United States District Court, D. Delaware
OPINION & ORDER
HONORABLE TIMOTHY B. DYK, UNITED STATES CIRCUIT JUDGE
4, 2017, Baxalta Inc. and Baxalta GmbH (together,
"Baxalta") filed suit against Genentech, Inc. and
Chugai Pharmaceutical Co., Ltd. alleging infringement of
claims 1, 4, 17, and 19 of U.S. Patent No. 7, 033, 590 patent
("the '590 patent"). Chugai was voluntarily
dismissed from this lawsuit pursuant to a stipulation of the
parties on September 19, 2018. Order Dismissing Chugai, ECF
alleged infringement is the manufacture, use, sale, offer to
sell, and importation of an antibody used to treat hemophilia
A and known as emicizumab, or ACE910, and marketed under the
brand name Hemlibra (hereinafter, "Hemlibra"). Now
before this court is the claim construction of six terms of
the '590 patent: antibody, antibody fragment,
bispecific antibody, isolated, binds Factor IX or Factor IXa
and increases, and increases the procoagulant
activity of Factor Ixa.
court held a Markman hearing on October 16, 2018, and
received expert testimony and argument regarding the
construction of the six terms. At an earlier preliminary
injunction hearing on June 13 and 14, 2018, the court also
received testimony and argument on construction of the terms
antibody and antibody fragment.
terms of the factual record, the court will consider oral
testimony given by experts at the Markman hearing, testimony
offered at the preliminary injunction hearing, and the
deposition testimony and reports and declarations of any of
those experts, but the court declines to consider the
declarations of experts who have not been subject to
cross-examination at either hearing.Whether or not such
declarations are considered makes no difference to the
constructions adopted by the court here.
4, 2017, Baxalta filed its complaint alleging infringement of
the '590 patent. Compl. ¶¶ 37-51, ECF No. 1. On
June 30, Genentech answered, denying Baxalta's
allegations and counterclaiming for declaratory judgment of
noninfringement and invalidity. Answer & Countercl.
¶¶ 37-51, 120-49, ECF No. 9.
December 14, 2017, Baxalta moved for a preliminary
injunction. Mot. Prelim. Inj. 2, ECF No. 41; Prop. Prelim.
Inj. Order 1, ECF No. 42-1. After an evidentiary hearing on
August 7, 2018, this court denied Baxalta's preliminary
injunction motion. Prelim. Inj. Order at 29, ECF No. 262. In
that connection, the court declined to construe
antibody and antibody fragment, concluding
that the parties had presented "substantial
arguments" on both sides. Id. at 13. Baxalta
did not appeal the denial of the preliminary injunction.
Discovery has been ongoing. Fact discovery is set to close on
December 14, 2018, and expert discovery is set to close on
April 19, 2019. Stip. & Order Amend. Sched. 1, ECF No.
Markman hearing, the parties presented expert testimony and
argued construction of six terms: antibody, antibody
fragment, bispecific antibody, isolated, binds Factor IX
or Factor IXa and increases, and increases
the procoagulant activity of Factor IXa. All these terms
appear in claims 1 and 4 of the '590 patent, which recite
1. An isolated antibody or antibody
fragment thereof that binds Factor IX or Factor
IXa and increases the procoagulant activity of
* * *
4. The antibody or antibody fragment according to claim 1,
wherein said antibody or antibody fragment is selected from
the group consisting of a monoclonal antibody, a chimeric
antibody, a humanized antibody, a single chain antibody, a
bispecific antibody, a diabody, and di-, oligo- or
'590 patent, col. 101, ll. 43-45, 51-56 (underlining
Coagulation & Hemophilia A
general, the term antibody is used to describe
glycoproteins that are "characterized by their ability
to bind both to antigens and to specialized cells or proteins
of the immune system." Strohl Decl. ¶22, ECF No.
112; accord Almagro Decl. ¶ 33, ECF No. 49.
