`
`United States Court of Appeals
`for the Federal Circuit
`______________________
`
`SANOFI-AVENTIS DEUTSCHLAND GMBH,
`Appellant
`
`v.
`
`MYLAN PHARMACEUTICALS INC.,
`Appellee
`______________________
`
`2019-1368, 2019-1369
`______________________
`
`Appeals from the United States Patent and Trademark
`Office, Patent Trial and Appeal Board in Nos. IPR2017-
`01526, IPR2017-01528.
`______________________
`
`Decided: November 19, 2019
`______________________
`
`ADAM BANKS, Weil, Gotshal & Manges LLP, New York,
`NY, argued for appellant. Also represented by ELIZABETH
`WEISWASSER, ANISH R. DESAI, ANDREW GESIOR, AARON L. J.
`PEREIRA; ROBERT T. VLASIS, III, Washington, DC.
`
` DOUGLAS H. CARSTEN, Wilson, Sonsini, Goodrich &
`Rosati, PC, San Diego, CA, argued for appellee. Also rep-
`resented by JEFFREY WILLIAM GUISE, ALINA LEONIDOVNA
`LITOSHYK, ELHAM FIROUZI STEINER, LORELEI WESTIN;
`NICOLE W. STAFFORD, Austin, TX; WENDY L. DEVINE, San
`
`
`
`2
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`Francisco, CA; ADAM WILLIAM BURROWBRIDGE, LORA
`MARIE GREEN, RICHARD TORCZON, Washington, DC.
` ______________________
`
`Before NEWMAN, TARANTO, and CHEN, Circuit Judges.
`Opinion for the court filed by Circuit Judge TARANTO.
`
`Dissenting opinion filed by Circuit Judge NEWMAN.
`TARANTO, Circuit Judge.
`Sanofi-Aventis Deutschland GMBH’s owns U.S. Patent
`Nos. 7,476,652 and 7,713,930, which describe and claim
`certain formulations of a particular kind of insulin. Mylan
`Pharmaceuticals Inc. sought and obtained from the Patent
`and Trademark Office (PTO) inter partes reviews of all
`claims of those patents under 35 U.S.C. §§ 311–319. In
`those reviews, the PTO’s Patent Trial and Appeal Board
`agreed with Mylan that the subject matter of the claims is
`unpatentable for obviousness. Sanofi appeals, challenging
`the Board’s findings that a relevant artisan would have
`had a motivation to combine prior-art references to arrive
`at the claimed inventions with a reasonable expectation of
`success, and also challenging the Board’s evaluation of
`Sanofi’s evidence of commercial success. We reject Sanofi’s
`challenges and affirm the Board’s decisions.
`I
`The ’930 patent issued from a continuation of the ap-
`plication that issued as the ’652 patent, and the two share
`a specification. The patents involve a genetically engi-
`neered form of insulin—insulin glargine (sometimes called
`simply “glargine”)—identified in the patent as “Gly(A21)-
`Arg(B31)-Arg(B32)-human insulin.” ’652 patent, col. 2,
`lines 56–57. The patents describe and claim formulations
`of glargine that include a nonionic surfactant—polysorb-
`ates or poloxamers in the ’652 patent, esters and ethers of
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`3
`
`polyhydric alcohols in the ’930 patent. Claim 7 of the ’652
`patent is illustrative for present purposes:
`7. A pharmaceutical formulation comprising
`Gly(A21), Arg(B31), Arg(B32)-human insulin,
`at least one chemical entity chosen from poly-
`sorbate and poloxamers;
`at least one preservative; and
`water,
`wherein the pharmaceutical formulation has a
`pH in the acidic range from 1 to 6.8.
`’652 patent, col. 11, lines 21–28.
`The parties accept that certain background facts were
`publicly known at the 2002 priority date for these patents.
