent as representative. Claim 29 recites an isolated human antibody that binds to IL-12 with a koff rate constant of 1×10-2 s-1 or less, as determined by surface plasmon resonance.

Centocor provoked an interference with Abbvie in 2007 which included arguments by Centocor that the Abbvie patents were invalid as obviousness. On August 6, 2009 the BPAI (predecessor to the PTAB) held in favor of Abbvie, that it was the senior party and that the Abbvie patents were not obvious.

So Abbvie promptly filed the instant action in the district of Massachusetts, alleging that Centocor was infringing the ‘128 and ‘485 patents. Centocor also appealed the Board’s decision under 35 USC § 146 in the Dist. Ct for the District of Columbia seeking a declaration of noninfringement and invalidity of the ‘128 and ‘485 patents. The two actions were consolidated in the Massachusetts court (Judge F. Dennis Saylor IV)

Following claim construction, the district court granted summary judgment to Abbvie that Centocor infringed the five asserted claims. However, on the issue of validity (written description, enablement, obviousness, and anticipation) the case was submitted to a jury, which determined that all the asserted claims were invalid on written description, enablement, and obviousness grounds. The jury found the claims invalid under the clear and convincing evidence standard. Since the § 146 action required a lower standard of proof, the court also entered judgment in Centocor’s favor on the § 146 claims. The court denied Abbvie’s motion for JMOL on the written description, enablement, and obviousness holdings.

Abbvie appealed, and the Federal Circuit panel (Lourie, O’Malley, and Chen, opinion by Lourie) upheld the district court in all respects. As a preliminary matter, footnote 5 in the opinion notes that Abbvie did not substantially challenge the obviousness finding at the district court, so the court notes that Abbvie’s patents are invalid for that reason. (This is presumably because one of the other issues here was whether the jury was properly instructed on the obviousness question.)  However, the court considered the written description question because the Federal Circuit is an “inferior court” so a “higher authority” (i.e., the Supreme Court) can find error on other bases in a more limited review. Since the Federal Circuit is not a court of last resort, its decision to decide a case on one of two alternative grounds does not strip the court of its power to decide the second question. Cardinal Chem. Co. v. Morton Int’l, Inc., 508 U.S. 83, 97–98 (1993).

The significance of this opinion is the holding that a genus described in functional language fails to meet the written description requirement where the genus exceeds the scope of the underlying structures disclosed in the specification. “The essence of the written description requirement” is that as part of the bargain with the public, a patent applicant must describe an invention so that the public will know what the invention is, and can show that the inventor was in possession of the claimed invention. Slip op. at 20–21.  “Possession” is shown by evidence in the disclosure (Ariad, 598 F.3d at 1351). When a patent claims a genus using functional language, the specification must demonstrate the applicant has invented species sufficient to support a claim to the functionally-defined genus. Slip op. at 22. The panel finds here that Abbvie failed to do that.

The panel analogizes this situation to a plot of land, where the property description is only to a corner of the plot, which is not a sufficient description to “distinctly circumscribe the outer boundaries” of the invention. Moreover, with the written description of a genus, merely drawing a fence around the invention isn’t good enough. A patent applicant has to also describe representative species encompassing the breadth of the genus. Otherwise the patent applicant has only described a research plan, leaving it to others to explore the unknown contours of the claimed genus. Slip op. at 23.

Here, despite the functional claiming that does not recite specific structural features, the panel limits Abbvie’s invention to a narrow set of antibodies derived from Joe-9 (“the described species are all of the similar type and do not qualitatively represent other types of antibodies encompassed by the genus”). Slip op at 24-25. However, the functional features are much broader and encompass Stelara.

This opinion takes a very harsh view of functional claiming: “the patents must at least describe some species representative of [the alleged infringing product].” Slip op. at 25. Thus, functional claims must have an underlying structural description that embraces the structural diversity of the claimed genus. Id. The opinion cautions that functional claims are vulnerable to written description attack unless the subject matter involves a “reasonable structure-function correlation.” Slip op. at 26. But in unpredictable technology fields (such as antibody function) this is difficult (and in the apparent view of the panel, unlikely to result in a valid claim). Moreover, The ‘128 and ‘485 patents do not describe any common features of the claimed antibodies. The claims attempt encompass any human antibody that achieves a desired result, including Stelara which is structurally very different from Abbvie’s Joe-9 –derived antibodies. Accordingly, the claims are invalid for lack of written description.

There were also procedural issues addressed in the case, pertaining to collateral estoppel over whether the BPAI decision in the interference proceeding was an actual litigation with a binding and final judgment with preclusive effect. The panel said “no,” and this question was also the subject of a concurring opinion by Judge O’Malley (who otherwise agreed with the “thoughtful” written description opinion).

Lessons: After this case, functional claiming will likely be treated with great caution by patent examiners and the courts. Patent applicants seeking to obtain functional claims should be careful to disclose a full range of underlying structural features that give rise to the functional features. The danger for applicants seeking to obtain functional claims is that someone will come along with a new structure (not contemplated by the original inventor) that exhibits the claimed function, and the patent could be lost. This case further underscores the policy preference for patents covering structural features of inventions, not functional features.

