In the United States Court of Appeals for the Federal … the United States Court of Appeals for the Federal Circuit ... Winston & Strawn LLP ... Winston & Strawn LLP (George C. Lombardi,

Nos. 2012-1340, -1341

In the United States Court of Appeals for the Federal Circuit

ALCON RESEARCH LTD., PLAINTIFF-APPELLANT

v.

BARR LABORATORIES, INC., DEFENDANT-CROSS APPELLANT

__________

ON APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE, CONSOLIDATED CASE NO. 09-CV-0318, HON. LEGROME D. DAVIS, PRESIDING

__________

BRIEF OF DEFENDANT-CROSS APPELLANT BARR LABORATORIES, INC.

__________

GEORGE C. LOMBARDI BRADLEY C. GRAVELINE WILLIAM P. FERRANTI Winston & Strawn LLP 35 West Wacker Drive Chicago, Illinois 60601 (312) 558-5600 [email protected] [email protected] [email protected] Counsel for Defendant-Cross Appellant Barr Laboratories, Inc.

CERTIFICATE OF INTEREST

Pursuant to Circuit Rule 47.4, undersigned counsel for Defendant-Cross Ap-

pellant Bar Laboratories, Inc., certifies the following:

1. The full name of every party or amicus represented by us is:

Barr Laboratories, Incorporated 2. The name of the real party in interest (if the party named in the caption is not

the real party in interest) represented by us is: Not applicable; the party named in the caption is the real party in interest

3. All parent corporations and any publicly held companies that own 10% or more of the stock of any party represented by us are:

Barr Laboratories, Inc. is a subsidiary of Barr Pharmaceuticals, LLC which is a subsidiary of Teva Pharmaceuticals USA, Inc., which is an indirect wholly owned subsidiary of Teva Pharmaceutical Industries, Ltd. Teva Pharmaceutical Industries, Ltd. is the only publicly traded company that in-directly owns, through the aforementioned chain, 10% or more of Barr La-boratories, Inc.

4. The names of all law firms and the partners or associates that appeared for

the parties now represented by us in the trial court or expected to appear in this court are:

Winston & Strawn LLP (George C. Lombardi, Bradley C. Graveline, Wil-

liam P. Ferranti, Brian J. Nisbet, Shawn M. Taylor, Ali Ahmed, Eric L. Broxterman, Keith R. Pozulp, Trang Hoang); Phillips, Goldman & Spence, P.A. (John C. Phillips, Jr., Megan C. Haney); Farman LLP (Brian E. Far-man)

Dated: September 28, 2012 /s/ William P. Ferranti

WILLIAM P. FERRANTI Counsel for Defendant-Cross Appellant Barr Laboratories, Inc.

ii

TABLE OF CONTENTS

TABLE OF AUTHORITIES ..................................................................................... v 

STATEMENT OF RELATED CASES ..................................................................... 1 

JURISDICTIONAL STATEMENT .......................................................................... 1 

STATEMENT OF THE ISSUES............................................................................... 2 

STATEMENT OF THE CASE .................................................................................. 3 

STATEMENT OF FACTS ........................................................................................ 5 

A.  The Castor Oil Patents broadly claim a method of using PECO to enhance the chemical stability of any prostaglandin. ....................... 6 

B.  Barr’s product does not use PECO to enhance chemical (as opposed to physical) stability, as claimed in the Castor Oil Patents.................................................................................................... 8 

1.  Alcon offered no direct evidence of infringement. .................... 9 

2.  The inventor himself (Mr. Schneider) testified that PECO enhances physical, not chemical, stability. .................. 10 

3.  Based on physical stability data from a solubility study, Alcon’s under-qualified expert (Dr. Levinson) offered an opinion that the district court found flawed. ....................... 11 

C.  The Castor Oil Patents are not enabled and lack adequate written description. .............................................................................. 19 

SUMMARY OF ARGUMENT ............................................................................... 22 

ARGUMENT ........................................................................................................... 26 

I.  Barr’s ANDA Product Does Not Infringe the Castor Oil Patents ................. 26 

A.  The district court accurately assessed the weakness of Alcon’s evidence, and agreed with Barr’s expert. ............................................ 27 

B.  The district court’s credibility assessment was not clear error. .......... 33 

iii

II.  The District Court Correctly Held The Castor Oil Patents Invalid For Lack of Enablement. ...................................................................................... 35 

A.  Guided by Wands, the district court made unimpeachable and well-supported findings of fact establishing that the Castor Oil Patents are not enabled. ....................................................................... 35 

1.  The Castor Oil Patents provide very little guidance to one skilled in the art. ................................................................ 37 

2.  The relevant technology is highly unpredictable, and there are several factors that impact chemical stability. .......... 40 

3.  The person of ordinary skill in the art possesses a relatively high level of skill, but the prior art teaches little about chemically stabilizing prostaglandins. .................. 41 

4.  The patent claims are extremely broad, and not supported by enough disclosure. ............................................. 42 

B.  The district court applied the correct legal standard in concluding that the Castor Oil Patents required “undue experimentation.” ................................................................................ 45 

III.  The District Court Correctly Held The Castor Oil Patents Invalid For Failing The Written Description Requirement. ............................................. 48 

IV.  Barr Is Entitled To A Judgment Of Non-Infringement On The ’383 And ’052 Patents. ........................................................................................... 51 

A.  Alcon alleged in its complaint, but failed to prove at trial, that Barr’s ANDA product infringes the ’383 and ’052 Patents. ............... 52 

1.  Alcon never removed these infringement claims from the litigation. ............................................................................ 52 

2.  This Court’s precedent does not require allowing Alcon to prevail by procedural error. ................................................. 55 

B.  The district court should have granted Barr’s motion for judgment as a matter of law because Alcon is estopped from arguing infringement of the ’383 and ’052 patents. ............................ 62 

iv

C.  In the absence of a judgment of non-infringement regarding the ’383 and ’052 patents, Alcon’s non-infringed patents may prevent Barr from entering the market. ............................................... 65 

CONCLUSION ........................................................................................................ 68 

v

TABLE OF AUTHORITIES

Page(s) CASES

800 Adept, Inc. v. Murex Sec., Ltd., 539 F.3d 1354 (Fed. Cir. 2008) .............................................................. 58, 59, 61

A.C. Aukerman Co. v. R.L. Chaides Const. Co., 960 F.2d 1020 (Fed. Cir. 1992) (en banc) .......................................................... 65

ACCO Brands, Inc. v. ABA Locks Mfrs. Co., 501 F.3d 1307 (Fed. Cir. 2007) .......................................................................... 51

Advance Transformer Co. v. Levinson, 1986 WL 84365 (N.D. Ill. June 17, 1986) .................................................... 57, 62

Advance Transformer Co. v. Levinson, 837 F.2d 1081 (Fed. Cir. 1988) ........................................... 56, 57, 58, 60, 61, 62

Alza Corp. v. Andrx Pharms., LLC, 603 F.3d 935 (Fed. Cir. 2010) ............................................................................ 37

Amgen, Inc. v. Chugai Pharma. Co., 927 F.2d 1200 (Fed. Cir. 1991) .......................................................................... 37

Amgen Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d 1313 (Fed. Cir. 2003) .................................................................... 26, 49

Ariad Pharma., Inc. v. Eli Lilly & Co., 598 F.3d 1336 (Fed. Cir. 2010) (en banc) .................................................... 49, 50

Atlas Powder Co. v. E.I. du Pont De Nemours & Co., 750 F.2d 1569 (Fed. Cir. 1984) .......................................................................... 47

Auto Techs. Int’l, Inc. v. BMW of N. Am., Inc., 501 F.3d 1274 (Fed. Cir. 2007) .......................................................................... 43

Bjorgung v. Whitetail Resort, LP, 550 F.3d 263 (3d Cir. 2008) ............................................................................... 55

vi

Boston Scientific Corp. v. Johnson & Johnson, 647 F.3d 1353 (Fed. Cir. 2011) .......................................................................... 49

Caraco Pharm. Labs., Ltd. v. Forest Labs., Ltd., 527 F.3d 1278 (Fed. Cir. 2008) .................................................................... 63, 66

CIF Licensing, LLC v. Agere Sys. Inc., 727 F. Supp. 2d 337 (D. Del. 2010) .................................................................... 56

DeGeorge v. Bernier, 768 F.2d 1318 (Fed. Cir. 1985) .......................................................................... 47

Denneny v. Siegel, 407 F.2d 433 (3d Cir. 1969) ............................................................................... 55

Egyptian Goddess, Inc. v. Swisa, Inc., 543 F.3d 665 (Fed. Cir. 2008) ...................................................................... 55, 60

Estate of Oliva ex rel. McHugh v. New Jersey, 604 F.3d 788 (3d Cir. 2010) ......................................................................... 54-55

G-I Holdings, Inc. v. Reliance Ins. Co., 586 F.3d 247 (3d Cir. 2009) ............................................................................... 64

Genentech Inc. v. Novo Nordisk A/S, 108 F.3d 1361 (Fed. Cir. 1997) .................................................................... 43, 48

Impax Labs., Inc. v. Aventis Pharm. Inc., 545 F.3d 1312 (Fed. Cir. 2008) .................................................................... 42-43

In re Bath & Kitchen Fixtures Antitrust Litig., 535 F.3d 161 (3d Cir. 2008) ............................................................................... 53

In re Wands, 858 F.2d 731 (Fed. Cir. 1988) ............................................. 35, 36, 37, 39, 40, 45

Krystal Cadillac-Oldsmobile GMC Truck, Inc. v. General Motors Corp., 337 F.3d 314 (3d Cir. 2003) ............................................................................... 64

Lear Auto. Dearborn, Inc. v. Johnson Controls, Inc., 528 F. Supp. 2d 654 (E.D. Mich. 2007) ....................................................... 52, 53

vii

LizardTech, Inc. v. Earth Resource Mapping, Inc., 424 F.3d 1336 (Fed. Cir. 2005) .......................................................................... 47

MagSil Corp. v. Hitachi Global Storage Techs., Inc., 687 F.3d 1377 (Fed. Cir. 2012) .......................................................................... 36

McKenna v. City of Philadelphia, 649 F.3d 171 (Fed. Cir. 2011) ............................................................................ 51

Minn. Mining & Mfg. v. Barr Labs., 289 F.3d 775 (Fed Cir. 2002) ............................................................................. 67

Mirror Worlds, LLC v. Apple, Inc., 784 F. Supp. 2d 703 (E.D. Tex. 2011) ................................................................ 56

Neman v. Quigg, 877 F.2d 1575 (Fed. Cir. 1989) .................................................................... 39, 40

Pfaff v. Wells Elect., Inc., 525 U.S. 55 (1998) .............................................................................................. 47

Pharm. Resources, Inc. v. Roxane Labs., Inc., 253 F. App’x 26 (Fed. Cir. 2007) ................................................................. 44, 45

Scanner Techs. Corp. v. ICOS Vision Sys. Corp. N.V., 528 F.3d 1365 (Fed. Cir. 2008) ..................................................56, 57, 58, 59, 61

Shire Labs., Inc. v. Corepharma, LLC, 2008 WL 4822186 (D.N.J. Nov. 3, 2008) .............................................. 63, 64, 65