`Glargine is a modified version of human insulin that, when
`injected as part of an acidic solution, acts for longer in a
`subject than does natural human insulin. Glargine stays
`in solution at relatively acidic pH levels, and in the prior-
`art glargine product (which lacked the surfactants claimed
`in the patents now at issue), it was injected into a patient
`as part of an acidic solution. Once the glargine-containing
`solution is in tissue under the skin, the higher, substan-
`tially neutral pH of the tissue causes glargine to precipitate
`out of solution and to aggregate into hexamers, which then
`act as a reservoir of glargine that is slowly released into the
`patient’s blood over twenty-four hours. Natural human in-
`sulin is more soluble than glargine at the neutral pH level
`of human tissue below an injection site. Natural human
`insulin is generally injected in a solution of comparably
`neutral pH; and when injected, it rapidly dissociates into
`monomers—the physiologically active form of insulin.
`Such rapid disassociation allows for faster processing by
`the body but also necessitates more frequent injections.
`Sanofi first commercially sold glargine in the U.S. in
`May 2001, under the trade name Lantus®, whose product
`
`
`
`4
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`label identifies, among other things, a pH of 4 and the in-
`clusion of some zinc. Physician’s Desk Reference at 709
`(55th ed. 2001) (Lantus® Label); J.A. 6690. Some patients
`soon began reporting problems with turbidity in the vials,
`i.e., before injection. Sanofi determined that the turbidity
`was caused by undesirable “non-native” aggregation of the
`glargine protein while still in solution. Non-native aggre-
`gation denatures the insulin protein and is substantially
`irreversible. By contrast, “native” aggregation preserves
`the structure of the insulin protein and is reversible.
`Glargine’s mechanism of action requires some amount of
`desirable native aggregation after injection under the skin
`for its slow-release property to take effect. Sanofi resolved
`the vial-turbidity problem by adding a nonionic surfactant
`to the glargine formulation to prevent non-native aggrega-
`tion.
`Mylan petitioned the PTO for inter partes reviews of
`all claims of the ’652 and ’930 patents, arguing unpatenta-
`bility for obviousness based on combining either the Lan-
`tus® Label or an article by Owens1 with one or more of
`three secondary references.2 The parties do not dispute
`that, for each claim, the asserted combinations of
`
`
`1 David R. Owens, et al., Pharmacokinetics of 125I-
`Labeled Insulin Glargine (HOE 901) in Healthy Men: Com-
`parison with NPH Insulin and the Influence of Different
`Subcutaneous Injection Sites, 23 DIABETES CARE 813 (2000)
`(Owens).
`three secondary references are: W.D.
`2 The
`Lougheed, et al., Physical Stability of Insulin Formula-
`tions, 32 DIABETES 424 (1983) (Lougheed); Farmaceutiska
`Specialiteter I Sverige, Summary of Product Characteris-
`tics Entry for Insuman Infusat (2000) (FASS); and Ulrich
`Grau & Christopher D. Saudek, Stable Insulin Preparation
`for Implanted Insulin Pumps: Laboratory & Animal Trials,
`36 DIABETES 1453 (1987) (Grau).
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`5
`
`references teach every claim limitation. The main dispute
`is whether a relevant artisan would have been motivated
`to combine these references in the way claimed in the two
`patents at issue, with a reasonable expectation of success.
`On December 13, 2017, the Board, acting as delegee of
`the PTO’s Director, 37 C.F.R. §§ 42.4, 42.108, instituted the
`two requested reviews. Mylan Pharm. Inc. v. Sanofi-
`Aventis Deutschland GmbH, IPR2017-01526, 2017 WL
`6403855 (P.T.A.B. Dec. 13, 2017) (covering the ’652 patent);
`Mylan Pharm. Inc. v. Sanofi-Aventis Deutschland GmbH,
`No. IPR2017-01528, 2017 WL 6403082 (P.T.A.B. Dec. 13,
`2017) (covering the ’930 patent). On December 12, 2018,
`the Board issued final written decisions in both proceed-
`ings, determining that all claims in both patents are un-
`patentable for obviousness based on combinations of
`Lantus® Label or Owens with Lougheed, FASS, and/or
`Grau. Mylan Pharm. Inc. v. Sanofi-Aventis Deutschland
`GmbH, IPR2017-01526, 2018 WL 6584915 (P.T.A.B. Dec.
