Astrazeneca Ab, Aktiebolaget Hassle, Astrazeneca Lp, Kbi Inc., and Kbi-Einc. v. Hanmi USA, Inc., Hanmi Pharmaceutical Co., Ltd., Hanmi Fine Chemical Co., Ltd.

N OTE: This disposition is nonprecedential.
United States Court of Appeals
for the Federal Circuit
______________________
ASTRAZENECA AB, AKTIEBOLAGET HASSLE,
ASTRAZENECA LP, KBI INC., AND KBI-E INC.,
Plaintiffs-Appellants,
v.
HANMI USA, INC., HANMI PHARMACEUTICAL
CO., LTD., HANMI FINE CHEMICAL CO., LTD.,
AND HANMI HOLDINGS CO., LTD.,
Defendants-Appellees.
______________________
2013-1490
______________________
Appeal from the United States District Court for the
District of New Jersey in No. 11-CV-0760, Judge Joel A.
Pisano.
______________________
Decided: December 19, 2013
______________________
J AY I. ALEXANDER , Covington & Burling LLP, of
Washington, DC, argued for plaintiffs-appellants. With
him on the brief were EINAR STOLE and ERIC R.
SONNENSCHEIN. Of counsel on the brief were HENRY J.
RENK, BRUCE C. HAAS , and J OSHUA I. ROTHMAN, Fitzpat-
rick, Cella, Harper & Scinto, of New York, New York.

-- 1 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 2
BLAIR M. J ACOBS , McDermott Will & Emery, LLP, of
Washington, DC, argued for defendants-appellees. With
him on the brief were CHRISTOPHER G. P AULRAJ and
P ATRICK J. STAFFORD. Of counsel on the brief were MARK
BOLAND, MICHAEL R. D ZWONCZYK, and RENITA S.
RATHINAM , Sughrue Mion, PLLC, of Washington, DC.
______________________
Before D YK, M OORE, and T ARANTO, Circuit Judges.
T ARANTO, Circuit Judge.
Plaintiffs AstraZeneca AB, Aktiebolaget Hassle,
AstraZeneca LP, KBI Inc., and KBI-E Inc. (collectively,
AstraZeneca) sued Defendants Hanmi USA, Inc., Hanmi
Pharmaceutical Co., Ltd., Hanmi Fine Chemical Co., Ltd.,
and Hanmi Science Co., Ltd., formerly Hanmi Holdings
Co., Ltd. (collectively, Hanmi). Invoking 35 U.S.C.
§ 271(e)(2), AstraZeneca alleged that a drug Hanmi
proposed to market falls within claims of U.S. Patent Nos.
5,714,504 and 5,877,192. After the district court con-
strued the claim terms “alkaline salt” in the ’504 patent
and “pharmaceutically acceptable salt” in the ’192 patent,
the parties consented to the entry of a final judgment of
noninfringement based on the constructions.
This appeal presents a single issue: whether the writ-
ten description limits “alkaline salt” in the ’504 patent to
certain specifically named salts. We hold that it does.
The written description describes the invention clearly
and narrowly as including only those salts, and Astra-
Zeneca points to nothing in the intrinsic record that is
sufficient to overcome that disclaimer.
BACKGROUND
Omeprazole is an “effective gastric acid secretion in-
hibitor[], and [is] useful as [an] antiulcer agent[].” ’504
patent, col. 1, lines 22-23. Two distinct molecules have
omeprazole’s molecular formula and sequence of bonded

-- 2 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 3
atoms. These “enantiomers” of omeprazole are mirror-
images, which cannot be superimposed on each other. A
mixture of the enantiomers in equal amounts is a “race-
mate” of omeprazole (or “racemic” omeprazole).
Creating a salt out of omeprazole can enhance stabil-
ity during storage and transportation, a useful property in
pharmaceutical compounds. See J.A. 5442 (describing the
increased stability of certain salts of the racemate). Salts
are chemical compounds composed of two oppositely
charged ions: one positive (the cation) and the other
negative (the anion). In an omeprazole salt, the omepra-
zole molecule is the anion. Several cations have proved
suitable for omeprazole, including the metals from Groups
IA and IIA of the Periodic Table. J.A. 5259.
AstraZeneca discovered that certain salts of an
omeprazole enantiomer, as opposed to the racemate, have
“improved pharmacokinetic and metabolic properties
which will give an improved therapeutic profile such as a
lower degree of interindividual variation.” ’504 patent,
col. 1, lines 51-54. Its original application for the ’504
patent, filed in 1995 as a continuation-in-part of a 1994
application, described and claimed particular salts, de-
fined by six identified cations: Na+ , Mg 2+ , Li + , K + , Ca 2+ , or
N + (R)4, where R is an alkyl with one to four carbons. Id.
col. 1, line 1, to col. 6, line 35; J.A. 82-85. (The last formu-
la denotes a class of ammonium cations, but for present
purposes we may refer to it with the singular “cation”—
making six cations in all.)
During prosecution, AstraZeneca conducted experi-
ments that led it to conclude that one of the two enantio-
mers gave particularly good results. J.A. 312-25. The
preferred enantiomer is known as “(-)-omeprazole” or “(S)
omeprazole,” sometimes written as “esomeprazole.” In
early 1997, in response to the Examiner’s rejection of
original claims, AstraZeneca filed amended claims to
focus on that enantiomer. J.A. 121, 296-309; see also ’192

