Astrazeneca Pharmaceuticals LP v. Teva Pharmaceuticals USA, Inc.

United States Court of Appeals for the Federal Circuit
2008-1480, -1481

ASTRAZENECA PHARMACEUTICALS LP and ASTRAZENECA UK LIMITED,

Plaintiffs-Appellees,

v.

TEVA PHARMACEUTICALS USA, INC.
and TEVA PHARMACEUTICAL INDUSTRIES, LTD.,

Defendants-Appellants,

and

SANDOZ, INC.,

Defendant-Appellant.

Henry J. Renk, Fitzpatrick Cella, Harper & Scinto, of New York, New York, argued for
plaintiffs-appellees. With him on the brief were Bruce C. Haas; Charles E. Lipsey, Finnegan,
Henderson, Farabow Garrett & Dunner, LLP, of Reston, Virginia, and Thomas A. Stevens,
AstraZeneca Pharmaceuticals LP, of Wilmington, Delaware.

Ira J. Levy, Goodwin Proctor LLP, of New York, New York, argued for defendants-
appellants Teva Pharmaceuticals USA, Inc., et al. With him on the brief were Henry C.
Dinger, Daryl L. Wiesen and John T. Bennett, of Boston, Massachusetts.

Douglass C. Hochstetler, Schiff Hardin LLP, of Chicago, Illinois, argued for defendant-
appellant Sandoz, Inc. With him on the brief were Jason G. Harp; and Beth D. Jacob, of New
York, New York.

Appealed from: United States District Court for the District of New Jersey

Judge Joel A. Pisano
 United States Court of Appeals for the Federal Circuit

2008-1480,-1481

ASTRAZENECA PHARMACEUTICALS LP and ASTRAZENECA UK LIMITED,

Plaintiffs-Appellees,

v.

TEVA PHARMACEUTICALS USA, INC.
and TEVA PHARMACEUTICAL INDUSTRIES, LTD.,

Defendants-Appellants,

and

SANDOZ, INC.,

Defendant-Appellant.

Appeal from the United States District Court for the District of New Jersey in consolidated
Case Nos. 05-CV-5333, 06-CV-1528, 07-CV-1632, and 07-CV-3001, Judge Joel A. Pisano.
___________________________

DECIDED: September 25, 2009
___________________________

Before NEWMAN, RADER, and PROST Circuit Judges.

NEWMAN, Circuit Judge.

Teva Pharmaceuticals USA, Inc., Teva Pharmaceutical Industries, Ltd., and Sandoz,

Inc. appeal from the grant, by the United States District Court for the District of New Jersey,

of AstraZeneca Pharmaceuticals LP and AstraZeneca UK Limited’s motion for summary
judgment of no inequitable conduct. 1 The district court ruled that the Appellants had not

presented evidence sufficient for a reasonable jury to find that, in prosecution of the subject

patent application in the Patent and Trademark Office (“PTO”), AstraZeneca made a

misrepresentation of material fact or an omission of material fact, with intent to deceive or

mislead the patent examiner into granting the patent. We affirm the district court’s ruling.

BACKGROUND

AstraZeneca is the assignee of United States Patent No. 4,879,288 (“the ’288

patent”) which claims the antipsychotic drug quetiapine, having the following structural

formula:

AstraZeneca markets quetiapine under the brand name “SEROQUEL®.” The ’288

patent explains that this product is an “atypical” antipsychotic drug, which means that,

unlike “typical” antipsychotics, it does not produce involuntary body movements including

torsion spasms, muscle spasms and dystonia of the face, neck, or back with protrusion of

1
AstraZeneca Pharm. LP v. Teva Pharm. USA, 567 F. Supp. 2d 683 (D.N.J.
2008).

2008-1480,-1481 2
the tongue, and tonic spasms of the limbs (dyskinesias). Such undesirable side effects

were not unusual for antipsychotic drugs. An earlier atypical antipsychotic drug, clozapine,

had become available in the 1970s, but it was withdrawn from the market after it was

discovered to cause a potentially deadly reduction in white blood cell count known as

agranulocytosis. The AstraZeneca product quetiapine was approved by the FDA in 1997.

The ’288 patent expires on September 26, 2011.