Sitrick v. Dreamworks, LLC, 516 F.3d 993 (Fed. Cir. 2008) ...................................................................... 36, 46

Smith & Nephew, Inc. v. Arthrex, Inc., 453 F. App’x 977 (Fed. Cir. 2011) ..................................................................... 55

Streck, Inc. v. Res. & Diagnostic Sys., Inc., 665 F.3d 1269 (Fed. Cir. 2012) .......................................................................... 36

Sw. Efuel Network, L.L.C. v. Transaction Tracking Tech., Inc., 2010 WL 963697 (E.D. Tex. Jan. 13, 2010) ...................................................... 56

viii

TecSec, Inc. v. Int’l Bus. Mach. Corp., 769 F. Supp. 2d 997 (E.D. Va. 2011) ................................................................. 56

Teva Pharm. USA, Inc. v. Novartis Pharm. Corp., 482 F.3d 1330 (Fed. Cir. 2007) .......................................................................... 66

Tol-O-Matic, Inc. v. Proma Produkt-Und Mktg. Gesellschaft m.b.H., 945 F.2d 1546 (Fed. Cir. 1991) .................................................................... 60, 61

STATUTES

21 U.S.C. § 355(j) .................................................................................................... 66

28 U.S.C. § 1295(a)(1) ........................................................................................... 2, 3

28 U.S.C. § 1331 ........................................................................................................ 1

28 U.S.C. § 1338(a) ................................................................................................... 1

28 U.S.C. § 2111 ...................................................................................................... 33

35 U.S.C. § 112 ........................................................................................................ 36

35 U.S.C. § 282 ........................................................................................................ 59

OTHER AUTHORITIES

Fed. R. Civ. P. 8(e) ................................................................................................... 61

Fed. R. Civ. P. 15 ................................................................................... 26, 53, 54, 61

Fed. R. Civ. P. 41 ................................................................................... 26, 53, 54, 61

STATEMENT OF RELATED CASES

Cross-appellant is not aware of any related cases.

JURISDICTIONAL STATEMENT

These cross-appeals follow final judgment in a patent infringement action

initiated by Alcon Research, Ltd. against Barr Laboratories, Inc. (and others) in the

U.S. District Court for the District of Delaware. Alcon claimed that Barr infringed

six patents by filing an abbreviated new drug application (“ANDA”) seeking FDA

approval for a generic version of Alcon’s name-brand product, Travatan Z®. The

district court had jurisdiction pursuant to 28 U.S.C. §§ 1331 and 1338(a).

The claims against the other defendants were dismissed by stipulation.

At trial, Barr stipulated to the infringement and validity of two patents

(whose expiration was imminent). Barr also moved for judgment as a matter of

law on two other patents (based on Alcon’s failure to introduce any evidence of in-

fringement), but that motion was denied on procedural grounds. A53-54. As for

the final two patents, the district court held the asserted claims not infringed and

invalid for both lack of enablement and failure to meet the written description re-

quirement. See A1-52. Final judgment was entered January 9, 2012. A56-57.

Both parties filed timely post-judgment motions, Alcon on February 3, 2012,

and Barr on February 6, 2012 (A6234, A6254), which the district court denied on

March 16, 2012 (A58-78). Both parties then filed timely notices of appeal, Alcon

2

on April 12, 2012 (A6325-27), and Barr on April 13, 2012 (A6328-29). This

Court has jurisdiction pursuant to 28 U.S.C. §1295(a)(1).

STATEMENT OF THE ISSUES

Alcon raises three issues on appeal:

I. Whether the district court clearly erred in finding that claim 12 of U.S.

Patent No. 5,631,287 and claim 19 of U.S. Patent No. 6,011,062 are not infringed

by Barr’s proposed generic product.

II. Whether the district court clearly erred in the findings of fact support-

ing its holding that these same claims are invalid for lack of enablement.

III. Whether the district court clearly erred in finding that these same

claims are also invalid for lack of a sufficient written description.

Barr’s cross-appeal adds a fourth issue:

IV. Whether a plaintiff is entitled to avoid a judgment of non-

infringement where that plaintiff pleaded infringement claims in its complaint, de-

clared its intention to seek a stipulated dismissal of those claims with prejudice, but

then neither followed through on that promise nor presented any evidence of in-

fringement at trial, at which the defendant—consistent with its own pleaded prayer

for relief—moved for judgment as a matter of law.

3

STATEMENT OF THE CASE

Travatan Z® is a prescription eye drop indicated to treat glaucoma and ocu-

lar hypertension. In July 2009, after Barr and Par Pharmaceutical, Inc. separately

sought approval for generic versions of that product, Alcon filed suit alleging in-

fringement of six patents: U.S. Patent Nos. 5,631,287 (the “’287 patent”),

6,011,062 (the “’062 patent”), 6,503,497 (the “’497 patent”), 6,849,253 (the “’253

patent”), 5,510,383 (the “’383 patent”) and 5,889,052 patent (the “’052 patent”).

The ’287 and ’062 patents (the “Castor Oil Patents”) (A6362-71, A6372-83) are

the subject of Alcon’s appeal. The ’383 and ’052 patents are the subject of Barr’s

cross-appeal.

In August 2009, Barr answered Alcon’s complaint. In addition to denying

Alcon’s allegations, Barr prayed for a declaration that all six patents raised in the

complaint are invalid and not infringed by Barr’s ANDA product. A177-92. On

April 28, 2010, the district court ordered that any motion to amend the pleadings

be filed on or before June 30, 2010. No such motion was filed.

On August 4, 2010, however, Alcon sent a letter to Barr and Par stating that

it had “decided to dismiss its claims asserting [the ’383 and ’052 patents] against

the Barr and Par products that are generic versions of … Travatan Z®.” A11527.

Alcon stated that it would “prepare and circulate for [Barr and Par’s] review and

signature a Stipulation of Dismissal with Prejudice of those claims.” Id. But the

4

parties never reached agreement on such a stipulation. Prior to trial, the parties

submitted pretrial orders in which Alcon stated that it was “no longer asserting in-

fringement” of the ’383 and ’052 patents, and had “dropped” those patents from

the case. A2295, A3507. Alcon, however, never sought leave to amend its com-

plaint, or dismissal of the claims related to the ’383 and ’052 patents.

After the first day of trial, Par and Alcon settled. A4980-81. Barr then en-

tered a stipulation that its ANDA product infringed the ’497 and ’253 patents

(which expired in May 2012), and trial continued on the other patents asserted in

Alcon’s complaint. A4982-84.

At the close of Alcon’s case, Barr moved for judgment as a matter of law on

Alcon’s claims with respect to the ’383 and ’052 patents. A5686. As Barr ex-

plained, Alcon pleaded claims for infringement of those patents, those claims were

never dismissed from the litigation, and Alcon had not offered any evidence of in-

fringement of either patent. Id. Barr, therefore, was entitled to a judgment “on the

grounds that plaintiff has failed to establish infringement of those claims [by a]

preponderance of the evidence.” Id. Alcon opposed, but acknowledged that its

failure to seek dismissal had created a “lingering ambiguity.” A5013.

Following the completion of trial and the parties’ submission of post-trial

briefs, the district court issued a thorough opinion, published at 837 F. Supp. 2d

364 (D. Del. 2011), holding that Barr’s ANDA product does not infringe the Cas-

5

tor Oil Patents, and that those patents are invalid due to non-enablement and lack

of adequate written description. See A1-52. But the court “decline[d] to make any

findings or draw any conclusions about the infringement or validity of the [’383

and ’052] patents” because “neither party presented any evidence on [those] pa-

tents at trial.” A54.

Both parties filed post-judgment motions. Barr asked the court to reconsider

its decision to withhold judgment of non-infringement on the ’383 and ’052 pa-

tents. A6257-80. Alcon sought amendment of several non-dispositive factual

findings with respect to invalidity of the Castor Oil Patents, and also asked the dis-

trict court to vacate its judgment of non-infringement on those patents. A6234-53.

The district court denied both motions (in an unpublished decision), concluding

that this Court’s case law precluded entering judgment on the ’383 and ’052 pa-

tents (A78), and noting that Alcon’s arguments contained “a blatantly false state-

ment,” were rooted in “unfounded speculation,” and were “wholly unpersuasive”

(A61, A62, A65). Both parties appealed.

STATEMENT OF FACTS

In challenging the rulings with respect to its Castor Oil Patents, Alcon over-

looks or distorts facts critical to understanding this case and the district court’s de-

cisions. Accordingly, we provide below a full account of the facts—as found by

the district court and reflected in the record.

6

A. The Castor Oil Patents broadly claim a method of using PECO to enhance the chemical stability of any prostaglandin.

As pertinent here, Alcon’s Castor Oil Patents claim “methods of enhancing

the chemical stability of prostaglandin-containing compositions by adding polye-

thoxylated castor oil (‘PECO’) to the compositions.” A2 (emphasis added). Alcon

asserted one claim from each patent, but “the two (2) claims in-suit are identical

except that Claim 12 of the ’287 patent is open to any PECO, while Claim 19 of

the ’062 patent limits the PECO to one selected from the group of PEG-5 to PEG-

200 hydrogenated castor oils.” A4.

1. There are an inordinate number of “prostaglandins”—a class of com-

pounds that includes prostaglandins, their analogues and derivatives, and pharma-

ceutically acceptable esters and salts. A134-35; A5998. In general, prostaglandins

are highly potent (meaning that when used in pharmaceuticals, the prostaglandin

dosage required is very small). A5862. They also have poor solubility in water

(meaning an appropriate and effective inactive ingredient typically must be in-

cluded in any aqueous prostaglandin formulation). A5914. These generally shared

characteristics notwithstanding, prostaglandins have unique chemistry, chemical

reactivity, chemical stability, and pharmacologic properties. See A5855-64.

For example, there is wide variety across prostaglandin structures. A5844-

45. Each contains a 5-membered ring, an “alpha chain” at the top, and an “omega

chain” at the bottom, but this basic structure can be altered in several ways.

7

A5827, A5829-30, A5855-57, A5859. In addition, analogues and derivatives may

contain other functional groups such as esters, amides, or salts. A5843-44. Re-

flecting this variety, prostaglandin compounds are classified into groups using a

nomenclature system that has four components: (1) “PG”; (2) a letter; (3) a sub-

scripted number; and, optionally, (4) a subscripted Greek letter. A5855-57. For

example, travoprost—the prostaglandin used in the products at issue here—is

PGF2α, which means it is a type F2α prostaglandin analogue having the following

structure:

A11363.

2. Central to the disputed factual issues on both infringement and inva-

lidity is the problem of “chemical stability / loss v. physical stability / loss.” A22;

see also A31. Chemical instability concerns a chemical change or reaction that re-

sults in the molecule forming degradation products or degradants. A5842. Pros-

taglandins vary in chemical stability. A5867. For example, PGI2 and PGE com-

pounds are generally considered chemically unstable, meaning they are more sus-

ceptible to chemical reactions that cause them to degrade. A5868. PGF2α com-

pounds, on the other hand, are generally and relatively stable, meaning they are

8

less susceptible to chemical reactions that cause degradants or degradation prod-

ucts. A5868.

