Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 1 of 45
`
`IN THE UNITED STATES DISTRICT COURT
`FOR THE DISTRICT OF MASSACHUSETTS
`
`
`
`TEVA PHARMACEUTICALS
`
`INTERNATIONAL GMBH and
`
`TEVA PHARMACEUTICALS
`
`USA, INC.,
`
`
`Plaintiffs,
`
`v.
`
`Civil Action No.
`1:18-cv-12029-ADB
`
`
`
`
`
`ELI LILLY AND COMPANY,
`
`Defendant.
`
`
`
`
`
`
`
`
`
`
`DECLARATION OF DR. K. CHRISTOPHER GARCIA
`
`
`
`
`
`
`
`
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 2 of 45
`
`TABLE OF CONTENTS
`
`EDUCATION AND EXPERIENCE .................................................................................. 4
`
`COMPENSATION AND PRIOR TESTIMONY ............................................................... 6
`
`LEGAL CONSIDERATIONS ............................................................................................ 7
`
`LEVEL OF ORDINARY SKILL IN THE ART AT THE TIME OF THE
`ALLEGED INVENTION ................................................................................................... 8
`
`TASK SUMMARY ............................................................................................................ 8
`
`TECHNICAL BACKGROUND ......................................................................................... 9
`
`
`I.
`
`II.
`
`III.
`
`IV.
`
`V.
`
`VI.
`
`VII. DISPUTED CLAIM TERMS ........................................................................................... 13
`
`A.
`
`“ANTI-CGRP ANTAGONIST ANTIBODY” OR “ANTI-CALCITONIN
`GENE-RELATED PEPTIDE (CGRP) ANTAGONIST ANTIBODY” ............... 13
`
`1.
`
`2.
`
`3.
`
`A Person of Ordinary Skill in the Art Would Have No Way to
`Determine with Reasonable Certainty Whether an Antibody Is or
`Is Not an “Anti-CGRP Antagonist Antibody” .......................................... 13
`
`Teva’s Own Data Show That Contradictory Results Can Be
`Produced From (1) A Particular Bioassay Depending on the
`Selected Conditions and (2) Different Bioassays ..................................... 17
`
`Teva’s Specification Fails to Specify a Particular Threshold That
`Must Be Met to Constitute “Inhibition” of CGRP Biological
`Activity ..................................................................................................... 22
`
`4.
`
`If Construed, Lilly’s Proposed Construction Should Be Adopted ............ 25
`
`B.
`
`“HUMANIZED . . . ANTIBODY” ....................................................................... 26
`
`1.
`
`A Person of Ordinary Skill in the Art Would Have No Way to
`Determine with Reasonable Certainty Whether an Antibody Is or
`Is Not a “Humanized . . . Antibody” ......................................................... 27
`
`2.
`
`If Construed, Lilly’s Proposed Construction Should Be Adopted ............ 32
`
`C.
`
`“SPECIFIC BINDING” OR “PREFERENTIALLY BINDS” ............................. 32
`
`1.
`
`A Person of Ordinary Skill in the Art Would Have No Way to
`Determine with Reasonable Certainty Whether an Antibody
`Exhibits “Specific Binding” or “Preferentially Binds” ............................. 33
`
`2.
`
`If Construed, Lilly’s Proposed Construction Should Be Adopted ............ 40
`
`
`
`2
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 3 of 45
`
`D.
`
`“HUMAN IgG HEAVY CHAIN” ........................................................................ 41
`
`1.
`
`A Person of Ordinary Skill in the Art Would Have Understood
`“Human IgG Heavy Chain” to Mean the Entirety of the IgG Heavy
`Chain is Human......................................................................................... 41
`
`
`
`3
`
`
`
`
`
`
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 4 of 45
`
`1.
`
`2.
`
`I, Kenan Christopher Garcia, hereby declare as follows:
`
`I am a Professor at the Stanford University School of Medicine in the Department
`
`of Molecular and Cellular Physiology, and the Department of Structural Biology.
`
`3.