-- 3 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 4
patent, col. 2, lines 28-34 (continuation-in-part filed in
April 1997, stating: “[O]ne of the enantiomers of omepra-
zole . . . is hereby claimed to be an improved alternative to
omeprazole in the treatment of gastric acid related dis-
eases resulting in higher dose efficiency and in less inter-
individual variation in plasma levels.”). The new claims,
now at issue, are all limited to pharmaceutical compounds
that contain certain esomeprazole salts as an active
ingredient; but the independent claims no longer express-
ly refer to the originally identified six cations, instead
claiming an “alkaline salt” or “pharmaceutically accepta-
ble salt.” See ’504 patent, col. 14, lines 5-49; ’192 patent,
col. 7, line 17, to col. 8, line 54.
AstraZeneca sells Nexium®, a product whose active
ingredient is the magnesium (Mg 2+ ) salt of esomeprazole,
magnesium being one AstraZeneca’s original six cations.
In December 2010, Hanmi filed an application with the
Food and Drug Administration under 21 U.S.C.
§ 355(b)(2) seeking approval to sell a product that con-
tains the strontium (Sr2+ ) salt of esomeprazole, strontium
not being one of AstraZeneca’s original six cations. The
application certified that the ’504 and ’192 patents are
invalid or would not be infringed by Hanmi’s proposed
product. On February 9, 2011, AstraZeneca filed suit,
alleging that Hanmi’s proposed product infringed the
claims of the ’504 and ’192 patents under 35 U.S.C.
§ 271(e)(2)(A).
On December 12, 2012, the district court construed
the term “alkaline salt” in the ’504 patent and “pharma-
ceutically acceptable salt” in the ’192 patent. AstraZeneca
argued that both terms have the same broad meaning:
any “basic” salt of esomeprazole that is suitable for use in
a pharmaceutical formulation. Hanmi argued that both
terms are limited to the disclosed “Na+ , Mg2+ , Li+ , K+ , Ca 2+
or N + (R)4 salts of the single enantiomers of omeprazole.”
’504 patent, col. 2, lines 42-44. The district court agreed
with Hanmi, concluding that the written description

-- 4 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 5
defines the invention as limited to the disclosed salts.
AstraZeneca AB v. Hanmi USA, Inc., No. 11-CV-0760,
2012 WL 6203602, at *3-4 (D.N.J. Dec. 12, 2012). And
because the court held that the ’192 patent incorporates
the ’504 patent’s disclosure, it construed “pharmaceutical-
ly acceptable salt” the same way. Id. at *6-7.
After the district court denied AstraZeneca’s motion
for reconsideration, the parties consented to the entry of a
final judgment that the Hanmi product does not infringe
under the district court’s claim construction. Consent
Order and Final Judgment, AstraZeneca AB v. Hanmi
USA, Inc., No. 11-CV-0760, Dkt. No. 338 (D.N.J. June 3,
2013). Hanmi stipulated that both patents are valid and
enforceable. Id. AstraZeneca timely appealed, and we
have jurisdiction under 28 U.S.C. § 1295(a)(1).
D ISCUSSION
The only issue on appeal is the proper construction of
the claim term “alkaline salt,” a question that we decide
de novo. E.g., Cybor Corp. v. FAS Techs., Inc., 138 F.3d
1448, 1454-56 (Fed. Cir. 1998). Because the written
description of the ’504 patent contains a clear disclaimer
of any salt except those using six enumerated cations, we
agree with the district court that “alkaline salt” is limited
to the Na+ , Mg 2+ , Li+ , K + , Ca 2+ , and N+ (R)4 salts of the
now-claimed enantiomer of omeprazole.1
Independent claim 1 of the ’504 patent claims
[a] pharmaceutical formulation for oral admin-
istration comprising a pure solid state alkaline
1 AstraZeneca agrees that the salt limitations in
both patents should be construed the same way. Br. of
Appellant at 8. Like the district court, therefore, we focus
on the ’504 patent.