For patent challengers, potential new avenues of attack on functional claims are opened up, so that by showing a scope of a functional genus that exceeds the scope of the disclosure, a patent can be invalidated.

Link to this post:




Nautilus v. Biosig – The Supreme Court Clarifies the Clarity Requirement

06.05.14 Posted in Supreme Court Opinions by

Contact the author: Andrew Berks
Nautilus, Inc. v. Biosig Instr.. Inc., No. 13-369 (S. Ct. 6/2/2014) Opinion by Ginsburg.

The definiteness requirement, 35 U.S.C. §112(b) (AIA, effective 9/12/2012; previously §112 second para.) requires that a patent  “specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.” The Federal Circuit standard for this requirement has long been that a “claim is indefinite only when it is not amenable to construction or insolubly ambiguous.” Teva Pharms. USA, Inc. v. Sandoz, Inc., 723 F.3d 1363, 1368 (Fed. Cir. 2013). The court here rejects this language and holds that a claim is indefinite if it fails to inform a person skilled in the art, “with reasonable certainty,” of the scope of the invention when read in light of the specification and the prosecution history.

The Nautilus case involves heart rate monitors on exercise equipment, which have metal electrodes that a user grasps to get a heart rate reading. The disputed claim term was that the electrodes are “mounted … in a spaced relationship with each other.” Slip op. at 4. The patent in dispute, US 5,337,753, owned by Biosig, was reexamined at Nautilus’ request when the dispute began. In the reexamination, the USPTO affirmed the patent, concluding that the ‘753 patent was not anticipated or obvious. In the ensuing patent infringement litigation at the District Court, Nautilus prevailed on summary judgment following claim construction. The District Court held that the term “spaced relationship” was indefinite, because the claim did not define what the space should be or specify any parameters for determining appropriate spacing.

The Federal Circuit reversed and remanded, holding that there was sufficient information in the intrinsic record to inform a person of skill in the art what the spacing should be under its standards of “amenable to construction” and not “insolubly ambiguous.”

The Supreme Court reversed and remanded.  The analysis starts with a discussion of the level of imprecision §112(b) tolerates. Ginsburg notes this is a “delicate balance,” of the inherent limitations of language, a modicum of uncertainty that is the “price of ensuring appropriate incentives for innovation,” and the appropriate audience for patents, which are those skilled in the relevant art. On the other hand, patents “must be precise enough to afford clear notice of what is claimed.”

Ginsburg asserts that the “amenable to construction” and “insolubly ambiguous” standards lack the precision required by the statute and can “breed lower court confusion.” A court cannot ascribe some meaning to a patent’s claims. The inquiry should focus on the understanding of a skilled artisan at the time of the patent application, not that of a court viewing the matter post-hoc. Ginsburg suggests that the “insolubly ambiguous” standard diminishes the public notice function and would “foster the innovation-discouraging ‘zone of uncertainty.'” Slip op. at 11-12. The court says that the “amenable to construction” or “insolubly ambiguous” terminology can “leave the courts and patent bar at sea without a reliable compass.”

The new standard is that §112(b) requires “that a patent’s claims, viewed in light of the specification and prosecution history, inform those skilled in the art about the scope of the invention with reasonable certainty.” Slip op. at 11. “The definiteness requirement, so understood, mandates clarity.” Id.

Comment. One could argue that the former Federal Circuit standard allowed courts to read clarity into an unclear claim. This practice is now prohibited. This new standard is consistent with the doctrine that courts may not  re-draft claims, and must construe the claims as written. Ecolab, Inc. v. FMC Corp., 569 F.3d 1335, 1344 (Fed. Cir. 2009); Chef Am., Inc. v. Lamb-Weston, Inc., 358 F.3d 1371, 1374 (Fed. Cir. 2004); see also Exxon Chem. Patents v. Lubrizol Corp., 64 F.3d 1553, 1563 (Fed. Cir. 1995) (“[A court] is not free to read the claims as they might have been drafted, even if as drafted they do not accomplish what the inventor may have intended.”).


Gilead v. Natco – Gilead’s patent invalid for obviousness-type double patenting

04.27.14 Posted in Federal Circuit Opinions by

Contact the author: Andrew Berks
Gilead Sciences, Inc. v. Natco Pharma Ltd., No. 2013-1418 (Fed. Cir. 4/22/2014)

Gilead owns U.S. Patents 5,763,483 and 5,952,375, directed to antiviral compounds and their methods of use. The two patents have the same inventorship and are said to have similar subject matter, but they do not claim priority to a common patent application. The ‘375 patent was filed earlier and issued later than the ‘483 patent. Because of the difference in the earliest file dates, the ‘483 patent expires 22 months later than the ‘383 patent. The filing, issue, and expiration dates are mapped as:


Natco filed an ANDA for a drug covered by the ‘483 patent, and Natco’s only defense when sued for infringement of the ‘483 patent was obviousness-type double patenting, i.e., the assertion that the ‘483 patent was an obvious variation over the ‘375 patent that improperly sought to extend the right to exclude past the expiration of the ‘375 patent.

