Amgen v. Sanofi : The fall of the newly characterized antigen test (if it ever truly,jurisprudentially existed in the first place)

The case at the district court and motion for new trial

In district court, Amgen proposed the contested jury instruction regarding the newly characterized antigen test. Sanofi unsuccessfully opposed it, arguing that (1) the Federal Circuit had never held that mere disclosure of a newly characterized antigen was sufficient for written description in the absence of a correlation between the structure of antibodies that bind to a particular binding site and their function (i.e., binding to that site); (2) Federal Circuit precedent required the disclosure of a novel protein for the newly characterized antigen test to apply, and (3) this test was not a substitute for proper written description support for a genus claim, which requires disclosure of “a representative number of species falling within the genus or structural features common to the members of the genus.” Defendants’ Memo. Regarding Disputed Jury Instructions, 1:14-cv-01317-RGA (D. Del. Mar. 13, 2016) at 7–10.

The jury found that Sanofi had not proven invalidity of the asserted patent claims. After the verdict, Sanofi moved for a new trial, citing the challenged jury instruction as one ground. Op. Br. in Support of Defendants’ Mot. for A New Trial, 1:14-cv-01317-RGA (D. Del. 5 May 2016) at 25–26. In its motion, Sanofi offered essentially the same three legal arguments, in addition to challenging the sufficiency of the evidence. First, Sanofi argued that the Federal Circuit had never held that the written description requirement was met for a functional antibody claim (e.g. “antibodies that bind to target X”) only on the basis of the newly characterized antigen test; the test was based on an example in a United States Patent and Trademark Office (USPTO or PTO) training publication, and experts disputed the legal and scientific soundness of the publication’s assumption that fully characterizing an antigen provided adequate information about the structure of antibodies that would bind to it. Id. at 25–26 and fn. 7 (citing USPTO, Written Description Training Materials 45–46 (Revision 1 25 March 2008), available at http://www.uspto.gov/web/menu/written.pdf). Second, Sanofi argued that the newly characterized antigen test should only apply where the applicant disclosed a novel protein and claimed both the protein and the antibody that binds to it. Id. at 26. Finally, Sanofi argued that the jury instruction conflated the newly characterized antigen test with the “common structural features” test for adequate description of a genus, by stating (without precedent, according to Sanofi) that the common structural features test could be satisfied by disclosing a newly characterized antigen if the production of antibodies was conventional or routine. Id. at 26–27. In addition to these three legal arguments, Sanofi argued that there was no evidentiary basis for the decision in Amgen’s favor, because Amgen’s argument rested on the idea that the “antigen” that had been newly characterized was the region of the previously known PCSK9 protein to which the antibodies bind—the “sweet spot”—rather than an entire macromolecule. Id. at 10–12.

In response to Sanofi’s first legal argument regarding lack of Federal Circuit precedent, particularly on functional claims, Amgen cited Centocor, 636 F.3d 1341; Enzo, 323 F.3d 956 (Fed. Cir. 2002); and Noelle v. Lederman, 355 F.3d 1343 (Fed. Cir. 2004), all of which contain language endorsing the newly characterized antigen test. ³ Plaintiffs’ Answering Br. in Opp. to Defendants’ Mot. for a New Trial, 1:14-cv-01317-RGA (D. Del. 18 May 2016) at 25–26. Amgen further pointed to Centocor’s separate analysis of representative species, common structural features, and newly characterized antigen to support the idea that each of these represents a separate test. Id. (citing Centocor, 636 F.3d at 1351–53 and “many PTAB/BPAI decisions on newly characterized antigens”). In response to Sanofi’s second legal argument, Amgen argued that Centocor does not expressly require a novel protein, or that the patent must claim both the protein and the antibody—Centocor references “newly characterized” antigens, not newly discovered antigens, and the Board of Patent Appeals and Interferences had “expressly applied the doctrine to newly characterized amino acid residues on known proteins.” Id. at 26. ⁴ In response to Sanofi’s third legal argument, Amgen argued that both Enzo and Noelle tie together the concept of newly characterized antigens and structure–function correlations. Id. at 27. Finally, in response to Sanofi’s argument of insufficiency of the evidence, Amgen argued that an antigen could be a portion of a molecule rather than an entire molecule, and that the new characterization of the “sweet spot” on PCSK9 was enough for patentability. Id. at 13–14.