`
`I have been retained by counsel for Defendant Eli Lilly and Company (“Lilly”) in
`
`this case to offer opinions as to the scope and meaning that would have been given to certain
`
`terms that appear in the asserted claims of U.S. Patent Nos. 8,586,045 (Ex. 1, “the ’045 Patent”);
`
`8,597,649 (Ex. 2, “the ’649 Patent”); 9,266,951 (Ex. 3, “the ’951 Patent”); 9,340,614 (Ex. 4,
`
`“the ’614 Patent”); 9,346,881 (Ex. 5, “the ’881 Patent”); 9,884,907 (Ex. 6, “the ’907 Patent”);
`
`9,884,908 (Ex. 7, “the ’908 Patent”); 9,890,210 (Ex., 8, “the ’210 Patent”); and 9,890,211 (Ex. 9,
`
`“the ’211 Patent”) (collectively, the Patents-in-Suit) by a person having ordinary skill in the art at
`
`the time of the alleged inventions.1
`
`I.
`
`EDUCATION AND EXPERIENCE
`
`4.
`
`I am currently a Professor at the Stanford University School of Medicine. I am
`
`also an Investigator of the Howard Hughes Medical Institute in Stanford, California. Since 1999,
`
`my laboratory at Stanford has focused on the biophysical characterization of protein structure
`
`and function, including protein engineering, antibody recognition and engineering, therapeutic
`
`antibodies, and receptor-ligand signaling. I have authored over 200 publications on subjects
`
`such as antibody-antigen interactions, protein engineering, immunology, and structural and
`
`molecular modeling.
`
`
`1 I understand that Teva Pharmaceuticals International GmbH and Teva Pharmaceuticals USA,
`Inc. (collectively, “Teva”) asserted the following claims against Lilly: claims 1, 3, 4, 8, 9, 12, 15-
`17, 19, 20, 24, 27, 30, and 31 of the ’045 patent; claims 7 and 9 of the ’649 patent; claims 17 and
`18 of the ’951 patent; claims 18 and 19 of the ’614 patent; claims 17 and 19 of the ’881 patent;
`claims 1, 4-7, 15, and 17 of the ’907 patent; claims 1, 4-7, 15, and 17 of the ’908 patent; claims
`11 and 13 of the ’210 patent; and claims 2, 12, and 14 of the ’211 patent.
`
`
`
`4
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 5 of 45
`
`5.
`
`I received my B.S. in Biochemistry from Tulane University in 1984 and my Ph.D.
`
`in Biophysics from Johns Hopkins University in 1992, studying antibody structure, anti-peptide
`
`and anti-hormone antibody recognition, and molecular immunology. I completed my first post-
`
`doctoral fellowship in the Departments of Protein Engineering and Molecular Biology at
`
`Genentech. My research was focused on recombinant protein expression, protein, peptide and
`
`antibody engineering, as well as protein structure and activity. I completed my second post-
`
`doctoral fellowship at the Scripps Research Institute. My research was focused on immunology,
`
`including the relationship between antibodies and antigens, the molecules to which antibodies
`
`bind.
`
`6.
`
`I was elected to the National Academy of Sciences, and the National Academy of
`
`Medicine, and have received several awards and recognition in my field, including from the
`
`American Heart Association and the Cancer Research Institute. I was named a Keck
`
`Distinguished Medical Scholar and a Pew Scholar. I am on the Scientific Advisory Board of
`
`Harvard Medical School’s Program in Cellular and Molecular Medicine. I have co-founded or
`
`served on Scientific Advisory Boards of several biotech companies working in the area of
`
`immunology.
`
`7.
`
`I serve on the editorial boards of several scientific journals, including Proceedings
`
`of the National Academy of Sciences of the United States of America, Immunological Reviews,
`
`Immunity, and Structure.
`
`8.
`
`For the past twenty years, I have taught courses in the fields of molecular biology,
`
`immunology, structural biology, microbiology, and molecular and cellular physiology.
`
`9.
`
`My curriculum vitae is attached as Appendix A, which includes a detailed list of
`
`publications.
`
`
`
`5
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 6 of 45
`
`10.