The district court held for Gilead, concluding that a later-issued but earlier-expiring patent cannot serve as a double patenting reference to an earlier-issued but later-expiring patent, and rejected Natco’s double patenting invalidity defense.

The Federal Circuit majority (Chen and Prost, opinion by Chen) reversed and remanded (Rader dissented). The majority’s reversal is predicated on an assumption that the ‘483 patent is an obvious variation of the ‘375 patent, but the details of that conclusion are not discussed. However, the substantive issue of whether the two patents were obvious variants was not the issue here. Rather, this case on appeal turned on the narrow legal question: “Can a patent that issues after but expires before another patent qualify as a double patenting reference for that other patent?” The answer: Yes.

Gilead’s defense of its patents was that the ‘375 patent does not extend the term of exclusivity for the ‘483 patent because the ‘375 patent issued later. The majority dismisses this argument, stating that in the post-June 8, 1995 Uruguay Round (“URAA”) era, patent expiration dates should control the analysis, not patent issue dates. The panel was concerned over “gamesmanship” in patent expiries if Gilead’s argument was accepted, such as filing serial applications on obvious variations. The panel also points to potential irrational outcomes under Gilead’s arguments. For example, if the ‘375 patent hypothetically issued one day before the ‘483 patent, under Gilead’s argument the extra 22 months of patent life of the ‘483 patent would be improper. This could encourage gamesmanship in applicants’ orchestrating patent issue dates, and is contrary to the intent of Congress in the URAA.

The majority states that the ‘483 patent should have been subject to a terminal disclaimer based on the ‘375 patent according to MPEP § 804.I.B.1.

The panel concludes that looking to the earliest patent expiration for patents that are obvious variants “best fits and serves the purpose of the doctrine of double patenting.” Accordingly, an earlier-expiring patent can qualify as an obviousness-type double patent reference for a later-expiring patent under the circumstances of this case. The expiration of the ‘375 patent should trigger “the public’s right to use the invention claimed in it and all obvious modifications of that invention.” Reversed and remanded.

Rader’s dissent asserts that majority decision is an unwarranted expansion of the judicially created doctrine of obviousness-type double patenting. Rader makes general remarks that the court should be more cautious before articulating a “new rule.”

Rader makes two good points. First, if the ‘375 patent never issued, there would be no question of the validity of the full term of the ‘483 patent. Second, Rader points out that the later filing date of the ‘483 patent is subject to ten months of intervening prior art. But Rader doesn’t really tie these points into an argument that the majority was wrong.

Rader also makes policy objections to the majority holding, including that a disclaimer on the ‘375 patent reduces the risk of multiple assignees asserting essentially the same claims against accused infringers, which is not a danger here, and that no one possesses an affirmative right to practice claimed subject matter.

Ultimately, it appears that Rader’s objection is whether the ‘483 patent was an obvious variant of the ‘383 patent. He defers to the examiner, stating that subsequent improvements, if satisfying the criteria for patentability, could “bar the public from practicing some subject matter encompassed by the expiring patents.” But then he concedes that patentably indistinct claims should not bar the public. Rader makes no specific objection to the majority’s reasoning and stops short of saying he would have affirmed the district court.

Comment: I don’t see how the court could have come to a different conclusion here. In the URAA (post June 8, 1995) era, patent expirations must control this kind of analysis. The district court’s reasoning (and Gilead’s defense of its patents) was essentially based on the old 17-year expiration mindset, that patent expiration followed patent issue dates in lock step. That of course is no longer how the system works.

On the specific facts here, it would be more satisfying if the issue of obviousness was discussed, but presumably that will be taken up on remand.


Zydus Not Infringing Lansoprazole ODT Particle Size Claim

03.29.14 Posted in Federal Circuit Opinions by

Contact the author: Andrew Berks
Takeda Pharm. Co. Ltd. v. Zydus Pharms. USA, Inc., No. 2013-1406 (Fed. Cir. 2/20/2014)

This Hatch-Waxman case pertains to particle size claims for the brand name drug Prevacid® SoluTab™. The product is an orally dissolving tablet (ODT) containing lansoprzaole. Only claim 1 of US6328994 is in dispute, which reads, in relevant part:

Claim 1. An orally disintegrable tablet which comprises (i) fine granules having an average particle diameter of 400 µm or less… and  (ii)… having a hardness strength of about 1 to about 20 kg…

(emphasis added)

The district court construed the particle size term to be  400 µm (±10%) or less. On this construction, the Zydus product apparently infringed. In addition, Zydus’ arguments that claim 1 was invalid were dismissed.

On appeal, the Federal Circuit panel (Prost, Plager, and Chen, opinion by Prost), reversed the claim construction of the lower court. The panel was troubled by the imputation of ±10% into the particle size claim. The purpose of the particle size limitation was to avoid a feeling of roughness in the mouth, and the panel believed the specification drew a sharp line between particles of 400 µm or less that had the desired mouth feel, and larger particles that did not. Also, there was language cited in the specification pertaining to a practical size of 425 µm or less with a 5% deviation, but then the specification defined “average particle diameter” to mean median particle diameter. This confused drafting created contradictory definitions, so the panel concluded that the inventors did not intend to incorporate a 10% deviation into the particle size limitation in claim 1. Takeda also made “lexicographer” arguments, asserting that the phrase “about 400 µm or less” was recited three times in the specification. However, the panel noted that each time this recitation appeared, it was immediately followed by the statement that the particle was size was preferably 300 to 400 µm, so this argument was rejected.

