Bayer Healthcare LLC v. Baxalta Inc.

989 F.3d 964

• Court: Court of Appeals for the Federal Circuit
• Decided: March 1, 2021
• Docket: 19-2418
• Status: Published

Opinion

Case: 19-2418 Document: 67 Page: 1 Filed: 03/01/2021

United States Court of Appeals for the Federal Circuit ______________________

BAYER HEALTHCARE LLC, Plaintiff-Cross-Appellant

v.

BAXALTA INC., BAXALTA US INC., Defendants-Appellants

NEKTAR THERAPEUTICS, Defendant-Appellee ______________________

2019-2418, 2020-1017 ______________________

Appeals from the United States District Court for the District of Delaware in No. 1:16-cv-01122-RGA, Judge Richard G. Andrews. ______________________

Decided: March 1, 2021 ______________________

BRADFORD J. BADKE, Sidley Austin LLP, New York, NY, argued for plaintiff-cross-appellant. Also represented by SONA DE, CHING-LEE FUKUDA; JOSHUA JOHN FOUGERE, RYAN C. MORRIS, PAUL ZEGGER, Washington, DC.

EDGAR HAUG, Haug Partners LLP, New York, NY, ar- gued for defendants-appellants and defendant-appellee. Also represented by KAITLIN ABRAMS, NICHOLAS F. GIOVE, JONATHAN HERSTOFF, ERIKA SELLI. Case: 19-2418 Document: 67 Page: 2 Filed: 03/01/2021

______________________

Before NEWMAN, LINN, and STOLL, Circuit Judges. STOLL, Circuit Judge. This patent infringement case presents various issues of claim scope, infringement, validity, and damages. Bayer HealthCare LLC sued Baxalta Inc. and Baxalta US Inc. (collectively, “Baxalta”) and Nektar Therapeutics, alleging that Baxalta’s biologic product Adynovate® infringes cer- tain claims of Bayer’s U.S. Patent No. 9,364,520. The jury found that the asserted claims were enabled and infringed, and that Bayer was entitled to reasonable-royalty dam- ages. The district court did not, however, send the question of willful infringement to the jury, instead holding as a matter of law that Baxalta’s conduct did not meet the re- quirements for willfulness. Baxalta now appeals the dis- trict court’s denial of its motions for judgment as a matter of law or a new trial on the issues of infringement, enable- ment, and damages, along with the court’s award of pre- verdict supplemental damages. Bayer cross-appeals, chal- lenging the district court’s denial of its motion for a new trial on willfulness. We affirm. BACKGROUND I The ’520 patent is directed to recombinant forms of hu- man factor VIII (or FVIII). FVIII is a protein that is pro- duced, and released into the bloodstream, by the liver. FVIII comprises 2,332 amino acids and has six different structural domains: A1-A2-B-A3-C1-C2. Most relevant to this case is the domain of FVIII known as the “B-domain.” As “a critical component of the intrinsic pathway of blood coagulation,” FVIII is useful in the treatment of he- mophilia A. ’520 patent col. 1 ll. 29–30. Hemophilia A is the “most common hereditary coagulation disorder” and is “caused by deficiency or structural defects in” naturally Case: 19-2418 Document: 67 Page: 3 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 3

occurring FVIII. Id. at col. 1 ll. 27–29. Because FVIII has a “short half-life” of “only about 11 hours,” as a therapeutic it “must be administered frequently.” Id. at col. 1 ll. 52–55. The ’520 patent recognizes that “[t]he need for frequent in- travenous injection creates tremendous barriers to patient compliance” and, thus, “[i]t would be more convenient for the patients if a FVIII product could be developed that had a longer half-life and therefore required less frequent ad- ministration.” Id. at col. 1 ll. 56–60; see id. at col. 2 ll. 47–49. Moreover, “the cost of treatment could be re- duced if the half-life were increased because fewer dosages may then be required.” Id. at col. 1 ll. 60–62. The patent explains that a process called “PEGylation” has “been demonstrated to increase the in vivo half-life of a protein.” Id. at col. 3 ll. 34–35. PEGylation is the conju- gation (or attachment) of a polymer called polyethylene gly- col (PEG) to a protein such as FVIII. See id. at col. 3 ll. 35–37; see also id. at col. 9 ll. 12–14. PEG is a type of polyalkylene oxide. See id. at col. 8 ll. 41–43. The patent further teaches that “random” modification of FVIII by PEGylation was known and that certain ran- domly PEGylated proteins had been approved as therapeu- tics. The inventors sought, however, “a more specific method for PEGylating FVIII” because the prior-art “ran- dom” approach had several drawbacks: Random modification of FVIII by targeting pri- mary amines (N-terminus and lysines) with large polymers such as PEG and dextran has been at- tempted with varying degree[s] of success . . . . This random approach . . . is much more problem- atic for the heterodimeric FVIII. FVIII has hun- dreds of potential PEGylation sites, including the 158 lysines, the two N-termini, and multiple histi- dines, serines, threonines, and tyrosines, all of which could potentially be PEGylated with rea- gents primarily targeting primary amines. Case: 19-2418 Document: 67 Page: 4 Filed: 03/01/2021

Id. at col. 3 l. 50–col. 4 l. 1, col. 4 ll. 19–20. The patent identifies as an “additional drawback to not controlling the site of PEGylation on FVIII” the “potential activity reduction if the PEG were to be attached at or near critical active sites, especially if more than one PEG or a single large PEG is conjugated to FVIII.” Id. at col. 4 ll. 7–11. Furthermore, “[b]ecause random PEGylation will invariably produce large amounts of multiply PEGylated products, purification to obtain only mono-PEGylated products will drastically lower overall yield.” Id. at col. 4 ll. 11–14. Finally, the patent explains that “the enormous heterogeneity in product profile will make consistent syn- thesis and characterization of each lot nearly impossible” and constitutes “a barrier to commercialization,” since “good manufacturing requires a consistent, well-character- ized product.” Id. at col. 4 ll. 14–19; see id. at col. 4 ll. 4–7 (“[H]eterogeneous processing of full length FVIII can lead to a mixture of starting material that leads to further com- plexity in the PEGylated products.”). Thus, the inventors recognized “a need for an improved FVIII variant that possesses greater duration of action in vivo . . . while retaining functional activity,” and that is de- sirably “produced as a homogeneous product in a consistent manner.” Id. at col. 4 l. 65–col. 5 l. 3. The patent purports to overcome the drawbacks of “random” PEGylation by us- ing “site-directed” PEGylation at the B-domain of FVIII: The present invention is based on the discovery that polypeptides having FVIII activity can be co- valently attached at a predefined site to a biocom- patible polymer that is not at an N-terminal amine, and that such polypeptides substantially retain their coagulant activity. Furthermore, these poly- peptide conjugates have improved circulation time and reduced antigenicity. The conjugates of the in- vention are advantageous over the prior art conju- gates that had random polymer attachments to Case: 19-2418 Document: 67 Page: 5 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 5

FVIII or attachments at an N-terminal. Site-di- rected attachment allows one to design modifica- tions that avoid the regions required for biological activity and thereby to maintain substantial FVIII activity. . . . Site-directed attachment also allows for a uniform product rather than the heterogene- ous conjugates produced in the art by random pol- ymer coupling. Id. at col. 8 ll. 15–30; see id. at col. 15 ll. 9–13 (explaining that the “retained activity” of the claimed conjugates was “surprising” given “the problems with past polymeric con- jugates causing nonspecific addition and reduced activity”). Claim 1 is the only asserted independent claim and re- cites: 1. An isolated polypeptide conjugate comprising a functional factor VIII polypeptide and one or more biocompatible polymers, wherein the functional factor VIII polypeptide comprises the amino acid sequence of SEQ ID NO: 4 or an allelic variant thereof and has a B-domain, and further wherein the biocompatible polymer comprises polyalkylene oxide and is covalently attached to the functional factor VIII polypeptide at the B-domain. Id. at col. 61 ll. 9–16 (emphases added to the disputed claim terms). II Bayer sued Baxalta and Nektar, Baxalta’s collaborator on its Adynovate® product, for patent infringement. Ady- novate® is a recombinant PEGylated FVIII product used to treat hemophilia A. Adynovate® is made by PEGylating to amino acids in FVIII that have amine groups, including the amino acid lysine. We hereafter refer to this process as “amine/lysine” PEGylation. Case: 19-2418 Document: 67 Page: 6 Filed: 03/01/2021