The district court denied Sanofi’s motion. Order, 1:14-cv-01317-RGA (D. Del. January 3, 2017). Because the issue of the newly characterized antigen test had already been briefed when Sanofi challenged Amgen’s jury instruction, the court declined to reconsider the issue. Id. at 28. ⁵

Appeal to the Federal Circuit

On appeal, Sanofi presented essentially the same arguments on the newly characterized antigen test as in the trial court—the same three legal arguments, and the argument regarding insufficient evidence—with the addition of an initial argument that a textual reading of Section 112, which requires a “written description of the invention” as “the fundamental ‘quid pro quo’ of the patent system,” does not allow a description of the antigen to substitute for a description of the antibody, which is the actual invention. Corrected Br. for Defendants-Appellants at 35–36 (quoting Ariad, 598 F.3d at 1345). Regarding the first legal argument, Federal Circuit endorsement of the newly characterized antigen test (particularly for functional antibody claims), Sanofi pointed out that the Federal Circuit had never actually found written description satisfied based on the newly characterized antigen test—it had been endorsed only in dicta (in Enzo and Noelle), narrowed (and rejected on the facts) in Centocor, and supported by an example in PTO training materials which had been removed in more recent versions of the PTO’s training materials. Reply Br. for Defendants-Appellants at 2, 12 (citing Enzo, 323 F.3d at 965, and PTO, Examination Guidance and Training Materials (Archived 2008), http://bit.ly/2kLrTLa); Corrected Br. for Defendants-Appellants at 41 (describing how following Centocor, the PTO revised the Training Materials, which now omit any mention of antibody examples as a means of satisfying the written description requirement). ⁶

Regarding the second legal argument—whether an applicant must claim both a novel protein and its antibody for the newly characterized antigen test to apply—Sanofi argued that the Federal Circuit’s decision in Centocor required that an applicant describe both a novel protein and a specific antibody that could be produced routinely, not merely a category of antibodies that bind to a previously known protein. Id. at 41–43. Because Centocor explained that the antibody example “presumes that the applicant is disclosing a novel protein and then claiming both the protein and an antibody that binds to it,” Sanofi argued, the jury instruction should not have included a claim to an antibody but not a protein. Id. (citing Centocor; emphasis added by Sanofi).

Regarding its third legal argument—whether the newly characterized antigen test provided a basis to conclude that the applicant had described common structural features across antibodies—Sanofi argued that the newly characterized antigen test is invalid because a description of an antigen does not allow a person of skill of art to infer the structure of the corresponding antibody. Sanofi argued that the newly characterized antigen test relies on the idea that there must be a correlation between structure and function, but that it is at least a contested factual question whether disclosure of a binding site on an antigen reveals anything about the structure of antibodies that bind to that site. Corrected Br. for Defendants-Appellants at 35–37; see also Br. of Plaintiffs-Appellees at 7–9 (describing how antibodies “dock” together with their antigens).

Finally, as in the district court, Sanofi argued that the evidence was insufficient to support a finding of disclosure of a “newly characterized antigen,” because the full PCSK9 protein was already known; Amgen had identified the site on PCSK9 where the antibody bound as the “newly-characterized antigen,” but Sanofi argued that the term “antigen” did not properly cover the binding site alone rather than a full protein. Id. at 43–45.