Additionally, the panel noted that claim 1 issued with no language of approximation modifying the particle size, but language of approximation did appear in claim 1 pertaining to the hardness (“about 1 to about 20 kg”), so the claim drafters weren’t bashful about language of approximation. Further, in the prosecution history, there was potentially invalidating prior art pertaining to lansoprazole particle size that was overcome by relying on a clear dividing line between particles of greater and lesser than 400 µm. Accordingly, the district court claim construction was reversed, and the claim was construed as requiring particles of precisely 400 µm or less.

The Zydus product had a particle size of 412 µm, outside of the claimed range as construed by the Federal Circuit. So there is no literal infringement of claim 1. This opinion does not discuss infringement under the doctrine of equivalents, which presumably will be left to the district court on remand.

Zydus also alleged that the ‘994 patent did not meet the requirements of 35 USC §112. Zydus alleged that the patent was indefinite for not specifying a particle size measurement method. However, the specification laser diffraction, and that was good enough for the panel, so this argument was dismissed. Zydus also alleged that the written description requirement was not met, because claim 1 addressed the particle size in the finished tablet, but the specification only discussed particle size pre-tableting. The panel dismissed this argument based on evidence from Takeda’s expert that the tableting process did not affect the particle size. Zydus also alleged that the enablement requirement was not met, because the specification did not teach a skilled artisan how to measure particle size with the “coulter counter” method without undue experimentation. This too was dismissed by the panel, which agreed with the lower court that a skilled artisan would know how to measure particle size with laser diffraction or optical microscopy.

Holding: Claim 1 is valid but not infringed. Remanded to the district court.

Take aways:

  • This is another case suggesting that particle size claims can be viable subject matter to to extend product life cycle.
  • There was a drafting problem in the specification, that strongly contributed to the panel holding limiting the patentee to a narrow claim construction.
  • There appeared to be an absence of teachings in the specification of methods of particle size measurement, which left the patent open  to §112 attacks.
  • This case emphasizes the value of language of approximation in claims reciting numerical ranges.


PTAB Obviousness Finding Reversed in Institut Pasteur GIIE Endonuclease Patents

01.05.14 Posted in Federal Circuit Opinions by

Contact the author: Andrew Berks

Inst. Pasteur v. Focarino, No. 2012-1485 (Fed. Cir. 12/30/2013)

Three patents were at issue, US6610545, US6833252, and US7309605, all based on an application originally filed 5/6/1992 and expired on 5/6/2012. The patents disclose group I intron encoded (GIIE) endonucleases. GIIE endonucleases are valuable research tools that are highly specific in selecting DNA cleavage sites. The claims were directed to methods of cleaving DNA, site directed genetic recombination, and recombinant mammalian chromosomes. The PTAB holding that the claims at issue were obvious was reversed in this appeal. (Before Newman, Clevenger, and Taranto, opinion by Taranto). This case also addresses jurisdictional issues of amendment of expired patent claims.

In 2009, Precision Biosciences, Inc., a co-appellee in this matter, filed an inter partes reexamination request for the three patents at issue. The request was granted, and Pasteur sought to amend the claims during the reexamination to be limited to chromosomal DNA in viable eukaryotic cells. However, the examiner rejected most claims as anticipated or obvious. The examiner was upheld on appeal to the Patent Trial and Appeal Board. The Board cited references from Quirk and Bell-Pedersen disclosing using a GIIE endonuclease to transfer DNA from a plasmid to non-chromosomal DNA in bacterial (i.e., prokaryotic) cells, and concluded based on other references that it would be obvious to substitute the eukaryotic DNA into the method of Quirk and Bell-Pedersen. The Board found a reasonable expectation of success in the Frey and Dujon references. For two of the patents, the Board considered Pasteur’s evidence that the claimed inventions were praised, copied, and licensed by the industry, but concluded that the evidence did not outweigh the “strong case of obviousness.”

On appeal to the Federal Circuit, the PTO sought to have all claims except claims 10 and 12 of the ‘545 patent dismissed as moot because the patents were all expired. Pasteur opposed dismissal of claim 14 of the ‘605 patent. These three claims were at issue in the appeal.

With respect to claim 14 of the ‘605 patent, Pasteur sought to reverse the Board holding that this claim was obvious. During the reexamination, Pasteur sought to substantively narrow claim 14, by limiting it chromosomal DNA. But under the pre-AIA statute, when a claim is substantively amended in an inter partes reexamination, the analysis of intervening rights after issuance treats an amendment as raising an irrebuttable presumption that the original claim was materially flawed, so there can be no liability for infringement before the amended claim issues. Since the ‘605 patent is expired, the claim amendments cannot be entered, because Pasteur cannot be granted rights beyond the term of the patent. The panel therefore dismissed Pasteur’s appeal with respect to the ‘605 patent.