Physical instability is different. A5842. Physical instability does not result

in degradation products; it results in loss by adsorption or absorption to the packag-

ing (plastic or glass), precipitation, or some other loss that is not caused by a chem-

ical reaction. A5842. One aspect of physical stability is solubility, which is the

ability of a solute (e.g., a prostaglandin) to dissolve in a solvent (e.g., an aqueous

composition). See A15-16; A5598, A5872-73, A6047. In pharmaceuticals, solubi-

lizers, including PECO, are used to increase the solubility of poorly soluble sub-

stances. A5881-82, A5976. As a result, solubilizers may protect prostaglandins

(which are typically poorly soluble) from physical loss by precipitation, or by ad-

sorption and absorption to the container or packaging. A5914-15.

B. Barr’s product does not use PECO to enhance chemical (as op-posed to physical) stability, as claimed in the Castor Oil Patents.

As noted, Travatan Z® is a prescription eye drop indicated to treat glaucoma

and ocular hypertension. The active ingredient in Travatan Z®—and Barr’s

ANDA product—is travoprost, which is a prostaglandin.

The only question with respect to infringement of the Castor Oil Patents is

“whether or not the amount of PECO in the composition of Barr’s ANDA product

is a ‘chemically-stabilizing amount’ such that Barr’s method of manufacturing its

ANDA product enhances the chemical stability of the composition.” A5 (citation

9

omitted). Based on its claim construction ruling (which Alcon does not challenge),

the district court further explained: “If the PECO in Barr’s ANDA product merely

enhances the physical stability of Travoprost, as opposed to enhancing its chemical

stability, at least in part, then Barr does not infringe Claim 12 of the ’287 patent

and Claim 19 of the ’062 patent, i.e., the asserted castor oil patent claims.” A5-6.

This question was the subject of conflicting expert testimony, and the district court

ultimately rejected Alcon’s expert and agreed in significant part with Barr’s.

1. Alcon offered no direct evidence of infringement.

Alcon offered no direct evidence of infringement. As the district court

pointed out, “Alcon could have determined, rather definitively, whether the PECO

in Barr’s ANDA product (or the PECO in Travatan Z®, its own product, which is

extremely similar, if not functionally identical, to Barr’s ANDA product) enhances

the chemical stability of the Travoprost in the composition by running the appro-

priate chemical stability tests. For whatever reason, Alcon did not perform this

testing.” A14 (internal citations and footnote omitted).

Instead, Alcon depended entirely on the testimony of a named inventor and

one expert witness—both formulators, not chemists. The district court acknowl-

edged that “Alcon is free to prove its case as it sees fit,” and that infringement can

be proved by circumstantial evidence. Id. But the court found that Alcon had

failed to carry its burden because “the method of proof [Alcon] chose merely in-

10

vites us to speculate that the PECO in Barr’s ANDA product chemically stabilizes

the Travoprost.” Id.

2. The inventor himself (Mr. Schneider) testified that PECO enhances physical, not chemical, stability.

Alcon’s first witness was L. Wayne Schneider, a named inventor on the Cas-

tor Oil Patents. Mr. Schneider was not identified as an expert witness, did not

submit an expert report, and Alcon stipulated that his testimony was not offered to

prove infringement. A22; A5431-32. The district court also found Schneider’s

“unpersuasive” testimony ineffective to show infringement in any event. A22.

First, Mr. Schneider outright admitted that PECO enhances the physical sta-

bility of prostaglandins, not chemical stability:

Q: That data [in the patents] also might indicate that it’s enhancing the physical stability of the composition in the prostaglandin, is that correct? A: Might. Yes, it might. I can’t deny that. But my feeling is that it

indicates the chemical stability because the polyethoxylated castor oil affords solubility and prevents loss of physical ab-sorption or physical – the polyethoxylated castor oil prevents the physical loss of the prostaglandin from the solution.

A5538-39 (quoted in critical part at A23) (emphasis added). Explaining why he

believed PECO enhances “chemical stability,” Mr. Schneider clearly described

physical stability, referencing PECO’s ability to enhance “solubility” and prevent

the “physical loss” of prostaglandin to packaging.

11

Second, Mr. Schneider testified about data from high pressure liquid chro-

matography (“HPLC”) tests that were designed to measure physical loss, i.e., the

total amount of prostaglandin lost from samples in various concentrations. A5562,

A5430. Notably, none of this data appears in the Castor Oil Patents (A24)—and

for good reason: Mr. Schneider did not test for, identify, or measure any degrada-

tion products, nor did he determine whether PECO inhibited the development of

prostaglandin degradation products. See A5561-64.

3. Based on physical stability data from a solubility study, Al-con’s under-qualified expert (Dr. Levinson) offered an opin-ion that the district court found flawed.

Alcon’s only other witness was Dr. Saul Levinson, an expert in pharmaceu-

tical formulation. Dr. Levinson was not qualified to opine that Barr’s ANDA

product infringes the Castor Oil Patents. First, he submitted a CV replete with er-

rors relating to patents on which he claimed to have worked. A60. Second, Dr.

Levinson’s experience includes limited involvement with only two prostaglandin

compositions. A5648-49. He has never formulated a prostaglandin composition or

supervised anyone else in creating an aqueous prostaglandin formulation. A5649.

Third, Dr. Levinson repeatedly testified that he is not a prostaglandin chemist.

A60. He was unable to discuss the chemical structure of any prostaglandin, chem-

ical changes to prostaglandins, prostaglandin degradation pathways, prostaglandin

degradation products, or the various chemical and physical properties of any of the

12

many prostaglandins claimed by the Castor Oil Patents. A5652, A5654-57,

A5659-60, A5665-66, A5669-70.

By contrast, Barr’s expert witness—Dr. John Kent—is an expert in both

prostaglandin chemistry and pharmaceutical formulation. A5832-38. Dr. Kent has

nearly 40 years’ experience in the pharmaceutical industry (A5815-24, A9058-61)

during which he has been involved in successfully developing several prostaglan-

din formulations, including two involving PGF2α—the very same type of prosta-

glandin as travoprost. See A5817-19, A5821-22, A5825, A9058-61. At Allergan,

Dr. Kent spent more than a decade developing Lumigan®, which is, just like Tra-

vatan Z®, a prostaglandin ophthalmic composition for treating glaucoma and ocu-

lar hypertension. A5829-30. Dr. Kent described numerous aspects of prostaglan-

din chemistry that are critical to chemical stability. A5852. He detailed degrada-

tion pathways and degradation products for prostaglandins—something Dr. Levin-

son could not do. A5671-72, A5873-77. Dr. Kent also described the different

structural, chemical, and physical properties for the many prostaglandins claimed

by the Castor Oil Patents—again, something Dr. Levinson could not do. A5655-

56, A5830-31, A5855-60, A5863-64, A5867-69.

Dr. Levinson did not test Barr’s ANDA product to determine whether PECO

was used to enhance the chemical stability of the travoprost contained therein. Nor

did he test Travatan Z® itself, or any product similar to either Travatan Z® or

13

Barr’s product. A13-15. In fact, Dr. Levinson did not undertake any testing or

point to any study tending to show that PECO enhances the chemical stability of

any prostaglandin, travoprost or otherwise. A5891-92, A5896-98, A5903-04.

Instead, Dr. Levinson based his opinion—that Barr’s ANDA product in-

fringed the Castor Oil Patents—on a single table, within a single Alcon develop-

ment record for travoprost. A15. He relied on nothing else. Id. The document is

a report entitled “Solubility of Travoprost in HCO-40 and Soaking Study of Pro-

pylene Material in Prototype Formulations” (the “Solubility Study”), and the table

is Table 7. Id.; A11238-53. That Study and Table, however, do not speak to

chemical stability. A20.

Significantly, Mr. Schneider, Dr. Levinson, and Dr. Kent all testified that

solubility and packaging compatibility relate to physical stability—not chemical

stability. A5537-38, A5676, A5917. And, as one might guess from its title, the

Solubility Study was designed to measure solubility and packaging compatibility—

not chemical stability. A15-16; A11238-53; A5905-06, A5917-18. The study’s ob-

jectives were clearly defined:

“Determine the equilibrium solubility of Travoprost in a Tris/borate/mannitol buffer system containing different concentrations of HCO-40 surfactant at pH 6.0 under ambient room temperature condi-tions.”

“A soaking study to evaluate the compatibility of 0.005% travoprost for-mulations with polypropylene packaging materials at 55° C in the pres-ence of varying percent levels of HCO-40 and BAK.”

14

A11240 (emphasis added). The Study’s conclusions reflect these goals, that is,

they relate to solubility and packaging compatibility, not chemical instability:

“Travoprost is relatively more soluble in a Tris / borate / mannitol buffer system at pH 6.1 than in water. An increase in the concentration of HCO-40 from 0.0% to 0.5% at pH 6.0 + 0.1 enhances the solubility of Travoprost from 0.02% to 0.1%, respectively.”

“Samples prepared with 0.5% HCO-40 and 0.1% BAC are compatible with polypropylene packaging material and yields good initial chromato-graphic recovery based on label claim. Under accelerated thermal study, samples stored in amber color glass amp[o]ules were not compatible.”

A11246 (emphasis added). As the district court noted, nowhere in the Solubility

Study do the authors conclude—or even mention—chemical instability, prosta-

glandin degradation products, or the proposition that PECO enhances the chemical

stability of prostaglandins. A15-16, A5675, A5906, A5921-22.

The Study’s Table 7, on which Dr. Levinson relied, shows the mass physical

amount of travoprost lost from two samples after eight weeks at an elevated tem-

perature (55°C). A11250. The first sample, containing no HCO-40 (a kind of

PECO), lost 12% of its travoprost. The second sample, containing 0.50% HCO-40,

had an 8% travoprost loss. Id.; A15. According to Dr. Levinson (but not the Solu-

bility Study researchers), the travoprost loss from both samples was the result of

chemical degradation, not physical loss, because physical loss in glass containers is

caused by adsorption (not absorption), which occurs rapidly, “certainly within

hours, most certainly within days.” A18-19; A5621. And, Dr. Levinson con-

15

cluded, because the second sample retained more travoprost, the HCO-40 must

have improved the prostaglandin’s chemical stability. A15, A27; A5620-22.

There were numerous problems with Dr. Levinson’s attempt to re-deploy the

Table 7 data to an issue the researchers did not examine and a purpose for which it

was never intended. First, the travoprost showed good stability, even in composi-

tions that included no PECO. A16. Second, “the data in Table 7 is subject to at

least some experimental error or uncertainty” as indicated by “anomalous data”

showing that “the [t]ravoprost concentration actually increases slightly over some

period of time in some of the samples.” A15-17 (internal citation omitted). Third,

travoprost stability is impacted by other ingredients in a composition (besides

PECO), such that even a “miniscule change” in concentration of the preservative

benzalkonium chloride (“BAC”) (0.01% versus 0.02%) resulted in a significant

change in travoprost loss, which “is particularly instructive in evaluating the proba-

tive value of the Table 7 data because all of the samples in Table 7 contain BAC,

while Barr’s ANDA product does not.” A17. Fourth, Dr. Levinson testified that

“data plotted linearly on a log scale … indicates a first-order chemical reaction,”

but “the data in Table 7 … is decidedly un-linear.” A18.