Amgen defended the jury instruction primarily on the basis of the same three Federal Circuit decisions that it relied upon at the district court: Enzo, Noelle, and Centocor. See Br. of Plaintiffs-Appellees at 41–43 (citing Enzo, 323 F.3d 956; Noelle, 355 F.3d 1343; Centocor, 636 F.3d 1341). Regarding Sanofi’s first argument—whether current precedent supports the continued vitality of the newly characterized antigen test—Amgen pointed out that Noelle and Enzo had explicitly recognized that Federal Circuit precedent adopted the newly characterized antigen test, and that the Manual of Patent Examining Procedure (MPEP), as well as current versions of the PTO’s training materials, still endorsed the test. Br. of Plaintiffs-Appellees at 42–43 (citing Noelle, 355 F.3d at 1349, Enzo, 323 F.3d at 964, and MPEP § 2163 ¶ II.A.3(a)); Noelle, 355 F.3d at 1349 (“If Noelle had sufficiently described the … antigen, he could have claimed its antibody by simply stating its binding affinity for the ‘fully characterized’ antigen”); Enzo, 323 F.3d at 964 (“We are persuaded by the Guidelines on this point and adopt the PTO’s applicable standard for determining compliance with the written description requirement,” including newly characterized antigen example); MPEP § 2163 ¶ II.A.3(a) (“[D]isclosure of an antigen fully characterized… provides an adequate written description of an antibody claimed by its binding affinity to that antigen, if generating the claimed antibody is so routine that possessing the antigen places the applicant in possession of an antibody.”) (citing Centocor; quote omitted).

With respect to Sanofi’s second legal argument that Centocor required patenting a novel protein, Amgen argued that Centocor rested on a holding that “generating the claimed antibodies was not routine” as of the priority date, not on whether the antigen protein was novel. See id. at 47 (citing Centocor, 636 F.3d at 1352–53). Amgen further argued that this argument did not make sense because naturally occurring proteins are patent-ineligible, and most antigens are naturally occurring proteins, implying that antibodies to natural proteins would not be patentable under such a standard. Id. at 46–47.

With respect to Sanofi’s third legal argument—that the newly characterized antigen test was unsound because of a lack of correlation between function and structure in antibodies—Amgen pointed to the “well defined structural characteristics for the five classes of antibody,” ⁷ and argued that once the “sweet spot” on PCSK9 had been characterized and exemplary antibodies had been developed, the production of the claimed antibodies of varying sequences was “a matter of routine.” Id. at 44–45 (citing Noelle, 355 F.3d at 1349). Because of the common structural characteristics of antibodies, Amgen argued, “specifically characterizing the antigen… demonstrate[s] possession of antibodies that bind to the antigen where, as here, production of the relevant antibodies would be routine.” Id. at 45. Amgen also emphasized the practical implications of the test, noting that because modern antibody technology makes it easy to make additional antibodies to a particular antigen that have different sequences, limiting antibody claims to a specific antibody sequence (rather than a function) would limit the efficacy of such claims to protect the invention from being copied. Id. at 46.

In response to Sanofi’s argument regarding the sufficiency of the evidence for a “newly-characterized” antigen, Amgen cited the patent’s definition for “antigen,” which included both molecules and the portions of molecules to which antibodies bind, to argue that its “newly-discovered antigen” could be a portion of a molecule (the “sweet spot”) rather than an entire protein. Id. at 7; see also Plaintiffs’ Answering Br. in Opp. to Defendants’ Mot. for A New Trial, 1:14-cv-01317-RGA (D. Del. May 18, 2016) at 13. Finally, Amgen argued that any error was harmless, as the claims also had adequate written description on the grounds of disclosure of representative species and common structural features. Br. of Plaintiffs-Appellees at 48.

