Endo Phamaceuticals Inc. et al. v. Amneal Pharmaceuticals LLC, et al.

Filing 155

TRIAL OPINION. Plaintiff should submit an agreed upon form of final judgment within two weeks. Signed by Judge Richard G. Andrews on 10/7/2016. Associated Cases: 1:14-cv-01382-RGA, 1:14-cv-01389-RGA(nms)

l

I

IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF DELAWARE

ENDO PHARMACEUTICALS INC. and
MALLINCKRODT LLC,
Plaintiffs,
Civil Action No. 14-1382-RGA

v.
AMNEAL PHARMACEUTICALS, LLC and
AMNEAL PHARMACEUTICALS OF NEW
YORK,LLC,
Defendants.

ENDO PHARMACEUTICALS INC. and
MALLINCKRODT LLC,
Plaintiffs,
V.

Civil Action No. 14-1389-RGA

TEVA PHARMACEUTICALS USA, INC.
and BARR LABORATORIES, INC.,
Defendants.

TRIAL OPINION
Jack B. Blumenfeld, Esq., Derek J. Fahnestock, Esq., Stephen J. Kraftschik, Esq., Morris,
Nichols, Arsht & Tunnell LLP, Wilmington, DE; Martin J. Black, Esq., Joseph J. Gribbin, Esq.,
Julie Latsko, Esq., Sharon K. Gagliardi, Esq., Dechert LLP, Philadelphia, PA; Jonathan D. Loeb,
Esq., Dechert LLP, Mountain View, CA; Robert D. Rhoad, Esq., Brian M. Goldberg, Esq.,
Dechert LLP, Princeton, NJ, attorneys for Plaintiff Endo Pharmaceuticals Inc.
Jack B. Blumenfeld, Esq., Derek J. Fahnestock, Esq., Stephen J. Kraftschik, Esq., Morris,
Nichols, Arsht & Tunnell LLP, Wilmington, DE; Jeffrey J. Toney, Esq., Paul G. Williams, Esq.,
Rodney R. Miller, Esq., Kasowitz, Benson, Torres & Friedman LLP, Atlanta, GA, Jonathan K.
Waldrop, Esq., Marcus A. Barber, Esq., Kasowitz, Benson, Torres & Friedman LLP, Redwood
Shores, CA, attorneys for Plaintiff Mallinckrodt LLC.

I

Mary B. Matterer, Esq., Richard K. Herrmann, Esq., Morris James LLP, Wilmington, DE; Jake
M. Holdreith, Esq., Kelsey J. Thorkelson, Esq., Robins Kaplan LLP, Minneapolis, MN, Oren D.
Langer, Esq., Robins Kaplan LLP, New York, NY, attorneys for Defendants Amneal
Pharmaceuticals, LLC and Amneal Pharmaceuticals of New York, LLC.
Richard L. Horwitz, Esq., David E. Moore, Esq., Bindu A. Palapura, Esq., Potter Anderson &
Corroon, LLP, Wilmington, DE; Huiya Wu, Esq., Jordan B. Weiss, Esq., Elizabeth J. Holland,
Esq., Daniel P. Margolis, Esq., Brigid M. Morris, Esq., Brian J. Robinson, Esq., Goodwin
Procter LLP, New York, NY, attorneys for Defendants Teva Pharmaceuticals USA, Inc. and Barr
Laboratories, Inc.

October

!I_, 2016
2

Plaintiffs brought these patent infringement actions against Amneal Pharmaceuticals,
LLC, Amneal Pharmaceuticals of New York, LLC (collectively, "Amneal"), Teva
Pharmaceuticals USA, Inc., and Barr Laboratories, Inc. (collectively, "Teva") in 2014. (D.I. 1). 1
On April 3, 2012, Amneal filed an Abbreviated New Drug Application ("ANDA"), seeking to
engage in the commercial manufacture, use, and sale of generic versions of Endo's Opana ER
CRF product.2 (D.I. 130, Ex. 1ii14). Teva filed an ANDA on April 17, 2012, with amendments
on May 4, 2012 and September 20, 2012, seeking to do the same. (Id. ii 16). Plaintiffs allege
that these ANDAs infringe U.S. Patent No. 8,871,779 ("the '779 patent").
These cases concern two molecules. The first is 14-hydroxydihydromorphinone, also
referred to as "oxymorphone" or "oxymorphone HCl."3 The other is 14-hydroxymorphinone,
also referred to as "oxymorphone ABUK." ABUK stands for alpha,beta-unsaturated ketone, an
organic compound having a double bond between the ketone's alpha and beta carbons.
Oxymorphone HCl was first patented in 1955 and first approved by the FDA in 1959.
(Trial Transcript ("Tr.") at 86: 1-5, 11-14). Prior to 2002, manufacturers of oxymorphone HCl
were aware of the presence of the impurity now known as oxymorphone ABUK. (Tr. at 229:9230:4; see also JTX-23). During the period before 2002, manufacturers regularly sold
oxymorphone HCl with oxymorphone ABUK levels in the range of hundreds of parts per million
("ppm"). (Tr. at 229:9-230:4). In 2002, the FDA informed Mallinckrodt and several other
manufacturers that it was concerned about the levels of ABUK in certain products. (Tr. 217:9-

1

Unless otherwise indicated, all docket citations, except those in the implied license section, are to C.A.
No. 14-1382. In the section on implied license, docket citations are to C.A. No. 14-1389.
2
"CRF" stands for "crush-resistant formulation." (Tr. 614:16-20).
3

Oxymorphone and oxymorphone HCl are actually different compounds, in that the latter is a salt formed when
chloride is added. In this opinion, however, they are used interchangeably, as the key distinction in this case is
between oxymorphone ABUK and oxymorphone without the ABUK double bond.

3

218:22). The FDA informed Mallinckrodt that it intended to impose limits on the levels of
ABUK, and that it might require limits as low as 0.001 percent (or 10 ppm) ABUK. (Tr. 110:19111 :21, 218:7-18). In 2004, the FDA mandated that opioid manufacturers lower the levels of
ABUK in opioid pharmaceuticals to less than 10 ppm. (Tr. 199:10-201:20, 218:7-18). In these
cases, oxymorphone HCl which contains less than 10 ppm of oxymorphone ABUK-and thus
complies with FDA's mandate-is called "low-ABUK oxymorphone."
In 2005, Mallinckrodt succeeded in reaching the low ABUK levels mandated by the FDA
for oxymorphone HCl. Mallinckrodt applied for a patent on its new low-ABUK oxymorphone
product. The application ultimately issued as the '779 patent. The asserted claims of the '779
patent4 are all composition claims directed to low-ABUK oxymorphone. (Tr. 88:22-89:8,
111:13-21; DTX-17 at 37:58-38:61).
Independent claim 1 of the '779 patent reads:
A hydrochloride salt of oxymorphone comprising less than 0.001%of14hydroxymorphinone.
(DTX-17 at 37:58-59). Dependent claim 2 limits the level of 14-hydroxymorphinone to less than
0.0005%. (Id. at 37:60-61). Dependent claim 3 claims a pharmaceutically acceptable form of
the hydrochloride salt in claim 1. (Id. at 37:62-63). Independent claim 4 reads:
A hydrochloride salt of a morphinan-6-one compound corresponding to Formula (2):

4

Plaintiffs assert that all six claims of the '779 patent are infringed.