4. Barr’s expert (Dr. Kent) disagreed with Dr. Levinson.

In addition to being intrinsically flawed, Dr. Levinson’s opinion met with

square disagreement from the more qualified Dr. Kent. Dr. Kent testified that, like

16

the rest of the Solubility Study of which it is a part, Table 7 has nothing to do with

chemical stability, does not measure for or identify any degradation products, and

cannot establish that Barr’s ANDA product infringes the Castor Oil Patents. A15-

16; A5677, A5911-12, A5915-18, A5921. Instead, the data reported in Table 7

comes from a “soaking study”—and such studies relate to solubility, not chemical

stability. A5677, A5911, A5917-18. Dr. Kent also explained that Table 7 meas-

ures total physical loss of travoprost over time, and—consistent with the goal and

design of this particular study—there is no data on any of travoprost’s major de-

gradation products, which had been identified by Alcon at least three years earlier.

A5895-98, A5921; A11352.

In any event, Dr. Kent testified, even if Table 7 contained chemical stability

data (which it does not), the formulations tested in the Solubility Study differ in

material respects from Barr’s ANDA product. A17-18, A20-21; A5982-84. For

example, the Study formulations contain BAC. Barr’s ANDA product does not.

Id. The formulations also have a different buffer than Barr’s ANDA product. Id.

Both of these factors can impact the chemical stability of a prostaglandin in an

ophthalmic formulation (A17-18, A21)—which means, even if PECO was chemi-

cally stabilizing in the study formulation, that still would reveal nothing meaning-

ful about Barr’s product. A21. Moreover, the formulations in the Solubility Study

have a different pH than Barr’s ANDA product. A20-21. Dr. Kent testified that a

17

slight difference in pH can have a serious impact on chemical stability. Indeed, pH

is the single most important factor in controlling chemical stability (not PECO),

and travoprost is a perfectly stable compound at the right pH. A5985; A11222.

For these reasons, Dr. Kent opined, Table 7 cannot support any conclusions

about whether PECO enhances the chemical stability of travoprost in Barr’s

ANDA product. A17-18, A20-21; A5910-12.

5. Alcon’s own documents contradicted Dr. Levinson.

Further, Dr. Levinson’s analysis was belied by U.S. Patent No. 6,235,781

(“the ’781 patent”) (A8751-62) and other Alcon documents.

Alcon’s ’781 patent discloses that PECO—including the PECO used in the

Solubility Study, HCO-40—inhibits the adsorption of prostaglandin onto container

walls over time. A19-20. The ’781 patent says nothing about chemical stability; it

deals with physical prostaglandin loss to different packaging materials specifically

caused by adsorption. A19-20; see also A5682-83. Dr. Levinson testified that Ta-

ble 7 reflects chemical loss because of the time period involved and the fact that

the study used glass containers. A5620-22. Undermining that analysis, however,

the studies depicted in Figures 1-4 of the ’781 patent show that the inventors

waited four weeks to measure prostaglandin loss due to adsorption, even in glass

containers, and that adsorption can cause a substantial amount of physical loss to

occur, even in glass containers. A19; A8752-55.

18

In addition, a number of internal Alcon documents show that PECO en-

hances solubility and physical stability, and does not enhance chemical stability.

In particular, in several development studies involving travoprost, Alcon scientists

measured travoprost loss in various packaging materials over time—like the loss

observed in Table 7 of the Solubility Study. They found no degradation products.

A5969-73. In other studies with similar results, Alcon scientists observed: “travo-

prost loss is known to be related to the packaging interaction. Chemistry data

shows no concurrent increase in degradation products which support this conclu-

sion.” A27 (quoting A8735) (emphasis added). Thus, Alcon scientists concluded

PECO solubilizes travoprost but does not inhibit growth of degradation products.

See A5904, A5969-73. Based on these materials, Dr. Kent testified and the district

court agreed, Alcon realized years ago that PECO solubilizes prostaglandins, and

has no impact on chemical stability. A26-28; A5966-79.

In sum, the district court found that Alcon failed to establish infringement

for two sets of reasons: First, the Table 7 results “could be attributed to a number

of factors other than PECO enhancing the chemical stability of the Travoprost,

e.g., experimental error or uncertainty, adsorption, precipitation, or other physical

loss.” A20. Second, even if Alcon were correct about Table 7, “the tested formu-

lations differ significantly from Barr’s ANDA product”—and “[s]ince variables

such as pH, buffer, buffer concentration, preservatives, chelating agents, and other

19

excipients can affect the chemical stability of prostaglandins in ophthalmic formu-

lations … the compositional differences between the Solubility Study formulations

and Barr’s ANDA product preclude … relying on the Solubility Study data to draw

any reliable inferences with respect to the stability of Barr’s ANDA product.”

A20-21.

C. The Castor Oil Patents are not enabled and lack adequate written description.

The district court also held that the Castor Oil Patents are not enabled and

lack adequate written description. A28-37.

1. To start, the asserted claims are incredibly broad. They claim that

nearly any PECO in the right amount can enhance the chemical stability of an in-

ordinate number of prostaglandin compounds, including every compound in the

PGE1, PGE2, PGE3, PGF1a, PGF2a, PGF3a, PGD2 and PGI2 classes and their

analogues and derivatives, including their pharmaceutically acceptable esters and

salts and their analogues and derivatives. A5843-44, A5999-6000.

The specification, however, provides little guidance to one skilled in the art

as to how to practice the full scope of this incredibly broad claimed invention.

A30-31. First, the specification does not contain any disclosure or discussion on

how the claimed invention works—how adding PECO inhibits any prostaglandin

degradation pathway or the development of any prostaglandin degradation prod-

ucts. A6003. Second, the ranges provided for the prostaglandin, PECO, and other

20

ingredients are so expansive as to be meaningless. A6003-04. For example, the

specification states:

“The prostaglandin compositions of the present invention contain one or more polyethoxylated castor oils in an amount effective to enhance the stability of the prostaglandin … However, other factors may limit the amount of polyethoxylated castor oil to be utilized in the compositions of the present invention. For example, too much polyethoxylated castor oil should not be used in order to avoid adversely affecting the prostaglan-din’s pharmacologic activity.” A6369, 6379 (emphasis added).

“In general compositions of the present invention will include one or more polyethoxylated castor oils in an amount between 0.02 and about 20.0 percent by weight (wt %)”—a 1,000x range—“and one or more prostaglandins in an amount between 0.00001 and about 0.2 wt %.”—a 20,000x range. A6369, A6379 (emphasis added).

“[T]he compositions of the present invention may further comprise vari-ous formulatory ingredients, such as antimicrobial preservatives, tonicity agents, and buffers.” A6369, A6380 (emphasis added).

“[C]ompositions may be formulated in various dosage forms suitable for topical ophthalmic delivery, including solutions, suspensions, emulsions, gels and erodible solid ocular inserts. The compositions are preferably aqueous, [and] have a pH between 3.5 and 8.0.” A6370, A6380 (empha-sis added).

Dr. Kent explained that the inadequate disclosure on these critical variables

(all of which influence chemical stability) make it impossible for one of ordinary

skill to determine how to enhance the chemical stability of a prostaglandin with a

chemically stabilizing amount of PECO without undue experimentation. A5991-

6010. More precisely, Dr. Kent testified to the number of “variables” involved in

practicing the claimed invention—including choice and amount of PECO, choice

21

and amount of prostaglandin, other excipients, preservation agents, choice and

concentration of buffer, pH, and dose—and explained that “any time you have a lot

[of variables] on top of one another, the experimentation gets out of control quick-

ly.” A6008-10. The district court agreed. A34.

2. Despite the claims’ breadth, the Castor Oil Patents do not include any

chemical stability data for any prostaglandin. Instead, they contain physical sta-

bility data for one compound, Compound 2, that is not even within the scope of the

claims. A5994-96. That data (in Figures 1-3 of the Patents) shows the total

amount of prostaglandin lost, without measuring, identifying, or quantifying any

degradation products or degradants of Compound 2. A5549-50, A5995. In fact,

there is no data in the patents (in Figures 1-3 or otherwise) that measures, identifies

or quantifies any prostaglandin degradation products, nor is there any data showing

PECO inhibiting the development of degradation products. A5549-53, A5996.

3. Even if the data in Figures 1-3 concerned chemical stability, that

would not teach a person of ordinary skill anything about the chemical stability of

the numerous prostaglandins, analogues and derivatives within the scope of the

claims. A5843-44, A5999-6000. Again, Compound 2 is not even within the class

of compounds claimed in the invention. A5998-99. It is structurally different—

and structural differences have a significant impact on chemical stability. A6000-

02. Additionally, Dr. Kent testified, it is not possible to predict the chemical sta-

22

bility of a given prostaglandin across the incredible range of prostaglandins

claimed by the Castor Oil Patents. A5996.

Based on this evidence, the district court found that persons of ordinary skill

in the art would not recognize that Mr. Schneider and his colleagues actually pos-

sessed the claimed invention—and even if they believed the invention existed,

such persons would not be able to practice the full scope of the claims without un-

due experimentation. A28-34. Accordingly, the district court held the asserted

claims of the Castor Oil Patents invalid for failure to meet the written description

requirement and for lack of enablement. A54.

SUMMARY OF ARGUMENT

Correctly applying straightforward principles of well-established law to

amply supported findings of fact, the district court ruled that the asserted claims of

the Castor Oil Patents are invalid and that, in any event, Barr’s ANDA product

does not infringe. Alcon has failed to establish any error in the district court’s

analysis—much less clear factual error. Indeed, all three issues (infringement,

enablement, and written description) were subject to a battle of experts, in which

the parties’ experts provide squarely conflicting assessments—and the district

court found Barr’s expert and arguments to be the more compelling.

With respect to non-infringement, the district court concluded: (1) Alcon

had no direct evidence that Barr’s ANDA product infringed the Castor Oil Patents.

23

(2) Table 7, the heart of Alcon’s case, does not contain chemical stability data. (3)

Even if it did, the Table 7 compositions differ from Barr’s ANDA product in ma-

terial respects. And (4) Alcon’s internal documents, studies, reports, and the ’781

patent demonstrate that PECO improves physical, not chemical, stability. A11-28.

Arguing in the teeth of the clear-error standard of review—and without re-

gard for the fact that the burden on this claim was Alcon’s, not Barr’s—Alcon in-

vites the Court to vacate the district court’s non-infringement ruling based on noth-

ing more than the fact that testimony from its own expert was not credited by the

district court. But Barr’s expert, Dr. Kent, offered a competing and more persua-

sive assessment. The district court did not clearly err in agreeing with Barr’s ex-

pert rather than Alcon’s—particularly given that Barr’s expert was better qualified,

and the inventor himself (Mr. Schneider) admitted that PECO enhances physical

(not chemical) stability of prostaglandins. A23.

With respect to invalidity, the district court found: (1) The claims cover an

inordinate number of prostaglandin compounds and nearly any PECO, to be used

in essentially any combination and concentration, and in compositions that contain

other ingredients that may impact chemical stability. (2) The patents contain only

three working examples, and no chemical stability data. (3) The only stability data

included in the patent goes to physical (not chemical) stability of a single com-

pound that is not even within the scope of the claims. (4) The patents do not ex-

24

plain how the invention works. (5) The art was highly unpredictable, and there are

several considerations that dictate the chemical stability of any given prostaglan-

din. (6) The prior art taught little about chemically stabilizing prostaglandins. (7)

The patents give little guidance on how to select a prostaglandin, PECO, or the

other components that must be included in the claimed invention. A28-37.