`12, 2018) (Decision); Mylan Pharm. Inc. v. Sanofi-Aventis
`Deutschland GmbH, IPR2017-01528, 2018 WL 6584640
`(P.T.A.B. Dec. 12, 2018).3 The Board found that a relevant
`artisan would have been motivated to make the required
`combination based on a recognition that insulins had an
`aggregation problem in vials with air space and that sur-
`factants (like the standard ones claimed here) offered a so-
`lution. Decision at *12–18. The Board also determined
`that, given the prior-art analysis, Sanofi’s evidence of com-
`mercial success was too weak to support a conclusion of
`nonobviousness. Id. at *18–20.
`
`
`3 The Board’s final written decisions are substan-
`tively identical for present purposes. In its appeal to this
`court, Sanofi has not made separate arguments regarding
`the two decisions. Accordingly, we hereafter discuss and
`cite only the decision in IPR2017-01526 (Decision), but our
`analysis applies equally to IPR2017-01528.
`
`
`
`6
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`Sanofi timely appealed under 35 U.S.C. §§ 141(c), 319.
`We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).
`II
` We review the Board’s compliance with legal standards
`de novo, Pride Mobility Products Corp. v. Permobil, Inc.,
`818 F.3d 1307, 1314 (Fed. Cir. 2016), and its underlying
`factual determinations for substantial evidence, Personal
`Web Technologies, LLC v. Apple, Inc., 848 F.3d 987, 991
`(Fed. Cir. 2017). Among the factual determinations in an
`obviousness analysis are “findings as to . . . the presence or
`absence of a motivation to combine or modify with a rea-
`sonable expectation of success[] and objective indicia of
`non-obviousness.” Ariosa Diagnostics v. Verinata Health,
`Inc., 805 F.3d 1359, 1364 (Fed. Cir. 2015).
`A
`Sanofi challenges the Board’s finding of a motivation to
`
`combine the prior-art references to arrive at the claimed
`glargine formulation with certain surfactants. Sanofi ar-
`gues that (1) KSR International Co. v. Teleflex Inc., 550
`U.S. 398 (2007), required the Board to find that the prior
`art disclosed an aggregation problem for glargine specifi-
`cally (not just insulins in general); (2) the Board improperly
`relied on each patent’s own (shared) specification in finding
`a motivation to combine; and (3) substantial evidence does
`not support the Board’s finding because key evidence cited
`by the Board concerned insulins in general rather than
`glargine specifically. The first two contentions assert legal
`errors, the third evidentiary insufficiency. We address the
`contentions in turn. We find each one unpersuasive.
`1
`Sanofi argues that the Board was required, under KSR,
`
`to find in the prior art a recognition of an aggregation prob-
`lem for glargine specifically, not just for insulins generally.
`In Sanofi’s view, KSR demands more than a factually sup-
`ported finding that recognition of an aggregation risk for
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`7
`
`insulins generally would have motivated a relevant artisan
`to address aggregation for this particular insulin. We re-
`ject Sanofi’s view of KSR.
`The Supreme Court in KSR explained that, “because
`inventions in most, if not all, instances rely upon building
`blocks long since uncovered, and claimed discoveries al-
`most of necessity will be combinations of what, in some
`sense, is already known,” “it can be important to identify a
`reason that would have prompted a person of ordinary skill
`in the relevant field to combine the elements in the way the
`claimed new invention does.” Id. at 418–19. But KSR
`stressed flexibility and realism over rigidity and formalism
`in assessing what such reasons might be:
`In KSR, the Supreme Court criticized a rigid
`approach to determining obviousness based on the
`disclosures of individual prior-art references, with
`little recourse to the knowledge, creativity, and
`common sense that an ordinarily skilled artisan
`would have brought to bear when considering com-
`binations or modifications. KSR, 550 U.S. at 415–
`22. Rejecting a blinkered focus on individual docu-
`ments, the Court required an analysis that reads
`the prior art in context, taking account of “demands
`known to the design community,” “the background
`knowledge possessed by a person having ordinary
`skill in the art,” and “the inferences and creative
`steps that a person of ordinary skill in the art
`would employ.” Id. at 418. This “expansive and
`flexible approach,” id. at 415, is consistent with our
`own pre-KSR decisions acknowledging that the in-
`quiry “not only permits, but requires, consideration
`of common knowledge and common sense.” DyStar
`Textilfarben GmbH & Co. Deutschland KG v. C.H.