4

R,

R

comprising less than 0.001 % measured by HPLC of an a,~-unsaturated ketone compound
corresponding to Formula (3):

R:

wherein the morphinan-6-one compound is oxymorphone and wherein Xis -N(R17)-;
R1 and R2 are hydrogen;
R3 is hydroxy;
Rio is hydrogen;
R14 is hydroxy; and
R11 is methyl.
(Id. at 38:16-57). Dependent claim 5 limits the level of 14-hydroxymorphinone to 0.0005%. (Id.

at 38:58-59). Dependent claim 6 claims a pharmaceutical formulation of the oxymorphone
chloride in claim 4. (Id. at 38:60-61).

5

The Court held a bench trial on July 11-13, 2016. Both Amneal and Teva concede that
their proposed products meet all the limitations of the '779 patent. (D.I. 150, Ex. 1 iii! 18-20).
Teva contends, however, that because it obtained an implied license from Mallinckrodt, it does
not infringe. Both defendants argue that the '779 patent is invalid as obvious.

I.

OBVIOUSNESS
A. Legal Standard
A patent claim is invalid as obvious under 35 U.S.C. § 103 "ifthe differences between

the subject matter sought to be patented and the prior art are such that the subject matter as a
whole would have been obvious at the time the invention was made to a person having ordinary
skill in the art to which said subject matter pertains." 35 U.S.C. § 103; see also KSR Int'! Co. v.

Teleflex Inc., 550 U.S. 398, 406--07 (2007). The determination of obviousness is a question of
law with underlying factual findings. See Kinetic Concepts, Inc. v. Smith & Nephew, Inc., 688
F.3d 1342, 1359-60 (Fed. Cir. 2012). "The underlying factual inquiries include (1) the scope and
content of the prior art; (2) the differences between the prior art and the claims at issue; (3) the
level of ordinary skill in the art; and (4) any relevant secondary considerations .... " Western

Union Co. v. MoneyGram Payment Sys., Inc., 626 F.3d 1361, 1370 (Fed. Cir. 2010) (citing
Graham v. John Deere Co., 383 U.S. 1, 17-18 (1966)).
A court is required to consider secondary considerations, or objective indicia of
nonobviousness, before reaching an obviousness determination, as a "check against hindsight
bias." See In re Cyclobenzaprine Hydrochloride Extended-Release Capsule Patent Litig., 676
F.3d 1063, 1078-79 (Fed. Cir. 2012). Relevant secondary considerations include commercial
success, long felt but unsolved needs, failure of others, praise, unexpected results, and copying,
among others. Graham, 383 U.S. at 17-18; Ruiz v. A.B. Chance Co., 234 F.3d 654, 662-63 (Fed.

6

Cir. 2000); Tex. Instruments, Inc. v. US. Int'! Trade Comm'n, 988 F.2d 1165, 1178 (Fed. Cir.
1993).
A party asserting that a patent is invalid as obvious must "show by clear and convincing
evidence that a skilled artisan would have been motivated to combine the teachings of the prior
art references to achieve the claimed invention, and that the skilled artisan would have had a
reasonable expectation of success in doing so." Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1361
(Fed. Cir. 2007). That "expectation of success need only be reasonable, not absolute." Id. at
1364. "Whether an ordinarily skilled artisan would have reasonably expected success .... is
measured as of the date of the invention[] .... " Amgen Inc. v. F. Hoffman-La Roche Ltd, 580
F.3d 1340, 1362 (Fed. Cir. 2009).

B. Findings of Fact
1. The level of ordinary skill in the art is either ( 1) a person with a Ph.D. degree in
medicinal chemistry, organic chemistry or a related discipline, and at least a few years of
experience in synthetic organic chemistry; or (2) a person with a lesser degree in one of those
fields, but with commensurately greater experience.
2. Weiss and Chapman are prior art.
3. Neither Weiss nor Chapman teach a person of ordinary skill that catalytic
hydrogenation could be used to create low-ABUK oxymorphone.
4. There was no simultaneous invention oflow-ABUK oxymorphone.
5. Low-ABUK oxymorphone would not have been obvious to one of ordinary skill in the
art.

7

t
f
I

I

I

I

i

I
i

l

C. Conclusions of Law

Defendants contend that the low-ABUK oxymorphone claimed in the '779 patent would
have been obvious to one of ordinary skill in the art. Specifically, Defendants argue that an
ordinary-skilled artisan would have had a reasonable expectation of success in using catalytic
hydrogenation to convert oxymorphoneABUK to oxymorphone HCl, thereby lowering the level
of oxymorphone ABUK to below 10 ppm. Defendants rely on Weiss, a paper published in 1957,
to demonstrate that a person of ordinary skill would understand that hydrogenation could be used
to convert oxymorphone ABUK to oxymorphone HCL (JTX-23). Defendants also rely on
Chapman, a 2005 patent application which claims priority to a provisional application filed on
March 30, 2004. (DTX-97; DTX-137). Defendants argue that the real-world experiment
described in Chapman "corroborates" the expectation of success instilled by Weiss. 5
The parties generally agree that the person of ordinary skill to whom the '779 patent is
directed is a person with "a Ph.D. degree in medicinal chemistry, organic chemistry or a related
discipline, and at least a few years of experience in synthetic organic chemistry" or a person with
a lesser degree in one of those fields, but with greater experience. (Tr. 361:2-15, 67:19-68:22;
D.I. 143 at p. 9 n.3). Plaintiffs' expert, Dr. Davies, opined that the person of ordinary skill would
"also need [experience with] process chemistry involving natural products or compounds of

5

In post-trial briefing, Defendants also rely on U.S. Patent No. 7 ,851,482 ("the Dung reference"). (DTX100; see, e.g., D.I. 139 at 22). Plaintiffs have moved to strike all discussions of Dung. (D.I. 183). At
trial, Defendants sought to move Dung into evidence. (Tr. 113:5-114:1). Plaintiffs objected on the
grounds that Dr. Heathcock had provided "nothing substantive about the document" in his report. (Tr.
114:2-7). The Court admitted the document into evidence for the limited purpose of"show[ing] that it
really exist[ed]." (Tr. 114:15-19). There was no testimony about Dung during trial. Thus, Plaintiffs
arguments regarding Dung are completely unsubstantiated by the trial record. Further, they seek to use
evidence admitted for one purpose for an entirely different purpose, in violation of Fed. R. Civ. P. 105.
Plaintiffs' motion to strike (D.1149; C.A. No. 14-1389, D.I. 183) is therefore GRANTED.