Teva and Sandoz each filed abbreviated new drug applications (“ANDAs”) for

approval to sell their production of quetiapine under 21 U.S.C. §355(j), certifying under

“paragraph IV” that the ’288 patent is invalid and/or not infringed. Paragraph IV

certifications are, by statute, an act of technical patent infringement designed to permit

litigation of patent issues for products subject to federal regulatory approval. 35 U.S.C.

§271(e)(2)(A). In accordance with the statutory procedures, AstraZeneca filed infringement

suits against Teva and Sandoz; the suits were consolidated in the District Court for the

District of New Jersey. See 21 U.S.C. §355(j)(5)(B)(iii).

The district court granted summary judgment that there was no inequitable conduct

in prosecution of the ’288 patent application. That is the only issue of this appeal.

DISCUSSION

The grant of summary judgment receives plenary review on appeal. Innogenetics,

N.V. v. Abbott Labs., 512 F.3d 1363, 1378 (Fed. Cir. 2008). “Although the premises of

inequitable conduct require findings based on all the evidence, a procedure that may

preclude summary determination, a motion for summary judgment may be granted when,

drawing all reasonable factual inferences in favor of the non-movant, the evidence is such

that the non-movant can not prevail.” ATD Corp. v. Lydall, Inc., 159 F.3d 534, 547 (Fed.

2008-1480,-1481 3
Cir. 1998) (citation omitted). On appellate review, we accordingly consider the evidence

and all reasonable factual inferences, in light of the applicable law and the burdens and

standards of proof. Id. at 540.

The issue presented in this case relates to the extent to which the patent applicant,

having fully disclosed the relevant prior art and having provided comparative data to the

satisfaction of the patent examiner, must also present any additional unpublished

information in the applicant’s possession concerning other less structurally similar

compounds, and must also synthesize additional compounds for comparative testing. The

district court reviewed the evidence in the dual contexts of the materiality of any withheld or

omitted information, and whether deceptive intent had been established. These are the

essential factual underpinnings of the charge of inequitable conduct, and both materiality

and deceptive intent must be established by clear and convincing evidence. See

Kingsdown Medical Consultants, Ltd. v. Hollister Inc., 863 F.2d 867, 872 (Fed. Cir. 1988)

(en banc) (both materiality and deceptive intent must be shown). Intent to deceive cannot

be inferred from a high degree of materiality alone, but must be separately proved to

establish unenforceability due to inequitable conduct. Id. at 876; see Star Scientific, Inc. v.

R.J. Reynolds Tobacco Co., 537 F.3d 1357, 1366 (Fed. Cir. 2008) (“[T]he fact that

information later found material was not disclosed cannot, by itself, satisfy the deceptive

intent element of inequitable conduct.”).

If both materiality and deceptive intent are established, the court shall balance these

findings, with cognizance of the underlying facts, and determine whether, in the specific

case, there was inequitable conduct in the prosecution of the patent application. Star

Scientific, 537 F.3d at 1365. Upon determining that there was inequitable conduct in

2008-1480,-1481 4
obtaining the patent, the district court may in its discretion declare the patent permanently

unenforceable. Kingsdown, 863 F.2d at 876.

The Prior Art

The references on which the Appellants base their argument of withholding

comparative data were all before the PTO. After filing of the ’288 patent application, before

the examiner’s first Office Action, AstraZeneca filed an Information Disclosure Statement

(“IDS”) listing several references including those described below, and stated: “Enclosed

herewith . . . is a list of references believed to be relevant to the subject matter of the

invention[.]” The relevant compounds in these references are Compound 21076, described

in the German-language publication Research Disclosure (1980) (“Research Disclosure”);

Compound 24028, described in U.S. Patent No. 3,539,573 to Schmutz et al. (“Schmutz II”);

Perlapine, described in U.S. Patent No. 3,389,139 to Schmutz et al. (“Schmutz I”); and

Fluperlapine, described in U.S. Patent No. 4,308,207 to Hunziker et al. (“Hunziker”). These

four compounds have the structural formulae shown below:

2008-1480,-1481 5
AstraZeneca had internal test data for these four compounds as well as for many other

compounds; data that were generated in the course of the research leading to quetiapine.

This internal information was not included in its IDS—that is the main basis of the

Appellants’ inequitable conduct argument, for they argue that AstraZeneca’s data showed

that some prior art compounds potentially exhibited atypical antipsychotic activity, and that

this information should have been reported to the patent examiner.