With respect to claims 10 and 12 of the ‘545 patent, Pasteur appealed the Board holding that the patents were obvious. The Federal Circuit panel reversed the Board’s holding that the claims were obvious, finding that (1) teachings in the prior art that targeting a cell’s chromosomal DNA could be toxic to the cell (teaching away) and (2) industry praise and licensing of Pasteur’s invention (secondary factors), were not properly considered by the Board. The opinion states that these errors were prejudicial.

Regarding the teachings of the prior art, the panel asserts that the Board “misread” (slip op. at 17) the Frey and Dujon references, relied upon to support the finding that the GIIE digestion of prokaryotic DNA could be extended to eukaryotic yeast cells. The Frey reference disclosed GIIE digestion of DNA extracted from yeast cells, but the claims required digestion of DNA within the cells. The Dujon reference did not clearly disclose digestion of cellular DNA. The Board therefore failed to establish a prima facie case of obviousness.

Also, the Board ignored teachings that targeting a GIIE endonuclease to chromosomal DNA in a living cell would be highly toxic. The panel did not use the words “teaching away” here, but rather says that such a teaching counts significantly against finding a motivation to take the claimed steps with a reasonable expectation of success. According to the panel, this was not a harmless error.

Also, the objective indicia of non-obviousness that the industry licensed, praised, and copied Pasteur’s inventions was erroneously disregarded by the Board. Accordingly, the panel reversed the Board’s rejection of claims 10 and 12 of the ‘545 patent.

The panel also remanded on the issue of the ‘252 patent claims. The ‘252 patent claims are effectively the first step to practicing claim 12 of the ‘545 patent. The Board rejected the ‘252 claims on the same basis that it rejected the ‘545 claims, but the opinion notes that the Board failed to consider whether other motivations could have made the ‘252 claims obvious, and also failed to consider objective-indicia evidence for the ‘252 patent.

Comment. This case is an inventor-friendly case holding that findings of obviousness in crowded arts (including genetic engineering) must be carefully weighed. The sloppy analysis all to often applied by the USPTO in finding patents obvious was reversed here.


Wyeth’s expansive claim construction results in finding of non-enablement

01.04.14 Posted in Federal Circuit Opinions by

Contact the author: Andrew Berks

Patents on rapamycin for restenosis held invalid for lack of enablement. Wyeth v. Abbott Laboratories, 720 F.3d 1380 (Fed. Cir. 2013).

Wyeth sued defendants for patent infringement. Defendants counterclaimed that the patents were invalid as not enabled and for lack of written description.  The specification disclosed only one rapamycin species (sirolimus). The alleged infringers marketed stent products with everolimus and zotarolimus, two rapamycin analogs. The district court adopted a broad construction of the term “rapamycin” as proposed by Wyeth, then held the claims non-enabled because the only species disclosed was sirolimus and because the chemical arts are unpredictable, because of the complexity of the invention, and because of limited knowledge of treatment of restenosis at the time of invention. Thus, practicing the full scope of the claims (i.e., with other rapamycin analogs besides sirolimus) required undue experimentation.

This finding was upheld by the Federal Circuit panel on appeal (Moore, Bryson, and Wallach, opinion by Moore). The claim scope is broad (id. at 1385). Wyeth attempted to broaden the background in the art (presumably to capture the infringing activity). But, the specification was silent on how to modify sirolimus in a way that would preserve utility for restenosis. Id. Such modification would require substantial experimentation. Id. The court found that the need to synthesize “at least tens of thousands of candidate compounds” would require undue experimentation. Id.

The opinion discussed factors for undue experimentation. Id. at 1386. Undue experimentation is a matter of degree, citing Chiron Corp. v. Genentech, Inc., 363 F.3d 1247, 1253 (Fed. Cir. 2004). A considerable amount of experimentation may be routine, but even routine experimentation is not without bounds. Id.

The court found that the specification was only an invitation for further research (“a starting point for further iterative research in an unpredictable and poorly understood field.”). Id. at 1386.

The district court finding that the patents were invalid as non-enabled was affirmed.

Comment. This case is interesting because patentees typically strive for broad claim construction to capture the activities of accused infringers, but here, the patentee was granted a scope of claim construction, at their own request, that put them in the territory of non-enabled subject matter.


Sloppy Claim Drafting Invalidates Teva Copaxone Claims

08.27.13 Posted in Federal Circuit Opinions by

Teva Pharms. USA, Inc. v. Sandoz, Inc.,, No. 2012-1567 (Fed. Cir. 7/26/2013) (no reporter cite) (Rader, Moore, and Benson (sitting by designation from the District of Utah). Opinion by Moore.

Teva sued Mylan and Sandoz for patent infringement of nine patents covering copolymer-1, marketed by Teva as “Copaxone®” for the treatment of multiple sclerosis. Copolymer-1 is a polymer of alanine, glutamic acid, lysine, and tyrosine in a 6:2:5:1 ratio with a molecular weight of 5–9 kD.