8

'~
l

l
f

~

I

i

l

I

I

I

I

related complexity." (Tr. 361: 15-22). I do not think this addition makes a difference. In
determining obviousness, I considered the person of ordinary skill upon which the parties agreed.
i. Scope and Content of the Prior Art
1. Weiss

Weiss generally describes the process of hydrogenating oxymorphone ABUK, thereby
converting it into oxymorphone HCL Weiss does not provide the all of the reaction conditions
required to reproduce the described reaction. (Tr. 174:11-175:6, 347:20-22, 388:24-389:14,
390:21-391 :11; see also JTX-23 at 1507). Specifically, Weiss lacks details about hydrogen
pressure, amount of acid, amount and composition of catalyst, and reaction time. 6 (Id.). It is
undisputed that Weiss does not provide any information about the level of oxymorphone ABUK
or other impurities remaining after hydrogenation. (Tr. 107:22-108:11, 380:11-15, 389:15-21;
see also JTX-23 at p. 1507). Further, analytical methods available at the time of Weiss would

only have been able to determine the remaining ABUK levels in the hundreds of ppm. (Tr.
145:18-22, 380:11-15). Between the publication of Weiss in 1957 and the date of invention in
2005, no other prior art reference mentioned oxymorphone ABUK. (Tr. 146:22-147:19).
2. Chapman

The Chapman reference is a United States patent application filed on March 30, 2005.
The parties dispute whether the Chapman reference is prior art. Defendants argue that Chapman
qualifies as 3 5 U.S. C. § 102(e) prior art. That section provides that an invention described in an
application for a U.S. patent filed before the invention under review is prior art. Since§ 102(e)
requires that the application predate "the invention," a patentee may "swear behind" a potential §

6

The parties agree that Weiss lacks these parameters. Defendants' expert, Dr. Heathcock, opines that these could all
be determined based on routine experimentation. (Tr. 213:3-215:23, 316:22-317:2). Dr. Heathcock stated that they
were not recited because they were so simple. (Tr. 174:11-175:6).

9

I

I
j

102(e) reference. The dispute centers on the proper date of the invention for the '779 patent's
claims, and whether Chapman is entitled to the filing date of an earlier provisional application.
Specifically, while the parties agree that claims 1, 2, 4, and 5 of the '779 patent are entitled to a
priority date of February 2, 2005, they disagree as to whether claims 3 and 6 are entitled to that
date or the date of filing-March 2, 2007. (D.1. 139 at 13; D.I. 143 at pp. 13-14). Additionally,
Plaintiffs argue that Defendants have not shown that Chapman is entitled to the filing date of the
provisional application-March 30, 2004.
Since I conclude that the Chapman reference does not render obvious the claims of the
'779 patent, I need not resolve these questions. I will accept that Chapman is valid§ 102(e) prior
art, and that the date of invention for all asserted claims is February 2, 2005.
Chapman does not discuss oxymorphone. Instead, Chapman describes a process for
using hydrogenation to convert 14-hydroxycodeinone ("oxycodone ABUK") into oxycodone.
Chapman uses a "double hydrogenation" process. (Tr. 382:12-384:3). This process involves an
initial step of hydrogenating oxycodone ABUK, resulting in oxycodone which still contains
relatively high levels of oxycodone ABUK. (Tr. 127:9-128:14, 382:12-383:6; DTX-97 at fig. 1,

if 13).

Then, the oxycodone product from the first step is hydrogenated again under specific

parameters, producing oxycodone with less than 25 ppm of oxycodone ABUK. (Tr. 127:9128:14, 383:7-20; DTX-97 if 20). 7

7

Chapman also states that the process may reduce the levels of oxycodone ABUK to below 15 ppm, 10
ppm, or 5 ppm. (DTX-97 ii 16). Chapman, in an experiment called Example 2, states that two different
analytical methods showed levels of oxycodone ABUK at less than 5 ppm. (Tr. 194:20-23; DTX-97 iii!
189-90). In Example 3, Chapman stated that two different analytical methods showed levels of
oxycodone ABUK at 5 ppm and 10 ppm, respectively. (Tr. 194:10-19; DTX-97 iii! 197-98).

10

I

I

i

I
iii. Comparing Prior Art and Claimed Subject Matter
Defendants' expert, Dr. Heathcock, opines that a hydrogenation action, like the one
described in Weiss, carried out to its completion, would eventually result in a low concentration
of the initial reactant-in this case oxymorphone ABUK. (Tr. 93:2-95:1, 96:9-99: 14).
Specifically, Dr. Heathcock testified that the driving force of hydrogenation would strongly
propel the conversion of oxymorphone ABUK into oxymorphone. (Tr. 89:9-91 :3). Dr.
Heathcock stated that the prior art shows that the equilibrium constant 8 of a hydrogenation
reaction is on the order of 1020 • (Tr. 94:3-98:15; DTX-114). Thus, according to Dr. Heathcock,
if the hydrogenation reaction were carried out to completion-its equilibrium-the resulting
mixture would contain 5 parts per million million million of oxymorphone ABUK, a level well
below the 10 ppm required by the FDA. (Tr. 93:2-95:1, 96:9-99:14). In other words, according
to Dr. Heathcock, if a person of skill in the art just ran a hydrogenation reaction for a sufficient
amount of time, one would ultimately end up with low-ABUK oxymorphone. To support this
conclusion, Dr. Heathcock relies on an illustration involving hydrogenating cyclohexene to
cyclohexane. (Tr. 97:18- 99:14). Dr. Heathcock refers to all of this as "basic chemistry." (Tr.
135:11-136:14, 137:8-12).
One problem with Dr. Heathcock's "basic chemistry" theory is that there is simply no
indication, and certainly no experimental evidence, that the hydrogenation procedure described
in Weiss could result inABUK levels below 10 ppm. 9 At the time of the invention, it was "very

8

An equilibrium constant is essentially the ratio of the concentrations of the products and reactants at
equilibrium.
9
To achieve low-ABUK oxymorphone, Mallinckrodt did not use hydrogenation. (Tr. 207:14-208:4,
211:24-213:2). Instead, Mallinckrodt used a process involving sodium bisulfite and sulfurous acid to
achieve low-ABUK oxymorphone. (Id.). Additionally, while two other manufacturers, Noramco and
Johnson Matthey, were successful in making low-ABUK oxymorphone, no evidence in the record
explains how or when these companies succeeded in making it. (Tr. 237:3-238:4).

11

unusual" and "very, very challenging" to remove impurities like ABUK to levels below 10 ppm.
(Tr. 360:1-6, 370:23-371:8). These levels were described as "remarkable." (Tr. 401:22-403:8).
Further, "[t]here are very few methods that will measure such low levels." (Tr. 371 :4-5).
Dr. Heathcock's "basic chemistry" theory does not account for the complexities involved
in reducing ABUK levels to below 10 ppm. Oxymorphone has numerous impurities, aside from
the oxymorphone ABUK, the most important of which, for purposes of this case, is
oxymorphone diol. Oxymorphone diol, or 8, 14-dihydroxy-7,8-dihydromorphinone, is formed
when water is added to the oxymorphone ABUK. Since oxymorphone diol lacks the ABUK
double bond, it, unlike oxymorphone ABUK, is not converted to oxymorphone upon
hydrogenation. (Tr. 350:8-21, 351:14-24, 414:22-415:16). When the diol becomes dehydrated, it
converts back into oxymorphone ABUK. (Tr. 414:15-415:11). According to Dr. Davies, this
creates a problem. Generally, to create oxymorphone, one first begins with a poppy straw, which
is converted into thebaine. (Tr. 152:23-153:14). Then, through an oxidizing process, the
thebaine is converted into oxycodone ABUK, which is then hydrogenated to form oxycodone.
(Tr. 154:5-155:10). Then, the oxycodone is 0-demethylated to form oxymorphone. (Tr. 420:911 ). During oxidation, "you will produce [oxycodone] diol because you have water present with
the Oxycodone ABUK." (Tr. 420:2-8). During 0-demethylation, the oxycodone diol that
formed will be converted in oxymorphone diol. (Tr. 420:9-421: 15).
Oxymorphone diol will be converted into oxymorphone ABUK ''under acid and heat."