`Patrick Co., 464 F.3d 1356, 1367 (Fed. Cir. 2006).
`Randall Mfg. v. Rea, 733 F.3d 1355, 1362 (Fed. Cir. 2013);
`see also Arctic Cat Inc. v. Bombardier Recreational Prod.
`
`
`
`8
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`Inc., 876 F.3d 1350, 1359 (Fed. Cir. 2017) (“The court
`should consider a range of real-world facts to determine
`whether there was an apparent reason to combine the
`known elements in the fashion claimed by the patent at is-
`sue.”) (citation and internal quotation marks omitted).
`The Board did not depart from KSR when it made, and
`relied on, findings that a relevant artisan would have rec-
`ognized a potential aggregation-in-the-vial problem with
`glargine as part of the general recognition of aggregation
`problems with insulins. Nothing in KSR demands the kind
`of prior-art identifications of a problem at the level of spec-
`ificity that Sanofi urges. The Board thus properly exam-
`ined the evidence in this particular case to determine
`whether a relevant artisan would have recognized an insu-
`lin aggregation problem in the prior art and expected
`glargine to share that problem. Decision at *14–16.
`Whether the Board was correct is a case-specific matter of
`evidentiary sufficiency—a matter we discuss more fully in-
`fra.
`
`2
`We also reject Sanofi’s contention that the Board com-
`mitted legal error when it cited the shared patent specifi-
`cation. The “background of the invention” portion of the
`specification includes the following passage:
`The specific preparation of insulin glargine,
`which leads to the prolonged duration of action, is
`characterized, in contrast to previously described
`preparations, by a clear solution having an acidic
`pH. Especially at acidic pH, insulins, however,
`show a decreased stability and an increased prone-
`ness to aggregation on thermal and physicome-
`chanical stress, which can make itself felt in the
`form of turbidity and precipitation (particle for-
`mation (Brange et al., J. Ph. Sci 86:517-525 (1997)).
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`9
`
`The proneness to aggregation can additionally
`be promoted by hydrophobic surfaces which are in
`contact with the solution (Sluzky et al., Proc. Natl.
`Acad. Sci. 88:9377-9381 (1991). Surfaces which can
`be considered as hydrophobic are the glass vessels
`of the preparations, the stopper material of the
`sealing caps or the boundary surface of the solution
`with the air supernatant. In addition, very fine sil-
`icone oil droplets can function as additional hydro-
`phobic aggregation nuclei in the taking of the daily
`insulin dose by means of customary, siliconized in-
`sulin syringes and accelerate the process.
`’652 patent, col. 2, line 66 through col. 3, line 17. The Board
`cited this material in finding that insulin was known to ag-
`gregate on hydrophobic surfaces, at the air/water interface
`of a container, and in acidic solutions. Decision at *14–15.
`Sanofi challenges the Board’s reliance on this material
`as legally improper, invoking our longstanding recognition
`that a tribunal should not “look[] to knowledge taught by
`the inventor . . . and then use[] that knowledge against its
`teacher.” Panduit Corp. v. Dennison Mfg. Co., 774 F.2d
`1082, 1092 (Fed. Cir. 1985), vacated on other grounds, 475
`U.S. 809 (1986); see also InTouch Techs., Inc. v. VGO
`Commc’ns, Inc., 751 F.3d 1327, 1351 (Fed. Cir. 2014). But
`the Board did not violate that principle, because it did not
`use the specification for its teachings about the inventor’s
`discovery. Rather, it used the specification for its teachings
`about prior-art knowledge, and that use of a specification
`is not just common, given patent drafters’ standard prac-
`tice of reciting prior art in setting out the background of
`the invention, but permissible. E.g., Smith & Nephew, Inc.