The court first looked at definiteness under 35 USC § 112(b). For this analysis, the court divided the claims into two groups. Group I claims were directed to copolymer-1 having a molecular weight of 5-9 kD, made by a process of reacting and purifying. On claim construction, the molecular weight was determined to be an average molecular weight. Three different measures of average MW’s are discussed, Mp (peak mol. wt), Mn (number average mol. wt), and Mw (weight average mol. wt). The patent challengers argued that Teva inconsistently defined MW as Mw and Mp during prosecution, reinforcing ambiguity. The panel agreed. “It is undisputed that Group I claims contain an ambiguity because their plain language does not indicate which average molecular weight measure is intended.” Slip op. at 8. Significantly, the panel noted inconsistent statements during prosecution over which average molecular weight measure should be used.

Also, the panel noted drafting failings in Fig. 1, in which the peaks of the curves do not correspond to the values denoted as “average molecular weight” in the figure’s legend. The peak (Mp) of the 7.7 curve is clearly less than 7.7 kD on the x-axis, and the peak of the 12.0 curve is clearly less than 12 kD on the x-axis. The opinion states that it is therefore difficult to conclude that Mp is the intended measure for the Group I claims. The Group I claims are therefore indefinite.
By contrast, Group II claims, which recite copolymer I having over 75% of its mole fraction within a MW range of 2–20 kD, are held to be definite. The court states that the Group II claims do not recite the molecular weight in terms of a statistical measure (average molecular weights), but rather as precise points on the “molecular weight” axis. Thus, the scope of the Group II claims are readily ascertainable and not invalid for indefiniteness.

The opinion also reviewed the parties’ arguments on enablement (35 USC § 112(b)), which came down to a battle of experts, and concluded that the district court did not err in concluding that the claims were enabled. The panel also concluded that the claims were not obvious, based both on the prior art and secondary considerations. The prior art expressed a preference for a higher molecular weight product and taught away from the claimed invention. Regarding secondary considerations, the panel found that Copaxone® is coextensive with the asserted claims that trigger an assumption of a nexus between the drug’s commercial success and the claimed invention.

The panel also looked at infringement. The district court construed the claimed amino acid ratios of copolymer-1 to be “approximately 6:2:5:1.” The district court converted this to percentages of 42.9%, 14.3%, 35.7%, and 7.1% (without explaining the methodology). The panel did not tamper with this calculation. Each accused product was found to vary, on a percentage basis, by 4.5%, which met the “approximately 6:2:5:1” limitation. The claims were therefore literally infringed by the accused products.

Holdings: the Group II claims were therefore valid and infringed. The panel remanded the Group I claims because of the finding of invalidity.

This case is a good illustration of the danger of sloppy drafting. I have heard patent attorneys in the past argue in favor of ambiguous and obfuscatory drafting to maximize claim scope.  Here, the drafting attorneys and client may have felt no need to specify which average molecular weight measure in the specification, in the belief that this detail might be limiting. But in a case like this that comes down to a knock-down litigation,  when the courts start doing the claim construction to resolve ambiguous terms, the applicant loses control of the process and it becomes very expensive. In addition, the flaws in Fig. 1 undoubtedly annoyed the courts involved and was specifically cited in support of the Fed. Cir. conclusion that certain claims were indefinite.

These problems are all correctable by careful drafting by patent attorneys who understand the technology (and take the time to figure it out). And, they are preventable during prosecution also by patent attorneys who take the trouble to understand the underlying technology.


Wellman, Inc. v. Eastman Chemical Co. – A Trade Secret Does Not Excuse Best Mode

06.13.11 Posted in Federal Circuit Opinions, News by

Wellman lost a patent infringement claim on the grounds that it failed to disclose a specific preferred formulation known at the time the patent was filed, and also because it attempted to hold back part of the invention as a trade secret.   Wellman, Inc. v. Eastman Chemical Co., No. 2010-1249 (Fed. Cir. 4/29/2011).

Wellman had two patents on PET resins for beverage bottles that had favorable characteristics. Wellman had a specific formula at the time the patents were filed, but the patents did not disclose the preferred formula (“Ti818”) or any other specific formulae.  Claim 1 of the ‘317 patent, stated to be exemplary, had generic chemical limitations and functional elements to a “crystallization exotherm peak temperature” and an “L* luminosity value.”  Wellman alleged that Eastman infringed its patents. Eastman responded that the patents were indefinite and didn’t meet the best mode requirement.

The district court granted-in-part Eastman’s motion for lack of best mode, holding that key claims were invalid for want of best mode. The district ct found that Ti818 was a preferred formula that the time of filing.  The district ct also found that the patents did not disclose the use of carbon black, and that Wellman sought to retain that as a trade secret.

The Federal Circuit panel (opinion by Rader) affirmed on the best mode holding.  The panel noted that the actual formula of Ti818 was outside the claimed ranges for some ingredients. The panel concluded that Wellman concealed its invention. “By masking what at least one inventor considered the best of these slow-crystallizing resins, Wellman effectively concealed its recipe for Ti818.” Slip op. at 18.  The attempt to hold back the carbon black additive as a trade secret was condemned as an intentional concealment. Slip op. at 18–19. Choosing to maintain part of the invention as a trade secret “does not excuse Wellman’s compliance with the best mode requirement.” Slip op. at 19. The indefiniteness claim was rejected by the panel.