Structurally, antibodies are T-shaped, with two arms that are
connected by disulfide bonds. Almagro Decl. ¶ 34; Strohl
Decl. ¶ 22. Each arm of the Y contains two
polypeptide chains known as the heavy ("H") chain
and the light ("L") chain. Almagro Decl. ¶ 34;
Strohl Decl. ¶ 22. The portions of the heavy chain and
light chain that are responsible for binding an antigen are
called variable domains, VH and VL
respectively. Almagro Decl. ¶ 34; Strohl Decl. ¶23.
The remaining portions of the antibody are made up of
constant regions. Almagro Decl. ¶ 34; Strohl Decl.
each variable domain, the antigen binding sequence of the
antibody is divided into three regions called
complementarity-determining regions ("CDRs").
Almagro Decl. ¶ 35; Strohl Decl. ¶ 25. The three
CDRs in each variable region-designated CDR1, CDR2, and
CDR3-determine the binding specificity of the antibody,
Almagro Decl. ¶ 35; Strohl Decl. ¶25. The CDR3
region of the heavy chain variable domain is "primarily
responsible for antigen binding specificity." Strohl
Decl. ¶ 25; accord Almagro Decl. ¶ 38.
the parties agree as to these characteristics of an antibody,
they disagree in at least one critical respect. Genentech
contends that, as used in the patent, the term
antibody standing alone has two heavy chains that
are identical and the two light chains that are identical.
Genentech Op. Br. 6-8, ECF No. 160; Strohl Decl. ¶ 50.
Baxalta, in contrast, argues that the heavy chains are not
necessarily identical to one another and the light chains are
also not necessarily identical to one another. Baxalta Op.
Br. 4-5, ECF No. 158. The resolution of this difference
appears to be determinative of infringement.
A and the process of blood coagulation are described at
length in the preliminary injunction order. See
Prelim. Inj. Order 3-4. Relevant here is one particular step
of the clotting cascade involving Factor Villa and Factor
IXa. See Aledort Decl. ¶ 13, ECF No. 46. In
healthy individuals, Factor Villa and Factor IXa form a
complex, which allows Factor IXa to activate Factor X.
See id.; Sheehan Decl. ¶36, ECF No. 111. In
patients afflicted with hemophilia A, Factor VIII is reduced,
defective, or absent. See Aledort Decl. ¶ 14;
Sheehan Decl. ¶ 42. This hinders the coagulation cascade
by limiting the body's ability to activate Factor X.
Aledort Decl. ¶ 14; Sheehan Decl. ¶ 42.
drug, Hemlibra, is directed to this step of the coagulation
cascade and functions by replacing Factor Villa. See
Krishnaswamy Decl. ¶ 61, ECF No. 47. Hemlibra does not
have both identical light and identical heavy chains.
See Strohl Decl. ¶¶ 38, 53; see
also Krishnaswamy Decl. ¶¶ 55, 60. It is a
bispecific antibody. The parties agree that in this patent a
bispecific antibody has non-identical light chains,
or non-identical heavy chains, or both.Markman Tr.
38:5-21, ECF No. 320. One arm of the Hemlibra antibody binds
to Factor IX (or IXa) and the other binds to Factor X.
See Krishnaswamy Decl. ¶¶ 55, 60; Strohl
Decl. ¶ 53. By doing so, Hemlibra allows Factor IX to
activate Factor X. See Krishnaswamy Decl. ¶ 61,
Strohl Decl.¶¶ 178-79.
proposed construction: A molecule having a specific
amino acid sequence comprising two heavy chains (H chains)
and two light chains (L chains).
proposed construction: An immunoglobulin molecule,
having a specific amino acid sequence that only binds to the
antigen that induced its synthesis or very similar antigens,
consisting of two identical heavy chains (H chains) and two
identical light chains (L chains).
construction: An immunoglobulin molecule, having a
specific amino acid sequence that only binds to the antigen
that induced its synthesis or very similar antigens,
consisting of two identical heavy chains (H chains) and two
identical light chains (L chains).
the claims requires resolution of the parties' primary
dispute that an antibody in the claims is required
to have two identical heavy chains and two identical light
chains. The parties agree that the requirement
that an antibody have two identical heavy chains and two
identical light chains would exclude Hemlibra from the scope
of the term antibody. Prelim. Inj. Tr. 9:15-24, ECF
No. 214-15; Markman Tr. 109:23-110:8. Hemlibra is not an
antibody under Genentech's definition.