-- 5 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 6
salt of the (-)-enantiomer of [omeprazole] and a
pharmaceutically acceptable carrier.
Id., col. 14, lines 6-10; see also id., col. 14, lines 21-26
(independent claim 6, a method of administering “a pure
solid state alkaline salt of the (-)-enantiomer of [omepra-
zole]”); id., col. 14, lines 27-34 (independent claim 7). It is
undisputed that the term “alkaline salt,” on its face and
outside the context of the ’504 patent, would not be lim-
ited to the six cations named in the district court’s claim
construction. But we agree with the district court that
this is a patent in which the written description, by clear
disclaimer, limits the claim scope to those cations.
The first sentence of the Detailed Description de-
clares:
The present invention refers to the new Na+ ,
Mg 2+ , Li +, K+ , Ca 2+ or N + (R)4 salts of the single en-
antiomers of omeprazole, where R is an alkyl with
1-4 carbon atoms, i.e. Na +, Mg 2+ , Li+ , K+ , Ca 2+ or
N + (R)4 salts of (+)-[omeprazole] and
(-)-[omeprazole], where R is an alkyl with 1-4 car-
bon atoms.
Id., col. 2, lines 42-49. That language clearly defines “the
present invention” not as salts of omeprazole, or salts of
single enantiomers of omeprazole, but as a particular set
of “new” salts of enantiomers of omeprazole, limited to the
six named cations. The Abstract, though not grammati-
cally a sentence, confirms the limiting disclaimer by
identifying what AstraZeneca said was “novel”: “The novel
optically pure compounds Na + , Mg 2+ , Li+ , K +, Ca 2+ or
N + (R)4 salts of (+)-[omeprazole] or (-)-[omeprazole], in
particular sodium and magnesium salt form thereof . . .”
Id., Abstract.
Those statements clearly confine the invention to the
six identified cations, disclaiming anything else. See, e.g.,
Verizon Servs. Corp. v. Vonage Holdings Corp., 503 F.3d

-- 6 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 7
1295, 1308 (Fed. Cir. 2007); Honeywell Int’l, Inc. v. ITT
Indus., 452 F.3d 1312, 1318-19 (Fed. Cir. 2006); SciMed
Life Sys., Inc. v. Advanced Cardiovascular Sys., Inc., 242
F.3d 1337, 1340-45 (Fed. Cir. 2001). The very specificity
and origin of the list of cations for salts confirm the plain
meaning of the disclaimer language. AstraZeneca’s
expert attested that a skilled artisan would presumptively
understand that all of the metals of Periodic Table Groups
IA and IIA, plus ammonium as an “honorary” member of
those Groups for these purposes, would be suitable for
forming a salt with the negatively charged enantiomer.
J.A. 5354. AstraZeneca chose only certain members of
that presumptively suitable class: the first five of the six
cations are metals in Groups IA and IIA, and the sixth is
ammonium. By conspicuously choosing only certain
members of the class, and using the language it did,
AstraZeneca conveyed a clear and definitive meaning that
it was disclaiming other members of the class—like
Hanmi’s chosen strontium, another metal from Group IIA,
immediately below calcium in the Periodic Table.
AstraZeneca advances three arguments for concluding
otherwise. These arguments do not suffice to overcome
the clear disclaimer language of the written description.
First, AstraZeneca asserts that one passage in the
written description shows an intent to cover salts other
than those of the listed six cations. The sentence Astra-
Zeneca relies on reads: “Alkaline salts of the single enan-
tiomers of the invention are, as mentioned above, beside
the sodium salts . . . and the magnesium salts . . . , exem-
plified by their salts with Li + , K + , Ca 2+ or N + (R)4, where R
is an alkyl with 1-4 C-atoms.” ’504 patent, col. 5, lines 7-
11. AstraZeneca reads the “exemplified” language to
mean that all six of the cations identified in column 2 are
just examples of a broader group that is the invention.
But the sentence does not say that. The “exemplified”
language applies directly to only four of the six cations