(Tr. 422:4-8; see also Tr. 414: 15-415: 11 ). When working-up 10 the reaction, "you filter the acidic
reaction mixture." (Tr. 422:17-19). Then, during purification, "you do a crystallization which

10

"Work[ -]up is the processing of a reaction product mixture in order to remove unwanted components
such as solvents and inorganic materials that may be resulting from the reaction, and to obtain the
intended products, normally the organic products." (Tr. 156:21-157:3).

12

involves heating it in a solvent .... [,] [a]nd then you heat dry the product." (Tr. 422:19-22). In
each stage of this process, the oxymorphone diol may regenerate the oxymorphone ABUK, even
iftheABUK had previously been reduced to extremely low levels. (Tr. 422:17-423:2; see also
415:17-416:21, 424:24-430:20). Thus, the diol can act like a "reservoir" for regenerating
oxymorphoneABUK. (Tr. 388:6-23, 415:2-11, 428:13-28). This was described at various times
asthe"reappearingABUK." (See, e.g., Tr. 170:2-7, 189:5-12). 11 This regeneration is
significant, since the FDA requires, and the '779 patent claims, such low levels of oxymorphone
ABUK.
Defendants argue that Weiss had removed the oxymorphone diol from his starting
material for the hydrogenation reaction, and that it would therefore have no impact on the levels
of the ABUK. Defendants rely on the statement in Weiss that "[t]he solid residue was kept at
room temperature for about 24 hr. with 60 ml. acetone, which dissolved the [oxymorphone diol]
present." (JTX-23 at p. 1506).
In response, Dr. Davies testified that "it's very hard to remove the Oxymorphone diol."
(Tr. 416:20-21). Dr. Davies testified that the acetone wash described in Weiss-a process called
trituration-would not "completely remove the compound you're trying to wash away." (Tr.
416:22-417:24). Since it is a "very crude technique," it could not be expected to completely
remove the oxymorphone diol. (Tr. 417: 14-418 :4 ).

11

This "reappearing ABUK" problem occurred in Chapman, in the context of hydrogenating oxycodone
ABUK to form oxycodone. In Example 5, oxycodone ABUK levels were undetectable before work-up,
and after work-up, were measured at 11 ppm. (Tr. 426:9-427:21; DTX-97 iii! 267-69). In Example 3, no
oxycodone ABUK was detected after work-up, "but during the purification[,] ... 5 or 10 ppm of ABUK
have reappeared." (Tr. 427:18-428:23; DTX-97 iii! 192-98).

13

I
Thus, while a person of ordinary skill would reasonably expect the Weiss hydrogenation
procedure to reduce the levels of oxymorphone ABUK, a person of ordinary skill would not
reasonably expect that the Weiss hydrogenation procedure to lower ABUK levels below 10 ppm.
Defendants contend that Chapman "corroborates" the hydrogenation procedure described
in Weiss. (Tr. 133:19-24; see also Tr. 136:5-14). The Chapman hydrogenation procedure differs
from Weiss in some critical respects, however. Most importantly, Chapman describes the
hydrogenation of oxycodone, rather than oxymorphone. (Tr. 186:9-187: 13, 382:2-11 ). While
oxycodone and oxymorphone are both morphinan-6-ones that may form an ABUK, the evidence
demonstrates that oxycodone and oxymorphone react in different ways. These differences are
attributable to certain structural variations. Oxycodone and oxycodone ABUK contain anisole, a
benzene ring with an OCH3 (methoxy) group attached. (Tr. 411: 1-6). Oxymorphone and
oxymorphone ABUK, on the other hand, contain phenol, a benzene ring with an OH (hydroxy)
group attached. (Id.).
Two prior art references illustrate how these structural variations result in reactivity
differences. A prior art patent from 1965, U.S. Patent No. 3,193, 584 ("the '584 patent"),
compares the hydrogenation of phenol with anisole. (PTX-90; Tr. 410:21-411:19). Table 1 of the
'584 patent indicates that, under basic, neutral, and acidic conditions, phenol hydrogenates faster
than anisole. (PTX-90; Tr. 411: 1-19). This means that the six-membered ring to which the
methoxy group or hydroxy group is attached, is reduced. 12 (Tr. 412: 18-413: 15). This reduction
fundamentally changes the molecule; it ceases to be oxymorphone or oxycodone. (Tr. 413:16-

12

In this context, a reduction occurs when each of the three double bonds in the six-membered ring is
reduced to a single bond, resulting in an additional hydrogen atom bonding to each carbon atom in the
ring. (See Tr. 412:18-413:19).

14

I

414:4). The '584 patent indicates that this is more likely to occur in oxymorphone than in
oxycodone. (Tr. 412:6-17, 413:8-414:14).
In Schmidhammer, a prior art paper published in 1990, a hydrogenation reaction was

performed on two ABUK molecules with similar structures to oxycodone ABUK and
oxymorphone ABUK. (PTX-79). When hydrogenating these two molecules, the anisole
compound yielded 92%, while the phenol compound yielded 76%. (PTX-79; Tr. 407:2-410:5).
In other words, the compound similar to oxycodone ABUK hydrogenated more efficiently than
the compound similar to oxymorphone ABUK. (Tr. 409: 1-410: 1). Dr. Davies therefore opined
that a person of ordinary skill is "less likely to succeed with [the] Chapman [process] on
Oxymorphone ABUK than ... on oxycodone ABUK." (Tr. 409:21-410:1).
Dr. Davies also explained that the higher amount of diol present in oxymorphone would
lead an ordinary-skilled artisan to believe that hydrogenation of oxymorphone ABUK would be
less effective than hydrogenation of oxycodone ABUK. Dr. Davies refers to oxymorphone diol
as an ABUK precursor, since it is formed by adding water to the ABUK double bond, and can
convert back into ABUK when dehydrated. (Tr. 414: 15-415:8). Weiss explains that, in alkaline
solution, oxymorphone ABUK is converted to oxymorphone diol "with unexpected ease." (JTX23 at p. 1507; Tr. 168:4-10, 169:8-170:7). In other words, oxymorphone diol is "produced very,
very easily from Oxymorphone ABUK." (Tr. 418:5-14). On the other hand, oxycodone ABUK
hydrates to form oxycodone diol "much less readily than" oxymorphone ABUK hydrates to form
oxmorphone diol. (JTX-23 at p. 1506; Tr. 170:8-171:2,405:7-407:1). As explained previously,
the oxymorphone diol "act[s] like a reservoir for regenerating ABUK." (Tr. 388:6-23, 428:1328). Therefore, "ABUK precursors"-the oxymorphone diols-would "just ... regenerate
ABUK at the end of the day." (Tr. 436:5-16).