The district court construed the claim term “an isolated polypeptide conjugate” in claim 1 to mean “a polypeptide conjugate where conjugation was not random.” Bayer Healthcare LLC v. Baxalta Inc., No. 16-cv-1122, 2018 WL 3212425, at *2 (D. Del. June 29, 2018) (Claim Construction Op.). The district court determined that, during prosecu- tion of the ’520 patent, Bayer had “clearly and unmistaka- bly disclaimed any ‘polypeptide conjugate where conjugation was random.’” Id. at *4. In addition, the district court construed the claim term “at the B-domain” in claim 1 to mean “attachment at the B-domain such that the resulting conjugate retains func- tional factor VIII activity.” Id. at *8. The district court re- jected Baxalta’s proposed construction of the term to mean “at a site that is not any amine or carboxy site in factor VIII and is in the B-domain,” finding that Bayer did not dis- claim PEGylation at amine or carboxy sites. Id. at *8–9. Baxalta moved for summary judgment of noninfringe- ment and invalidity for lack of enablement. The district court denied Baxalta’s motions due to genuine factual dis- putes. See Bayer HealthCare LLC v. Baxalta Inc., No. 16-cv-1122, 2018 WL 6727054, at *3–7 (D. Del. Dec. 21, 2018) (Summ. J. Op.). Baxalta also moved in limine re- questing that the district court clarify the meaning of “ran- dom” in its construction of “an isolated polypeptide conjugate.” The district court denied Baxalta’s motion, “again reject[ing] [Baxalta’s] argument that [Bayer] de- fined ‘random’ conjugation as ‘any conjugation at amines or carboxy sites.’” Bayer HealthCare LLC v. Baxalta Inc., No. 16-cv-1122, 2019 WL 10890386, at *1 (D. Del. Jan. 3, 2019) (Mot. in Lim. Order). Prior to trial, Baxalta moved to exclude the testimony of Bayer’s damages expert, Dr. Addanki, regarding his pro- posed reasonable-royalty rate. In his expert report, Dr. Ad- danki opined that Bayer was entitled to a royalty rate of 23.75%—the midpoint of the bargaining range of 5.1% to Case: 19-2418 Document: 67 Page: 7 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 7

42.4%—based on the Nash Bargaining Solution. The dis- trict court concluded that the expert failed to tie his 50/50 split to the facts of the case, and thus excluded his “opinion that a reasonable royalty rate is ‘the mid-point of the bar- gaining range’ . . . , including any subsequent opinions that rely on that mid-point rate.” Bayer HealthCare LLC v. Bax- alta Inc., No. 16-cv-1122, 2019 WL 330149, at *8 (D. Del. Jan. 25, 2019) (Daubert Order). The district court denied, however, Baxalta’s request to prohibit Dr. Addanki from testifying as to his proposed bargaining range of 5.1% to 42.4%. The case then proceeded to a jury trial. The district court granted Baxalta’s pre-verdict motion for JMOL of no willful infringement. Thereafter, the jury found that Bax- alta infringed asserted claims 1–3 and 8 of the ’520 patent, and that none of those claims were invalid for lack of ena- blement. J.A. 1889–90, 1892. The jury also found that Bayer was entitled to $155,190,264 in reasonable-royalty damages for the time period from June 14, 2016 through November 30, 2018 based on a 17.78% royalty rate applied to a $872,836,128 royalty base. J.A. 1893. Following the verdict, Baxalta moved for JMOL or a new trial on the issues of infringement, enablement, and damages. The district court denied Baxalta’s motions. See generally Bayer Healthcare LLC v. Baxalta Inc., 407 F. Supp. 3d 462 (D. Del. 2019) (JMOL Op.). For its part, Bayer moved under Rule 59 of the Federal Rules of Civil Procedure for an award of pre-verdict supplemental damages for the time period from December 1, 2018 through February 8, 2019, the date of the district court’s judgment. The district court granted Bayer’s motion and awarded Bayer supplemental damages in the amount of $18,324,562 based on the actual sales data for that period and the jury’s 17.78% royalty rate. Bayer Healthcare LLC v. Baxalta Inc., No. 16-cv-1122, 2019 WL 4016235, at *6 (D. Del. Aug. 26, 2019) (Rule 59 Op.); Final Judgment at 2, Bayer Healthcare LLC v. Baxalta Inc., No. 16-cv-1122 Case: 19-2418 Document: 67 Page: 8 Filed: 03/01/2021

(D. Del. Sept. 10, 2019), ECF No. 507. Bayer also moved for a new trial under Rule 59(a) on the issue of willfulness, but the district court denied its motion. Rule 59 Op., 2019 WL 4016235, at *8–9. Bayer and Baxalta appeal. We have jurisdiction pur- suant to 28 U.S.C. § 1295(a)(1). DISCUSSION This case presents various issues on appeal and cross- appeal. We start by addressing Baxalta’s challenges to the district court’s construction of the claim term “at the B-do- main” and its interpretation of the word “random” in its construction of the claim term “an isolated polypeptide con- jugate.” We then turn to Baxalta’s challenges to the dis- trict court’s judgments of infringement and enablement, along with the court’s awards of damages and pre-verdict supplemental damages. Finally, we address Bayer’s chal- lenge to the district court’s judgment of no willful infringe- ment. I We first address Baxalta’s challenge to the district court’s construction of the disputed term “at the B-domain” in claim 1. Baxalta contends that the district court erred in failing to construe the term to exclude amine/lysine con- jugation. Neither the parties nor the district court consid- ered extrinsic evidence in construing this claim term. “Claim construction based on the intrinsic evidence is a question of law that this court reviews de novo.” Hologic, Inc. v. Minerva Surgical, Inc., 957 F.3d 1256, 1269 (Fed. Cir. 2020) (citing Trs. of Columbia Univ. v. Symantec Corp., 811 F.3d 1359, 1362 (Fed. Cir. 2016)). “The con- struction of claim terms based on the claim language, the specification, and the prosecution history are legal deter- minations.” Id. (quoting Trs. of Columbia Univ., 811 F.3d at 1362). Based on our review of the claim language, spec- ification, and prosecution history, we conclude that the Case: 19-2418 Document: 67 Page: 9 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 9

district court did not err in construing the disputed claim term. A The plain claim language recites “[a]n isolated polypep- tide conjugate” in which PEG “is covalently attached to the functional factor VIII polypeptide at the B-domain.” The plain language of claim 1 does not require PEGylation at any particular amino acid sites in the B-domain, let alone exclude from its scope any specific amino acids as attach- ment sites in the B-domain. Rather, claim 1 more broadly requires PEGylation at the B-domain as a region. B Statements in the specification support this view of the claims and indicate that site-directed PEGylation target- ing the B-domain, rather than particular amino acid sites within it, achieved the desired benefit of retaining coagu- lation activity. See, e.g., ’520 patent col. 4 ll. 7–9 (“An ad- ditional drawback to not controlling the site of PEGylation on FVIII is a potential activity reduction if the PEG were to be attached at or near critical active sites . . . .”); id. at col. 8 ll. 23–26 (“Site-directed attachment allows one to design modifications that avoid the regions required for bi- ological activity and thereby to maintain substantial FVIII activity.”). We acknowledge that the specification discloses embodiments of site-directed PEGylation at cysteine amino acids within the B-domain. See, e.g., id. at col. 5 ll. 46–57, col. 20 l. 56–col. 21 l. 5. But the specification does not state that this is the only embodiment or otherwise indicate that the invention is limited to site-directed PEGylation at cys- teine amino acids within the B-domain and, as noted above, the claim language is not so limited. On appeal, Baxalta contends that the specification dis- parages amine/lysine conjugation. In particular, it points to the “Background of the Invention,” which discloses that “[r]andom modification of FVIII by targeting primary Case: 19-2418 Document: 67 Page: 10 Filed: 03/01/2021