-- 7 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 8
and is not preceded by the word “also” or any other lan-
guage that might have more affirmatively suggested an
intent to treat all six identified cations as merely exem-
plary. The context is also important: the sentence comes
after two paragraphs describing the salts from Periodic
Table Group IA (involving sodium) and the salts from
Group IIA (involving magnesium) and says that it is just
repeating information that was “mentioned above.” In
context, we think that the AstraZeneca-highlighted
sentence is best read as summarizing the two preceding
paragraphs, neither of which suggests that the invention
is a broader class of salts than the group identified in
column 2. This sentence does not negate the clear dis-
claimer language quoted above.
AstraZeneca next turns to the prosecution history, but
it too does not overcome the clear disclaimer language in
the patent. When AstraZeneca filed the application that
issued as the ’504 patent, the language of the claims
aligned perfectly with the written description’s clear
language about the scope of the “present invention”: the
broadest of the claims were limited in terms to salts using
the six identified cations, combined with either one of the
two omeprazole enantiomers. J.A. 82. After the Examin-
er rejected those claims for anticipation and obviousness,
J.A. 117-18, AstraZeneca shifted the focus to unexpected
benefits achieved by using the (-)-enantiomer rather than
the (+)-enantiomer. Whereas the original claims and
written description treated the two enantiomers with
parity, AstraZeneca now distinguished the prior art by
amending the claims to cover only esomeprazole, which it
argued “unexpectedly exhibits a different and more ad-
vantageous pharmacokinetic profile than the racemic
mixture or the (+)-enantiomer of omeprazole.” J.A. 300.
To support that assertion, AstraZeneca submitted clinical
studies that, it explained to the Examiner, “involved both
the monovalent sodium salt and the divalent magnesium
salt of the (-)-enantiomer of omeprazole, thus supporting

-- 8 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 9
the full scope of the genus of alkaline salts disclosed in the
application and as claimed herein.” Id. (emphasis added).
AstraZeneca relies on the italicized language to argue
for a broadening that overcame the disclaimer in the
patent itself. But we think that AstraZeneca is making
more of this passage than the language supports. In the
context of an amendment that otherwise narrowed the
rejected claims, and a written description that so clearly
limited the invention to the identified salts, the statement
on which AstraZeneca relies is, at a minimum, not clear
enough to overcome the limitation.
Neither this statement nor anything else in the prose-
cution history states that there is a substantial expansion
being undertaken. The prosecution history does not
mention, or include data for, any salt beyond the salts
identified in the written description. And contrary to
AstraZeneca’s contention, the reference to “the full scope
of the genus of alkaline salts . . . as claimed” does not
establish a disclaimer-overriding expansion to a wide
class of alkaline salts beyond the alkaline salts identified
in column 2. For one thing, the term “genus” can refer
simply to an enumerated collection, without an inde-
pendently unifying characteristic of the collection’s mem-
bers—as when “genus” is used for a Markush-type claim
that recites a group whose members may have nothing in
common but their membership in the group. See, e.g., Eli
Lilly & Co. v. Teva Parenteral Meds., Inc., 689 F.3d 1368,
1373 (Fed. Cir. 2012) (“generic, Markush-style claims”);
Merck & Co. v. Mylan Pharms., Inc., 190 F.3d 1335, 1340
(Fed. Cir. 1999) (“Markush genus”); Abbott Labs. v. Baxter
Pharm. Prods., Inc., 334 F.3d 1274, 1280 (Fed. Cir. 2003)
(“‘A Markush group is a sort of homemade generic expres-
sion . . .’”). In any event, the full language at issue refers
to the “genus of alkaline salts disclosed in the application
and as claimed herein.” (Emphasis added.) That lan-
guage is naturally understood to limit the genus being
described to the particular salts “disclosed in the applica-

-- 9 of 10 --

ASTRAZENECA AB v. HANMI USA, INC. 10
tion,” i.e., those based on the six enumerated cations. It
certainly is not clear in conveying a broader meaning.
Finally, AstraZeneca presses an argument based on
claim differentiation, noting that each independent claim
reciting an “alkaline salt” has a dependent claim that
differs only by the addition of “wherein the alkaline salt is
a Na +, Mg 2+ , Li+ , K+ , Ca 2+ or N+ (R)4 salt.” See ’504 patent,
col. 14, lines 14-15; id., col. 14, lines 48-49. But “the
doctrine of claim differentiation does not . . . override
clear statements of scope in the specification.” The Toro
Co. v. White Consol. Indus., Inc., 199 F.3d 1295, 1302
(Fed. Cir. 1999). Here, what otherwise might be an
inference from differences in claim language cannot
override the unmistakable limitation of “alkaline salt” set
out in the written description.
CONCLUSION
The written description of the ’504 patent contains a
clear disclaimer of claim scope, and no other aspect of the
intrinsic record clearly points the other way. We there-
fore conclude that the district court’s construction of
“alkaline salt” was correct, and we affirm the judgment of
noninfringement based on that construction.
AFFIRMED

-- 10 of 10 --