`
`I have been retained by Lilly to offer my opinion as to whether or how a person of
`
`ordinary skill in the art (“POSITA”), as defined below, would have understood the following
`
`terms as used in the claims of the Patents-in-Suit:
`
`•
`
`•
`
`•
`
`•
`
`“anti-Calcitonin Gene-Related Peptide (CGRP) antagonist antibody” or “anti-
`
`CGRP antagonist antibody,”
`
`“humanized . . . antibody,”
`
`“specific binding” or “preferentially binds,” and
`
`“human IgG heavy chain.”
`
`11.
`
`I reserve the right to supplement and/or amend my opinions in this declaration
`
`based on future positions taken by the parties, their experts, additional documents, testimony, or
`
`other information provided by the parties or their witnesses, any orders from the Court, or as
`
`otherwise necessary.
`
`II.
`
`COMPENSATION AND PRIOR TESTIMONY
`
`12.
`
`I am being compensated my usual consulting rate of $1,500 per hour for my work
`
`related to this matter. My compensation is in no way dependent on the outcome of this dispute
`
`or the testimony or opinions that I may provide. I have no financial interest, beneficial or
`
`otherwise, in the patents or any of the parties.
`
`13. Within the past four years, I have testified as an expert in Juno Therapeutics, Inc.
`
`et al v. Kite Pharma, Inc., Case No. CDCA-2-17-cv-07639 and in Bayer Healthcare
`
`Pharmaceuticals Inc. v. Biogen Idec Inc., Case No. DNJ-2-10-cv-02734.
`
`
`
`6
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 7 of 45
`
`III. LEGAL CONSIDERATIONS
`
`14.
`
`I am not a legal expert and offer no legal opinions. I have, however, been
`
`informed by counsel of the legal requirements related to claim construction and indefiniteness,
`
`and I have applied them in arriving at my conclusion.
`
`15.
`
`I have been informed that the patent laws require that a patent claim be construed
`
`according to the meaning it would have had to a POSITA at the time of the alleged invention. I
`
`have also been informed that the construction of a patent term should focus primarily on the
`
`intrinsic evidence of record, which consists of the claims, the patent specification, and the
`
`prosecution history. Further, I have been informed that the patent specification is highly relevant
`
`to the understanding of claim terms: if an inventor clearly defines a claim term in the patent
`
`specification, the inventor’s definition controls, even if the inventor’s definition differs from the
`
`term’s ordinary and customary meaning.
`
`16.
`
`I have been informed that a claim fails to satisfy the definiteness requirement of
`
`the patent laws if the claim does not inform a POSITA with reasonable certainty about the scope
`
`of the claimed invention when read in light of the patent specification and the prosecution
`
`history. I have also been informed that the patent must be precise enough to provide clear notice
`
`of what is claimed so that the public is certain about what is still open to them, i.e., what would
`
`or would not infringe the patent claim. Further, I have been informed that a claim term is
`
`indefinite if it can be measured in different ways that would (at least in some instances) yield
`
`different results, with no express guidance from the patent or its prosecution history as to which
`
`measurement to use.
`
`
`
`7
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 8 of 45
`
`IV.
`
`LEVEL OF ORDINARY SKILL IN THE ART AT THE TIME OF THE
`ALLEGED INVENTION
`
`17.
`
`I understand that a person having ordinary skill in the art may be defined by
`
`factors such as: (a) the levels of education and experience of the inventor and other persons
`
`actively involved in the field; (b) the types of problems encountered in the field; (c) prior art
`
`solutions to those problems; (d) rapidity with which innovations are made; and (e) the
`
`sophistication of the technology.
`
`18.
`
`I have been instructed to assume that, for the purposes of this declaration, the
`
`relevant timeframe for assessing the claims of the Patents-in-Suit is November 14, 2005. Unless
`
`otherwise specified, all of my opinions expressed herein regarding a POSITA apply to such a
`
`person as of November 14, 2005.
`
`19.