Given these findings, the district court correctly held the asserted claims

invalid for non-enablement. A54. Alcon concedes that the district court “articu-

lated” the proper standard, but complains that the court conducted a “freestanding

inquiry” into whether the claims are “too broad” relative to the specification. Al-

con Br. at 24-25. Not so. Alcon distorts the legal standard and the district court’s

analysis, both. Nor is Alcon correct in its improbable assertion that “Barr did not

show—indeed, it introduced no evidence—that the person of ordinary skill in the

art would have been unable to practice the asserted claims without resorting to un-

due experimentation.” Id. at 30. In fact, Barr’s expert squarely addressed this

point, testifying to the numerous variables that impact chemical stability, and ex-

plaining that, when “you have a lot of variables on top of one another, the experi-

mentation gets out of control quickly.” A6009.

Alcon likewise fails to establish clear error with respect to the district court’s

finding that the Castor Oil Patents fail the written description requirement. Alcon

insists that the patent specifications “describe the precise invention covered by the

25

claims.” Alcon Br. at 43. But saying does not make it so. There is nothing “pre-

cise” about the invention here, which encompasses countless prostaglandin com-

pounds, a 20,000-fold range of prostaglandin concentrations, a wide range of PE-

COs, and a 1,000-fold range of PECO concentrations. See A5797, A5863.

Nor is Alcon fairly characterizing the record with its contention that Barr

failed to adduce “any evidence that the person of ordinary skill would fail to rec-

ognize that Mr. Schneider possessed the invention he claimed.” Alcon Br. at 45.

This notion is incredible on its face. Indeed, not only did Barr’s expert directly ad-

dress this point (A5848), but the district court correctly recognized that the patent

claims fail the written description requirement for essentially the same reasons they

fail the enablement requirement—including that the specifications offer physical

stability data from one compound, outside the scope of the claims, to support the

novel proposition that PECO enhances the chemical stability of all prostaglandins.

See A34-37. What is more, the inventor himself explained his supposed invention

in terms of physical stability. A23. Even today, it does not appear that Mr.

Schneider has possession of the claimed invention.

Though central to trial and the decision below, in its validity arguments (Al-

con Br. at 38-45) Alcon does not address the issue of physical versus chemical sta-

bility. That failure, together with a record that amply supports the district court’s

findings of fact, dooms Alcon’s quest for reversal.

26

Finally, Barr is entitled to a judgment of non-infringement on the ’383 and

’052 patents because Alcon brought infringement claims in its complaint, never

withdrew those claims, and failed to introduce any evidence of infringement at tri-

al. The district court was concerned that Alcon was unfairly surprised when Barr

moved at trial for judgment as a matter of law. But the only unfairness (or sur-

prise) would be to Barr if this judgment is withheld—and to the public interest if

Barr is precluded from marketing its product when Alcon no longer may assert any

of the six patents raised in this litigation. Alcon promised to have its ’383 and ’052

claims dismissed with prejudice. After failing to follow through, Alcon should

have been estopped from opposing Barr’s motion for judgment. And in any event,

a party may not informally abandon pleaded claims when Fed. R. Civ. P. 15 and 41

permit amendment or voluntary dismissal only upon motion.

ARGUMENT

I. Barr’s ANDA Product Does Not Infringe the Castor Oil Patents

“After a full bench trial, infringement is a question of fact” that is reviewed

for clear error. Amgen Inc. v. Hoechst Marion Roussel, Inc., 314 F.3d 1313, 1339

(Fed. Cir. 2003). Here, the district court did not clearly err in finding that Barr’s

ANDA product does not infringe the asserted claims of the Castor Oil Patents; in-

deed, a contrary conclusion would have been utterly unsustainable on the record.

Alcon contends that its own expert (Dr. Levinson) provided an opinion that would

27

have supported a finding of infringement, and that the district improperly rejected

that testimony because Dr. Levinson is a formulator, not a chemist. Alcon Br. at

26-27. But Alcon gives its own evidence undue weight, and its dismissal of Barr’s

competing evidence both misstates that evidence and fundamentally misunder-

stands the burden of proof on a claim of infringement—which was Alcon’s to bear,

not Barr’s. Alcon’s credibility complaint also fails.

A. The district court accurately assessed the weakness of Alcon’s evidence, and agreed with Barr’s expert.

Alcon had no evidence directly related to whether Barr’s ANDA product in-

fringed the Castor Oil Patents, and no evidence directly related to whether PECO

enhances the chemical stability of any prostaglandin. It would have been easy to

do so, but Alcon did not test Barr’s ANDA product, Travatan Z®, or any similar

formulation to determine whether PECO enhances the chemical stability of travo-

prost. A14. For that matter, Alcon did not present any study proving that any

PECO enhances the chemical stability of any prostaglandin. Id.

Rather than investigate and prove infringement directly, Alcon offered ex-

pert testimony from Dr. Levinson, who rested his opinion on a single table (Table

7) in a single study (the Solubility Study) that was designed to test a fundamentally

different pair of issues (solubility and packaging compatibility) and provides data

on physical (not chemical) stability. Thus, “instead of having a relatively clear an-

swer on the infringement question, [the district court was] left drawing inferences

28

and making educated guesses.” A14. Though Alcon is free to prove its case as it

sees fit, “the method of proof it chose merely invites … speculat[ion] that the

PECO in Barr’s ANDA product chemically stabilizes the Travoprost.” Id.

On appeal, Alcon argues that the district court clearly erred by not agreeing

with Dr. Levinson that Table 7 demonstrates enhanced chemical stability insofar as

it shows more travoprost loss in the no-PECO composition as compared to the

0.5% PECO composition. Alcon Br. at 45-50. Dr. Levinson asserted that this loss

resulted from chemical instability because the travoprost was lost over time (while

adsorption is sudden and rapid) and absorption does not occur in glass (the tests

were carried out in glass ampoules). A5620-22. And Dr. Levinson extrapolated

his interpretation of Table 7 to Barr’s product because, he opined, the compositions

contain the same type of prostaglandin (i.e., travoprost) and the same type and

amount of PECO. A5609, A5620-29.

Dr. Levinson, however, was wrong—both in his analysis of Table 7, and in

his attempt to extrapolate to Barr’s product. The data in Table 7 simply does not

speak to whether Barr’s ANDA product infringes the Castor Oil Patents, and the

district court identified several reasons why.

To start, even if Table 7 did contain chemical stability data, it still would not

show that PECO enhances the chemical stability of Barr’s ANDA product. Barr’s

ANDA product is not the same as the formulations in Table 7. A20-21. Those

29

formulations contain BAC (a preservative). Barr’s ANDA product does not. A21.

Also, the buffer in Barr’s ANDA product is different than the buffer in the formu-

lations in Table 7. Id. Further, the formulations in Table 7 have a different pH

than Barr’s ANDA product. Id. Dr. Kent testified that pH is the most important

contributor to chemical stability and any change will significantly impact the

chemical stability of a prostaglandin in an ophthalmic formulation. A5985. As the

district court found—and Alcon admits—BAC, the buffer, and pH can all have a

substantial impact on chemical stability. A20-21; Alcon Br. at 39; see also A5539-

40, A5985.

As a result, the district court concluded that the Table 7 data has no bearing

on whether Barr’s ANDA product infringes the Castor Oil Patents because the

formulations in Table 7 are meaningfully different from Barr’s ANDA product.

A20-21. Other than a generalized (and misplaced) complaint that the district court

unfairly questioned Dr. Levinson’s credibility (Alcon Br. at 50), Alcon does not

address this point, which alone provides sufficient grounds to affirm even if Alcon

were correct about Table 7.

And Alcon is not correct about Table 7.

First, the Solubility Study was designed to test travoprost solubility. A15-

16. As Mr. Schneider, Dr. Levinson, and Dr. Kent all testified, solubility relates to

physical stability, not chemical stability. A5537-39, A5598, A5872-73, A6047.

30

Table 7 reports results from a soaking study—again, solubility, not chemical stabil-

ity. A15-16; A5905-06. The researchers did not measure or report chemical de-

gradation or travoprost degradation products, much less show that PECO enhances

the chemical stability of travoprost in Barr’s ANDA product, which is materially

different from any of the study formulations. A5917-18.

Second, the district court found that the Table 7 data might be attributable to

experimental error. A16-17. As the district court observed, in some samples, the

travoprost concentration actually increased over time. “No one has suggested to us

that the PECO actually generates Travoprost over time, so experimental error or

uncertainty most likely explains the anomalous data.” A17. Alcon does not ad-

dress this point.

Third, the district court noted several considerations that contravened Dr.

Levinson’s testimony that adsorption occurs rapidly and not in glass. For example,

Dr. Levinson admitted that it was possible for “some adsorption” to occur in glass

(A5621-22), and did not provide any authority or documentary evidence to support

his assertion that adsorption would occur rapidly. A19. Moreover, the district

court found that one of Alcon’s own patents, the ’781 patent, directly contradicted

Dr. Levinson’s conclusions. A19-20. As the district court described, the ’781 pa-

tent “says nothing about chemical stability, but rather deals with prostaglandin loss

to different packaging materials.” A19. And, the court further observed, Figures

31

1-4 of the ’781 patent show that its “inventors waited four (4) weeks before mea-

suring prostaglandin loss,” and that “a substantial amount of prostaglandin physical

loss may occur, even in glass containers.” A19.

Fourth, Barr’s expert, Dr. Kent, made clear that Table 7 did not speak to

chemical stability and could not be used to show that PECO enhanced the chemical

stability of Barr’s ANDA product. A5677, A5911-12, A5915-18, A5921. Surpri-

singly, Alcon complains that Dr. Kent “never testified that there was no enhance-

ment of chemical stability shown in the Table 7 experiment.” Alcon Br. at 47.

That is incorrect. Regarding Table 7, Dr. Kent specifically testified:

Q: And based on your expert opinion and evaluation of this table, does this table have any information with regard to chemical stability of travoprost?

A: No, it does not. *** Q: And is any of the missing Travoprost shown in [column containing

the 0.5% PECO composition data] due to chemical degradation? A: Not that – that – not that they’ve demonstrated in the – in the analysis,

not at all.

A5912; A1916-17.

Fifth, Dr. Kent also explained, and the district court agreed, that in addition

to the ’781 patent, two Alcon internal documents showed that PECO worked by

physically stabilizing travoprost, not by chemically stabilizing it. A26. In one

1996 stability study, Alcon scientists reported that travoprost loss “is now known

32

to be related to the packaging interaction. Chemistry data shows no concurrent in-

crease in degradation products which supports this conclusion.” A27 (emphasis

added). In another 1997 stability study, Alcon scientists measured a 7% loss in

travoprost after eight weeks—just like they did in the Solubility Study—and found

no travoprost degradation products. A26. “In other words, 7% of the travoprost

went missing, but it was not due to chemical degradation.” A26. These documents

conclusively refute Dr. Levinson’s testimony (that Table 7 contains “immutable

data” that chemical degradation is going on), and demonstrate the PECO impacts

physical stability, not chemical stability.