amines (N-terminus and lysines) with large polymers such as PEG” has “been attempted with varying degrees of suc- cess,” and that “[t]his random approach, however, is much more problematic” because FVIII has “hundreds of poten- tial PEGylation sites.” Appellants’ Br. 24 (quoting ’520 pa- tent col. 3 l. 50–col. 4 l. 1). To support its disparagement argument, Baxalta cites Indivior Inc. v. Dr. Reddy’s Labor- atories, S.A., 930 F.3d 1325 (Fed. Cir. 2019), and SciMed Life Systems, Inc. v. Advanced Cardiovascular Systems, Inc., 242 F.3d 1337 (Fed. Cir. 2001). Although Baxalta pre- sents a close question as to disparagement, the facts here differ from those of cases finding disclaimer based on the specification. In Indivior, the asserted patent claimed pharmaceuti- cal films with uniform distribution of the active ingredient. 930 F.3d at 1331. The claims identified “drying” as a pa- rameter that contributes to film uniformity. Id. at 1332. The specification taught that using conventional drying methods, which applied hot air to the top of the film, pro- duced nonuniform films. Id. The specification also dis- closed controlled drying methods that differed from conventional techniques. Id. The district court construed the drying limitation to mean “dried without solely employ- ing conventional convection air drying from the top” based on its conclusion that the patentee disclaimed drying films using solely conventional top air drying. Id. at 1336. We agreed with the district court’s construction, concluding that “the specification repeatedly disparages conventional top air drying because such drying does not produce uni- form films, the central object of the claimed invention.” Id. at 1337. The “specification also provide[d] examples that demonstrate[d] the failure of conventional drying methods to achieve uniformity.” Id. Similarly, we concluded in SciMed that the shared specification of the patents-in-suit disclaimed an embodi- ment from the scope of the asserted claims. The patents were drawn to balloon dilatation catheters containing two Case: 19-2418 Document: 67 Page: 11 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 11

passageways, or lumens. 242 F.3d at 1338–39. Two ar- rangements of the lumens were practiced in the art: the dual lumen configuration and the coaxial lumen configura- tion. Id. at 1339. The district court construed the asserted claims to be limited to catheters with coaxial lumens, and not to read on catheters with dual lumens. Id. We agreed with the district court’s construction because the specifica- tion unequivocally limited “all embodiments” of the claimed invention to a coaxial lumen configuration. Id. at 1340, 1343–44. In particular, the specification repeat- edly defined “the invention” as catheters having a coaxial lumen structure and discussed the disadvantages of prior- art catheters having a dual lumen configuration. Id. at 1342–44. We likewise determined that the patent owner in Gaus v. Conair Corp., 363 F.3d 1284 (Fed. Cir. 2004), unequivo- cally disclaimed a particular embodiment from its claims’ scope and therefore could not rely on the doctrine of equiv- alents to cover that embodiment. The patent at issue claimed a device, such as a hairdryer, comprising “an elec- trical operating unit and a pair of spaced-apart electrically exposed conductive probe networks.” Id. at 1287. The specification “made clear that it is essential to [the] inven- tion that the pair of probe networks be separate from the voltage-carrying components of the appliance, i.e., the ‘elec- trical operating system.’” Id. at 1291. By contrast, the specification “criticized prior art in which the protective de- vice relied on the fluid coming in contact with the voltage- carrying portions of the system.” Id. Thus, we concluded that the patent owner “disavowed coverage of devices in which the two components are not separate and in which the protective cut-off mechanism is not triggered until the water reaches the electrical operating system.” Id. Each of these cases presented a strong case of dis- claimer of a particular feature from the scope of a claim because the specification made clear that the invention did not include that feature. See Indivior, 930 F.3d at 1337; Case: 19-2418 Document: 67 Page: 12 Filed: 03/01/2021

SciMed, 242 F.3d at 1342–44; Gaus, 363 F.3d at 1291. The specification of the ’520 patent disparages random PEGyla- tion of FVIII, including random PEGylation targeting amines like lysines, but nowhere disparages non-random, site-directed amine/lysine PEGylation at the B-domain. Thus, we agree with the district court that Bayer did not unequivocally disclaim non-random amine/lysine PEGyla- tion in the specification. C We turn next to the prosecution history, which can be helpful in understanding the proper claim construction. We are not persuaded that the prosecution history includes a clear and unmistakable surrender of claims directed to non-random amine/lysine PEGylation. Nor are we per- suaded that Bayer’s statements during prosecution other- wise require a claim construction that would exclude such conjugates. During prosecution of the ’520 patent, the Examiner rejected the pending claims as inherently anticipated by a patent application filed by Nektar: Bossard. 1 On appeal to the Patent Trial and Appeal Board, Bayer argued that Bossard did not teach the claimed conjugates because Bossard’s “alleged showing of B-domain attachment is ran- dom PEGylation and does not ensure that attachment oc- curs at the B-domain.” J.A. 3507. Bayer also argued that its then-rejected claim 58, which ultimately issued as claim 1, “excludes compositions of randomly PEGylated factor VIII in which some conjugates have PEGylation at the B-domain but a large number do not.” Id. Bayer con- tinued by stating that: The Patent Office . . . relies on passages in Bossard that show random PEGylation at any one of nu- merous amino acid residues in the 2,332-amino

1 U.S. Patent. Pub. No. 2004/0235734. Case: 19-2418 Document: 67 Page: 13 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 13

acid protein that have the disclosed functionality. Specifically, the Patent Office asserts that Bossard et al. “teach many polymer attachment sites such as ‘lysine, cysteine and/or arginine’, ‘amine groups’ and/or ‘carboxyl groups’ for the site of attachment to factor VIII, wherein said many polymer attach- ment sites are presented in the B-domain . . . .” These types of sites are present in the B-domain, but they are also present in many other domains of factor VIII. This description does not ensure that attachment occurs at the B-domain, and does not disclose isolated conjugates with attachment at the B-domain. J.A. 3507–08 (second ellipsis in original) (citations omitted) (emphases added). In response to the Examiner’s Answer and specifically the Examiner’s arguments based on Bossard, Bayer argued in its reply brief to the Board that: Much of the Patent Office’s prior arguments relied upon possible conjugation at amines or carboxy sites, which are present not only in the B-domain but in other domains. Any conjugation with these reactive groups is random and does not ensure that attachment occurs at the B-domain. J.A. 4599 (emphases added). Baxalta’s disclaimer arguments based on Bayer’s state- ments to the Board have some merit. Ultimately, however, we agree with the district court’s view as to how a person of ordinary skill in the art would have understood the pros- ecution history as a whole. Specifically, we agree with the district court that when the “entire passage” from Bayer’s reply brief is “read together, it is clear that ‘[a]ny conjuga- tion’ refers to the conjugations allegedly disclosed by Bossard.” Summ. J. Op., 2018 WL 6727054, at *4. Specif- ically, though Bossard “may have taught conjugations at various sites on factor VIII, such as amines or carboxy sites,” Bayer “argued that the conjugations were not Case: 19-2418 Document: 67 Page: 14 Filed: 03/01/2021

targeted to particular sites” and, thus, “described ‘[a]ny conjugation’ in Bossard as random.” Id. As the district court reasoned, it “does not follow” that Bayer “defined all conjugations with amines or carboxy sites as random.” Id.; see also Claim Construction Op., 2018 WL 3212425, at *9 (“[Bayer] stated that conjugation at amines and carboxy sites cannot ensure PEGylation at the B-domain—not that conjugation cannot occur at amines and carboxy sites.”). We agree with the district court that Bayer distinguished the prior art on the ground that it did not teach non-ran- dom PEGylation at the B-domain, and that Bayer did not clearly and unmistakably disclaim all PEGylation at amines and carboxy sites. Additional sections of the prosecution history support our conclusion. First, the prosecution history reveals that Bossard randomly PEGylated at both lysines and cysteines on FVIII, which undermines Baxalta’s position that Bayer disclaimed the former but not the latter as sites for PEGylation. Second, the Board reversed the Examiner’s anticipation rejection based on a separate prior-art refer- ence, Rostin. 2 The Examiner argued that Rostin’s PEG at- tached to the lysine in Rostin’s shortened B-domain, and thus inherently anticipated Bayer’s claims. The Board dis- agreed, holding that Rostin’s PEG did not “necessarily at- tach[] to the single lysine located in the [shortened] B-domain” because there were many other lysines through- out FVIII. J.A. 3589–90. These passages suggest that the Examiner and the Board understood that lysine PEGyla- tion at the B-domain was within the scope of Bayer’s claims. Finally, Baxalta’s reliance on the applicants’ state- ments during prosecution of the ’520 patent’s European