Eli Lilly, Pfizer, and Ipsen submit amicus briefs

Two amicus briefs were submitted on the topic of the newly characterized antigen test in support of Sanofi’s position, one from Eli Lilly and Company, and another from Pfizer Inc. and Ipsen Pharma S.A.S. Eli Lilly’s brief argued that the application of the newly characterized antigen rule in the PTO’s training materials essentially means that the claim is described so long as it is enabled, contradicting the Federal Circuit’s stance requiring written description separate from enablement (e.g. in Ariad, 598 F.3d 1336). Eli Lilly further argued that patents claiming antibodies “solely by their function(s), e.g. by reference to their ability to bind to a naturally occurring biological target,” “chill and tax development” of antibody drug products by enabling their holders “to monopolize a naturally occurring biological target and thereby preempt an entire therapeutic antibody market from any competition.” Br. of Amicus Curiae Eli Lilly and Co. (“Eli Lilly Am. Br.”) at 2–3. Id. at 6–8. Its brief argued in favor of Sanofi’s position that antibody structure cannot be scientifically determined from a description of the antigen, stating that regardless of whether Amgen argued that the claims were enabled, the fact that there is “no known function/structure relationship between antibodies and… binding to a specified antigen,” the public was not put on notice of the boundaries of the patent, and would be “left to conduct research to actually invent the claimed subject matter.” Id. at 10. Eli Lilly also walked through Enzo and Noelle, arguing that neither squarely addressed the issue of whether the newly characterized antigen test is appropriate. Id. at 11–15. It argued that Enzo treated the existence of an applicable function/structure relationship as a question of fact, and did not question whether the PTO’s Training Materials were factually accurate in describing a function/structure relationship for antibodies, as well as noting that although the Enzo court endorsed the PTO’s Guidelines, they did not explicitly adopt its Training Materials. Id. at 13. Eli Lilly finally argued that neither party challenged the newly characterized antigen test in Noelle, and that because both Enzo and Noelle related to nucleotide sequences rather than antibodies, any discussion of antibodies was dicta. Id. at 14–15.

Pfizer and Ipsen also cited Ariad, arguing that the written description requirement is necessary to “ensur[e] that patent applicants are unable to preempt future innovators by obtaining claims that are far broader than their actual invention,” particular for claims where a patent application claims an invention based on its function rather than its structure. Br. of Amici Curiae Pfizer Inc. and Ipsen Pharma S.A.S. in Support of Defendants-Appellants and Reversal of the District Court’s Decision on Written Description at 3. They further argued that the claims at issue did not disclose a representative number of species falling within the scope of the genus or adequate structural features common to the members of the genus. Id. at 5–15. In addition, they emphasized the non-binding nature of the PTO guidance cited in previous Federal Circuit decisions. Id. at 22–23. They also argued that broad protection for large molecules under the newly characterized antigen rule was not necessary, and would in fact stifle competition by preempting future innovation into the “large genus of antibodies” that might bind to a specific region of a protein. Id. at 16–17, 23–24. Eli Lilly, Pfizer, and Ipsen all urged the court to reject the idea that a new antigen could include one or more particular amino acids of a previously known protein. Id. at 18–21; Eli Lilly Am. Br. at 15–19.

Federal Circuit decision

The Federal Circuit held that the district court’s instruction was improper, and agreed that it “effectively permitted the jury to dispense with the required finding of a ‘written description of the invention’” by “improper[ly] equat[ing]” enablement with written description. Op. at 1377–78. It criticized the instruction for allowing a jury to find that any antibody encompassed by the claim was adequately described merely because some antibody to the newly characterized antigen—not any particular antibody—could be easily made. Id. Emphasizing the fact that a written description using functional terminology is only adequate where “the art has established a correlation between structure and function,” the court further noted that it has been “hotly disputed” in newly characterized antigen cases whether knowledge of the chemical structure of an antigen gives sufficient “structure-identifying information about the corresponding antibodies.” Id. at 1378. It further held that the newly characterized antigen test “flouts basic legal principles of the written description requirement”, because it “allows patentees to claim antibodies by describing something that is not the invention, i.e., the antigen.” Id. at 1378–79.