`
`For purposes of this declaration, I have assumed that a POSITA with respect to
`
`designing and optimizing “anti-CGRP antagonist antibodies” would have generally possessed a
`
`Ph.D. in immunology, molecular biology, or pharmacology with several years of post-doctoral
`
`research experience focused on antibody engineering, and/or antibody pharmacology.
`
`V.
`
`TASK SUMMARY
`
`20.
`
`I have been asked to review the Patents-in-Suit. I have been instructed that the
`
`Patents-in-Suit share the same patent specification, are part of the same patent family, and claim
`
`priority to the same priority document. My citations to the specifications of the Patents-in-Suit
`
`in this declaration are based on the ’045 patent or another patent if the cited claim term is not
`
`used in the ’045 patent claims. I have been asked to provide my opinion from the perspective of
`
`a POSITA having knowledge of the relevant art as of November 14, 2005, of how such a person
`
`would have interpreted the claim terms set forth below.
`
`
`
`8
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 9 of 45
`
`21.
`
`In preparing this declaration, I have considered the Patents-in-Suit in their
`
`entirety, portions of the prosecution histories from this patent family of the Patents-in-Suit, and
`
`the general knowledge of a POSITA as of November 14, 2005. In addition, I have reviewed
`
`references cited herein, the references cited in Appendix B, and the parties’ joint claim
`
`construction statement.
`
`VI.
`
`TECHNICAL BACKGROUND
`
`22.
`
`Proteins play a crucial role in the biological functions of the human body.
`
`Proteins are assembled with twenty building blocks called amino acids. Amino acids may be
`
`strung together to form a polypeptide chain. This sequence of amino acids is considered the
`
`primary structure of the chain. Amino acids in this sequence (referred to as “amino acid
`
`residues” or simply “residues”) interact with other adjacent amino acid residues to form local
`
`arrangements, also called secondary structure. The overall three-dimensional shape of the chain
`
`is referred to as tertiary structure. The sequence of amino acids determines the secondary and
`
`tertiary structures, which in turn are responsible for its biological function.
`
`23.
`
`Antibodies (also called immunoglobulins or Igs) are complex proteins found in
`
`the blood that recognize and bind to target molecules called antigens (e.g., calcitonin gene-
`
`related peptide or CGRP). The basic structure of an IgG antibody (discussed below) is
`
`symmetrical and consists of four polypeptide chains, two longer heavy chains and two shorter
`
`light chains.
`
`24.
`
`The four chains form the overall shape of an antibody, which resembles the letter
`
`“Y.” The top tip of each arm of the Y is responsible for binding the target antigen. Amino acids
`
`from both the heavy and light chains contribute to the antigen binding site. Different antibodies
`
`bind to different target antigens and have different heavy and light chains (and thus different
`
`primary sequences, secondary structures, and tertiary structures). Each heavy chain pairs with a
`
`
`
`9
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 10 of 45
`
`light chain to form one arm of the antibody structure, and disulfide bonds as well as other amino
`
`acid interactions link the heavy and light chain together to form the two arms of the antibody.
`
`25.
`
`A simplistic schematic of an IgG antibody is provided in Figure 1, which is not
`
`meant to reflect the complex secondary or tertiary structures of antibodies. The schematic shows
`
`the two heavy chains in dark green and two light chains in light green. Each contains a distinct
`
`variable region (patterned) and constant region (solid).
`
`Figure 1
`
`
`
`26.
`
`There are five classes of antibody molecules in humans—IgG, IgM, IgD, IgA, and
`
`
`
`IgE—based on the structure of the constant region of the human heavy chain. There are also
`
`subtypes within each of these classes of antibody molecules, e.g., IgG1, IgG2, IgG3, and IgG4.
`
`There are two different classes of human light chains, lambda and kappa.
`
`27.
`
`The heavy and light chains are each composed of segments of amino acids that
`
`fold into domains. A heavy chain consists of one variable domain (VH) and three constant
`
`domains (CH1, CH2, CH3). A light chain consists of one variable domain (VL) and one constant
`
`domain (CL). The variable domains of the heavy and light chains are responsible for antigen
`
`binding.