2 Rostin, J., et al., B-Domain Deleted Recombinant Coagulation Factor VIII Modified with Monomethoxy Poly- ethylene Glycol, Bioconjugate Chem., 11:387–96 (2000). Case: 19-2418 Document: 67 Page: 15 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 15

counterpart 3 is misplaced. To overcome an anticipation re- jection, the applicants argued that the asserted prior-art reference “teaches random PEGylation at any lysine on the FVIII molecule and therefore does not anticipate the pres- ently claimed isolated conjugates in which essentially all the molecules have a polymer attached at the B-domain.” J.A. 7702 (emphasis added). A skilled artisan would not read this statement as a clear and unmistakable exclusion of non-random lysine PEGylation at the B-domain from the asserted ’520 patent claims’ scope. Rather, when read in context of the entire response, this statement appears to distinguish the process claimed in the European applica- tion from the process of random PEGylation using lysine in the prior art. Indeed, the applicants argued that “none of the references cited by the Examiner anticipate the present claims, as they disclose only random modification and not specific conjugation at the B-domain.” J.A. 7701–02. More- over, the district court correctly noted this court’s admoni- tion that “varying legal and procedural requirements for obtaining patent protection in foreign countries might ren- der consideration of certain types of representations inap- propriate for consideration in a claim construction analysis of a United States counterpart.” 4 Claim Construction Op.,

3 Eur. Pat. Appl. No. 11153287.4. 4 We also consider unpersuasive Baxalta’s reliance on Bayer’s statements during prosecution of U.S. Patent Application No. 13/748,983, a continuation of the applica- tion that issued as the ’520 patent. Similar to the ’520 pa- tent specification, Bayer told the Examiner that “[r]andom modification of Factor VIII such as by targeting primary amines with large polymers such as PEG had been at- tempted,” but that this “random approach, however, re- sults in a heterogenous product with polymer attachment possible at many potential sites.” J.A. 9093. The claims of the continuation application recited polymer attachment to Case: 19-2418 Document: 67 Page: 16 Filed: 03/01/2021

2018 WL 3212425, at *5 (quoting AIA Eng’g Ltd. v. Ma- gotteaux Int’l S/A, 657 F.3d 1264, 1279 (Fed. Cir. 2011)). Although Baxalta’s challenge presents close questions, we conclude that the claim language, specification, and prosecution history are more aligned with the district court’s construction. II We next consider Baxalta’s challenge to the district court’s refusal to define the term “random” in its construc- tion of “an isolated polypeptide conjugate” in claim 1. Bax- alta contends that the district court’s decision not to construe “random” improperly delegated to the jury the task of determining claim scope and warrants a new trial on infringement. We disagree. “When the parties present a fundamental dispute re- garding the scope of a claim term, it is the court’s duty to resolve it.” O2 Micro Int’l Ltd. v. Beyond Innovation Tech. Co., 521 F.3d 1351, 1362 (Fed. Cir. 2008). “This duty re- sides with the court because, of course, ‘the ultimate ques- tion of construction [is] a legal question.’” Eon Corp. IP Holdings v. Silver Spring Networks, Inc., 815 F.3d 1314, 1318 (Fed. Cir. 2016) (alteration in original) (quoting Teva Pharms. USA, Inc. v. Sandoz, Inc., 135 S. Ct. 831, 842

FVIII “at only the heavy chain,” not “at the B-domain.” J.A. 9091, 9094. Thus, the cited statement is not relevant to the construction of the ’520 patent’s different claim term. See Microsoft Corp. v. Multi-Tech Sys., Inc., 357 F.3d 1340, 1349 (Fed. Cir. 2004) (“[T]he prosecution history of one pa- tent is relevant to an understanding of the scope of a com- mon term in a second patent stemming from the same parent application.” (emphasis added)). Even if this state- ment were relevant, it still demonstrates, at most, a dis- claimer of random PEGylation. Case: 19-2418 Document: 67 Page: 17 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 17

(2015)). We have observed, however, that “a sound claim construction need not always purge every shred of ambigu- ity.” Id. (quoting Acumed LLC v. Stryker Corp., 483 F.3d 800, 806 (Fed. Cir. 2007)). We conclude that the district court did not violate its duty to interpret the claims in declining to provide a de- tailed interpretation of the term “random” in its claim con- struction instructions to the jury. The district court construed the claim term “an isolated polypeptide conju- gate” to mean “a polypeptide conjugate where conjugation was not random.” The district court resolved the parties’ controversies as to the meaning of “random.” Specifically, the district court addressed and rejected Baxalta’s two ar- guments: (1) that “random” conjugation means any conju- gation at amines or carboxy sites; and (2) that “random” conjugation means all heterogenous conjugation. In construing the claim term “an isolated polypeptide conjugate,” the district court rejected Baxalta’s argument that the specification disclaimed “heterogeneous conju- gates” by disparaging such conjugates. Claim Construction Op., 2018 WL 3212425, at *3. The district court observed that the patent “never distinguishes B-domain pegylated conjugate products” having “any degree” or “even the slightest degree” of heterogeneity. Id. The district court explained that, instead, the specification “disparage[s] products with a high degree of heterogeneity.” Id.; see id. at *4 (concluding that the specification disparages random conjugation in stating that such conjugation is “problem- atic” for FVIII because it yields “enormous heterogeneity in its product profile” (quoting ’520 patent col. 8 ll. 28–30)). The district court reiterated its conclusion when it de- nied Baxalta’s motion for summary judgment of nonin- fringement. Specifically, the district court stated that it had “determined during claim construction that non-ran- dom conjugation does not require total homogeneity,” and that it had disagreed with Baxalta that Bayer “disclaimed Case: 19-2418 Document: 67 Page: 18 Filed: 03/01/2021

all heterogeneous conjugation.” Summ. J. Op., 2018 WL 6727054, at *4. The district court made clear that it was referring to heterogeneity among FVIII conjugates in a product and heterogeneity within each individual FVIII conjugate. Id. In other words, according to the district court, non-random conjugation neither required that each FVIII protein in a product such as Adynovate® be PEGylated in the same places (homogeneity among conju- gates in the product) nor required that every PEG on each FVIII protein be in the B-domain (homogeneity within each conjugate). Thus, it rejected Baxalta’s noninfringement po- sition that “the fact that some pegylation occurs outside the B-domain is sufficient to show that Adynovate is made by random conjugation.” Id. The district court further reiterated this conclusion in its orders denying Baxalta’s pre-trial motions. First, the district court denied Baxalta’s motion in limine seeking further construction of “random” to mean “any conjugation at amines or carboxy sites” or, alternatively, that PEG “may attach at any available site of reaction throughout the protein.” J.A. 29244. The district court found it unnec- essary “to further construe [its] construction by explicitly defining ‘random,’” given its earlier holdings at claim con- struction and summary judgment. Mot. in Lim. Order, 2019 WL 10890386, at *1. Furthermore, in denying Bax- alta’s motion to exclude Bayer’s damages expert, the dis- trict court reiterated once again that “infringement does not require complete homogeneity—that is, some PEGs may attach outside the B-domain.” Daubert Order, 2019 WL 330149, at *9. Having resolved this issue, the district court then left it to the parties to present evidence to the jury as to whether or not Adynovate® is the result of random or non-random conjugation. That the district court resolved this issue is further supported by the fact that Baxalta did not ask the district court to give additional jury instructions on the meaning of “random.” J.A. 1595–96 (Tr. 1374:18–1375:25). In fact, Case: 19-2418 Document: 67 Page: 19 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 19

Baxalta “object[ed] to anything going to the jury about claim construction other than [the district court’s] claim construction order.” J.A. 1595 (Tr. 1374:19–21). Moreover, Baxalta’s argument as to the district court’s use of “random” is another attempt to arrive at the same construction Baxalta sought for the claim term “at the B-domain” because it attempts to equate “random” with amine/lysine PEGylation. For the reasons discussed above regarding the district court’s construction of “at the B-do- main,” we again reject Baxalta’s argument that amine/ly- sine PEGylation is outside the scope of the asserted claims. III We next turn to Baxalta’s challenge to the jury’s in- fringement finding, which is a question of fact that we re- view for substantial evidence. Eon Corp., 815 F.3d at 1318. We hold that substantial evidence supports the jury’s in- fringement verdict and, therefore, the district court did not err in denying Baxalta’s motion for JMOL of noninfringe- ment. This court reviews the denial of a motion for JMOL un- der regional circuit law, here, Third Circuit law. Idenix Pharms. LLC v. Gilead Scis. Inc., 941 F.3d 1149, 1153 (Fed. Cir. 2019) (citing Tr. of Boston Univ. v. Everlight El- ecs. Co., 896 F.3d 1357, 1361 (Fed. Cir. 2018)). Applying Third Circuit law, we “exercise plenary review over a dis- trict court’s rulings on motions for JMOL, applying the same standard as the district court.” Id. (quoting Agrizap, Inc. v. Woodstream Corp., 520 F.3d 1337, 1341–42 (Fed. Cir. 2008)); see Gagliardo v. Connaught Labs., Inc., 311 F.3d 565, 568 (3d Cir. 2002). “A grant of JMOL is ap- propriate ‘where a party has been fully heard on an issue during a jury trial and the court finds that a reasonable jury would not have had a legally sufficient evidentiary ba- sis to find for the party on that issue.’” Idenix, 941 F.3d at 1153–54 (quoting Agrizap, 520 F.3d at 1342); see also Fed. R. Civ. P. 50(a). The court should grant JMOL Case: 19-2418 Document: 67 Page: 20 Filed: 03/01/2021