Comment: mixing trade secrets and patented inventions is not a good idea.  Especially in this case where the carbon black appeared to be essential for practicing the invention to get PET with the desired characteristics.


Stanford v. Roche – When is an assignment not an assignment?

06.12.11 Posted in Supreme Court Opinions by

This case was a dispute over conflicting assignments by an inventor to Stanford and a private lab where he did supporting work. When infringement litigation erupted between the parties, the defendant asserted it was a co-owner so the plaintiff had no standing to sue. The underlying research was federally funded, and Stanford asserted that the Bayh-Dole Act gave it superior rights.  This contention was rejected by the Supreme Court, but the decision did not squarely address the contract law issue of which assignment had priority.  Stanford Univ. v. Roche Molecular Systems, Inc. , No. 09-1159 (S.Ct. June 6, 2011) (563 U.S. __ (2011)).

Mark Holodniy was hired by Stanford University in 1988 to work on HIV research supported by federal funds.  Dr. Holodniy signed an agreement when hired to assign his “entire right, title, and interest” in inventions resulting from his employment to the university. Holodniy then collaborated with Cetus, a private firm (whose successor in interest is Roche) and worked at a Cetus facility learning “PCR,” a special technology for the research.  As a condition of working with Cetus, Holodniy signed an agreement (referred to as the “VCA” in the opinion) stating that he “will assign and hereby does assign (to Cetus) his right title and interest in each of the ideas, inventions and improvements made as a consequence of his access to Cetus.” (slip op. at 2, internal quotes and punctuation omitted).  Holodniy returned to Stanford and continued his research, which eventually resulted in valuable patents.

Stanford sued Roche in 2005 alleging that HIV test kits marketed by Roche infringed its patents.  Roche’s defense was that it was a co-owner of the technology, and that Stanford therefore lacked standing to sue Roche for patent infringement.  Stanford countered that Holodniy had no rights to assign because Holodniy’s research was federally funded.  Stanford claimed superior rights because of the Bayh-Dole Act and because of the assignment to Stanford.

The district court sided with Stanford, holding that Bayh-Dole Act conveys title in an invention to the inventor only after the government and contracting party (in this case, Stanford) have declined title (which, of course, was not the case here). Thus, the court held that Holodniy had no rights to convey to Cetus.

The Federal Circuit reversed, holding that the Stanford assignment was a “mere promise” to assign, but the Roche VCA assignment was a present assignment, based in the Federal Circuit’s rule in FilmTec Corp. v. Allied-Signal, Inc., 939 F. 2d 1568, 1572 (Fed. Cir. 1991).  (Breyer’s dissent, slip op. at 6–7). The Federal Circuit concluded that the Cetus assignment trumped the Stanford assignment.  Additionally, the court held that the Bayh-Dole Act “does not automatically void ab initio the inventors rights in government-funded inventions.” Majority op. at 5. The Supreme Court majority (opinion by Roberts) upheld the Federal Circuit.

The majority asserts that Stanford argues that the Bayh-Dole Act “reorders” the normal assignment process of conveying inventors rights to employers by contract, by vesting title to inventions in the federal contractor, rather than the employee-inventor.  Majority op. at 8. The court rejects this argument. There is a lengthy discussion of the Bayh-Dole Act in the context of the assumption by the majority that the Cetus assignment had priority over the Stanford assignment.  The majority opinion does not fully address the contract law issue of why the Cetus assignment trumps the Stanford assignment.  Rather, the majority appears to address a Stanford argument that even if Cetus had a superior assignment, the Bayh-Dole Act gives Stanford full rights the invention.  The court is probably correct that Stanford loses on that argument, but the failure to fully analyze why Stanford’s assignment failed is disappointing.

Sotomayer, in a concurrence with the majority, suggests that Stanford’s appeal to the Supreme Court did not challenge the Federal Circuit’s Filmtec rule, which explains the majority holding.

Breyer wrote a substantial dissent. He would have sent the case back to the Federal Circuit for further argument on whether an assignment can be “terminated unilaterally … through a separate agreement purporting to assign the inventor’s rights to a third party.” Breyer dissent at 1, quoting the question presented. Breyer questions the Federal Circuit conclusion that the Cetus agreement words “do hereby assign” trumped the Stanford agreement words “agree to assign,” because the Cetus words “operated on the invention automatically,” while the Stanford words did not. Id. at 6.  Breyer calls this “slight linguistic differences” that “makes too much of too little.”  Id. at 7. Breyer concludes the language is functionally equivalent, and that both agreements give rise only to equitable interests, and that Stanford should probably take priority because Stanford came first in time and obtained a postinvention assignment from Holodniy.   Id.

Breyer asserts that the Filmtec rule changed many years of precedent, but without explanation.  Breyer further states that Filmtec undercuts the objectives of the Bayh-Dole Act.  Id. at 8. Breyer also states that the Federal Circuit reasoning (in this case) brings about an interpretation contrary to the Stanford agreement, running “counter to the drafters’ reasonable expectations” of how the contract would be interpreted.  Still further, the Bayh-Dole Act and Executive Order 10096 are controlling rules for federally funded research, that in effect require assignments of federally funded research to the federal contractor. Id. at 8–9.