In fact, Alcon does not contest Dr. Kent’s observation that Alcon’s internal

reports demonstrated that PECO is “involved in the physical stabilization of travo-

prost,” and from them, it “is not evident … that [PECO was] involved in chemical

stabilization.” A5977. Rather, Alcon claims that Dr. Kent did not establish that

enhancing physical stability precluded enhancing chemical stability, and therefore,

his testimony “did not establish that the [PECO] added to Barr’s product does not

chemically stabilize the prostaglandin.” Alcon Br. at 47 (emphasis added).

Alcon’s double-negative is telling. This argument is fundamentally at odds

with both the district court’s factual conclusions and the burden of proof on claims

of infringement. Barr had no responsibility to prove non-infringement—i.e., that

enhancing physical stability precludes enhancing chemical stability. Rather, Alcon

33

had the burden to prove infringement—i.e., that adding PECO to Barr’s product

chemically stabilized the prostaglandin. Alcon failed to do so.

B. The district court’s credibility assessment was not clear error.

Finally, Alcon complains that the district court’s decision not to accept Dr.

Levinson’s testimony “was based, at least in significant part, on the court’s mistak-

en belief that Dr. Levinson needed to have the skills of a prostaglandin chemist in

order to present credible testimony regarding the chemical stability of the prosta-

glandin in Barr’s product.” Alcon Br. at 48-49.

Alcon misstates the district court’s holding. The court noted that “Dr. Le-

vinson admitted numerous times that he is not an expert in prostaglandin chemistry

or chemical changes in prostaglandins. To some extent, this undermines our faith

in his ability to reliably assess whether or not the PECO in Barr’s ANDA product

chemically stabilizes Travoprost.” A21-22. Even so, the district court made clear

that all of its factual findings and legal conclusions “would remain the same if [it]

were to adopt the [person of ordinary skill in the art] definition espoused by Al-

con.” A4-5. Thus, even if the district court erred in expressing a concern with the

“reliab[ility]” of Dr. Levinson’s assessment, that error was harmless and no basis

for vacatur. See 28 U.S.C. §2111.

But there was no error. The district court properly rejected Dr. Levinson’s

testimony because it was not based in relevant fact and turned entirely on evidence

34

(the data in Table 7 of the Solubility Study) that had no bearing on whether Barr’s

ANDA product infringed the Castor Oil Patents.

In marked contrast to Dr. Levinson, Dr. Kent, is an expert in prostaglandin

chemistry and pharmaceutical formulations. A5832-38. He has 40 years’ expe-

rience in the pharmaceutical industry (A5815-24, A9058-61), and spent more than

a decade developing Lumigan®, a product highly similar to those at issue here, i.e.,

a prostaglandin ophthalmic composition for treating glaucoma and ocular hyper-

tension (A5829-30). Dr. Kent described numerous aspects of prostaglandin chemi-

stry, prostaglandin structures, degradation pathways, and degradation products that

Dr. Levinson could not. A5671-72, A5873-77. The district court reasonably and

appropriately determined that Dr. Kent’s opinion—that Barr’s ANDA product does

not infringe the Castor Oil Patents—was more credible than Dr. Levinson’s.

Alcon does not and cannot argue that the expert opinions were consistent.

They were diametrically opposed. For example, Alcon complains that Dr. Levin-

son testified that the differences between the Table 7 formulations and Barr’s

ANDA product “‘are not likely to impact’ enhancement of chemical stability.” Al-

con Br. at 50 (quoting A5617). But Dr. Kent testified to precisely the opposite.

See A17-18, A20-21; A5982-84. Even if the district court was not strictly required

to choose the more convincing opinion from the more qualified Dr. Kent, it was

not clear error for the court to do so.

35

* * * *

In sum, the district court made two independently dispositive findings:

First, the “extremely small difference” in results for the two Table 7 samples

“could be attributed to a number of factors other than PECO enhancing the chemi-

cal stability of the Travoprost, e.g., experimental error or uncertainty, adsorption,

precipitation, or other physical loss.” A20. Second, and in any event, “the tested

formulations differ significantly from Barr’s ANDA product”—and “[s]ince va-

riables such as pH, buffer, buffer concentration, preservatives, chelating agents,

and other excipients can affect the chemical stability of prostaglandins in ophthal-

mic formulations, the compositional differences between the Solubility Study for-

mulations and Barr’s ANDA product preclude … relying on the Solubility Study

data to draw any reliable inferences with respect to the stability of Barr’s ANDA

product.” A20-21. Either will do to affirm, but for both of these reasons, Alcon

failed to prove by a preponderance of the evidence that Barr’s ANDA product in-

fringes the Castor Oil Patents.

II. The District Court Correctly Held The Castor Oil Patents Invalid For Lack of Enablement.

A. Guided by Wands, the district court made unimpeachable and well-supported findings of fact establishing that the Castor Oil Patents are not enabled.

The enablement requirement “prevents both inadequate disclosure of an in-

vention and overbroad claiming that might otherwise attempt to cover more than

36

was actually invented.” MagSil Corp. v. Hitachi Global Storage Techs., Inc., 687

F.3d 1377, 1381 (Fed. Cir. 2012). “Thus, a patentee chooses broad claim language

at the peril of losing any claim that cannot be enabled across its full scope of cov-

erage.” Id. “Whether a claim satisfies the enablement requirement of 35 U.S.C. §

112, ¶ 1 is a question of law, reviewed de novo, based on underlying facts, which

are reviewed for clear error.” Sitrick v. Dreamworks, LLC, 516 F.3d 993, 999

(Fed. Cir. 2008).

Alcon attempts to reduce the analysis to whether one skilled in the art can

practice a single embodiment covered by the claimed invention without undue trial

and error in a laboratory setting. Alcon Br. at 30-34. In fact, “[t]he full scope of

the claimed invention must be enabled …. The scope of the claims must be less

than or equal to the scope of the enablement.” Sitrick, 516 F.3d at 999 (emphasis

added); accord Streck, Inc. v. Res. & Diagnostic Sys., Inc., 665 F.3d 1269, 1288

(Fed. Cir. 2012) (“To be enabling, a patent’s specification must teach those skilled

in the art how to make and use the full scope of the claimed invention without un-

due experimentation.”).

Furthermore, evaluating enablement “is not a single, simple factual determi-

nation, but rather is a conclusion reached by weighing many factual considera-

tions.” In re Wands, 858 F.2d 731, 737 (Fed. Cir. 1988). Accordingly, in Wands,

the Court set forth eight factual considerations to guide the enablement analysis:

37

(1) the quantity of experimentation necessary, (2) the amount of direc-tion or guidance presented, (3) the presence or absence of working ex-amples, (4) the nature of the invention, (5) the state of the prior art, (6) the relative skill of those in the art, (7) the predictability or unpredic-tability of the art, and (8) the breadth of the claims.

Id. The district court here made factual findings on seven of these factors, and cor-

rectly concluded that “one skilled in the art would have to perform undue experi-

mentation to practice the full scope of the asserted castor oil patent claims.” A31,

34; see Alza Corp. v. Andrx Pharms., LLC, 603 F.3d 935 (Fed. Cir. 2010) (affirm-

ing finding of non-enablement when “the district court found that seven of the

eight Wands factors weighed in favor of a finding that undue experimentation

would be required to enable the full scope of the claims”); Amgen, Inc. v. Chugai

Pharma. Co., 927 F.2d 1200, 1213 (Fed. Cir. 1991) (“it is not necessary that a

court review all the Wands factors to find a disclosure enabling”).

1. The Castor Oil Patents provide very little guidance to one skilled in the art.

As the district court recognized, “the castor oil patents provide very little

guidance to one skilled in the art.” A31. Dr. Kent testified, and the district court

agreed, that these patents claim a method of improving chemical stability for “a ra-

ther inordinately large number of potential [prostaglandin and PECO] com-

pounds.” A33; A5844-45. Despite that breadth, the patents contain just three

working examples, with stability data on “a single prostaglandin compound

(‘Compound No. 2’),” and “only two (2) PECO compounds.” A31; see also

38

A6364-66, A6374-76.

Moreover—and critically—the Castor Oil Patents do not contain any chemi-

cal stability data for even a single prostaglandin. See A6002. The only stability

data in the Castor Oil Patents is for Compound 2. That data appears in Figures 1-3

in the Castor Oil Patents, which depict the total amount of prostaglandin lost over

time. See A6054. There is no dispute about this. But that data, the district court

found, goes to physical stability, not chemical. A25-26. That is the only reasona-

ble explanation for the fact that the data on Compound 2 concerns only total pros-

taglandin lost; there is no data on Compound 2’s degradation products. A5549-50,

A5995. Indeed, there is no data anywhere in the patents on prostaglandin degrada-

tion products or inhibiting the development of prostaglandin degradation products

by adding PECO to the composition. A5549-53, A5996.

Further, Compound 2 is not even among the prostaglandin classes within the

scope of the claimed invention. A5998-99. Thus, even if the data in Figures 1-3

concerned chemical stability (which they do not), such information about Com-

pound 2 would not teach a person of ordinary skill anything about the chemical

stability of prostaglandins falling within the district court’s claim construction of

that term. See A5843-44, A5999-6000. For this reason, as Dr. Kent testified, and

Alcon admits, it is not possible to predict chemical stability for any given prostag-

landin based on data from Compound 2. A5996; Alcon Br. at 41 n.7. That data is

39

simply not applicable to other prostaglandins.

The district court also properly concluded that “while the castor oil patents

baldly assert that PECOs chemically stabilize prostaglandins, the patents do not

disclose how.” A31 (emphasis original). Alcon argues that it “is not a requirement

of patentability that an inventor correctly set forth, or even know how or why the

invention works.” Alcon Br. at 35 (quoting Neman v. Quigg, 877 F.2d 1575, 1581-

82 (Fed. Cir. 1989)). In the very next line of Newman, however, the Court added

that a patent applicant is not “relieved of the requirement of teaching how to

achieve the claimed result” and that a patent application is not enabling if the

“claimed device and method do not produce the claimed result, following the

teachings of the specification.” Neman, 877 F.2d at 1581-82.

That is exactly the problem with the Castor Oil Patents. Although Alcon de-

clines to address the physical-versus-chemical issue in its enablement argument,

the district court observed that “the patents do not discuss how PECOs might inhi-

bit or prevent prostaglandins from chemically degrading,” which is “particularly

problematic in light of the apparent difficulty in distinguishing between” chemical

stability, physical stability, and the causes of prostaglandin loss in ophthalmic

compositions. A31-32. That is—given the breadth of claims, dearth of disclosure,

and the further Wands findings addressed below—just as in Newman, it is not clear

here that the “claimed … method” (adding PECO to a prostaglandin composition)

40

produces the “claimed result” (enhanced chemical stability of that prostaglandin)

“following the teachings of the specification[s].” See 877 F.2d at 1582.

2. The relevant technology is highly unpredictable, and there are several factors that impact chemical stability.

Consistent with Dr. Kent’s testimony, the district court found that “the tech-

nology at issue here is highly unpredictable.” A32. As the district court observed,

“even small variations in the composition of a prostaglandin-containing solution

can, and do, have a noticeable effect on the stability of a prostaglandin” (id.); “pH,

buffer, buffer concentration, preservatives, chelating agents, and other excipients

may affect the chemical stability of prostaglandins in ophthalmic formulations”;

and the prostaglandin’s chemical structure too can influence chemical stability.