The Meaning of the Term Antibody in the Patent as
clear from the '590 patent's specification that, as
originally drafted, the term antibody in the claims
required identical heavy and identical light chains.
on the evidence, I find that the term antibodies
does not have a single fixed meaning in the art. The word
antibody can denote different meanings to a person
skilled in the art depending on the context in which it
appears. For example, antibody standing alone may
connote a different meaning than when it is part of a larger
term that defines its structure-e.g., bispecific
parties agree that the term antibody standing alone
without other structural terms can have different meanings to
those skilled in the art. See Markman Tr.
174:21-175:24. One definition is Baxalta's definition
(hereinafter the "broader" definition), requiring
only a molecule with a specific amino acid sequence and
comprising two heavy chains and two light chains. The other
definition is Genentech's definition (hereinafter the
"narrower" definition), requiring a pair of
identical heavy chains and a pair of identical light chains.
Baxalta argues that its broader definition should apply
because it would have been utilized by persons of ordinary
skill in the art. Baxalta Op. Br. 5. But in its opening
preliminary injunction brief, Baxalta itself used the
narrower definition, stating that "[a]n antibody
comprises two identical heavy chains and two identical light
chains." Baxalta Op. Prelim. Inj. Br. 13 n.7, ECF No.
42. Similarly, Dr. Almagro, Baxalta's expert, described
an antibody in his declaration as "a glycoprotein that
has a specific 'Y' shape" that "has two
pairs of identical polypeptide chains, which are linked
together by disulfide bonds." Almagro Decl. ¶ 34.
Genentech's expert agreed that such a definition
"represents the plain and ordinary meaning of
'antibody' and is consistent with standard textbook
definitions of immunoglobulin molecules dating from prior to
the 1999 priority date of the '590 [p]atent." Strohl
CI. Const. Decl. ¶¶ 41-42, ECF No. 161; accord
Id. ¶ 44.
references cited on the face of the '590 patent use a
similar definition. The Roitt reference describes an antibody
as "a unit consisting of two identical light polypeptide
chains and two identical heavy polypeptide chains." Ivan
Roitt et al., Immunology 72 (5th ed. 1998), Strohl Decl. Ex.
F, ECF No. 112-6. And the Harlow and Lane reference says that
"[e]ach Y contains four polypeptides[:] [t]wo identical
copies of a polypeptide known as the heavy chain and two
identical copies of a polypeptide called the light
chain." Ed Harlow & David Lane, Antibodies: A
Laboratory Manual 7 (1988), Strohl Decl. Ex. E, ECF No.
112-5. Baxalta does not point to any references cited in the
'590 patent that use a broader definition.
the parties agree that Baxalta's broader definition was
also known to those skilled in the art. See Markman
Tr. 172:6-176:6. During the Markman hearing, Dr. Strohl,
Genentech's expert testified that the broader definition
was a "common language definition." Id. at
175:12-13. Baxalta only points to Dr. Strohl's testimony
as evidence of the understanding of someone skilled in the
patent specification, the applicant chose the narrower
definition. In relevant part, the summary of the invention
Antibodies are immunoglobulin molecules having a specific
amino acid sequence which only bind to antigens that induce
their synthesis (or its immunogen, respectively) or to
antigens (or immunogens) which are very similar to the
former. Each immunoglobulin molecule consists of two types of
polypeptide chains. Each molecule consists of large,
identical heavy chains (H chains) and two light, also
identical chains (L chains).