In its discussion of the three primary cases cited by the parties (Enzo, Noelle, and Centocor), the Federal Circuit noted that in Enzo (which involved functionally claimed genetic material rather than antibodies), the court had not held that all functional descriptions of genetic material would be sufficient to meet the written description requirement, but instead cited to the PTO’s Guidelines on written description for the idea that functional characteristics, coupled with a known correlation between function and structure, could satisfy the requirement. Id. at 1376–77. In dicta, the Enzo court said that

the PTO would find compliance with 112, ¶ 1, for a claim to an isolated antibody capable of binding to antigen X, notwithstanding the functional definition of the antibody, in light of the well-defined structural characteristics for the five classes of antibody, the functional characteristics of antibody binding, and the fact that the antibody technology is well developed and mature.

The Federal Circuit therefore concluded that the jury instruction in this case had impermissibly conflated the enablement requirement with the written description requirement. Id. at 1377–78. Even if the disclosure of a newly characterized antigen would mean that the production of corresponding antibodies was “routine or conventional,” and therefore presumably enabled, the Federal Circuit was unwilling to draw the conclusion that characterizing the antigen would provide written description. Id. at 1378. It found that the proposition that “knowledge of the chemical structure of an antigen gives the required kind of structure-identifying information about the corresponding antibodies” was “hotly disputed,” and therefore not “so settled as to be entitled to judicial notice.” Instead, it required “fact finding” on the issue of whether an adequate “correlation between structure and function” had been demonstrated in order to show that an enabling disclosure also provided adequate written description. Id. (quote omitted).

Amgen’s petition for rehearing en banc and supporting briefs

Amgen filed a petition for rehearing en banc. It emphasized the longstanding use of the newly characterized antigen test, arguing (despite the panel’s view of the earlier case law) that Enzo, Centocor, and Noelle were binding precedent rather than dicta. Amgen’s Pet. for Rehearing En Banc at 2, 5–9. Amgen argued that in each of these cases, the Federal Circuit “articulated the newly-characterized-antigen test; applied that test to the case before it; and rendered judgment based on the test,” and that “[t]he ‘rationale upon which the Court based the results of its earlier decisions’ is as binding as ‘the result[s].’” Id. at 7 (quoting Seminole Tribe, 517 U.S. at 66–67). In addition, Amgen argued that the newly characterized antigen test was grounded in PTO guidance—both the MPEP and the PTO’s training materials—and that the test still appears in the MPEP. Id. at 6 (“[T]he MPEP declares that ‘disclosure of an antigen fully characterized by its structure, formula,’ etc., ‘provides an adequate written description of an antibody claimed by its binding affinity to that antigen, if “generating the claimed antibody is so routine that possessing the [antigen] places the applicant in possession of an antibody.”’”) (quoting MPEP §2163 ¶ II.A.3(a)). Because of the “myriad patents” issued under the newly characterized antigen test, Amgen argued, the reliance interests of innovators in this space weighed against abandoning the test. Id. at 9–11. Citing the $2 billion that Amgen alone invested into the claimed invention, Amgen argued that pharmaceutical innovators would not invest in new therapeutic agents without “confidence they can take [the Federal Circuit’s] precedents at their word,” which would be undermined by the panel’s decision. Id. at 10. Amgen further argued that because of the ease of generating new antibodies once an antigen’s structure is known, “[w]ithout genus claims, patent protection for antibodies would be nearly worthless,” diminishing the incentive to invest in targeting antigens (particularly those which are difficult to target, which Amgen argued included PCSK9). Id. at 10–11.

Three biopharmaceutical companies—Bristol-Myers Squibb Company, Bavarian Nordic, and Enzo Biochem, Inc.—filed a joint amicus brief in support of the petition for rehearing en banc. These companies expressed their worry that the Federal Circuit’s abandonment of the newly characterized antigen test would diminish their previously held patent protection and undermine the efficacy of the patent system at incentivizing the development of antibody therapeutics. Br. of Amicus Curiae Bristol-Myers Squibb Company, Bavarian Nordic, and Enzo Biochem, Inc. in Support of Appellee’s Pet. for En Banc Rehearing, at 1–4. Defending the statement in Noelle that the newly characterized antigen test is valid, and arguing that disclosure of the structure of an antigen” puts the full genus of antibodies in the hands of the public” and “show[s] possession” of the invention, the companies argued that the panel decision in the current case improperly overruled Noelle without an en banc decision. Id. at 4–7.