15

I
Weiss does not, on its own, disclose low-ABUK oxymorphone. (Tr. 380:6-15). That is,
it does not teach that the hydrogenation procedure described would result in the low-ABUK
oxymorphone claimed in the '779 patent. "Although published subject matter is 'prior art' for all
that it discloses, in order to render a claimed apparatus or method obvious, the prior art must
enable one skilled in the art to make and use the apparatus or method." 13 In re Kumar, 418 F.3d
1361, 1365 (Fed. Cir. 2005) (quoting Beckman Instruments, Inc. v. LKB Produkter AB, 892 F.2d
1547, 1551 (Fed. Cir. 1989)); see also In re Payne, 606 F.2d 303, 314-15 (C.C.P.A. 1979).
Since Weiss does not disclose low-ABUK oxymorphone, Defendants must "establish that a
person of ordinary skill would have nonetheless been able to make [the claimed invention]."

Geo. M Martin Co. v. Alliance Mach. Sys. Int'! LLC, 618 F.3d 1294, 1303 (Fed. Cir. 2010); see
also Rockwell Int'! Corp. v. United States, 147 F.3d 1358, 1365 (Fed. Cir. 1998). If the Weiss
hydrogenation procedure would not actually produce low-ABUK oxymorphone, it cannot be said
that the prior art enables those of skill in the art to make low-ABUK oxymorphone. While Dr.
Heathcock opines that the hydrogenation would result in the claimed invention, Dr. Davies
opines that an experiment would be required to verify that prediction. (Tr. 94:3-99:14, 389:22390:4). Dr. Heathcock did not run any experiments to confirm that a hydrogenation process
would indeed result in low-ABUK oxymorphone. (Tr. 151: 1-8, 174:4-10, 390:5-391: 11 ).
Therefore, Defendants have failed to show that a person of ordinary skill in the art could make
low-ABUK oxymorphone using hydrogenation.

13

"Under§ 103, ... a reference need not be enabled; it qualifies as a prior art, regardless, for whatever is
disclosed therein." Amgen Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d 1313, 1357 (Fed. Cir. 2003);
see also Symbol Techs., Inc. v. Opticon, Inc., 935 F.2d 1569, 1578 (Fed. Cir. 1991) ("While a reference
must enable someone to practice the invention in order to anticipate under § 102(b), a non-enabling
reference may qualify as prior art for the purpose of determining obviousness under§ 103."); Beckman
Instruments, Inc. v. LKB Produkter AB, 892 F.2d 1547, 1551 (Fed. Cir. 1989) ("Even if a reference
discloses an inoperative device, it is prior art for all that it teaches."). Whether the prior art enables one
skilled in the art to produce the claimed invention is, however, a different question.

16

Even ifthe Weiss hydrogenation procedure could produce low-ABUK oxymorphone,
however, Defendants would still have much of their work ahead of them. The prior art must
actually "suggest to one of ordinary skill in the art how to [make the claimed apparatus] with a
reasonable likelihood of success." Rockwell, 147 F.3d at 1365. Neither Weiss, nor Chapman,
disclose low-ABUK oxymorphone. Additionally, Defendants have not proven that the
combination of those references would enable a person of ordinary skill to make low-ABUK
oxymorphone with a reasonable expectation of success. See Geo. M Martin, 618 F.3d at 1303;
Rockwell, 147 F.3d at 1365. In fact, the Chapman inventors, when seeking to lower ABUK

levels in oxycodone, found that a single hydrogenation reaction was insufficient to reach the
desired ABUK levels. (Tr. 396:5-397:6; see also Tr. 32:2-384:3). "[T]here can be little better
evidence negating an expectation of success than actual reports of failure." Boehringer
Ingelheim Vetmedica, Inc. v. Schering-Plough Corp., 320 F.3d 1339, 1354 (Fed. Cir. 2003).

Although the Chapman inventors succeeded in creating low-ABUK oxycodone with a double
hydrogenation process, due to the differences between oxycodone and oxymorphone, that would
not suggest to a person having ordinary skill that the same process would have been effective in
creating low-ABUK oxymorphone. (Tr. 403:2-13, 414:5-14). Thus, Defendants have not shown
that a person of ordinary skill ''would have had a reasonable expectation of success in"
"achiev[ing] the claimed invention .... " Pfizer, 480 F.3d at 1361.
Defendants discuss, at length, the Federal Circuit's recent decision in Purdue Pharma
L.P. v. Epic Pharma., LLC, 811 F.3d 1345 (Fed. Cir. 2016). To the extent the conclusions in
Purdue are relevant to this case, 14 they do not suggest that low-ABUK oxymorphone would have

14

Defendants repeatedly cite to facts described in Purdue. This is improper. "The reports of [decisions]
may be referred to as expositions of law upon the facts there disclosed, but they are not evidence of those
facts in other cases." Mendenhall v. Cedarapids, Inc., 5 F.3d 1557, 1570 (Fed. Cir. 1993) (alteration in
original) (quoting MacKay v. Easton, 86 U.S. (19 Wall.) 619, 632 (1873)).

17

I

been obvious to one of ordinary skill in the art. In Purdue, the patentee-in applications which
continued from Chapman-claimed low-ABUK oxycodone. "The district court found that the
prior art taught that oxidation ofthebaine produced [oxycodone ABUK] and that it was well
known in the art that [oxycodone ABUK] could be removed using hydrogenation." Id. at 1351.
The patentee, in arguing for the patent's validity, argued that the discovery of the source of
oxycodone ABUK-the 8a isomer of oxycodone diol-rendered its solution non-obvious. The
Federal Circuit confirmed that the discovery of 8a as the source of the ABUK was not necessary
to the claimed invention, which was directed to low-ABUK oxycodone as an end product. "One
need not know that the [oxycodone ABUK] was derived from 8a" to know that it was obvious to
use hydrogenation to remove the oxycodone ABUK. Id. at 1353.
I fail to see the relevance of Purdue Pharma. Purdue's validity position hinged on
discovering the source of the oxycodone ABUK. Plaintiffs here make no analogous argument.
Additionally, the Federal Circuit's conclusion that low-ABUK oxycodone was obvious does not
command a conclusion that low-ABUK oxymorphone is obvious. As stated above, the evidence
reveals significant differences between oxycodone and oxymorphone, such that an ordinaryskilled artisan would not reasonably expect that what had been successful with oxycodone would
have been successful with oxymorphone.
I conclude that Defendnats have failed to make aprimafacie showing that the '779
patent would have been obvious to one of ordinary skill in the art.
iv. Secondary Considerations
"[S]econdary considerations, when present, must be considered in determining
obviousness." Ruiz, 234 F.3d at 667; see also Cyclobenzaprine, 676 F.3d at 1076 ("[E]vidence
on these secondary considerations is to be taken into account always, not just when the

18

decisionmak:er remains in doubt after reviewing the art." (internal quotation marks omitted)
(quoting Cable Elec. Prods. v. Genmark, Inc., 770 F.2d 1015, 1026 (Fed. Cir. 1985))). Here,
Plaintiff did not present any evidence on any secondary considerations. Defendants, however,
argue that there is evidence of near-simultaneous invention by others in the industry.
"Independently made, simultaneous inventions, made 'within a comparatively short space of
time,' are persuasive evidence that the claimed apparatus 'was the product only of ordinary
mechanical or engineering skill."' Geo. M. Martin Co. v. Alliance Mach. Sys. Int'/ LLC, 618
F.3d 1294, 1305 (Fed. Cir. 2010) (quoting Concrete Appliances Co. v. Gomery, 269 U.S. 177,
184 (1925)).
Defendants assert that Chapman's invention oflow-ABUK oxycodone through a process
involving hydrogenation is a "near-simultaneous invention." (D.I. 139 at 35; DTX-97 iii! 18590). Since low-ABUK oxycodone is not low-ABUK oxymorphone, I do not think there is any
evidence of simultaneous invention. Thus, there are no secondary considerations to be
contemplated here.
Having considered the scope and content of the prior art, the differences between the
prior art and the claims at issue, and the level of ordinary skill in the art, I conclude that
Defendants have not carried their burden of showing that "the differences between the subject
matter sought to be patented and the prior art are such that the subject matter as a whole would
have been obvious at the time the invention was made to a person having ordinary skill in the
[pertinent] art." 35 U.S.C. § 103.