“sparingly” and “only if, viewing the evidence in the light most favorable to the nonmovant and giving it the ad- vantage of every fair and reasonable inference, there is in- sufficient evidence from which a jury reasonably could find liability.” Marra v. Phila. Hous. Auth., 497 F.3d 286, 300 (3d Cir. 2007) (citations omitted). Bayer presented documentary evidence and expert wit- ness testimony demonstrating that Adynovate® meets the limitations of claim 1 of the ’520 patent as construed by the district court. Bayer’s biochemistry and protein chemistry expert, Dr. Ploegh, testified about Baxalta’s letter to the U.S. Food and Drug Administration seeking support for Orphan Drug designation for its recombinant FVIII prod- uct BAX 855, which it later named Adynovate®. J.A. 654–55 (Tr. 433:9–434:25). In its submission, Baxalta represented to the FDA that “BAX 855 is a modified recom- binant FVIII protein created by the pre-translational inclu- sion of the B-domain, and controlled, targeted chemical addition of 20 [kilodalton] PEG conjugates to this FVIII B-domain.” J.A. 36942 (emphases added). Baxalta’s sub- mission also stated that “evidence suggests that the B-do- main is more surface-exposed than the other domains, and becomes accessible to the PEG-reagent under the specific reaction conditions utilized for the controlled chemical PEGylation of the FVIII molecule.” J.A. 36944 (emphasis added). Dr. Ploegh explained that Baxalta’s statements show that it “exert[s] control [over] how [it] chemically modif[ies] this protein and the net result is the attachment of 20 kilodalton preferable to the B-domain of Factor VIII,” J.A. 655 (Tr. 434:7–11), and that Baxalta made specific re- action condition choices to “allow[] [it] to target the PEGylation reaction to the B-domain” of FVIII, id. (Tr. 434:20–25). Dr. Ploegh also walked the jury through the reaction condition choices Baxalta made to achieve con- trolled PEGylation at the B-domain. J.A. 656–61 (Tr. 435:1–440:25); J.A. 670 (Tr. 449:13–22). Case: 19-2418 Document: 67 Page: 21 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 21

Bayer’s expert, Dr. Ravetch, a professor of molecular genetics and immunology and expert on protein structure, function, and modification, also testified about Baxalta’s letter to the FDA. In his view, Baxalta’s representations showed that “the reaction was controlled in such a way to result in the targeted addition of th[e] PEG to the B-do- main and that is a nonrandom process.” J.A. 756 (Tr. 535:2–4). Dr. Ravetch opined that random PEGyla- tions are neither “controlled” nor “targeted,” and that based on its representations to the FDA, “Baxalta use[d] conjugation that was not random.” Id. (Tr. 535:6–12). Dr. Ravetch testified that based on the FDA documents, testimony from Baxalta and Nektar scientists, and Dr. Ploegh’s testimony, in his opinion Adynovate® met the “isolated polypeptide conjugation element” because Bax- alta “uses a non-random conjugation method to generate Adynovate.” J.A. 760 (Tr. 539:7–11). Dr. Ravetch further opined that Adynovate® contains PEG attached at the B-domain based on test results re- ported in Baxalta’s Biologics License Application for Ady- novate®, which indicated to him that “there’s a consistent PEGylation in a particular region predominantly on B-do- main” and “B-domain PEGylation was observed as a pre- dominant aspect of the product.” J.A. 774–75 (Tr. 553:4– 554:3); see also J.A. 38063 (“[T]he consistency of the region- specific PEGylation predominantly on the B-domain was confirmed.”). Dr. Ravetch also discussed additional FDA submissions in which Baxalta provided the FDA with data showing that 40 out of 55 (or 73%) of the PEG attachment sites are in the B-domain, which supported his opinion that Adynovate® is formed by a non-random process in which PEGylation is region specific and predominant at the B-do- main. J.A. 775–76 (Tr. 554:12–555:13 (discussing J.A. 37097)). In one such submission, a technical report identifying the PEGylation sites of BAX 855, Baxalta stated: “PEGylation sites are distributed over the whole rFVIII molecule but clustered on the B-domain. . . . The Case: 19-2418 Document: 67 Page: 22 Filed: 03/01/2021

basis for preferential PEGylation of the B-domain is the higher density of lysine residues in the B-domain and in addition the higher flexibility and surface exposure. By preferential PEGylation the B-domain carries now a new function in BAX 855 – the prolongation of action of rFVIII.” J.A. 37077 (emphases added). We conclude that this is substantial evidence to sup- port the jury’s finding that Adynovate® meets the claim limitations. Baxalta’s expert, Dr. Zalipsky, provided con- trary testimony to Drs. Ploegh and Ravetch, but the jury was in the best position to determine whether it found Dr. Zalipsky or Bayer’s experts more persuasive. See Mo- bileMedia Ideas LLC v. Apple Inc., 780 F.3d 1159, 1168 (Fed. Cir. 2015) (“[W]hen there is conflicting testimony at trial, and the evidence overall does not make only one find- ing on the point reasonable, the jury is permitted to make credibility determinations and believe the witness it con- siders more trustworthy.”). We discern no error in the dis- trict court’s conclusion that substantial evidence supported the jury’s infringement verdict. IV We turn next to Baxalta’s argument that no reasonable jury could find that the ’520 patent enables the full scope of the claims because it does not enable non-random lysine PEGylation. “Whether a claim satisfies the enablement re- quirement is a question of law that we review de novo. However, ‘in the context of a jury trial, we review the factual underpinnings of enablement for substantial evi- dence.’” Idenix, 941 F.3d at 1154 (quoting Everlight, 896 F.3d at 1361)). On the record before us, we conclude that substantial evidence supports the jury’s verdict that Baxalta failed to prove by clear and convincing evidence that the asserted claims are invalid for lack of enablement. “Enablement requires that ‘the specification teach those in the art to make and use the invention without un- due experimentation.’” Id. (quoting In re Wands, 858 F.2d Case: 19-2418 Document: 67 Page: 23 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 23

731, 737 (Fed. Cir. 1988)). “A claim is not enabled when, ‘at the effective filing date of the patent, one of ordinary skill in the art could not practice their full scope without undue experimentation.’” Id. (quoting Wyeth & Cordis Corp. v. Abbott Labs., 720 F.3d 1380, 1384 (Fed. Cir. 2013)). Factors for assessing whether a disclosure would require undue experimentation include: “(1) the quantity of exper- imentation necessary, (2) the amount of direction or guid- ance presented, (3) the presence or absence of working examples, (4) the nature of the invention, (5) the state of the prior art, (6) the relative skill of those in the art, (7) the predictability or unpredictability of the art, and (8) the breadth of the claims.” Wands, 858 F.2d at 737. The specification of the ’520 patent includes detailed instructions as to the reaction conditions required to prac- tice the claimed invention using cysteine PEGylation, and includes a working example for non-random cysteine PEGylation at the B-domain. See ’520 patent col. 20 l. 56–col. 21 l. 5. Additionally, Bayer presented evidence to the jury bridging the gap between the patent’s disclosures about cysteine PEGylation on the one hand and non-ran- dom PEGylation, including non-random lysine PEGyla- tion, on the other. For example, ’520 patent inventor Dr. Murphy testified that at the time of the invention, he did not believe that cysteine conjugation would be the only way to achieve the claimed B-domain PEGylation of FVIII. J.A. 574 (Tr. 353:5–8). In addition, multiple witnesses testified that random lysine PEGylation was known at the time of the invention. E.g., J.A. 1248 (Tr. 1027:8–16) (’520 patent inventor Dr. Pan testifying that nonspecific “lysine PEGylation has been around for 20, 30 years” and was a “very old technol- ogy”); J.A. 1563 (Tr. 1342:20–22) (Dr. Ravetch testifying that lysine PEGylation had been known in the art “20, 30 years prior to these patents[,] [s]o a long time”); J.A. 1456 (Tr. 1235:21–25) (Dr. Zalipsky testifying that random lysine PEGylation had been performed by 2005). Case: 19-2418 Document: 67 Page: 24 Filed: 03/01/2021