`v. Rea, 721 F.3d 1371, 1378–79 (Fed. Cir. 2013); Phar-
`maStem Therapeutics, Inc. v. ViaCell, Inc., 491 F.3d 1342,
`1362 (Fed. Cir. 2007); cf. WesternGeco LLC v. ION Geophys-
`ical Corp., 889 F.3d 1308, 1329–30 (Fed. Cir. 2018) (speci-
`fication confirmed Board’s understanding of prior art in
`anticipation context).
`
`
`
`10
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`The Board understood the patent specification, on this
`issue, to be addressing what was already known—a read-
`ing that is reasonable given the language used and cita-
`tions to prior art. Moreover, the Board used the cited
`material not as the sole support for any finding but in con-
`junction with support from other sources. The Board found
`evidence of insulin aggregation on hydrophobic surfaces
`and at air/water interfaces in a handful of other prior-art
`references. Decision at *14–15. The Board cited four addi-
`tional references to support the finding that insulin was
`known to aggregate in acidic solutions. Id. at *15. The
`Board’s use of the patent specification, we conclude, did not
`rest on legal error.
`
`3
` We further conclude that the Board’s finding of a moti-
`vation to combine is supported by substantial evidence.
`While the Board must provide “a reasoned basis” for its ac-
`tions, “‘we will uphold a decision of less than ideal clarity if
`the agency’s path may reasonably be discerned.’” In re
`NuVasive, Inc., 842 F.3d 1376, 1383 (Fed. Cir. 2016) (quot-
`ing Bowman Transp., Inc. v. Ark.–Best Freight System,
`Inc., 419 U.S. 281, 285, 286 (1974)). The Board “must ar-
`ticulate a reason why a [relevant artisan] would combine
`the prior art references.” Id. at 1382. And the finding of
`such a reason must be supported by substantial evidence,
`which is “such relevant evidence as a reasonable mind
`might accept as adequate to support a conclusion.” Id. at
`1380 (citation and internal quotation marks omitted); see
`also Intelligent Bio-Systems, Inc. v. Illumina Cambridge
`Ltd., 821 F.3d 1359, 1366 (Fed. Cir. 2016) (explaining that
`review for substantial evidence “requires examination of
`the record as a whole, taking into account evidence that
`both justifies and detracts from an agency’s decision”) (ci-
`tation and internal quotation marks omitted).
`The Board’s findings with respect to the motivation to
`combine are detailed and well supported. The Board found
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`11
`
`that insulins “had a known tendency to aggregate in the
`presence of hydrophobic surfaces” and at air-water inter-
`faces and that a relevant artisan would have expected
`glargine to behave similarly to other insulins when in con-
`tact with hydrophobic surfaces and at air-water interfaces.
`Decision at *14. The Board also found that nonionic sur-
`factants, including the claimed ones, were well known and
`had been used successfully to stabilize insulin formula-
`tions, and so would have been looked to by a relevant arti-
`san concerned about aggregation in glargine. Id. at *11–
`12, *17. The record contains substantial evidence to sup-
`port those findings.
`Two references by Brange disclose that insulins with a
`variety of amino acid structures each display some degree
`of aggregation. J.A. 6762; J.A. 6797. Likewise, as already
`discussed, the shared specification of the ’652 and ’930 pa-
`tents itself indicates, in a discussion introduced by discuss-
`ing glargine, that
`insulins tend to aggregate on
`hydrophobic surfaces (like the glass of vials), especially in
`acidic solutions like those used for glargine. See ’652 pa-
`tent, col. 2, line 66 through col. 3, line 17. Mylan’s expert
`explained, with citations to prior art, that “insulin aggre-
`gation is a well-established problem in the field and de-
`scribed in detail by numerous references.” J.A. 6475.