`
`28.
`
`Each variable domain is made up of hypervariable regions, also known as
`
`complementarity-determining regions (CDRs). Each variable domain has three CDRs,
`
`
`
`10
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 11 of 45
`
`designated as CDR1, CDR2, and CDR3, respectively. Amino acid sequence variability is higher
`
`in the CDRs than in any other portion of an antibody. The three CDRs are interspersed between
`
`four regions that serve as a framework to support the CDRs (hence “framework regions” or
`
`“FRs”). The four FRs are designated as FR1, FR2, FR3, and FR4, respectively. As shown in
`
`Figure 2, each heavy and light variable domain is composed of four FRs and three CDRs in the
`
`following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4.
`
`
`
`
`
`
`
`Figure 2
`
`11
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 12 of 45
`
`29.
`
`In the three-dimensional structure of an antibody, the CDRs cluster together at the
`
`tip of the Y to form the antigen binding site. (See Figure 2.) The CDRs that form the antigen
`
`binding site fold into loop structures and present an exposed protein surface to which the antigen
`
`can bind. The FRs influence the structure of the antigen-binding pocket formed by the CDRs,
`
`and the structures of the CDR loops themselves, and thus can also affect antigen binding even
`
`though they might not directly contact the antigen.
`
`30.
`
`Numerous techniques, such as surface plasmon resonance, enzyme-linked
`
`immunosorbent assay, radioimmunoassay, competition assays, and fluorescence measurements,
`
`might be used to assay (assess) the many characteristics that contribute to an antibody’s function,
`
`such as its selectivity (ability to differentiate between different antigens) and binding properties.
`
`Each technique is based on a different scientific principle and relies on a different readout.
`
`31.
`
`There are many known methods for making antibodies. Classic immunization
`
`techniques to produce antibodies involved presenting a non-human animal, such as a mouse,
`
`with a desired antigen. Immune cells in the animal, called B cells, start to produce antibodies in
`
`response to the antigen. Those cells may then be fused with immortal cancer-derived cells to
`
`form hybridoma cell lines that can continuously produce the antibodies.
`
`32.
`
`One of the early problems with using antibodies generated in non-human animals
`
`to treat human disease was that the human patients generated immune responses against the
`
`foreign, non-human antibody (immunogenicity). The response can cause undesirable effects,
`
`such as interfering with treatment or allergic reactions. Thus, more human or fully human
`
`antibodies have been made to address these problems.
`
`33.
`
`Since the 1980s, scientists have been building non-human antigen-binding sites
`
`into human antibodies. In building these antibodies, portions of a human antibody were replaced
`
`
`
`12
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 13 of 45
`
`with portions of a non-human antibody (e.g., a mouse antibody) targeting a desired antigen, with
`
`the goal of producing an antibody with similar binding characteristics as the non-human antibody
`
`but without the undesirable immune effects caused by administering a non-human antibody to a
`
`human. There were a variety of known techniques for making partially or completely human
`
`antibodies that provided decreased risk of immunogenicity relative to a non-human antibody.
`
`VII. DISPUTED CLAIM TERMS
`
`A.
`
`“ANTI-CGRP ANTAGONIST ANTIBODY” OR “ANTI-CALCITONIN
`GENE-RELATED PEPTIDE (CGRP) ANTAGONIST ANTIBODY”
`
`Term
`“anti-CGRP
`antagonist
`antibody” or
`“anti-
`Calcitonin
`Gene-Related
`Peptide
`(CGRP)
`antagonist
`antibody”
`
`
`
`
`Lilly’s Construction
`Indefinite.
`
`In the alternative, Lilly proposes
`construing the terms as:
`
`“An antibody that is able to bind to any
`form of calcitonin gene-related peptide
`(CGRP) (including variants thereof that
`retain at least partial activity) from any
`mammalian species and inhibit CGRP
`biological activity and/or downstream
`pathway(s) mediated by CGRP
`signaling.”
`
`
`Teva’s Construction
`No construction needed in light of the
`specification.