Moreover, Dr. Russell, Bayer’s expert in protein modifica- tion, including PEGylation, testified that lysine PEGyla- tion can be either random or non-random, and that skilled artisans would have known that various factors, including reaction conditions and process steps, can be manipulated to achieve either random or non-random PEGylation. 5 See J.A. 1527–28 (Tr. 1306:23–1307:10). Consistent with this, the jury was also entitled to rely on Dr. Ravetch’s testi- mony that, at the time of the invention, “protein chemistry [wa]s a well-defined field” going back 150 years and that the “sophistication and knowledge” in this field regarding proteins and amino acids had “accumulated over time.” J.A. 1564 (Tr. 1343:4–18). Similarly, in response to Dr. Zalipsky’s testimony that ordinarily skilled artisans would not have known from reading the specification how to non-randomly PEGylate FVIII with any amino acid other than cysteine, Dr. Ravetch opined that skilled arti- sans would have understood that “some amino acids are more amenable than other amino acids for modification” based on the amino acids’ known reactivities under differ- ent conditions. J.A. 1564–65 (Tr. 1343:4–1344:1). In short, Dr. Ravetch explained to the jury that the advanced knowledge concerning amino acids—along with the fact that only a certain subset could be used for PEGylation— would have aided skilled artisans in applying the teachings in the patent to prepare a non-random lysine PEGylated conjugate without undue experimentation. Baxalta, in

5 Though Dr. Russell referenced Baxalta’s Ady- novate® product in his discussion of random and non-ran- dom lysine PEGylation, his testimony, when read in context, more generally concerned what was known in the art at the time of the invention based on his experience in protein modification. See J.A. 1528 (Tr. 1307:3–10) (re- sponding to the question: “what determines whether you have random PEGylation as opposed to non-random PEGylation?”). Case: 19-2418 Document: 67 Page: 25 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 25

turn, chose not to cross-examine Dr. Ravetch. J.A. 1565 (Tr. 1344:21). Based on the record before us, we conclude that sub- stantial evidence supports the jury’s verdict, and the jury was not obliged to credit Baxalta’s expert Dr. Zalipsky over Bayer’s contrary evidence. Bayer presented substantial ev- idence from which a reasonable juror could find that the specification’s disclosure of instructions as to the reaction conditions required to practice the claimed invention using cysteine PEGylation were sufficient to enable not only non- random cysteine PEGylation at the B-domain, but also non-random lysine PEGylation at the B-domain. Baxalta contends that the district court legally erred in resorting to the knowledge of a person of ordinary skill as a “substitute for the specification’s lack of a basic enabling disclosure.” Appellants’ Br. 59. Baxalta focuses on the ab- sence of working examples of lysine PEGylation in the specification. As the district court correctly recognized, however, that the “novel aspect” of the asserted claims is non-random PEGylation at the B-domain “does not mean the specification must disclose an embodiment for non-ran- dom pegylation at each amino acid in the B-domain.” JMOL Op., 407 F. Supp. 3d at 472. As this court has re- peatedly made clear, the specification need not include a working example of every possible embodiment to enable the full scope of the claims. See Alcon Rsch. Ltd. v. Barr Labs., Inc., 745 F.3d 1180, 1189–90 (Fed. Cir. 2014); Amgen Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d 1313, 1336–37 (Fed. Cir. 2003). We are thus not persuaded that the district court erred in considering the knowledge of an ordinarily skilled artisan to determine whether undue ex- perimentation would have been required to practice the in- vention with lysine. Indeed, whether a patent is enabled— or requires undue experimentation—are questions that must be viewed “from the perspective of one of ordinary skill in the art.” Falko-Gunter Falkner v. Inglis, 448 F.3d 1357, 1365 (Fed. Cir. 2006) (concluding that the Board of Case: 19-2418 Document: 67 Page: 26 Filed: 03/01/2021

Patent Appeals and Interferences did not err in concluding that claims directed to a poxvirus-based vaccine were ena- bled by the specification’s detailed teachings of a herpesvi- rus-based vaccine and the knowledge of a skilled artisan regarding the two viruses); see also McRO, Inc. v. Bandai Namco Games Am. Inc., 959 F.3d 1091, 1102 (Fed. Cir. 2020) (“An ‘artisan’s knowledge of the prior art and routine experimentation can often fill gaps, interpolate between embodiments, and perhaps even extrapolate beyond the disclosed embodiments, depending upon the predictability of the art,’ and a ‘patent need not teach, and preferably omits, what is well known in the art.’” (first quoting AK Steel Corp. v. Sollac & Ugine, 344 F.3d 1234, 1244 (Fed. Cir. 2003); and then quoting Spectra-Physics, Inc. v. Coherent, Inc., 827 F.2d 1524, 1534 (Fed. Cir. 1987))). Baxalta also contends that it was legal error for the dis- trict court to rely on testimony by Bayer’s experts concern- ing post-priority lysine PEGylation used in Baxalta’s Adynovate®’s manufacture. See Appellants’ Br. 62–63. We agree that post-priority knowledge about the “reaction con- ditions” for the accused product cannot support the jury verdict of enablement. But any error was harmless be- cause the district court also cited other testimony, includ- ing Dr. Ravetch’s testimony regarding the depth of knowledge in the art about protein chemistry and PEGyla- tion at the time of the invention, J.A. 1563–65 (Tr. 1342:3–1344:1), and Dr. Russell’s testimony concern- ing the knowledge regarding reaction conditions and pro- cess steps, among other factors, that determine whether PEGylation is random or non-random, J.A. 1527–28 (Tr. 1306:23–1307:10). Furthermore, contrary to Baxalta’s assertion, the specification’s purported “teaching away” from random PEGylation at amines such as lysines does not render the asserted claims non-enabled as to non-ran- dom lysine PEGylation. For all these reasons, we conclude that the district court did not err in its legal analysis. In addition, we Case: 19-2418 Document: 67 Page: 27 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 27

cannot say that a reasonable jury could not find the claims enabled. Accordingly, the district court correctly denied Baxalta’s motion for JMOL of invalidity for lack of enable- ment. V We next turn to the issue of damages. Baxalta con- tends that the district court correctly held that Bayer’s damages expert’s (Dr. Addanki’s) 50-50 split royalty rate was not properly tied to the facts of the case. According to Baxalta, however, the district court thereafter committed several errors. We address each argument in turn. First, Baxalta argues that the district court errone- ously permitted Bayer to rely on a flawed and speculative methodology—namely, asking the jury to pick a rate be- tween the range of feasible rates presented by Dr. Addanki as the reasonable rate. We review a district court’s deci- sion to admit expert testimony for abuse of discretion. Amgen Inc. v. Hospira, Inc., 944 F.3d 1327, 1341 (Fed. Cir. 2019) (citing Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997)); see also SmithKline Diagnostics, Inc. v. Helena Labs. Corp., 926 F.2d 1161, 1164 (Fed. Cir. 1991) (noting that “subsidiary decisions underlying a damage[s] theory,” such as “the methodology for arriving at a reasonable roy- alty,” are reviewed for abuse of discretion (citations omit- ted)). The district court properly exercised its discretion in allowing Bayer to ask the jury to select a rate between the range presented. While an expert must use reliable meth- odology for determining the range of possible hypothetical negotiation royalty rates, we are aware of no precedent that requires an expert to provide a single proposed royalty rate. As an initial matter, a jury is “entitled to choose a damages award within the amounts advocated by the op- posing parties” and is “not bound to accept a rate proffered by one party’s expert but rather may choose an intermedi- ate royalty rate.” Powell v. Home Depot U.S.A., Inc., Case: 19-2418 Document: 67 Page: 28 Filed: 03/01/2021