`Sanofi argued that the prior art discloses aggregation
`only in insulin pumps, but the Board disagreed, finding in-
`stead that “it is the air-water interfaces and interactions
`with hydrophobic surfaces that promote insulin aggrega-
`tion, and not the type of device used to deliver the insulin
`formulation.” Decision at *15. Prior art supports the
`Board’s determination. See, e.g., J.A. 6796 (noting that in-
`sulin has a tendency to aggregate on hydrophobic surfaces);
`J.A. 14535 (“It has been suggested that insulin is destabi-
`lized by adsorption at hydrophobic interfaces (air-water or
`water-pump materials). . . .”); J.A. 6906; J.A. 6951. The
`Board also reasonably understood Mylan’s expert to testify
`that aggregation “was known in the art not to be unique to
`
`
`
`12
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`[insulin] pumps,” J.A. 12246 (quoted in Decision at *15),
`and found that Sanofi’s expert, in suggesting otherwise, re-
`lied on evidence that went no further than indicating that
`insulin pumps showed a greater tendency for aggregation
`than other container types, Decision at *15.
`Other evidence reasonably supports the Board’s find-
`ing that a relevant artisan would have understood glargine
`to come within the general recognition of an aggregation
`problem for insulins. The Lantus® Label discloses glargine
`formulated as a solution with an acidic pH, J.A. 6690, and
`both the Lantus® Label and Owens teach glargine formu-
`lations in vials known to contain hydrophobic surfaces and
`an air-water interface, J.A. 6693; J.A. 6699–700. There
`was evidence, too, that, while insulin exists in equilibrium
`as monomers, dimers, and hexamers, an acidic environ-
`ment shifts the equilibrium toward monomers, which are
`more susceptible to aggregation. J.A. 6769–70; J.A. 6798–
`99; J.A. 6830; J.A. 14535. And relatedly, although Lantus®
`contains zinc, which can affect rates of aggregation, the ev-
`idence supports the Board’s findings, Decision at *15, that
`zinc does not bind to insulin in an acidic solution, like the
`Lantus® solution, J.A. 13741, and, more generally, that
`zinc in the Lantus® solution would not have led a relevant
`artisan to see glargine as immune from the general prob-
`lem of insulin aggregation in vials.
`The evidence also supports the Board’s finding that the
`prior art taught use of nonionic surfactants like those
`claimed in the present patents to address the aggregation
`problem. For example, Lougheed teaches the addition of
`polysorbate 20 or polysorbate 80 to insulin formulations to
`reduce aggregation. J.A. 6706 (“[A]ggregate formation [in
`insulin formulations] was inhibited by the nonionics . . .
`Tween 20, [and] Tween 80.”). Both FASS and Grau teach
`the use of a poloxamer to stabilize an insulin formulation.
`J.A. 6725 (“Addition of a stabilizer poly(oxyethylene, oxy-
`propylene), glycol, prevents precipitation and flocculation
`of the insulin.”); J.A. 6732 (“Genapol, a surface-active
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`13
`
`polyethylene-polypropylene glycol, effectively prevents ad-
`sorption of insulin to hydrophobic surfaces.”). Mylan’s ex-
`pert declaration provides further support when it points,
`with citations to prior art, to “the routine use of polysorb-
`ates and poloxamers in insulin formulations for inhibiting
`protein aggregation.” J.A. 6475–76.
`Sanofi points to our non-precedential decision in No-
`vartis Pharmaceuticals Corp. v. Watson Laboratories, Inc.,
`611 F. App’x 988 (Fed. Cir. 2015), but that decision does
`not undermine the Board’s finding here. In Novartis, we
`affirmed a district court’s determination of non-obvious-
`ness where the prior art teaching was reasonably found to
`differ significantly from the claimed invention. Id. at 995–
`96 (concluding that it would not be obvious to modify ri-
`vastigmine in the way claimed to solve the well-known
`problem of oxidative degradation with physostigmine, be-
`cause the prior art taught that rivastigmine had “greater
`chemical stability” than physostigmine). That ruling does
`not help Sanofi in challenging the Board’s determination of
`obviousness based on findings that the glargine compound
`is similar to other insulins in the respects relevant to the
`obviousness analysis.