The examiner, in the first Office Action, cited several references showing compounds

of close structural similarity to quetiapine. The examiner cited Schmutz II as the primary

reference, in view of U.S. Patent No. 4,097,597 to Horrom (“Horrom”), which was another

reference included in AstraZeneca’s IDS, and other combinations of secondary references.

In particular, the examiner identified a compound described in Schmutz II to which the

2008-1480,-1481 6
parties refer as “Schmutz X,” and a compound described in Horrom. These structures are

as follows:

During the prosecution the applicant and the examiner discussed the prior art

compounds, for it was not disputed that the structural differences were small among these

various compounds. The examiner based the rejection on structural similarity alone, and

the applicant in response pressed the unpredictability of the critical physiological property of

atypicality, and that the prior art provided no reason to make the particular compound

quetiapine for the purpose of obtaining atypical antipsychotic properties. AstraZeneca

pointed to the long-felt need for such a drug, because the use of clozapine was severely

limited and no suitable replacement was available in the United States.

The examiner persisted in the rejection, stating that in accordance with the law of

obviousness as to the claimed structure it is not necessary that the prior art described or

suggested that a specific structure would impart a specific property, if in fact the prior art

product had the desired property. The examiner stated that in order to overcome the

structural obviousness rejection AstraZeneca must provide proof that the prior art

compounds:

2008-1480,-1481 7
 do not necessarily or inherently possess the characteristics of the claimed
product. . . . In other words, once a condition of prima facie structural
obviousness has been made out, it must be overcome by a side-to-side
comparison with the closest art compound(s). In this case, one would test
both the prior art species ([Horrom compound] and [Schmutz X]) and the
claimed specie for their ability to avoid e.g. tardive dyskinesia (or whatever
undesirable side effect applicant wishes to focus on).

PTO Paper No. 10, December 2, 1988. The examiner referred to the IDS and stated:

[T]he cited references are noted. None disclose [sic] compounds closer than
the species already discussed.

Id.

In response to the examiner’s position, AstraZeneca submitted the Declaration of

one of the inventors, Dr. Migler, with data for the Horrom compound and a Schmutz

compound. The prosecuting attorney explained that AstraZeneca did not have psychotic

test data for Schmutz X, and that such data would be “very expensive to generate now,”

and as a substitute AstraZeneca offered pre-existing internal data for Schmutz B, which the

inventors believed was closer structurally than Schmutz X because the hydroxyethyl side-

chain of Schmutz B is more similar to quetiapine’s side-chain than is the ethyl side-chain of

Schmutz X. These structures are:

Dr. Migler’s declaration contained psychotic test data showing that the Horrom

compound’s antipsychotic properties were “typical.” The declaration also contained

2008-1480,-1481 8
psychotic test data for Schmutz B, and showed that it too was “typical.” Dr. Migler also

included test data for another compound described in Schmutz II that the parties call

Schmutz A, which differs from quetiapine in that Schmutz A’s left benzene ring is

substituted with chlorine (whereas quetiapine’s left ring is unsubstituted), and Schmutz A's

side chain is methyl (as compared to quetiapine's hydroxyethoxyethyl). The test data

showed that Schmutz A was inactive for antipsychotic activity; AstraZeneca explained that

in light of this inactivity no test for atypical side-effects was conducted for Schmutz A.

The Appellants’ inequitable conduct arguments before the district court, and again

before this court, focus on the fact that AstraZeneca did not submit to the PTO its internal

test data for perlapine, fluperlapine, compound 21076, and compound 24028, shown supra.

The Appellants state that it was a material withholding to provide test data only for the

compounds on which the examiner relied, stating that AstraZeneca’s internal test data

showed that compounds other than quetiapine possessed potential atypical antipsychotic

activity. Thus, the Appellants argued that the Migler Declaration was deliberately

misleading.

The district court found that AstraZeneca properly addressed the closest prior art, in

response to the examiner’s specific requests, and that the premises of the factual

allegations of material withholding with deceptive intent had not been shown sufficiently to

avoid the grant of summary judgment, citing that inequitable conduct in patent prosecution

requires proof by clear and convincing evidence of both (1) an affirmative misrepresentation

of material fact, a failure to disclose material information, or submission of false material

information, and (2) an intent to deceive the examiner by such material falsity.