A17-18, A23. Indeed, Alcon itself admits, “different prostaglandins are different

from each other; some will be more chemically stable than others; and one cannot

predict for certain in advance which ones will be more chemically stable than oth-

ers in an aqueous composition.” Alcon Br. at 41 n.7.

Alcon thus concedes the district court’s basic point (that several parameters

impact a prostaglandin’s chemical stability) but attempts to skirt its conclusion by

asserting that such factors only matter in optimizing chemical stability. See Alcon

Br. at 39. But the district court was not concerned with optimization; it was

troubled by the fact that these omitted parameters necessarily impact whether

PECO enhances the chemical stability of a prostaglandin at all. See A17-18, A23,

41

A26, A32. Further, these parameters impact how much PECO constitutes a chemi-

cally stabilizing amount. See A6008-10. As Dr. Kent testified, when “you have a

lot of variables on top of one another, the experimentation gets out of control

quickly.” A6009; see also A6122 (“[T]he number of experiments would be—I

would hate to calculate the number, but it would be very large in order to end up at

the desired composition”).

These facts—and Dr. Kent’s testimony in particular—give lie to Alcon’s as-

sertion that there “is simply no evidence that a person of ordinary skill … would

require any (let alone undue) experimentation to practice the invention.” Alcon Br.

at 37; accord id. at 19.

3. The person of ordinary skill in the art possesses a relatively high level of skill, but the prior art teaches little about chemically stabilizing prostaglandins.

The district court expected that a person of ordinary skill in the art would

possess a relatively high level of skill. A33. Even so, the court found that he or

she would have to undertake undue experimentation to practice the claimed inven-

tion. Id. In part this is because, as the district court further found, “the prior art

teaches very little about chemically stabilizing prostaglandins, much less by using

PECO in ophthalmic formulation.” A32. This gave the court “all the more reason

to insist on an adequate disclosure in the castor oil patents.” Id. Alcon does not

claim that the district court clearly erred in this factual finding.

42

4. The patent claims are extremely broad, and not supported by enough disclosure.

Finally, the district court correctly observed that the Castor Oil Patent claims

are “extremely broad”: they “encompass a rather inordinately large number of po-

tential compounds,” and nearly every PECO, and are “open to a myriad of other

components typically present in ophthalmic compositions … in any concentration

or ratio, and at any pH.” A33.

Without disputing the breadth of its claims, Alcon attempts to rewrite the

Castor Oil Patents as thorough documents with precise recipes to practice the in-

vention. Alcon pretends that they “expressly describe how much [PECO] to use to

practice the invention,” and that “the patents’ specifications provide a great deal of

detail about how to practice the claimed invention.” Alcon Br. at 36. Alcon’s fic-

tion is plagued by the facts, however—and no reimagining here can add to the li-

mited disclosure in the Castor Oil Patents or cabin their overly broad claims. In

reality, as reflected by the district court’s analysis, these patents broadly claim a

method of stabilizing an “inordinately large number” of prostaglandins. A33.

Concerning the ranges disclosed, the Castor Oil Patents state “the present

invention will include one or more polyethoxylated castor oils in an amount be-

tween 0.02 and about 20.0 percent by weight”—at least a 1,000x range—and “one

or more prostaglandins in an amount between 0.00001 and 0.2 wt %”—a 20,000x

range. A6369, A6379. These ranges are meaningless, not enabling. Cf. Impax

43

Labs., Inc. v. Aventis Pharm. Inc., 545 F.3d 1312, 1315 (Fed. Cir. 2008) (no

enablement where the patent’s “dosage guidelines are broad and general without

sufficient direction or guidance to prescribe a treatment regimen”).

The Castor Oil Patents also include numerous other ingredients, such as “an-

timicrobial preservatives, tonicity agents, and buffers” that may comprise the

claimed invention, and do not limit the composition’s pH. A6369, A6380. As the

district court found, “this compositional breadth poses a huge challenge to one

skilled in the art because … these variables may affect the stability of a prosta-

glandin-containing composition.” A33-34.

Alcon relies on the testimony of Mr. Schneider to support “the conclusion

that the ‘full scope’ of the claimed invention is enabled.” Alcon Br. at 37. But

“[i]t is the specification, not the knowledge of one skilled in the art, that must

supply the novel aspects of an invention in order to constitute adequate enable-

ment.” Genentech Inc. v. Novo Nordisk A/S, 108 F.3d 1361, 1366 (Fed. Cir. 1997).

Mr. Schneider’s testimony cannot supplement the specification. Id.; accord Auto

Techs. Int’l, Inc. v. BMW of N. Am., Inc., 501 F.3d 1274, 1283 (Fed. Cir. 2007)

(“the novel aspects of an invention must be enabled in the patent”).

Moreover, the district court concluded that Mr. Schneider “lacks credibility

on the critical ‘chemical stability / loss v. physical stability /loss’ issue in this

case”:

44

[Mr. Schneider] appears confused about whether PECO enhances the chemical stability of prostaglandins or merely enhances their physical stability. In his deposition Schneider was asked whether the data in his patents might indicate that the PECO enhances the physical stabili-ty of the prostaglandin compositions. He answered: “Might. Yes, it might. I can’t deny that. But my feeling is that it indicates the chemi-cal stability because the [PECO] affords solubility and prevents loss of physical absorption or physical – the polyethoxylated castor oil prevents the physical loss of the prostaglandin from the solution.”

A22-23 (citation omitted). Alcon slices and dices Mr. Schneider’s testimony to

falsely suggest that he studied the chemical stability of “scores … maybe hun-

dreds” of prostaglandin formulations (Alcon Br. at 37) (in fact, that’s how many he

prepared; there is no evidence he studied that many), but has no answer for his

fundamental confusion. It does not matter that Mr. Schneider thinks his invention

works for numerous prostaglandins when he apparently does not understand the

difference between chemical and physical stability.

Finally, contrary to Alcon’s contention (at 41), the record in this case is

similar to that in Pharm. Resources, Inc. v. Roxane Labs., Inc., 253 F. App’x 26

(Fed. Cir. 2007). In Roxane, which concerned a highly unpredictable art, the pa-

tent contained only three working examples, allowed “the choice of any surfactant

in any concentration,” suggested “that the claims encompass hundreds of possible

surfactants,” and did not limit “the concentration of surfactant.” Id. at 29-30.

Here, as discussed above, the patents contain inapplicable stability data, on one

compound that is outside the scope of the claims, in an unpredictable art, claiming

45

a method of use involving an inordinate number of compounds and nearly all PE-

COs, to be used in essentially any combination and concentration, and in composi-

tions that contain other ingredients that may impact chemical stability. Like the

patents in Roxane, the Castor Oil Patents are not enabled.

B. The district court applied the correct legal standard in concluding that the Castor Oil Patents required “undue experimentation.”

The district court’s factual findings create insurmountable problems for Al-

con. Unable to show clear error, Alcon resorts to inventing law, arguing that the

Wands factors “are simply irrelevant” without a “threshold showing that experi-

mentation is necessary to practice the invention.” Alcon Br. at 34-35. Alcon has

no legal basis or support for this assertion. It cannot cite a single case holding—or

even suggesting—that the Wands factors “come into play only when a party chal-

lenging a patent has put forward evidence that some experimentation is needed to

practice the claims.” Id. at 34. Alcon is just wrong. The whole point of the

Wands factors is to guide an assessment of whether the amount of experimentation

needed to practice the claimed invention is undue. See Wands, 858 F.2d at 737

(identifying eight factors “to be considered in determining whether a disclosure

would require undue experimentation”). It would be senseless, and is not the law,

to require a challenger to make some sort of threshold proffer or prima facie case

of “experimentation” to trigger a full analysis of whether there is “undue experi-

mentation.”

46

In any event, even if there were such a rule, Dr. Kent testified that, given the

possible number of interactions between the broad range of prostaglandins, castor

oils, and other variables encompassed by the patents, the number of experiments

“would be very large in order to end up at the desired composition.” A6122.

Again, the record unequivocally contradicts Alcon’s false cry of “no evidence.”

Alcon Br. at 23.

Nor did the district court create a new rule arbitrarily demanding greater de-

tail in a specification to support a broad claim, or believe “that the burden some-

how shifts to the patentee” to demonstrate enablement. Alcon Br. at 31. In fact,

the district court (1) observed that “the accused infringer must prove lack of

enablement by clear and convincing evidence,” (2) analyzed the patents consistent

with the mountain of Federal Circuit case law (some of which is cited above at 35-

37) warning that “[a] patentee who chooses broad claim language must make sure

the broad claims are fully enabled,” Sitrick, 516 F.3d at 999, (3) made factual find-

ings well-supported by the evidence, and (4) determined that the limited disclosure

in these patents did not enable their overly broad claims. A28-34.

Alcon asserts that a patent need not “spell out every detail,” that “an inven-

tion may be patented before it is reduced to practice,” and that “prophetic exam-

ples” can be enabling. Alcon Br. at 28-30 (citations omitted). But none of these

unremarkable propositions describes Alcon’s specification, or excuses its short-

47

comings. Indeed, to confirm that Alcon’s “disclose nothing and claim the world”

approach to enablement is untenable, the Court need look no further than the cases

on which Alcon itself relies—LizardTech, Inc. v. Earth Resource Mapping, Inc.,

424 F.3d 1336, 1345-46 (Fed. Cir. 2005) (holding “that the description of one me-

thod for creating [the invention] does not entitle the inventor of the ’835 patent to

claim any and all means for achieving that objective”); DeGeorge v. Bernier, 768

F.2d 1318, 1324 (Fed. Cir. 1985) (concluding, in the context of an interference

proceeding, that the enablement requirement “was satisfied by disclosure of de-

tailed, claimed TCCPI circuitry” despite the absence of “detailed disclosure of all

related, unclaimed circuitry with which the TCCPI might be interfaced”); Atlas

Powder Co. v. E.I. du Pont De Nemours & Co., 750 F.2d 1569, 1576-77 (Fed. Cir.

1984) (recognizing that that many inoperative combinations in a patent might

render the claims invalid); Pfaff v. Wells Elect., Inc., 525 U.S. 55, 61-62 (1998)

(excusing the fact that Alexander Graham Bell “never actually transmitted tele-

graphically spoken words” before applying for his patent because the specification

“describe[d] … the exact electrical condition that must be created to accomplish

his purpose”—and “with sufficient precision to enable one of ordinary skill in such

matters,” to make and use the full scope of his invention) (emphasis added).

Time and again, the Court has reminded patentees that the rule cited by Al-

con—the rule “that a specification need not disclose what is well known in the

48

art”—is “merely a rule of supplementation, not a substitute for a basic enabling

disclosure.” E.g., Genentech, 108 F.3d at 1366. This rule may excuse the “omis-

sion of minor details”; it does not allow a patentee to rely on the knowledge or in-

ferences of a person of skill in the art to “supply the patentable [or novel] aspects

of the invention.” Id. That is what Alcon attempted to do here, and the district

court correctly refused to allow it. The district court applied the correct legal stan-

dard in reaching its conclusion that the Castor Oil Patents require “undue experi-

mentation” and therefore are not enabled. A34. That conclusion is amply sup-

ported by the record and should be affirmed.