`
`B
`Sanofi also challenges the Board’s finding that a rele-
`
`vant artisan would have had a reasonable expectation of
`success in adding the claimed surfactants to the existing
`glargine preparation in the way claimed in the patents at
`issue here. Its focus on this issue, as on the related moti-
`vation-to-combine issue, is the contention that the Board
`looked at insulins generally and did not make adequately
`supported findings about glargine specifically. We reject
`Sanofi’s challenge.
`
`1
`As a preliminary matter, we address Sanofi’s argument
`
`that the Board improperly relied, in its reasonable-
`
`
`
`14
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`expectation-of-success analysis, on evidence submitted by
`Mylan in reply to Sanofi’s patent owner’s response. We re-
`view the Board’s decisions regarding the scope of proper re-
`ply material for an abuse of discretion. Ericsson Inc. v.
`Intellectual Ventures I LLC, 901 F.3d 1374, 1379 (Fed. Cir.
`2018). We see no abuse of discretion in the present IPRs.
`Under the governing IPR rules, there is no impropriety
`when the Board considers reply evidence to the extent that
`the evidence is offered to show why a patent owner’s re-
`sponse is wrong in its criticisms of the sufficiency of the
`petition’s case for unpatentability, including where the pa-
`tent owner’s response introduces what amounts to a new
`defense to an otherwise-sufficient case of unpatentability
`in the petition. See, e.g., Idemitsu Kosan Co. v. SFC Co.,
`870 F.3d 1376, 1381 (Fed. Cir. 2017) (reply evidence may
`respond to teaching-away contention in patent owner’s re-
`sponse). Here, Mylan’s petitions made its case for finding
`a reasonable expectation of success, see, e.g., J.A. 384; J.A.
`457, and after Sanofi made arguments against such a find-
`ing in its patent owner’s response, Mylan’s reply included
`rebuttal argument and evidence addressing Sanofi’s
`points, J.A. 1819–37; J.A. 12231–91 (excerpts of reply ex-
`pert declaration); see J.A. 2414–18 (excerpts of Sanofi’s
`specification of objected-to passages). The Board allowed
`Sanofi to file at least one sur-reply on the issue of reasona-
`ble expectation of success, as well as several motions to ex-
`clude, but the Board found all of Sanofi’s objections either
`unpersuasive, because Mylan’s reply evidence was proper
`rebuttal evidence, or moot, because the Board had not re-
`lied on particular objected-to evidence. See Decision at *5–
`6; J.A. 15304–06. We see no abuse of discretion in the
`Board’s rulings in this regard.
`2
`On the merits, Sanofi argued to the Board that, alt-
`hough surfactants were known to stabilize insulins gener-
`ally, a relevant artisan would not have expected the same
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`15
`
`result for glargine specifically because its mechanism of ac-
`tion depends on some favorable native aggregation. To the
`extent that Sanofi contends that the Board did not consider
`this argument, Sanofi is incorrect. The Board thoroughly
`considered Sanofi’s argument but found it unpersuasive.
`To the extent that Sanofi contends that there is no substan-
`tial evidence to support a finding of reasonable expectation
`of success for glargine specifically, we conclude that Sanofi
`is incorrect in that contention as well.
`The Board began its reasonable expectation of success
`analysis by finding that a number of nonionic surfactants—
`including the claimed nonionic surfactants—were shown in
`the prior art to have been successfully used to prevent ag-
`gregation of various types of insulins and other peptides.
`Decision at *17. The prior art supports this determination.
`See, e.g., J.A. 6706–07 (“[A]ggregate formation [in insulin
`formulations] was inhibited by the nonionic[] [surfac-
`tants],” including polysorbate 20 and polysorbate 80.); J.A.
`6725 (“Addition of a stabilizer poly(oxyethylene, oxypropyl-
`ene), glycol,” a poloxamer, “prevents precipitation and floc-
`culation of the insulin.”). Mylan’s expert declared that a
`relevant artisan, when considering which nonionic surfac-
`tants to use in a glargine formulation, would look to
`nonionic surfactants (such as polysorbates) approved by
`the Food and Drug Administration (FDA) for use in other
`protein formulations, and the Board, after its prior-art rec-
`itation, credited that statement. Decision at *17.