2008-1480,-1481 9
AstraZeneca, 567 F. Supp. 2d at 691 (citing Cargill, Inc. v. Canbra Foods, Ltd., 476 F.3d

1359, 1364 (Fed. Cir. 2007)).

Materiality

Several standards for establishing the fact of materiality in connection with patent

prosecution have been proposed. See, e.g., Litton Indus. Products, Inc. v. Solid State

Systems Corp., 755 F.2d 158, 166 n.19 (Fed. Cir. 1985) (identifying “four tests: (1)

objective ‘but for’; (2) subjective ‘but for’; (3) ‘but it may have been’; (4) PTO Rule 1.56(a),

i.e., whether there is a substantial likelihood that a reasonable examiner would have

considered the omitted reference or false information important in deciding whether to allow

the application to issue as a patent.”) While a uniform standard has not been rigorously

applied in the courts, the fourth test of whether a reasonable examiner would have

considered the information important in deciding whether to grant the patent, even when

the omitted information does not negate patentability, is most often employed. See, e.g.,

Upjohn Co. v. Mova Pharm. Corp., 225 F.3d 1306, 1312 (Fed. Cir. 2000). “However, a

reference need not be provided to the examiner if it is merely cumulative to or less material

than other references before the examiner.” Id.

Appellants contend that AstraZeneca misrepresented that the atypical properties of

quetiapine were unexpected, stating that AstraZeneca presented internal test data about

similar compounds that were typical while omitting internal test data about similar

compounds that were potentially atypical. AstraZeneca states that it properly focused on

the prior art compounds that were structurally closest to quetiapine and that its internal test

data show that these closest compounds were not atypical. AstraZeneca states that this is

the comparison requested by the examiner, and the comparison most directly in line with

2008-1480,-1481 10
the applicable law. AstraZeneca states that it never represented to the examiner that

quetiapine’s atypical properties made it completely unique among structurally similar

compounds, and points out that in its IDS AstraZeneca identified the references describing

the compounds that the Appellants here assert are potentially atypical antipsychotics, i.e.,

perlapine, fluperlapine, compound 24028, and compound 21076.

No relevant reference is asserted to have been withheld. Various references stated

that certain prior art compounds were known to be atypical antipsychotics, such as

clozapine and fluperlapine. For example, the cited Hunziker reference identifies

fluperlapine as a “non-classical neuroleptic” (i.e., antipsychotic) “like clozapine.” 2 U.S.

Patent No. 4,308,207 col.3 ll.5-9. AstraZeneca points out that since the Hunziker reference

describes fluperlapine as having atypical properties, AstraZeneca’s internal test data on

whether fluperlapine is atypical is, at most, cumulative information.

With all of the references before him, the Examiner made a prima facie obviousness

rejection based on the structural similarity between quetiapine and the two prior art

compounds that he identified as structurally closest to quetiapine; that is, Schmutz X and

the Horrom Compound. As we have discussed, AstraZeneca responded that it had test

data for the Horrom Compound, but that it did not have test data for Schmutz X, and

provided test data for Schmutz B; AstraZeneca told the examiner that the test data on

Schmutz X would be “very expensive to generate now” and that Schmutz B was actually

2
AstraZeneca requests that we take judicial notice that Sandoz, in a document
filed in a separate proceeding before the United States District Court for the District of New
Jersey, acknowledged that the Hunziker patent describes fluperlapine as an atypical
antipsychotic. AstraZeneca Pharm. LP v. Sandoz, Inc., Civil Action No. 09-CV-870 (JAP)
(TJB) (D.N.J. 2009). We take judicial notice of this statement, which merely confirms an
uncontested fact, as Sandoz’s brief already states “Hunziker described fluperlapine as
being a non-classical antipsychotic like clozapine.”

2008-1480,-1481 11
structurally closer to quetiapine than Schmutz X. Appellants charge that both of these

statements to the examiner are material misrepresentations. AstraZeneca replies that both

statements are accurate and true.