In response, Sanofi defended the panel’s holding that the newly characterized antigen test “is not legally sound and is not based on any binding precedent” and that written description requires the patent applicant to describe the invention itself (here, the antibody, rather than the antigen). Resp. to Pet. for Rehearing En Banc at 1 (quoting Op. at 1376; emphasis omitted); id. at 7–8 (citing Ariad, 598 F.3d at 1345). It re-emphasized that none of the decisions in Enzo, Noelle, or Centocor upheld the validity of an antibody patent on the basis of the newly characterized antigen test, and that there was therefore no conflict in result between the panel’s decision here and in any previous case. Id. at 4–6. Finally, Sanofi reiterated that the PTO’s guidance endorsing the newly characterized antigen test had been withdrawn. Id. at 6–7.

On 23 February 2018, the Federal Circuit denied the petition for rehearing en banc, without discussion.

Petition for Certiorari

On 23 July 2018, Amgen filed a petition for writ of certiorari. Petition for a Writ of Certiorari, No. 18-127 (23 July 2018) (“Pet.”). There, Amgen argued that the Federal Circuit’s separate and distinct standards for enablement and written description were inconsistent with the text of § 112, which states that the specification “shall contain a written description” of the invention “in such full, clear, concise, and exact terms as to enable.” Id. at 2–5. Amgen argued that the Federal Circuit’s test for written description—whether the inventor was “in possession” of the invention at the time of filing—contradicted the text and history of § 112 and defied precedent. Id. at 16–27. Amgen further argued that resolution of this issue was extremely important, as the Federal Circuit’s allegedly “[e]ver-[c]hanging” possession test impeded innovation, particularly in the biotech field, by making it difficult for innovators to know what needed to be disclosed in order to obtain patent protection. ⁸ Id. at 27–29.

On 27 August 2018, several amici submitted a joint brief in support of Amgen—Bristol-Myers Squibb Company, Morphosys AG, Bavarian Nordic A/S, and UCB Biopharma SPRL. Brief of Amici Curiae Bristol-Myers Squibb Company, Morphosys AG, Bavarian Nordic A/S, and UCB Biopharma SPRL in Support of Petitioners, No. 18-127 (27 August 2018). ⁹ These amici endorsed Amgen’s arguments, including a statement that the Federal Circuit’s separate standards for written description and enablement “is divorced from [§ 112]’s language and fails to serve the statute’s purpose,” and that § 112 should instead be assessed according to a “unitary standard” of “enabl[ing] a skilled artisan to make and use the invention.” Id. at 10, 12. In addition, they emphasized the importance of patent incentives in the biopharmaceutical industry as well as the importance of therapeutic antibodies in modern clinical research. Id. at 1–8. Their brief further argued that the Federal Circuit was placing increasingly rigid restrictions on patentees through the “representative number of examples” test, ¹⁰ which they argued was overly strict in practice for reasons including the difficulty of knowing whether the patentee had disclosed a sufficient number of examples. Id. at 18–21.