`
`To the extent construction is needed,
`Teva proposes construing the terms as:
`
`“An antibody that is able to bind to
`CGRP and inhibit CGRP biological
`activity and/or downstream pathway(s)
`mediated by CGRP signaling.”
`
`1.
`
`A Person of Ordinary Skill in the Art Would Have No Way to
`Determine with Reasonable Certainty Whether an Antibody Is or Is
`Not an “Anti-CGRP Antagonist Antibody”
`
`34.
`
`All of the asserted claims in the Patents-in-Suit recite an “anti-CGRP antagonist
`
`antibody” or an “anti-Calcitonin Gene-Related Peptide (CGRP) antagonist antibody.” These
`
`terms are used interchangeably and differ only as to whether “CGRP” is abbreviated or spelled
`
`out in full. There is no material difference between these terms and the Patents-in-Suit likewise
`
`treat them interchangeably by providing a single definition for “CGRP” and “calcitonin gene-
`
`
`
`13
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 14 of 45
`
`related peptide.” (Ex. 1, ’045 patent at 13:55-61.) As such, my opinion applies equally to both
`
`terms.
`
`35.
`
`The specification refers to an “anti-CGRP antagonist antibody” as “an antibody
`
`that is able to [1] bind to CGRP and [2] inhibit CGRP biological activity and/or downstream
`
`pathway(s) mediated by CGRP signaling.” (Id. at 13:62-66.) In other words, as defined in the
`
`specification, the term “anti-CGRP antagonist antibody” must satisfy two requirements: it must
`
`be able to (1) bind to CGRP and (2) inhibit CGRP biological activity and/or downstream
`
`pathway(s) mediated by CGRP signaling. The specification, however, does not provide
`
`sufficient guidance for determining if either requirement is met.
`
`36.
`
`A POSITA would have known that activity and inhibition are not all-or-none
`
`phenomena for CGRP and its receptor (a G protein-coupled receptor). (See, e.g., Ex. 15,
`
`Vauquelin 2005 at 52-53.) Rather, the determination depends on the bioassay used and the
`
`threshold set, which are not specified in Teva’s patent specification. An antibody may bind to
`
`certain portions of CGRP and leave certain other portions of CGRP exposed, which may or may
`
`not bind to a receptor. If an exposed portion of CGRP binds to a receptor, it may activate the
`
`receptor (“agonist”), it may elicit a partial response from the receptor (“partial agonist”), it may
`
`elicit no response from the receptor but block others from binding the receptor (“neutral
`
`antagonist”), it may partially inhibit the receptor (“weak antagonist”), or it may produce any
`
`number of other responses. (See, e.g., id.; see also id. at 47 (Figure 2).) A POSITA would have
`
`understood these characterizations to depend on the bioassay, bioassay conditions, comparator,
`
`and threshold used. None of these parameters is specifically provided in Teva’s patent
`
`specification.
`
`
`
`14
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 15 of 45
`
`37. With respect to the first requirement (bind to CGRP), the specification is silent as
`
`to which assay should be used to determine whether or not an antibody binds to CGRP, let alone
`
`a particular degree of binding that the antibody must exhibit to meet the requirement. The lack
`
`of guidance is particularly problematic in view of the breadth of the term “CGRP,” which the
`
`patents define as “any form of calcitonin gene-related peptide and variants thereof that retain at
`
`least part of the activity of CGRP,” including α- and β-CGRP, from “all mammalian species”
`
`including “human, canine, feline, equine, and bovine.” (Ex. 1, ’045 patent at 13:55-61.) Table 6
`
`of the specification, for example, reports that Antibody M58 was able to bind to human α-CGRP
`
`(KD = 12.19 nM) but not rat α-CGRP (“No binding”). (Id. at cols. 60-65 (Table 6 (reproduced
`
`excerpt below)).) In the absence of guidance in the specification as to which assay to use to
`
`assess binding and the threshold for determining whether or not an antibody binds, a POSITA
`
`would not be able to determine with reasonable certainty whether or not a given antibody
`
`satisfies the first requirement of binding CGRP.
`
`38.