First, the Appellants argue that AstraZeneca falsely stated that generating test data

on Schmutz X would be “very expensive.” Appellants contend that expense was mere

pretext, asserted because AstraZeneca believed that Schmutz X was atypical based on the

structural similarity between Schmutz X and Compound 24028, which AstraZeneca knew to

have potential atypical properties. Appellants do not assert that AstraZeneca had data for

Schmutz X and withheld it; nor do they argue that Schmutz X is in fact atypical, and no data

are presented to this effect. Instead the Appellants seem to argue that because of the

structural similarity between Schmutz X and Compound 24028, Schmutz X could be

atypical, and thus should have been synthesized and its antipsychotic properties tested.

The Appellants’ position appears to be that AstraZeneca’s failure to do so was a material

withholding or omission, with deceptive intent.

Although there may be situations in which the failure to conduct specific tests of

specific compounds can be criticized, in this case there was no evidence that the

information gleaned, if such tests had been conducted, would have been material to

patentability. It was not disputed that it was unpredictable whether a given compound

would exhibit atypical antipsychotic properties. The record demonstrates that structural

similarity is not a predictor of whether antipsychotic behavior would be typical or atypical.

As AstraZeneca points out, the properties of these structurally similar compounds vary

significantly with minor structural changes. 3 The Appellants made no showing as to

3
During the course of this litigation, the district court collaterally estopped Teva

2008-1480,-1481 12
whether the structural similarity between Schmutz X and Compound 24028 would establish

whether Schmutz X would have atypical properties. In the context of the knowledge in this

field, as reflected in the various references, AstraZeneca’s provision of its existing test data

for Schmutz B, instead of preparing and testing Schmutz X, cannot constitute a material

misrepresentation. The district court correctly so held.

The Appellants also argue that AstraZeneca should have submitted to the examiner

its existing test data for Compound 24028 because Compound 24028 and Schmutz B are

“equally close“ to quetiapine. The Appellants state that the data should have been

submitted although not identified or requested by the examiner.

To ascertain what is “equally close,” in identifying the structurally closest prior art,

the compounds are viewed as they would be perceived by persons experienced in the

particular field of science. Precedent suggests that as a starting point it is useful to

ascertain the common elements of the claimed invention and the prior art. See In re

Merchant, 575 F.2d 865, 868 (CCPA 1978) (determining the common elements). Thus the

Appellants point out that Compound 24028, Schmutz B, and quetiapine differ only in their

respective side chains. That is correct. However, AstraZeneca is also correct in pointing

out that Schmutz B’s hydroxyethyl side chain is structurally closer to quetiapine’s

hydroxyethoxyethyl side chain, both in length of the chain and in the hydroxyl end group,

than is Compound 24028’s methyl side chain, as shown below:

from challenging the unpredictability of atypical antipsychotic properties because Teva had
agreed to be bound by the decision of the district court in Eli Lilly & Co. v. Zenith Goldline
Pharms., Inc., 364 F. Supp. 2d 820, 831 ¶14 (S.D. Ind. 2005), which involved many of the
same issues regarding the development of an atypical antipsychotic. See AstraZeneca
Pharm. LP v. Teva Pharm. USA, Civil Action No. 05-CV-05333 (JAP) (TJB), Order Granting
Motion for Summary Judgment Based on Collateral Estoppel (D.N.J. Oct. 12, 2007).

2008-1480,-1481 13
The patent examiner did not dispute AstraZeneca’s scientific position, and accepted the

proffered data on Schmutz B as the closest prior art. Accordingly, AstraZeneca’s

substitution of Schmutz B in place of Schmutz X was not a material misrepresentation, and

the non-provision of the data on Compound 24028, which is structurally less similar to

quetiapine, was not a material omission.

The appellants also argue that AstraZeneca’s submission of its internal test data of

Schmutz B along with its internal test data of Schmutz A was an implied misrepresentation

because it omitted other internal test data of potentially atypical compounds including

fluperlapine, perlapine, Compound 21076, and Compound 24028. We agree with the

district court that the Migler declaration, which was directed to the closest prior art

compounds, was not an implied misrepresentation. AstraZeneca has not asserted that no

prior art compound is atypical. Rather, AstraZeneca demonstrated, as the examiner

required, that the structurally closest prior art compounds did not possess the same

properties as quetiapine. The prosecuting attorney explained that Schmutz A was on the

same list of antipsychotics as Schmutz X. A reasonable examiner would not have

understood the Migler declaration as stating that no prior art product had the atypical

property shown by quetiapine, for it was known that other atypical antipsychotics existed,

2008-1480,-1481 14
including fluperlapine and clozapine. A reasonable examiner would understand

AstraZeneca’s statements to refer to the closest prior art compounds, not all prior art

compounds. AstraZeneca’s inclusion of the Schmutz A data is not an “implied”

misrepresentation.