On 19 November 2018, Sanofi filed its Brief in Opposition to Amgen’s petition. Br. in Opp., No. 18-127 (19 November 2018). As a procedural matter, Sanofi argued that Amgen had never challenged the Federal Circuit’s separate written description and enablement requirements in the district court or at the Federal Circuit and could not raise the issue for the first time in its petition for certiorari. Id. at 1–2. Even though the law of written description that Amgen was attempting to challenge was settled Federal Circuit law, Sanofi argued that Amgen “still had an obligation to preserve the issue before the district court and Federal Circuit panel,” particularly in light of the Federal Circuit’s general willingness to “reconsider its precedents en banc, even when the prior precedent was itself an en banc decision.” Id. at 16. In addition, Sanofi argued, Amgen had “consistently cited Ariad with approval in the district court and the Federal Circuit,” and had endorsed the newly characterized antigen test—a biotechnology-focused subtest of written description of the type that Amgen’s certiorari brief now criticized. Id. at 17. Particularly because Ariad was decided almost a decade ago, Sanofi urged the Court to grant review on this issue “only in a case where the Federal Circuit has at least had an opportunity to provide its most recent thinking on the question.” Id. at 20.

Sanofi further noted that the question presented “is not dispositive to the outcome of this case and may well be mooted by the parties’ upcoming trial in February 2019.” Id. at 2. Sanofi pointed out that Amgen did not contend that its interpretation of the written description requirement would result in a determination that the Amgen patents were valid, and that the Federal Circuit had remanded for a new trial on enablement as well as written description. Id. at 21. In addition, the decision being appealed was interlocutory, with the proceedings in the trial court “moving forward expeditiously toward a February 2019 trial date,” creating a risk of mootness. Id. at 22–23. As a result, Sanofi argued, this case would not be an efficient use of judicial resources. Id. at 21, 23. ¹¹

Substantively, Sanofi argued, no reason exists to disrupt “more than half a century of settled patent law” setting forth the written description requirement. Id. at 24. In fact, it contended, Amgen itself had supported Ariad’s decision on the written description requirement in an amicus brief in 2009. Id. Sanofi argued that it was unclear what Amgen was challenging—the existence of separate written description and enablement requirements, or specifically the Federal Circuit’s “possession” standard—but argued that both standards were substantively correct. Id. at 25–34. Sanofi cited Ariad’s interpretation of the text of § 112, which requires both a “written description of the invention” and a “written description… of the manner and process of making and using it”, to support the idea that the requirements are separate according to both the language and the intent of the statute. Id. at 27–29 (citing 598 F.3d at 1344–45). Sanofi further argued that the courts’ settled understanding of the distinct written description requirement, as well as Congress’s silence on the issue while amending the statute in the face of this settled understanding, support the existence of a separate written description requirement, as does language in certain of the Supreme Court’s decisions. Id. at 29–34 (“In Festo [Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722 (2002)], for instance, this Court succinctly summarized the three separate requirements of §112: the specification must ‘describe, enable, and set forth the best mode of carrying out’ the invention. 535 U.S. at 736.”). ¹²

Sanofi additionally argued that “the settled nature of the law in this area weighs heavily against disrupting the longstanding understanding of the written description requirement,” and that any change would “repudiate decades of law and practice and require significant changes in USPTO’s public guidance and its examination procedures.” Id. at 32–33. Finally, Sanofi challenged Amgen’s contention that the Federal Circuit’s existing written description law had produced instability and provoked criticism, suggesting that Amgen had not cited any recent support for such a claim. Id. at 34. To the extent that any developments since Ariad created issues for innovation in biotechnology, Sanofi argued, these arguments should have been presented to the courts below. Id. at 34–35.

In reply, Amgen argued that review was warranted because the Federal Circuit’s “possession” test defied the statutory text of § 112. Rep. for Pets., No. 18-127 (4 December 2018). Specifically, Amgen argued that § 112 imposes a single requirement of “a written description” and that the phrases “of the invention,” “of the manner and process of making and using” the invention and “in such full, clear, concise, and exact terms as to enable” skilled artisans to practice the invention all modify “written description,” resulting in one unitary requirement for a description of the invention. Id. at 3 (citing § 112). Amgen argued that the comma after “of the manner and process of making and using it,” separates that phrase from the phrase “in such full, clear, concise, and exact terms as to enable,” showing that Congress intended the latter standard “to apply to all the antecedents instead of only to the immediately preceding one.” Id. at 3. Amgen additionally argued that this construction also reduced surplusage, because a description of the invention is necessary to enable artisans to make and use it. Id. at 4. Finally, Amgen argued that there is nothing in the statutory text that supported a “possession” standard, rather than a “description” standard, as Amgen advocated for, so any requirement of a “written description” should be understood as being part of a single requirement for an enabling description. Id. at 4–5.