`
`The specification similarly is deficient with respect to the second requirement of
`
`inhibiting CGRP biological activity and/or downstream pathway(s) mediated by CGRP
`
`signaling. Rather than defining what it means by “CGRP biological activity” and/or
`
`
`
`
`
`15
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 16 of 45
`
`“downstream pathway(s) mediated by CGRP signaling,” the specification merely identifies non-
`
`limiting aspects of “CGRP biological activity” (identified in bold):
`
`An anti-CGRP antagonist antibody encompasses antibodies that
`block, antagonize, suppress or reduce (including significantly)
`CGRP biological activity, including downstream pathways
`mediated by CGRP signaling, such as receptor binding and/or
`elicitation of a cellular response to CGRP. For purpose of the
`present invention, it will be explicitly understood that the term
`“anti-CGRP antagonist antibody” encompasses all the previously
`identified terms, titles, and functional states and characteristics
`whereby the CGRP itself, an CGRP biological activity (including
`but not limited to its ability to mediate any aspect of headache),
`or the consequences of the biological activity, are substantially
`nullified, decreased, or neutralized in any meaningful degree.
`(Id. at 13:66-14:12.)2
`
`39.
`
`Because the specification does not specify any particular aspect of CGRP
`
`biological activity that one must consider, a POSITA would need to perform multiple different
`
`assays (in vitro, in vivo, and perhaps even in humans) measuring various aspects of CGRP
`
`biological activity, spanning from “receptor binding” to its ability to mediate “any aspect of
`
`headache.” (Id.)
`
`40.
`
`Elsewhere the specification offers similarly vague guidance for assessing
`
`candidate antibodies—any characteristic, any function, any assay, and any outcome—rather than
`
`specifying a particular means to make this determination. For instance, the specification states
`
`that “the activity of a candidate anti-CGRP antagonist antibody can be further confirmed and
`
`refined by bioassays, known to test the targeted biological activities.” (Id. at 31:57-60.) The
`
`specification provides non-limiting examples of “bioassays” such as “stimulation of cAMP in the
`
`cell” and “animal models, such as measuring skin vasodilatation induced by stimulation of the
`
`
`2 It is also unclear how an anti-CGRP antibody can inhibit “downstream pathway(s) mediated by
`CGRP signaling” independently of inhibiting “CGRP biological activity” and vice versa.
`
`
`
`16
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 17 of 45
`
`rat saphenous nerve.” (Id. at 31:60-67.) The patent specification does not specify which
`
`bioassay a POSITA should use. Likewise, no statements are made clarifying the specific testing
`
`conditions that should be used for any given bioassay or how to assess any resulting data. As
`
`such, a POSITA would be uncertain as to which or how many bioassays to conduct or what
`
`conditions to employ.
`
`41.
`
`Even if a POSITA were to conduct many of these bioassays under multiple
`
`conditions, the POSITA would not be able to determine the scope of the claims with reasonable
`
`certainty because those bioassays can and do lead to conflicting conclusions.
`
`2.
`
`Teva’s Own Data Show That Contradictory Results Can Be Produced
`From (1) A Particular Bioassay Depending on the Selected Conditions
`and (2) Different Bioassays
`
`42.
`
`A POSITA would also have known that (1) selecting different conditions for a
`
`particular bioassay can produce contradictory results and that (2) different bioassays can provide
`
`contradictory readouts of whether an antibody does or does not qualify as an anti-CGRP
`
`antagonist antibody according to the patent specification. (See Figure 3 below.) Teva’s own
`
`data in the patent specification illustrates this problem.
`
`
`
`17
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 18 of 45
`
`Figure 3
`
`
`
`43.
`
`Example 2 in Teva’s patent specification describes two different cAMP bioassays
`
`
`
`for testing potential “anti-CGRP antagonist antibodies” that provide opposite results. (Ex.