19

II.

IMPLIED LICENSE
Teva concedes that its ANDA meets every limitation of the asserted claims of the '779

patent, but maintains, as an affirmative defense, that Plaintiffs' infringement claims are barred by
its implied license defense. (D.I. 150 if 16).
A. Legal Standard

"[A]n implied license, like an express license, is a defense to patent infringement."
Carborundum Co. v. Molten Metal Equip. Innovations, Inc., 72 F.3d 872, 878 (Fed. Cir. 1995).

A license may be inferred based on "[a]ny language used by the owner of [a] patent, or any
conduct on his part exhibited to another from which that other may properly infer that the owner
consents to his use of the patent in making or using it, or selling it .... " Wang Labs., Inc. v.
Mitsubishi Elecs. Am., Inc., 103 F.3d 1571, 1580 (Fed. Cir. 1997) (quoting De Forest Radio Tel.
Co. v. United States, 273 U.S. 236, 241 (1927)). The Federal Circuit has acknowledged that

there are "various avenues to an implied license." Id. "[I]mplied licenses arise by acquiescence,
by conduct, by equitable estoppel (estoppel in pais ), or by legal estoppel." Id. "[ J]udicially
implied licenses are rare under any doctrine." Id. at 1581.
In Wang Laboratories, the Federal Circuit confirmed the existence of an implied license
where:
the jury necessarily found that (1) a relationship existed between [the parties], (2) within
that relationship, [the patentee] granted to [the accused infringer] a right to use its ...
inventions, (3) [the patentee] received valuable consideration for that grant of right, (4)
[the patentee] denied that [the accused infringer] had an implied license, and (5) [the
patentee's] statements and conduct created the impression that [the patentee] consented to
[the accused infringer] making, using, or selling [the] patented inventions ....
Id. at 1579. "Courts grant implied licenses to preclude patent holders from suing purchasers for

infringement where, at the time of sale, the patentee led the purchaser to believe that his
manufacture, use, or sale of the patented article was permissible." Monsanto Co. v. Good, 2004
20

WL 1664013, at *7 (D.N.J. July 23, 2003). "A mere sale," however, "does not import a license
except where the circumstances plainly indicate that the grant of a license should be inferred."
Bandag, Inc. v. Al Bolser's Tire Stores, Inc., 750 F.2d 903, 925 (Fed. Cir. 1984). "[T]he alleged

infringer ... ha[ s] the burden of establishing the existence of an implied license as an affirmative
defense." Carborundum, 72 F.3d at 878. Whether an implied license exists, based on the
underlying facts, is a question oflaw. Anton/Bauer, Inc. v. PAG, Ltd., 329 F.3d 1343, 1348 (Fed.
Cir. 2003).

B. Factual Background
Teva and Mallinckrodt, in 2008, entered into a supply agreement for the supply of nonlow-ABUK oxymorphone to Teva. (D.I. 154 iJ 7). 15 That agreement expired in 2009. (Id.). In
late 2010 and early 2011, Teva purchased two batches oflow-ABUK oxymorphone API 16 from
Mallinckrodt pursuant to stand-alone purchase orders. (Id.

iii! 11-12). Those purchase orders

were dated October 31, 2010 and February 3, 2011, respectively. (Id.). Mallinckrodt shipped
the requested quantities oflow-ABUK oxymorphone API, and Teva paid Mallinckrodt the
amount due. (Id.

iii! 13-14). Since those purchase orders, Teva has not purchased any low-

ABUK oxymorphone API from Mallinckrodt. (Id.

iJ 16).

Mallinckrodt maintains a Drug Master File ("DMF") with the FDA, which contains
confidential and proprietary information about its low-ABUK oxymorphone APL (D.I. 154 iJ
17). This DMF is numbered 14502. (D.I. 154 iJ 17). In February 2012, Teva requested a Letter

15

For the implied license phase of the trial, the parties submitted a joint stipulation of facts. (D.I. 154).
Since the underlying facts material to Teva's implied license defense are not in dispute, the stipulation is
adopted as the Court's findings of fact.
16
API means "active pharmaceutical ingredient." (D.I. 154 ~ 1). Mallinckrodt is in the business of,
among other things, manufacturing and selling API for use in pharmaceutical products. (Id.). Teva is in
the business of, among other things, manufacturing and selling finished dosage forms which contain APL
(Id.~ 3).

21

of Authorization ("LOA") for low-ABUK oxymorphone API from Mallinckrodt. (D.I. 154 ii
18). On March 8, 2012, Mallinckrodt sent a copy of an LOA for low-AB UK oxymorphone API
to Teva. (D.I. 154 ii 20; DTX-501). This LOA allowed the FDA to review, in connection Teva's
ANDA filing, the information contained in Mallinckrodt's DMF, without Mallinckrodt having to
share that information with Teva. (Tr. 564:2-13; 694:1-695:2). In other words, the LOA
provided a mechanism whereby Teva could cross-reference information about Mallinckrodt's
low-ABUK oxymorphone API in its ANDA, without Mallinckrodt having to reveal that
information to Teva. As explained by Teva's industry expert, Dr. Fabian, the LOA
accomplished two things: (1) Mallinckrodt authorized Teva to incorporate by reference the
information from its DMF into an ANDA, and (2) Mallinckrodt authorized the FDA to review its
DMF when considering Teva's ANDA application. (Tr. 703:17-704:3). Mallinckrodt only knew
that Teva sought to use DMF No. 14502 in a product; it did not know the particular product for
which Teva sought the LOA. (Tr. 748:23-749:15).
On April 17, 2012, Teva submitted an ANDA to the FDA, requesting approval for its
generic version ofEndo's Crush-Resistant Formulation of Opana ER ("Teva CRF ANDA").
(D.I. 154 ii 22). Teva incorporated DMF No. 14502 into the Teva CRF ANDA. (Id.

ii 24).

Mallinckrodt is the only supplier oflow-ABUK oxymorphone API referenced in the Teva CRF
ANDA. (Id.

ii 25).

Teva has the ability, however, to amend its ANDA to qualify additional

suppliers oflow-ABUK oxymorphone APL (Id.

ii 26).