Amgen also argued that the “possession” test had no historical basis – even Sanofi dated the “possession” standard only to “at least 1967.” Id. at 6. Amgen also dismissed Sanofi’s argument that Congress’s failure to “disturb” the “possession” test was a tacit acceptance of that test, and that Sanofi had not identified even one instance where Congress had even considered the “possession” test. Id. Amgen further argued that Supreme Court jurisprudence had always viewed written description was being “governed by the ‘full, clear, concise, and exact terms as to enable’ standard”—i.e. a unitary standard—and that the Federal Circuit’s decision to find that that phrase only modifies “the written description … of the manner and process of making and using” the invention flew in the face of Supreme Court precedent. Id. at 6–8. Finally, Amgen argued that, even if the “possession” test was entrenched, the longevity of an error is no barrier to review. Id. at 8.

Amgen further argued that the question was exceptionally important and thus ripe for Supreme Court review, noting that the question had been reviewed en banc at the Federal Circuit level and prompted amici papers from four leading biopharmaceutical innovators. Id. at 8–9. Amgen argued that “[u]nmoored from § 112’s text, the “possession” standard has spawned ever-changing sub-tests with a disparate impact on biotechnology,” which needed to be addressed by the Supreme Court. Id. at 9.

Finally, Amgen argued that this case was an appropriate vehicle for review, as the decision below turned on the Federal Circuit’s application of existing written description law. Id. at 10–11. Nor would the Federal Circuit’s remand weigh against review—if the Supreme Court were to grant the petition, Amgen would agree to a stay of trial below so that it could be conducted under the correct standard, thus saving judicial resources. Id. at 11–12.

On 7 January 2019, the Supreme Court denied certiorari.

Fallout: PTO guidance and updates to the MPEP

As a result of the Federal Circuit’s decision, there has been additional fallout at the PTO. On 22 February 2018, in response to the Federal Circuit’s panel decision, the PTO released guidance entitled “Clarification of Written Description Guidance For Claims Drawn to Antibodies and Status of 2008 Training Materials,” where the PTO withdrew the “newly characterized antigen” test from being used in examining antibody-related patents. ¹³ The guidance summarized the Federal Circuit’s decision in Amgen, and then noted that “In view of the Amgen decision, adequate written description of a newly characterized antigen alone should not be considered adequate written description of a claimed antibody to that newly characterized antigen, even when preparation of such an antibody is routine and conventional. Id. The Amgen decision will be added to the MPEP in due course.” The memorandum instructed patent examiners to “continue to follow the guidance in the MPEP regarding written description … except insofar as MPEP § 2163 indicates that disclosure of a fully characterized antigen may provide written descriptive support of an antibody to that antigen. The MPEP will be updated in due course to reflect these changes.” See discussion of MPEP § 2163, supra. As of the publishing of this article, it does not appear that the MPEP has been updated to reflect the Federal Circuit’s decision.

Conclusion

With the Supreme Court’s denial of certiorari, it appears that the Federal Circuit’s longstanding precedent of separate written description and enablement requirements under § 112 will continue undisturbed. In addition, the Federal Circuit’s rejection of the newly characterized antigen test for biotechnology patents will remain controlling law, despite concerns by Amgen and amici. If and when the MPEP is updated to reflect the Federal Circuit’s now-final decision in Amgen, additional clarity from the PTO regarding its interpretation of Amgen may be useful to applicants seeking to define the bounds of their antibody patents. Despite the undisturbed precedent regarding the separate written description requirement, we can expect continuing litigation over this area of patent law so long as antibody-based therapeutics retain their momentum within the biopharmaceutical industry.