`
`1, ’045 patent at 51:5-27 (Table 2); 52:1-27 (Table 3); 53:36-55:25 (Example 2).) Specifically,
`
`in the first assay (measuring antagonist activity using human α-CGRP and SK-N-MC cells),
`
`Antibodies 6H2, 10A8, and 14E10 inhibit activation of cAMP. (Id. at 51:5-27 (“Yes” in Table 2,
`
`reproduced below, for Antibodies 6H2, 10A8, and 14E10); 54:56-58 (“antibodies that inhibit
`
`activation of cAMP by α-CGRP were identified (referred to as ‘yes’) in Tables 2 and 3”).) But
`
`when the same antibodies are tested in a cAMP bioassay using different conditions (measuring
`
`antagonist activity using rat α-CGRP and L6 cells), activation of cAMP is not inhibited. (Id.
`
`52:1-27 (“No” in Table 3, reproduced below, for Antibodies 6H2, 10A8, and 14E10); 54:56-58.)
`
`
`
`18
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 19 of 45
`
`
`
`44.
`
`These contradictory readouts confound the interpretation of whether the antibody
`
`is an “anti-CGRP antagonist antibody” as the term is used in the specification. For example,
`
`Antibody 6H2 would be considered an “anti-CGRP antagonist antibody” if assessed by the
`
`cAMP bioassay using SK-N-MC cells (Table 2). But the same antibody would not be considered
`
`an “anti-CGRP antagonist antibody” if assessed based on the cAMP bioassay using L6 cells
`
`(Table 3).
`
`45.
`
`These contradictory results for Antibody 6H2 cannot be explained by the other
`
`data presented in Table 3. (Id. at 52:1-27.) For instance, Antibody 6H2 appears to bind more
`
`strongly to rat α-CGRP (KD of 54 nM) than Antibodies 8B6 and 7D11 (KD of 75 nM and 218
`
`
`
`19
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 20 of 45
`
`nM, respectively).3 (Id.) Yet according to the cAMP bioassay data in Table 3, Antibodies 8B6
`
`and 7D11 would be considered “anti-CGRP antagonist antibodies” while Antibody 6H2 would
`
`not. (Id.)
`
`46. While the specification states that antibodies “having KD (determined at 37° C) of
`
`about 47 nM or less to rat α-CGRP showed antagonist activity in [the cAMP] assay” (id. at
`
`54:60-64), as noted above, Antibodies 8B6 and 7D11 with KD values higher than 47 nM to rat α-
`
`CGRP are identified as having antagonist activity (“Yes”) (id. at 52:1-27) . Thus, the
`
`specification’s generalized statement does nothing to resolve any uncertainty.
`
`47.
`
`For the same reason, the recited binding affinity values such as 47 nM to rat α-
`
`CGRP cannot be used as a threshold for determining whether or not an antibody is able to inhibit
`
`CGRP biological activity and/or downstream pathway(s) mediated by CGRP signaling. (See
`
`also infra, § VII.A.3.)
`
`48.
`
`Conflicting conclusions also result when comparing outcomes from the cAMP
`
`assays and a rat saphenous nerve assay, another “bioassay” identified in the specification. (Ex.
`
`1, ’045 patent at 31:65-66.) As explained in the specification, the rat saphenous nerve assay
`
`“measur[es] skin vasodilation induced by stimulation of the rat saphenous nerve.” (Id.) Notably,
`
`Table 3 indicates that Antibody 6H2 did not block vasodilation in this model. (Id. at 52:1-27
`
`(“No” in Table 3).) Thus, Antibody 6H2 would not be considered an “anti-CGRP antagonist
`
`antibody” based on the rat saphenous nerve assay but would be considered an “anti-CGRP
`
`antagonist antibody” based on the cAMP assay using SK-N-MC cells discussed above.
`
`49.
`
`Similarly, based on Teva’s own data, Antibody 14E10 would be considered an
`
`“anti-CGRP antagonist antibody” according to the patent specification if assessed based on the
`
`
`3 As explained below, a higher KD value corresponds to a lower binding affinity.
`
`
`
`20
`
`

`

`Case 1:18-cv-12029-ADB Document 68 Filed 09/11/20 Page 21 of 45
`
`cAMP bioassay using SK-N-MC cells (Ta