In August 2012, Mallinckrodt and Teva began negotiating a supply agreement for lowABUK oxymorphone APL (D.1. 154 ii 30). On August 3, 2012, Ayne Klein of Teva sent
Stephanie Bucalo and Nick Litzsinger of Mallinckrodt a draft supply agreement via email. (Id.
31; JTX-300; JTX-301). Mr. Litzsinger, on September 5, 2012, emailed Ms. Klein a counter-

22

ii

proposal. (D.I. 154 if 32; JTX-302; JTX-303). Teva did not respond with any further proposals.
(D.I. 154 if 33). Several months later, on or around November 29, 2012, Ms. Klein sent an email
to Mr. Litzsinger which memorialized a discussion which had occurred the previous day. (Id.

if

34). In the email, Ms. Klein wrote: "3) Oxymorphone-we agreed we would complete
Morphine supply agreement and then tackle the Oxymorphone." (Id.

if 34; DTX-542). The

parties never reached an agreement for the supply oflow-ABUK oxymorphone APL (D.I. 154 if
36).
In 2012 and 2013, Mallinckrodt and Endo were parties to Patent Interference No. 105,893
in the PTO, which related to U.S. Patent Application No. 11/915,606-which issued as the '779
patent-and U.S. Patent No. 7,851,482 ("the '482 patent"). 17 (Id.

if 38). On May 15, 2013, Endo

filed a patent infringement lawsuit against Mallinckrodt, alleging that a Mallinckrodt ANDA
filing infringed the '482 patent. (Id.

if 39). On December 16, 2013, Endo and Mallinckrodt

settled the interference proceedings and the district court litigation, and entered into two license
agreements. (Id.

if 41; JTX-3; PTX-10). Pursuant to the agreement settling the interference

proceedings, Mallinckrodt granted Endo an exclusive license to the patent which ultimately
issued as the '779 patent. (Id.

if 42; JTX-3). After settling with Endo, Mallinckrodt did not

withdraw or modify the LOA it had issued to Teva. (Id.

iii! 47-51).

Because of its agreement

with Endo, Mallinckrodt is unwilling to sell low-ABUK oxymorphone API to Teva for use in the
product described in the Teva CRF ANDA. (Id.

if 44). Mallinckrodt remains willing to sell low-

ABUK oxymorphone API to Teva for use in an immediate release oxymorphone product. (Id.
45).

17

Endo had previously acquired U.S. Patent No. 7,851,482 from Johnson Matthey. (Id.~ 40).

23

if

On November 7, 2014, less than two weeks after the issuance of the '779 patent, Endo
and Mallinckrodt sued Teva for infringement of the '779 patent. (Id.

if 43).

D. Conclusions of Law
In a Hatch-Waxman case, a plaintiff's infringement claim is based on the accused
infringer's future conduct, rather than past acts of infringement. 18 "The filing of an ANDA is
considered an act of infringement under§ 271(e)(2)(A), but this 'act' is merely a vehicle 'to
create case or controversy jurisdiction to enable a court to promptly resolve' a dispute
concerning infringement that will happen in the future." Bayer AG v. Etan Phann. Research

Corp., 212 F.3d 1241, 1249 (Fed. Cir. 2000) (quoting Glaxo, Inc. v. Novopharm, Ltd., 110 F.3d
1562, 1569 (Fed. Cir. 1997)). Thus, for an implied license defense to succeed, the accused
infringer must demonstrate that the patentee consented to its use of the claimed invention in "the
ANDA product that is likely to be sold following FDA approval." Spectrum Phann., Inc. v.

Sandoz Inc., 802 F.3d 1326, 1336 (Fed. Cir. 2015). Here, Teva has failed to make such a
showing.
Teva's implied license defense relates only to API which may be supplied by
Mallinckrodt. (D.I. 175 at p. 3 n.2). 19 According to Teva, since Mallinckrodt is the only API
supplier identified in the Teva CRF ANDA, the product that is likely to be sold-should Teva's
ANDA be approved-will contain Mallinckrodt APL Put another way, Teva argues that, to the
extent any infringing low-ABUK oxymorphone API ends up in its product, that API will come
from Mallinckrodt. Based on Mallinckrodt's past conduct, Teva argues that it must be entitled-

18

Otherwise infringing acts, undertaken in connection with the development and submission of an
ANDA, are immunized from liability. 35 U.S.C. § 271(e)(l).
19
Teva concedes this and also acknowledges that "[n]o case or controversy exists today regarding
whether the sale of a hypothetical future product that includes API purchased from a supplier other than
Mallinckrodt would infringe the '779 patent." (Id.).

24

or, impliedly licensed-to use low-ABUK oxymorphone API that it receives from
Mallinckrodt. 20 Under these circumstances, to succeed in an implied license defense, Teva must
show that Mallinckrodt consented to Teva's use of the patented invention in the product likely to
be sold. Under the facts here, there is no such license.
Teva's argument focuses on the 2010 and 2011 purchase orders, and the 2012 LOA. The
2010 and 2011 purchase orders concerned discrete, stand-alone purchases. (D.I. 154 iii! 11-15;
Tr. 602:18-605:11). On each, Teva included written terms and conditions. No evidence suggests
that Mallinckrodt objected to these terms and conditions. (D.I. 154 if 15). Each purchase order
included an integration clause which provided that, "[e]xcept as expressly set forth in writing
executed by [Teva], the terms and conditions set forth in this order constitute the entire
agreement between the parties regarding the subject matter .... " (DTX-502; DTX-503). Thus,
these terms and conditions constitute the entirety of the agreement with respect to each purchase
order. Under the terms and conditions in each order, Mallinckrodt agreed "to exonerate,
indemnify and hold harmless [Teva] from and against any and all liability ... which may accrue
to, or be sustained by [Teva] on account of any claim ... brought against [Teva] ... for ...
infringement of any patent ... by reason of the manufacture of goods covered by this order .... "
(DTX-502; DTX-503). By the terms of the purchase orders, Mallinckrodt granted Teva
permission to use the low-ABUK oxymorphone API however it wished. As the terms and
conditions of the purchase orders make clear, however, the scope of that permission does not
extend beyond the "manufacture of goods covered by th[e] [purchase] order." (DTX-502; DTX503). The only material covered by the terms and conditions is the low-ABUK oxymorphone

20

There is some intuitive appeal to the argument that Mallinckrodt should not be able to sue a buyer for
using the product it sold. In such a scenario, however, Teva would likely obtain a license not by virtue of
Mallinckrodt's past conduct, but through the terms of a later transaction. The question here is whether the
conduct hitherto undertaken by Mallinckrodt suffices to show that Mallinckrodt licensed the '779 patent.