'590 patent, col. 5, ll. 56-63. The Federal Circuit has
recognized that use of the verb "is" may
"signify that a patentee is serving as its own
lexicographer." Sinorgchem Co. v. Int'l
Trade Comm'n, 511 F.3d 1132, 1136 (Fed.
Cir. 2007) (quoting Abbott Labs. v. Andrx Pharms.,
Inc., 473 F.3d 1196, 1210 (Fed. Cir. 2007)). The use of
the term "are" here is the equivalent of the term
"is." Here, the specification unequivocally states
what "[a]ntibodies are." '590 patent, col. 5,
ll. 56-63. This definition is also clearly defining the term
antibodies covered by the claims of this patent
because the definition immediately follows and immediately
precedes references to "the inventive antibodies and
antibody derivatives." '590 patent, col. 5, 1. 53;
id col. 6, 1. 1. The fact that the applicants chose
to include the narrower definition in the specification over
a broader definition confirms that the applicants intended
the narrow definition apply to the term antibody
argues that the specification in other places uses the
broader definition. Baxalta Op. Br. 6-7. For example, the
specification and claims disclose bispecific antibodies,
which do not have identical heavy and light chains. See
id; '590 patent col. 6, ll.. 1-5; id. col.
7, ll.. 32-35; id. col. 101, ll.. 51-56 (claim 4).
Baxalta also points to the inclusion of IgM antibodies and
IgA antibodies in claims 3 and 20 as well as throughout the
specification. '590 patent, col. 6, ll. 35-38;
id. col. 12, ll.. 25-26; id col. 14, 1.
22-col. 15, 1.4 (Example 4); id. col. 30, 1.
ll.-col. 31, 1. 10 (Example 13); id. col. 101, ll.
49-50 (claim 3); id. col. 104, ll.. 6-7 (claim 20).
IgM and IgA antibodies can have more than two heavy chains
and more than two light chains. Markman Tr. 71:11.-13,
156:12-159: 11. Baxalta thus contends that this limitation of
the narrow definition of antibody is inappropriate
in the context of the '590 patent. Baxalta Op. Br. 6-7.
But all these embodiments were initially listed as falling
within "antibodies or antibody derivatives."
See, e.g., '590 patent, col. 5, 1. 51; U.S.
Patent App. No. 09/661, 992, at 10-11, Strohl Decl. Ex. D,
ECF No. 112-4. Thus, as originally drafted, the claims
covered antibodies as more broadly defined, but not because
they fell within the term antibody but because they
fell within the term antibody derivative.
such circumstances, the Federal Circuit has held that the
specification's choice of definition governs. See
Sinorgchem, 511 F.3d at 1136-40 ("Where, as here,
multiple embodiments are disclosed, we have previously
interpreted claims to exclude embodiments where those
embodiments are inconsistent with unambiguous language in the
patent's specification or prosecution history.");
Irdeto Access, Inc. v. Echostar Satellite Corp., 383
F.3d 1295, 1300 (Fed. Cir. 2004) (concluding that the scope
of the claim terms was controlled by the specification even
in the absence of express definitions where applicant
admitted to the examiner that the terms had "no accepted
meaning in the art" and were "adequately described
in the specification"). Indeed, given the
specification's clarity, the definition included in
column 5 of the '590 patent would govern even if it were
contrary to an ordinary meaning of the term. See Thorner
v. Sony Comp. Entmt. Am. LLC, 669 F.3d 1362, 1365-66
(Fed. Cir. 2012) ("[T]he inventor's written
description of the invention, for example, is relevant and
controlling insofar as it provides clear
lexicography" (alterations and emphasis in
original) (quoting C.R. Bard, Inc. v. U.S.
Surgical Corp., 388 F.3d 858, 862 (Fed. Cir. 2004))).
turning to the prosecution history, it is important to
ascertain the meaning of the term antibody
derivative in the patent as initially drafted. The
parties agree that antibody derivative is not a term
that is commonly used in the art. Markman Tr. 119:21-120:3.