663 F.3d 1221, 1241 (Fed. Cir. 2011) (quoting Spectralytics, Inc. v. Cordis Corp., 649 F.3d 1336, 1347 (Fed. Cir. 2011)). In addition, we have previously held that a jury’s damages award that fell within the range suggested by the pa- tentee’s damages expert was supported by substantial evi- dence. See Rembrandt Wireless Techs., LP v. Samsung Elecs. Co., 853 F.3d 1370, 1382 (Fed. Cir. 2017). In grant- ing-in-part Baxalta’s Daubert motion, the district court de- termined that Dr. Addanki’s expert report included “substantial analyses to determine ‘end points of the bar- gaining range for the hypothetical negotiation,’” including “deriv[ing] a maximum royalty rate from the incremental profits Baxalta would expect to earn from Adynovate, and a minimum royalty rate from the profits Bayer would ex- pect to lose by granting a license to Baxalta.” Daubert Or- der, 2019 WL 330149, at *7. Dr. Addanki’s testimony further demonstrates that he considered and discussed the appropriate Georgia-Pacific factors at length in determin- ing the range of reasonable royalties, see J.A. 962–1020, and Bayer in its closing statement explained to the jury that its damages award should fall within that range, J.A. 1728 (Tr. 1507:16–25). Moreover, the district court permitted Baxalta to cross-examine Dr. Addanki on his end points and range and to present the testimony of its own damages expert. Under these circumstances, we conclude that the district court did not err in allowing the jury to hear Dr. Addanki’s testimony regarding a range of possible hypothetical reasonable royalty rates instead of a single proposed royalty rate. Second, Baxalta argues that the district court should have excluded Dr. Addanki’s testimony that the entire range of rates was reasonable because such a conclusion was not in his expert report. The district court rejected Baxalta’s argument that Dr. Addanki presented new opin- ions at trial in violation of Rules 26(a)(2)(B) and 37(c) of the Federal Rules of Civil Procedure. Such evidentiary rulings are also reviewed for an abuse of discretion. See Siemens Case: 19-2418 Document: 67 Page: 29 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 29

Med. Sols. USA, Inc. v. Saint-Gobain Ceramics & Plastics, Inc., 637 F.3d 1269, 1286–87 (Fed. Cir. 2011). We agree with the district court that Dr. Addanki’s opinions in his expert reports are consistent with his trial testimony about a range of royalty rates and were not limited to the single rate that the district court had excluded. For instance, Dr. Addanki had opined in his reply expert report that “as a matter of economics, any of these royalty rates [between 5.1% and 42.4%] would be a feasible outcome to the negoti- ation” and the “reasonable royalty is the likely outcome within that range.” J.A. 21887 (emphasis added). The dis- trict court exercised its sound discretion in permitting Dr. Addanki’s testimony. J.A. 233 (Tr. 12:6–10). Third, Baxalta appears to challenge the jury’s determi- nation of the amount of damages, which is an issue of fact that we review for substantial evidence. Amgen, 944 F.3d at 1341 (citing Lucent Techs., Inc. v. Gateway, Inc., 580 F.3d 1301, 1310 (Fed. Cir. 2009)). “A jury’s damages award ‘must be upheld unless the amount is grossly exces- sive or monstrous, clearly not supported by the evidence, or based only on speculation or guesswork.’” Id. (quoting Lucent, 580 F.3d at 1310). We agree with the district court that substantial evidence supports the jury’s 17.78% roy- alty rate, which was “within the range encompassed by the record as a whole.” JMOL Op., 407 F. Supp. 3d at 480 (quoting Unisplay, S.A. v. Am. Elec. Sign Co., 69 F.3d 512, 519 (Fed. Cir. 1995)). Dr. Addanki provided ample guid- ance to the jury to help it determine the royalty rate. E.g., J.A. 962–67 (explaining the Georgia-Pacific factors, the concept of the hypothetical negotiation, and determining maximum and minimum points in a range), J.A. 967–68, 970–1005, 1008–13 (explaining how he applied the hypo- thetical negotiation analysis and determined the specific end points of 5.1% and 42.4% in this case), J.A. 1014–20 (discussing each of the Georgia-Pacific factors and explain- ing whether and how each one applied in his analysis). Case: 19-2418 Document: 67 Page: 30 Filed: 03/01/2021

Finally, we note that Baxalta’s criticisms about the jury’s royalty rate are more appropriate for cross-examina- tion of the expert. Indeed, Baxalta cross-examined Dr. Ad- danki about his methodology and presented testimony of its own damages expert, Dr. Rausser, who opined that a reasonable-royalty rate would be 1%. J.A. 1485–93 (Tr. 1264:22–1272:14). While Baxalta complains that it could not cross-examine Dr. Addanki on his prior selection of a single, mid-point rate, Baxalta still could have—and in fact did—cross-examine Dr. Addanki on why the upper end of his proposed range would not result in an absurd rate in an effort to undermine Dr. Addanki’s credibility. A party need not present expert testimony on damages or, as a cor- ollary, on every aspect of damages, such as a single royalty rate. See Dow Chem. Co. v. Mee Indus., Inc., 341 F.3d 1370, 1382 (Fed. Cir. 2003) (“[S]ection 284 is clear that expert testimony is not necessary to the award of damages, but rather ‘may [be] receive[d] . . . as an aid.’” (annotations in original) (quoting 35 U.S.C. § 284)). A party runs the risk, however, of loss to its expert’s credibility on cross-examina- tion if the expert does not identify a single royalty rate. But “[w]here the methodology is sound and the evidence relied upon is sufficiently related to the case, disputes over the expert’s credibility or over the accuracy of the underlying facts are for the jury.” Summit 6, LLC v. Samsung Elecs. Co., 802 F.3d 1283, 1299 (Fed. Cir. 2015) (citing i4i Ltd. P’ship v. Microsoft Corp., 598 F.3d 831, 852 (Fed. Cir. 2010)). Here, Baxalta cross-examined Dr. Addanki and, ul- timately, the jury evaluated his opinions and adopted a rate within his proposed range. The district court did not err in denying Baxalta’s motion for JMOL or a new trial on damages. VI We turn to the final issue raised in Baxalta’s appeal— whether the district court violated Baxalta’s Seventh Amendment right to a jury trial by amending its judgment under Rule 59 to award Bayer pre-verdict supplemental Case: 19-2418 Document: 67 Page: 31 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 31

damages. We disagree with Baxalta that the Seventh Amendment right to a jury trial attaches to the award of pre-verdict supplemental damages in this case. We also conclude that the district court acted within its discretion in awarding supplemental damages. This court reviews “the constitutional question of whether a party is entitled to a jury trial” de novo. TCL Commc’n Tech. Holdings Ltd. v. Telefonaktiebolaget LM Ericsson, 943 F.3d 1360, 1371 (Fed. Cir. 2019) (quoting Te- gal Corp. v. Tokyo Electron Am., Inc., 257 F.3d 1331, 1339 (Fed. Cir. 2001)). The Seventh Amendment provides that, “[i]n Suits at common law, where the value in controversy shall exceed twenty dollars, the right of trial by jury shall be preserved . . . .” U.S. CONST. amend. VII. Under the Patent Act, “[u]pon finding for the claimant the court shall award the claimant damages adequate to compensate for the infringement,” and “[w]hen the damages are not found by a jury, the court shall assess them.” 35 U.S.C. § 284. “The methodology of assessing and computing damages un- der 35 U.S.C. § 284 is within the sound discretion of the district court.” TWM Mfg. Co. v. Dura Corp., 789 F.2d 895, 898 (Fed. Cir. 1986) (citations omitted). The jury was asked to calculate the royalty rate, roy- alty base, and total damages for the time period between June 14, 2016 through November 30, 2018. The jury ren- dered its verdict on February 4, 2019, and the district court entered judgment on February 8, 2019. Bayer thereafter sought pre-verdict supplemental damages for the period from December 1, 2018 through February 8, 2019. The dis- trict court granted Bayer’s motion and determined that supplemental damages should be based on the actual sales data for that period—which Baxalta had produced after trial—and the jury’s 17.78% royalty rate. Rule 59 Op., 2019 WL 4016235, at *6. The district court rejected Bax- alta’s contentions that Bayer should have used the pro- jected sales data Baxalta had produced before trial and Case: 19-2418 Document: 67 Page: 32 Filed: 03/01/2021

that awarding supplemental damages would violate Bax- alta’s Seventh Amendment right to a jury trial. We have on occasion reviewed decisions by district courts awarding pre-verdict supplemental damages with- out directly addressing the constitutionality of such awards. See, e.g., Finjan, Inc. v. Secure Computing Corp., 626 F.3d 1197 (Fed. Cir. 2010) (addressing a challenge to the district court’s refusal to award additional supple- mental damages for post-verdict sales up until the date of its injunction while noting that pre-verdict supplemental damages had been awarded); Godo Kaisha IP Bridge 1 v. TCL Commc’n Tech. Holdings Ltd., 967 F.3d 1380, 1382 (Fed. Cir. 2020) (affirming the district court’s grant of pa- tent owner’s request for “supplemental damages and an ac- counting of infringing sales of all adjudicated products through the date of the verdict”). Here, Baxalta contends that the district court’s award of pre-verdict supplemental damages “constitutes an impermissible additur.” Appel- lants’ Br. 82. Baxalta cites Dimick v. Schiedt, 293 U.S. 474 (1935), for the proposition that “a court has no power to consider new facts to award additional damages that are not part of the verdict, as it would invade the constitutional right to a trial by jury.” Appellants’ Br. 81–82. In Dimick, the plaintiff moved the district court for a new trial, argu- ing that the jury’s damages award was inadequate. 293 U.S. at 475. The district court entered a “conditional order” for a new trial whereby it would grant the plaintiff’s request “unless [the defendant] would consent to an in- crease of the damages to the sum of $1,500.” Id. at 475–76. The Supreme Court held that this practice of additur vio- lates the Seventh Amendment. Id. at 486–87. Calculating pre-verdict supplemental damages in this case merely required applying the jury’s royalty rate to the undisputed actual infringing sales base. That royalty rate was based on a hypothetical negotiation as of the date of first infringement (June 14, 2016), and was applied by the jury to the actual infringing sales base from the date of first Case: 19-2418 Document: 67 Page: 33 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 33