25

API Teva actually purchased. No evidence suggests that, by selling low-ABUK oxymorphone
API to Teva on two occasions, Mallinckrodt consented to Teva's commercial sale of products
embodying the '779 patent through the use oflow-ABUK oxymorphone API not covered by the
purchase orders. The two purchases oflow-ABUK oxymorphone API do not create an implied
license.
Aside from the purchase orders, Teva relies heavily on the LOA issued by Mallinckrodt.
An LOA is a regulatory document. In an LOA, a DMF holder grants an ANDA applicant
"permission to incorporate their API into the ANDA." (Tr. 697:21-698:6). As conceded by
Teva's expert, Dr. Fabian, these documents create no binding commercial obligations. (Tr.
706:4-707:5). Dr. Fabian testified that "the issue of successful or unsuccessful consummation of
a Supply Agreement is a completely independent issue as to whether or not permissions have
been granted based on actions of certain parties." (Id.). Further, Dr. Fabian testified that "there
is no obligation to purchase [API] because an LOA has been received." (Tr. 698:19-699:6).
Thus, while Mallinckrodt is the only supplier listed on the Teva CRF, it is not under any
obligation to supply low-ABUK oxymorphone API to Teva. Similarly, no witness testified that
the LOA itself confers a patent license.
Teva argues that because Mallinckrodt chose not to withdraw or modify the LOA,
Mallinckrodt "suggested [to Teva] that it could file ANDA 204324 without being sued for
infringing the '779 patent." (D.I. 175 at p. 10). This inference draws on a misapprehension
about the significance of a LOA. Since a LOA places no binding obligation on Mallinckrodt, it
would have no reason to withdraw the LOA. Additionally, leaving the LOA in place allows the
FDA to review Teva's ANDA while Teva seeks to find, and qualify, additional suppliers oflowABUK oxymorphone APL (Tr. 724: 16-726: 1). This is standard within the industry, as ANDA

26

applicants seek to qualify multiple suppliers for pharmaceutical products. (Tr. 617:13-618:3,
699:8-22). Indeed, Dr. Fabian testified that, when an API supplier is asked for a LOA from a
manufacturer, the API supplier "would have no reason to believe" that "they would be the
primary supplier to the applicant." (Tr. 699:8-22). This is because "API suppliers realize that
ANDA sponsors mitigate [their] risk ... by including more than a single supplier in their
ANDA." (Id.). Thus, while an ANDA applicant might request an LOA so that it can eventually
"market a dosage form on the market with [the DMF holder's] API in that dosage form," that is
not the only possible scenario. (Tr. 698 :2-18).
Teva has amassed considerable evidence for the unremarkable proposition that Teva
could use the low-ABUK oxymorphone API supplied by Mallinckrodt in pursuing its ANDA.
Mallinckrodt did not place any limits on the low-ABUK oxymorphone API Teva purchased in
2010 and 2011, and indeed, explicitly agreed to "exonerate, indemnify and hold harmless" Teva
from any patent infringement liability. (DTX-502; DTX-503). Ms. Klein testified, based on the
preceding facts, that Teva believed it was authorized to include Mallinckrodt's low-ABUK
oxymorphone API in its ANDA. (Tr. 574:2-16). Additionally, to the extent Teva used the lowABUK oxymorphone API for a purpose reasonably related to its ANDA, such activity is
protected by§ 271(e)(l). Mallinckrodt is not, however, suing Teva for using the low-ABUK
oxymorphone API it previously supplied. Rather, Mallinckrodt is suing Teva for including lowABUK oxymorphone API in the product that is likely to be sold in the future. Teva has not
pointed to any evidence demonstrating that Mallinckrodt granted Teva a license to include lowABUK oxymorphone API in the CRF product it ultimately sells. That is a fatal shortcoming.
In support of its theory of implied license, Teva introduced evidence about the way these
sorts of interactions might ordinarily proceed between an entity like Teva and an entity like

27

Mallinckrodt. Ms. Klein summarized the relevant process as follows: "You order a product. You
test the product. You put it into your drug product towards the ultimate goal of submitting to
FDA, having a review, having the product approved and selling it commercially. And the fact
that somebody ... sells you the API and gives you that Letter of Authorization, this is the
process." (Tr. 574:7-16). Ms. Klein testified that she "never had anyone provide [her] with a
letter of authorization, put them in the [ANDA] and then have them sue me for using their
product." (Tr. 575:7-10). Mr. Litzsinger similarly testified that Mallinckrodt had never "given
an LOA to a customer and then sued that customer for filing an ANDA that included the API
related to that LOA." (Tr. 829:13-18). Dr. Fabian also stated that, in his experience, he had
never "seen a case where ... a DMF holder ... provided [a] LOA to the ANDA applicant and
then subsequently ... sued them." (Tr. 729:4-22). I think this testimony suggests that, in the
majority of circumstances, a party in Teva's shoes could expect to eventually enter into an API
supply agreement with a party in Mallinckrodt's shoes. That did not happen here. Teva cannot
pretend that it did in order to sustain an implied license defense. This is perhaps an unusual
situation, but it is not one where an implied license arises. While Teva may have hoped that
Mallinckrodt would eventually supply it low-ABUK oxymorphone API, despite Mallinckrodt's
assertions for the last twenty months that it would not (D.I. 154 iJ 44; Tr. 646:21-647:13), "an
implied license cannot arise out of unilateral expectations or even reasonable hopes of one
party." Stickle v. Heublein, Inc., 716 F.2d 1550, 1558 (Fed. Cir. 1983).
Teva briefly argues that its .product may not contain low-ABUK oxymorphone after all.
Plaintiff relies on the FDA's requirement that an opioid manufacturer must show either that (1)
the drug contains less than 10 ppm of ABUK, or (2) that levels above that amount are not

28

genotoxic. (D.I. 166 at pp. 13-14). 21 This argument fails for at least two reasons. First, the
specification in the Teva CRF ANDA limits the amount of oxymorphone ABUK to less than 10
ppm. (D.I. 154 if 24). Second, Teva never raised this argument in the Final Pretrial Order, and it
is therefore waived.
Teva argues that it has proven the five facts that the Federal Circuit found sufficient to
create an implied license in Wang Laboratories. I disagree. In Wang Laboratories, the Federal
Circuit concluded that the jury necessarily found that the patentee granted the accused infringer a
right to use the claimed invention. Wang Laboratories, 103 F.3d at 1579. Here, as discussed
above, Mallinckrodt only granted Teva the right to use the low-ABUK oxymorphone API which
was the subject of the two purchase orders.
Additionally, in Wang Laboratories, the patentee had "received valuable consideration
for [the] grant of [a] right [to use the claimed invention]." Id. Here, the only consideration paid
to Mallinckrodt was the purchase price of the two stand-alone purchases. In short, Teva paid for
two quantities oflow-ABUK oxymorphone API; it did not pay for rights regarding future sales
oflow-ABUK oxymorphone APL Teva argues that Mallinckrodt received consideration for the
LOA "in the form of potential future sales of commercial quantities of [low-AB UK]
[o]xymorphone APL" (D.I. 166 at pp. 9-10). As Dr. Fabian testified, however, commercial
supply is an issue entirely separate from a LOA authorization. (Tr. 706:4-707:5). Since a LOA
does not create a binding commercial obligation, Mallinckrodt is under no obligation to sell any
low-ABUK oxymorphone API to Teva. "[W]here the promisor may perform or not, solely on
the condition of his whim, his promise will not serve as consideration." Wallach v. Eaton Corp.,
125 F. Supp. 3d 487, 493-94 (D. Del. 2015) (quotation marks omitted) (quoting 3 Samuel
21

This argument is premised on JTX-4, which was not admitted into evidence or supported by any
testimony.

29

Williston & Richard A. Lord, Williston on Contracts§ 7.7 (4th ed. 1992)), appeal filed, No. 153320 (3d Cir. Sept. 29, 2015).
I therefore conclude that Teva has failed to demonstrate the existence of an implied
license.

III.

CONCLUSION

Defendants failed to prove by clear and convincing evidence that any of the asserted
claims of the '779 patent are invalid. Teva failed to prove its affirmative defense of implied
license by a preponderance of the evidence.
Plaintiffs should submit an agreed upon form of final judgment within two weeks.

30