But Dr. Almagro, Baxalta's expert, admitted during the
preliminary injunction hearing, and again during the Markman
hearing, that antibodies that have been altered in some
significant way are "sometimes... called
derivatives." Prelim. Inj. Tr. 413:4-15; accord
Markman Tr. 120:6-11 (Dr. Almagro agreeing with previous
testimony that "significant variants" of antibodies
are "sometimes.. . called derivatives"). Also in
support of this understanding, Dr. Almagro, a person of skill
in the art, repeatedly described Hemlibra, a bispecific
antibody, as being "derived" from other
antibodies. Almagro Decl. ¶ 54 n. 5, ¶ 87 n.
17; Markman Tr. 122:8-24; see also id 120:6-21 (Dr.
Almagro agreeing that "[t]alking in plain English"
Hemlibra" is "derived from a Factor IX antibody and
a Factor X antibody"). This definition of antibody
derivatives-an antibody that has been altered in some
significant way-is consistent with the specification. First,
the specification makes clear that the group consisting of
antibodies and antibody derivatives
includes bispecific antibodies and other structures that do
not have identical light and heavy chains. Since bispecific
antibodies are not within the definition of
antibodies, they must be within the definition of
understanding is further supported by the uses of
antibody derivative in those places in the patent
specification where antibody derivative is used
separately. Antibody derivative is used separately
in three instances that inform the interpretation of the
term: (1) in Example 10, which is entitled
"Structure and Procoagulant Activity of Antibody
Derivatives Derived from Anti-FIX/FIXa-antibodies; Subcloning
Antibody Variable Domains from Hybridoma Cell Lines,"
id. col. 19, ll.. 3-7; (2) in the body of Example
11, where the patent includes in a list of examples of
antibody derivatives "scFv, Fab, etc.,"
id. col. 20, 1. 36; and (3) again in Example 13,
where the patent refers to "antibody derivatives such as
Fab, F(ab)2, scFv, etc.," id. col.
30, ll.. 16-17. See also Markman Tr. 13:19-15:6.
uses make clear that Fab, F(ab)2, and scFv are all
antibody derivatives. As Dr. Almagro and Dr. Strohl agree,
Fab and F(ab)2 are the sort of canonical
antibody fragments (a subset of derivatives) that a
person of skill in the art would unquestionably have
understood as such. Markman Tr. at 117:7-14 (Dr. Almagro),
155:19-23 (Dr. Strohl). Each of these can be derived from an
existing antibody as defined in the specification. A
Fab comprises "the complete light chain paired with
the full variable and a portion of the constant domain of
the heavy chain" and can be excised from an existing
antibody. See Strohl Claim Const. Decl. ¶ 31
& Fig. 2, ECF No. 161; accord Markman Tr.
151:18-152:24 (Dr. Strohl). A F(ab)2, comprises
"two Fab fragments linked with disulfide bonds" and
also "can be generated from an antibody by cleaving off
the other portions" to leave the fragment remaining.
Markman Tr. 152:25- 153:11 (Dr. Strohl); accord
Strohl Decl. ¶ 32 & Fig. 3. Based on inclusion of
Fab and F(ab)2 it is clear that antibody
fragments are antibody derivatives.
specification makes clear that an scFv is not an antibody
fragment using the definition of antibody from
the specification. Rather, it is called a single-chain
variable fragment and is synthetically created by linking
with a stretch of synthetic peptide "a truncated
fragment comprising only the [variable heavy] domain" of
an antibody with a truncated fragment comprising only the
variable light region of an antibody. Strohl Claim Const.
Decl. ¶ 33 & Fig. 4; accord Almagro Decl.
¶ 35; Markman Tr. 60:2-6, 201:17-22. Because of the
reference to scFv as an antibody derivative, the
term antibody derivative was clearly meant to
include antibodies that have been altered in some
find that the term antibody derivative was used in
the patent to denote antibodies within the column 5
definition that had been altered in some significant way. As
initially drafted, there was no inconsistency between ...