We conclude, as did the district court, that the evidence cannot support a finding that

AstraZeneca misrepresented or omitted material information.

Deceptive Intent

Intent to deceive is an independent element of inequitable conduct, and must be

independently established by clear and convincing evidence. Star Scientific, 537 F.3d at

1365; Cargill, 476 F.3d at 1364; see Molins PLC v. Textron, Inc., 48 F.3d 1172, 1181 (Fed.

Cir. 1995) (“[C]lear and convincing evidence must prove that an applicant had the specific

intent to accomplish an act that the applicant ought not to have performed, viz., misleading

or deceiving the PTO. In a case involving nondisclosure of information, clear and

convincing evidence must show that the applicant made a deliberate decision to withhold a

known material reference.”); Allen Organ Co. v. Kimball Int’l, Inc., 839 F.2d 1556, 1567

(Fed. Cir. 1988) (“[m]ateriality does not presume intent, which is a separate and essential

component of inequitable conduct”).

Appellants state that they have shown a “high degree of materiality,” and that they

therefore need a proportionally lesser showing of intent to deceive to establish the requisite

threshold level of intent. That is incorrect. Evidence of mistake or negligence, even gross

negligence, is not sufficient to support inequitable conduct in patent prosecution.

Kingsdown, 863 F.2d at 876. To establish the requisite deceptive intent, “the involved

conduct, viewed in light of all the evidence, including evidence indicative of good faith, must

2008-1480,-1481 15
indicate sufficient culpability to require a finding of intent to deceive.” Id. While the court

must, as the final step, weigh and balance the findings of materiality and intent, this

presupposes that a threshold level of both of these elements has already been established

by clear and convincing evidence. See Manville Sales Corp. v. Paramount Systems, Inc.,

917 F.2d 544, 551 (Fed. Cir. 1990) (“Inequitable conduct requires proof by clear and

convincing evidence. A threshold showing of both materiality and intent to mislead or

deceive must be first established, and then those fact-findings are balanced to make the

determination whether ‘the scales tilt to a conclusion that inequitable conduct occurred.’”).

The only evidence of intent offered by the Appellants is AstraZeneca’s internal

knowledge of certain compounds of this structural class that were atypical, without including

this information in the IDS. The Appellants argue that it is irrelevant that the omitted

compounds are not the closest structural analogs, and irrelevant that the patent examiner

asked for comparative data for other, closer compounds. The Appellants offer no evidence

or suggestion of deceptive intent, other than the fact that this information was not provided.

The law is clear that “inequitable conduct requires not intent to withhold, but rather intent to

deceive. Intent to deceive cannot be inferred simply from the decision to withhold

[information] where the reasons given for the withholding are plausible.” Dayco Products,

Inc. v. Total Containment, Inc., 329 F.3d 1358, 1367 (Fed. Cir. 2003). As argued by

Appellants, an applicant would not know how much of its research effort must be filed with

the PTO, although of no interest to the examiner, or run the risk of accusation of

wrongdoing no matter where the line is drawn.

As we have discussed, the omitted test data were not material because the

compounds were not the structurally closest compounds, whereas AstraZeneca and the

2008-1480,-1481 16
examiner focused on the structurally closest compounds. Although the Appellants argue

that this is an inadequate reason, no evidence of bad faith has been proffered concerning

the omission of data for the less similar compounds, AstraZeneca presented plausible

reasons for its presentation of arguments and data during the prosecution, and although

the Appellants dispute AstraZeneca’s reasoning, intentional withholding for the purpose of

deceiving the examiner is unsupported by evidence sufficient to avert summary judgment.

We reach the same conclusion as did the district court, that the Appellants have not

provided evidence sufficient to establish the threshold facts of material withholding with the

intent to deceive. The grant of summary judgment of no inequitable conduct is affirmed.

COSTS

Each party shall bear its own costs.

AFFIRMED.

2008-1480,-1481 17