`The Board found “unpersuasive [Sanofi’s] arguments
`that an ordinarily skilled artisan would not have reasona-
`bly expected success when adding a nonionic surfactant to
`insulin glargine in view [of] their success stabilizing other
`insulins and proteins.” Id. For example, Sanofi contended
`that adding a nonionic surfactant to a strong acid had the
`potential to cause undesirable hydrolysis or saponification.
`But the Board explained that Sanofi did not put forth any
`evidence that the prior-art glargine compounds existed in
`a strong acid, and it pointed to evidence that polysorbates
`
`
`
`16
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`had in fact been used in pharmaceutical formulations at
`acidic pH (3.0 to 4.0). Id. at *18 (citing J.A. 7450–51; J.A.
`12907).
`The Board also credited Mylan’s evidence that the
`presence of phenols in a glargine formulation would not
`have dissuaded a relevant artisan from expecting success
`in using nonionic surfactants. Id. The Board reasonably
`did so. The Board noted that other pharmaceutical formu-
`lations include both nonionic surfactants and phenols. De-
`cision at *18 (citing, e.g., J.A. 12911). There also was
`evidence, including from Sanofi’s expert, that phenols in
`insulin formulations stabilize hexamers, whereas surfac-
`tants prevent irreversible denaturation of monomers but
`do not prevent hexamer formation. J.A. 14249–53; J.A.
`14387; see J.A. 6732; J.A. 6910. Moreover, the testimony
`of Sanofi’s expert about a problem was carefully limited,
`stating only that nonionic surfactants in a glargine formu-
`lation “could” disrupt the native aggregation that phenols
`promote. J.A. 14307–09. Mylan’s expert, in contrast,
`stated unequivocally that a nonionic surfactant’s potential
`interference with phenols would not dissuade a relevant
`artisan from using both in a formulation. J.A. 12298.
`The Board did not expressly address Sanofi’s argu-
`ments about the potential for discoloration or peroxide for-
`mation. But the Board rejected them implicitly as bases
`for finding no reasonable expectation of success: those ar-
`guments were within the pages of the patent owner’s re-
`sponse
`that
`recited
`various
`potential negative
`consequences that the Board addressed collectively, find-
`ing Sanofi’s arguments in those pages unpersuasive
`whether considered with respect to motivation to combine
`or reasonable expectation of success. Decision at *18. The
`Board is not required to “expressly discuss each and every
`negative and positive piece of evidence lurking in the rec-
`ord.” Novartis AG v. Torrent Pharm. Ltd., 853 F.3d 1316,
`1328 (Fed. Cir. 2017). Sanofi has not shown that its evi-
`dence on these two particular potential consequences
`
`
`
`SANOFI-AVENTIS DEUTSCHLAND v. MYLAN PHARMACEUTICALS
`INC.
`
`17
`
`undermines the Board’s finding that, considering all rele-
`vant factors, an ordinary artisan would have had a reason-
`able expectation of success in adding a nonionic surfactant
`to a glargine formulation. Decision at *18. We conclude
`that the Board’s finding is supported by substantial evi-
`dence.
`
`C
`Lastly, Sanofi challenges the Board’s analysis of com-
`mercial success. The Board accepted that Sanofi’s product
`was a commercial success. Decision at *19. The Board
`found that Sanofi’s commercial success evidence was ulti-
`mately “weak” so as not to warrant an ultimate conclusion
`on obviousness different from the one strongly indicated by
`the motivation-to-combine and reasonable-expectation-of-
`success analysis. Decision at *19 n.14, *20. We reject
`Sanofi’s challenge to the Board’s reasoning—whether it is
`viewed as a factual finding of only a weak nexus of com-
`mercial success to the claimed invention or as part of the
`ultimate legal weighing to determine obviousness. See In-
`tercontinental Gre