However, the issue of whether the Filmtec rule is appropriate where the Bayh-Dole Act is applicable was apparently not fully addressed by either the Federal Circuit or in the Supreme Court briefs.  Breyer therefore dissented and would have remanded the matter for further arguments in the Federal Circuit on this issue.

Comment: I agree with Breyer that the two agreements convey essentially identical rights by a plain reading.  Splitting hairs over contract language as was done in this case does not seem appropriate.  In my view, the Cetus language “hereby does assign,” which the Filmtec rule gives so much weight to, is actually meaningless, because no invention had yet been made at the time of the Cetus agreement.  I have not read the Filmtec decision, but the real issue is which of the two seemingly conflicting purported assignments should control, and that issue was not squarely addressed in this decision.



Billups-Rothenberg – Another Biotech Patent Invalid for Lack of Written Description

05.17.11 Posted in Federal Circuit Opinions by

In Billups-Rothenberg, Inc. v.  Assoc’d Regional Univ. Pathologists, Inc., No. 2010-1401 (Fed. Cir. 4/29/2011), U.S. Patent Nos. 5,674,681 (the ’681 patent) and 6,355,425 (the ’425 patent) describe genetic tests for Type I hereditary hemochromatosis, an iron disorder characterized by excessive iron absorption by the body. The defective gene in hemochromatosis is the High Fe (“HFE”) gene. The claims at issue are directed to a mutation in the HFE gene, known as the C282Y mutation.

At the time the original patent applications were filed, the applicant claimed methods of detecting mutations responsible for hemochromatosis but had not yet identified any specific disease-causing mutations. Following this filing, other groups published relevant research on hemochromatosis. On 3/26/1999, Billups inventors filed a second application that matured to the ‘425 patent claiming a method of diagnosing an iron disorder by determining the presence of a specific mutation in the HFE gene.

In 2009, Billups sued ARUP for patent infringement. After a Markman hearing and claim construction, the court granted summary judgment of invalidity for lack of a written description to ARUP. The district court found that the DNA sequence of the hemochromatosis gene and/or sequence of the C282Y mutation were not expressly specified in the ’681 patent. Slip op. at 8. The court further stated that “[t]he invention claimed in the ’681 patent is a method to test for a DNA mutation, not a test for a defective protein.” Id. The district court also concluded that defandants’ enablement arguments were persuasive, but declined to rule on them in light of its ruling that the claims failed the written description requirement.  Id.  The district court also held that the ’425 patent was invalid as anticipated by US 6,025,130 (the ‘130 patent), which was § 102(e) prior art. Id. at 9. The ’130 patent disclosed the exact genetic sequences for the three mutations at issue in this case.

Billups appealed.  The Federal Circuit panel (Gajarsa, Linn, and Moore, opinion by Gajarsa) upheld the district court in all respects.

Gajarsa concludes that the ‘681 patent was only a research plan, and failed the written description test.  The ‘681 patent disclosed the mutation as a general location “within less than a 300 base pair region of a defined exon of a well studied multi-gene family,” but the specification disclosed neither the hemochromatosis gene sequence nor any specific mutations within that gene. Slip op. at 10-11. Billups conceded these were “adequately described the envisioned, but then unknown, mutations.”  Id. at 11. Gajarsa concludes that “Billups cannot satisfy the written description requirement merely through references to later-acquired knowledge. This case is like Regents [of the Univ. of Cal. v. Eli Lilly & Co.], and Fiers [v. Revel], in which the DNA sequences at issue were unknown in the art.”

Furthermore, the ’681 patent had unsupported genus claims to methods of identifying a genus of unknown genetic mutations.  Id. at 11. But under Ariad, a patent must set forth “either a representative number of species falling within the scope of the genus or structural features common to the members of the genus,” which the ‘681 patent did not do.  Id. at 12.  The later discovery of the the C282Y mutation within the claimed genus is not adaquate written description.  Id.

Additionally, the ‘681 patent failed the written description requirement because the claims have functional claim language without a correlation between structure and function.  Id. at 12. The general location disclosed in the ‘681 patent was too “imprecise to constitute structural features necessary to meet the written description requirement.” Id.

With regard to the holding that the ‘425 patent anticipated, Billups argued that the ‘130 patent taught away from the genetic sequences relevant to the ‘425 patent.  Gajarsa rejects this argument, holding that “whether a reference ‘teaches away’ from the invention is inapplicable to an anticipation analysis.”  Id. at 15, citing Celeritas Techs., Ltd. v. Rockwell Int’l Corp., 150 F.3d 1354, 1361 (Fed. Cir. 1998).  Thus, since the ’130 patent disclosed a diagnostic test for hemochromatosis that included identification of the S65C mutation, the ’425 patent’s claims are anticipated.  Furthermore, Billups failed to present evidence of nonenablement of the ‘130 patent. Accordingly, the ‘130 patent is an anticipatory reference. Id. at 16.


Stay In Touch!
Subscribe to our Newsletter!

Enter your email address to subscribe to this blog and receive notifications of new posts by email.