infringement through November 30, 2018. Under these circumstances, we are not persuaded that the district court’s award constitutes an impermissible additur or an otherwise “bald addition of something which in no sense can be said to be included in the verdict.” Id. at 486. Nor are we convinced that our conclusion conflicts with WEGCO, Inc. v. Griffin Services, Inc., 19 F. App’x 68, 73–74 (4th Cir. 2001), where the district court’s erroneous supple- mental damages award was based on its “own findings re- garding what monies were owed” to the plaintiff and was “not based on any findings made by the jury.” Similarly inapposite is Ohio-Sealy Mattress Manufacturing Co. v. Sealy, Inc., 669 F.2d 490 (7th Cir. 1982), which con- cerned a plaintiff’s ability to obtain future damages in an antitrust context, and which was silent as to the Seventh Amendment. Moreover, the district court acted within its discretion in awarding supplemental damages. In the district court’s view, “it was reasonable for Bayer to limit its damages at trial to the period for which it had actual sales data.” Rule 59 Op., 2019 WL 4016235, at *6. The district court also observed that the “projected sales were substantially higher than the actual sales” and, thus, Baxalta would have faced higher damages if Bayer had used projected sales. Id. Baxalta cites TransPerfect Global, Inc. v. Mo- tionPoint Corp., No. 10-cv-2590, 2014 WL 6068384 (N.D. Cal. Nov. 13, 2014), to support its argument that “fail[ing] to present the jury with extrapolated damages de- spite having the opportunity to do so” warrants denying pre-verdict supplemental damages. Appellants’ Br. 83. In TransPerfect, the district court denied the patentee’s mo- tion to amend the judgment to award pre-verdict supple- mental damages, noting that the patentee “could have sought to ask the jury to extrapolate post-2012 damages from the pre-2012 financial records and analysis.” 2014 WL 6068384, at *4. Unlike this case, it was “not even clear” to the district court “that the jury did not award Case: 19-2418 Document: 67 Page: 34 Filed: 03/01/2021

damages for the full [pre-verdict] period,” and unlike Bayer, the patentee had not sought updated financial in- formation during discovery. Id. Based on the facts of this case, we conclude that the district court did not abuse its discretion in awarding pre-verdict supplemental damages. VII Lastly, we address Bayer’s challenge in its cross-appeal of the district court’s judgment as a matter of law that Bax- alta did not willfully infringe the asserted claims. In Bayer’s view, it presented sufficient evidence of willfulness at trial to send the question to the jury. We disagree. Willful infringement is a question of fact. Polara Eng’g Inc. v. Campbell Co., 894 F.3d 1339, 1353 (Fed. Cir. 2018). To establish willfulness, the patentee must show the ac- cused infringer had a specific intent to infringe at the time of the challenged conduct. See Halo Elecs., Inc. v. Pulse Elecs., Inc., 136 S. Ct. 1923, 1933 (2016). As the Supreme Court stated in Halo, “[t]he sort of conduct warranting en- hanced damages has been variously described in our cases as willful, wanton, malicious, bad-faith, deliberate, con- sciously wrongful, flagrant, or—indeed—characteristic of a pirate.” Id. at 1932. A patentee needs to show by a pre- ponderance of the evidence the facts that support a finding of willfulness. Id. at 1934. In granting Baxalta’s motion for JMOL of no willful in- fringement, the district court concluded that Bayer failed to present sufficient evidence of the “state of mind” neces- sary for a finding of willfulness. J.A. 1355–56 (Tr. 1134:24–1135:3). According to the district court, there was no dispute that Baxalta was “aware of the ’520 patent,” and that Bayer merely “assume[d] that [Baxalta] knew Adynovate infringed because it involved pegylation at the B-domain of factor VIII.” Rule 59 Op., 2019 WL 4016235, at *9. The district court concluded that this was not “enough for a reasonable jury to find that infringement was Case: 19-2418 Document: 67 Page: 35 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 35

‘either known or so obvious it should have been known.’” Id. (quoting Halo, 136 S. Ct. at 1930). On appeal, Bayer identifies evidence that purportedly satisfies the state of mind requirement for willfulness. Specifically, Bayer presented testimony of Baxalta and Nektar witnesses concerning their awareness of the patent application that issued as the ’520 patent. J.A. 634–38. Additionally, Bayer presented evidence of Baxalta’s repre- sentations to the FDA that Adynovate®’s activity was due to controlled, targeted PEGylation at the B-domain. J.A. 654–55 (Tr. 433:9–434:25); J.A. 718–20 (Tr. 497:7– 499:5); J.A. 755–60 (Tr. 534:1–539:11); J.A. 774–77 (Tr. 553:12–556:18); J.A. 897–98 (Tr. 676:14–677:17). Bayer further contends that the jury heard evidence show- ing that Baxalta’s internal documents described Ady- novate®’s PEGylation process as controlled and consistent. J.A. 1238–43 (Tr. 1017:9–1022:21); J.A. 38368–90. Bayer also presented testimony by Baxalta and Nektar witnesses that the companies’ initial approach with FVIII conjugates was random PEGylation, and that they switched to tar- geted, B-domain PEGylation. J.A. 713–16 (Tr. 492:8– 495:19); J.A. 722–23 (Tr. 501:1–502:13); J.A. 1121–23 (Tr. 900:12–902:19). In Bayer’s view, this evidence shows that “Baxalta knew from prior dealings that random pegylation had failed, found out about Bayer’s B-domain pegylation work that underpins the ’520 patent, and con- sciously redirected its own research to B-domain pegyla- tion after learning about Bayer’s invention.” 6 Cross- Appellant’s Reply 4.

6 As additional support for its willfulness case, Bayer identifies a poster regarding FVIII PEGylation that Bayer had purportedly presented to a conference attended by Baxalta scientists in 2007 and that Baxalta subsequently circulated internally. The district court initially excluded Case: 19-2418 Document: 67 Page: 36 Filed: 03/01/2021

Even when accepting Bayer’s evidence as true and weighing all inferences in Bayer’s favor, we conclude that the record is insufficient to establish that Baxalta’s “con- duct rose to the level of wanton, malicious, and bad-faith behavior required for willful infringement.” SRI Int’l, Inc. v. Cisco Sys., Inc., 930 F.3d 1295, 1309 (Fed. Cir. 2019). The evidence adduced at trial merely demonstrates Bax- alta’s knowledge of the ’520 patent and Baxalta’s direct in- fringement of the asserted claims. Knowledge of the asserted patent and evidence of infringement is necessary, but not sufficient, for a finding of willfulness. Rather, will- fulness requires deliberate or intentional infringement. Eko Brands, LLC v. Adrian Rivera Maynez Enters., Inc., 946 F.3d 1367, 1378 (Fed. Cir. 2020). Accordingly, we conclude that the district court did not err in granting Baxalta’s motion for JMOL of no willfulness or denying Bayer’s motion for a new trial. CONCLUSION We have considered the parties’ other arguments, but we do not find them persuasive. For the foregoing reasons, we affirm the district court’s judgments of infringement, enablement, damages, pre-verdict supplemental damages, and no willfulness. AFFIRMED

this poster on relevance grounds. The district court then stated that it was “going to keep th[e poster] out for now,” and that Bayer could “bring it up again in a day or two if things have changed.” J.A. 536 (Tr. 315:21–23). Bayer did not, however, raise the issue of the poster’s admissibility again. On appeal, Bayer does not argue that the district court abused its discretion in excluding the poster. We do not consider the poster here. Case: 19-2418 Document: 67 Page: 37 Filed: 03/01/2021

BAYER HEALTHCARE LLC v. BAXALTA INC. 37

COSTS No costs.