III. The District Court Correctly Held The Castor Oil Patents Invalid For Failing The Written Description Requirement.

Nor did the district court clearly err in holding that the “castor oil patent

claims also fail the written description requirement for essentially the same reasons

that they fail the enablement requirement: the art in question is highly unpredicta-

ble and the claims are extremely broad, but the written description is relatively li-

mited.” A36. Alcon asserts that the same supposed problems with the district

court’s enablement analysis infect the written description ruling too. Alcon Br. at

26, 45. As shown above, however, Alcon’s arguments fail to grapple with the dis-

trict court’s full rationale. In particular, Alcon ignores Barr’s evidence and the dis-

trict court’s findings with respect to chemical versus physical stability.

49

To meet the written description requirement, a patent’s “description must

‘clearly allow persons of ordinary skill in the art to recognize that [the inventor]

invented what is claimed.’ In other words, the test for sufficiency is whether the

disclosure of the application relied upon reasonably conveys to those skilled in the

art that the inventor had possession of the claimed subject matter as of the filing

date.” Ariad Pharma., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1351 (Fed. Cir.

2010) (en banc) (internal citation omitted). This assessment “requires an objective

inquiry into the four corners of the specification.” Id. “The level of detail required

to satisfy the written description requirement varies depending on the nature and

scope of the claims and on the complexity and predictability of the relevant tech-

nology.” Id. Whether a claim meets the written description requirement is a ques-

tion of fact. Amgen, 314 F.3d at 1330. A district court’s decision on the adequacy

of a written description is reviewed for clear error. Id.

As the district court recognized, this Court has set forth several factors for

evaluating the adequacy of the disclosure, including “the existing knowledge in the

particular field, the extent and content of the prior art, the maturity of the science

or technology, [and] the predictability of the aspect at issue.” A36 (quoting Boston

Scientific Corp. v. Johnson & Johnson, 647 F.3d 1353, 1363 (Fed. Cir. 2011)).

As described above—and in the district court’s findings of fact—the Castor

Oil Patents are exceptionally broad, covering a method for enhancing the chemical

50

stability of innumerable prostaglandins by adding to them PECO in an endless

number of combinations and concentrations. A28-37. The disclosures on the other

hand are limited, to say the least. They do not contain any meaningful range to

guide the selection of the prostaglandin(s) and PECO(s), alone or in combination,

or their concentrations. A31-34. More importantly, the specifications do not con-

tain any chemical stability data whatsoever, any data on prostaglandin degradation

products, or any data on prostaglandin degradation pathways. A28-37; A5549-53,

A5994-96, A6002. “[T]he purpose of the written description requirement is to en-

sure that the scope of the right to exclude, as set forth in the claims, does not over-

reach the scope of the inventor’s contribution to the field of art as described in the

patent specification. It is part of the quid pro quo of the patent grant.” A35 (quot-

ing Ariad, 598 F.3d at 1353-54 (internal citations and quotations omitted). Here,

however, the claims “overreach” and flunk the written description requirement. Id.

In challenging that ruling, Alcon makes the same mistakes it did with

enablement. Alcon again conjures a patent disclosure that simply does not exist—

one containing detailed descriptions, lists, limits, testing, examples, compounds,

PECOs, and excipients. Alcon Br. at 42-45. Nor does the patent disclosure con-

tain the results of HPLC tests demonstrating enhanced chemical stability, as Alcon

asserts. Id. at 43. Rather, the Castor Oil Patents contain physical stability data—

and only for a single compound that is not even within the scope of the claims. See

51

A25-26; A5538-39, A5994-96. Even if the data in question were chemical stabili-

ty data, therefore, it could not be used to predict chemical stability for any other

prostaglandin. A20-21; A5996-97. Alcon even admits as much. Alcon Br. at 41

n.7. Moreover, Mr. Schneider—a named inventor—gave testimony that was at

best confused about the nature of the claimed invention. See A23 (quoted above at

44). Even today, it is not clear he has possession of the claimed invention.

Given these facts, and the “apparent difficulty in distinguishing between

chemical stability/loss and physical stability/loss of prostaglandins in ophthalmic

compositions” (A31-32)—a difficulty Alcon declines to address in this part of its

brief—it is neither surprising nor clear error that the district court concluded that

the disclosure failed to meet the written description requirement.

IV. Barr Is Entitled To A Judgment Of Non-Infringement On The ’383 And ’052 Patents.

The district court did, however, make a procedural mistake in denying Barr’s

motion for a judgment of non-infringement on the ’383 and ’052 patents. This

Court reviews “[t]he grant or denial of a motion for judgment as a matter of law …

under the law of the regional circuit in which the appeal from the district court

would usually lie.” ACCO Brands, Inc. v. ABA Locks Mfrs. Co., 501 F.3d 1307,

1311 (Fed. Cir. 2007) (internal citation omitted). Accordingly, under Third Circuit

law, the district court’s denial of Barr’s motion is subject to de novo review by this

Court. McKenna v. City of Philadelphia, 649 F.3d 171, 176 (Fed. Cir. 2011).

52

Barr is entitled to a judgment of non-infringement on the ’383 and ’052 pa-

tents because, as the district court expressly found, Alcon alleged infringement in

its complaint, never dismissed those claims, and failed to introduce any evidence

of infringement at trial. See A53. It is as simple as that. At trial, Barr moved for

judgment as a matter of law (id.), and that motion should have been granted. In

addition, by representing that it withdrew the ’383 and ’052 patents from the case

and would seek dismissal with prejudice, Alcon was estopped from opposing

Barr’s motion. Withholding judgment from Barr, moreover, would reward Alcon’s

litigation gamesmanship and thwart the purposes of the Hatch-Waxman Act.

A. Alcon alleged in its complaint, but failed to prove at trial, that Barr’s ANDA product infringes the ’383 and ’052 Patents.

1. Alcon never removed these infringement claims from the litigation.

Alcon brought claims for infringement—and then neither proved those

claims nor pursued any of the “customary ways” to remove them from the case.

Lear Auto. Dearborn, Inc. v. Johnson Controls, Inc., 528 F. Supp. 2d 654, 669-70

(E.D. Mich. 2007). Even at trial, when Barr moved for judgment as a matter of

law, still Alcon did not seek dismissal of its claims. Instead, Alcon stated that

“those claims have been now withdrawn or are no longer at issue in the case.”

A5686. But that is wrong. Indeed, as Alcon itself acknowledged, its conduct—

announcing that it intended to have these claims dismissed, but never actually

53

doing so—resulted in a “lingering ambiguity.” A5013; see also A9, A53 (also ac-

knowledging this failure on Alcon’s part).

The Federal Rules of Civil Procedure expressly provide for withdrawing

claims by stipulation, or motion for amendment or dismissal. See Fed. R. Civ. P.

15(a), 41(a). But Alcon decided not to take any such step—and Barr had every

right to rely on that decision. Otherwise, what is the purpose of requiring a party

to obtain a stipulation or file a motion? What is the purpose of Rules 15 and 41?

Those Rules prohibited Alcon from unilaterally amending its complaint or dismiss-

ing its claims related to the ’383 and ’052 patents. See Fed. R. Civ. P. 15(a)(1)-(2)

( “a party may amend its pleading only with the opposing party’s written consent

or the court’s leave” when the amendment is not made as a “matter of course”); In

re Bath & Kitchen Fixtures Antitrust Litig., 535 F.3d 161, 165 (3d Cir. 2008) (de-

fendant may “cut[] off” plaintiff’s right to end the case under Rule 41(a) by filing

an answer or motion for summary judgment).

“Simply stated, once a plaintiff has pleaded a claim and the defendant has

served an answer, Rule 41(a) dictates that the plaintiff cannot voluntarily dismiss

its claim absent either the defendant’s stipulation or an order of the court. Much as

a plaintiff might wish to abandon a claim, then, the Rule makes clear there is no

absolute entitlement to unilaterally do so once the defendant has answered the

claim.” See Lear Auto., 528 F. Supp. 2d at 672 n.10 (even though plaintiff averred

54

it “has not asserted, and will never assert” the patent against defendant, holding

that patent infringement claim remained part of the case because “of the express

allegation [of infringement] in [plaintiff’s] complaint”).

By allowing Alcon to informally abandon its claims so late in the litigation,

the district court nullified these Rules and punished Barr for Alcon’s procedurally

improper course. As discussed further below, to obtain FDA approval of its

ANDA product, Barr needs a judgment of non-infringement on the ’383 and ’052

patents. At no time during the litigation, however, did Barr have any reason to as-

sert a declaratory judgment counterclaim—and this was because of decisions Al-

con made: If Alcon had omitted these claims from its complaint, Barr would have

asserted a counterclaim for a declaratory judgment of non-infringement. If Alcon

had sought leave to amend its complaint under Rule 15 to delete these claims, Barr

would have opposed and, in the alternative, sought leave to amend its answer to

assert a declaratory judgment counterclaim. If Alcon had sought voluntary dismis-

sal without prejudice under Rule 41, Barr would have opposed and, in the alterna-

tive, sought leave to amend to assert a declaratory judgment counterclaim.

Given the manifest prejudice to Barr of allowing Alcon to avoid a judgment

reflecting the undisputed fact that Barr’s ANDA product does not infringe the ’383

or ’052 patents, any such opposition or request on Barr’s part would have been

well-taken. E.g., Estate of Oliva ex rel. McHugh v. New Jersey, 604 F.3d 788,

55

803 (3d Cir. 2010) (“[A] district court may exercise its discretion to deny leave to

amend a complaint in situations in which the plaintiff has delayed seeking leave to

amend if the delay is undue, motivated by bad faith, or prejudicial to the opposing

party.”) (citation omitted); Bjorgung v. Whitetail Resort, LP, 550 F.3d 263, 268-69

(3d Cir. 2008) (denying leave to amend complaint due to prejudicial delay). In-

stead, Alcon advised that it intended to obtain a stipulation of dismissal, failed to

do so, and failed to take any other procedurally proper step to resolve these claims.

Alcon attempted to informally abandon these claims during discovery—but that

should not entitle it to victory by default. Barr’s timely, procedurally proper, well-

founded motion for judgment as a matter of law should have been granted.

2. This Court’s precedent does not require allowing Alcon to prevail by procedural error.

All agree that, despite having the burden of proof, Egyptian Goddess, Inc. v.

Swisa, Inc., 543 F.3d 665, 679 (Fed. Cir. 2008), “Alcon did not present any evi-

dence whatsoever” regarding infringement of any claim of the ’383 and ’052 pa-

tents. A53. Barr, therefore, was entitled to judgment as a matter of law. See Smith

& Nephew, Inc. v. Arthrex, Inc., 453 F. App’x 977, 980-81 (Fed. Cir. 2011) (re-

versing district court’s denial of defendant’s motion for judgment as a matter of

law where the patentee “failed to present any evidence” that the allegedly infring-

ing device contained the claimed elements of the invention); Denneny v. Siegel,

407 F.2d 433, 439-40 (3d Cir. 1969) (a motion for judgment as a matter of law

56

should be granted when the record “is critically deficient of that minimum quan-

tum of evidence from which a jury might reasonably afford relief”).

Indeed, courts routinely enter a judgment of non-infringement when the pa-

tentee fails to present any evidence of infringement. E.g., Mirror Worlds, LLC v.

Apple, Inc., 784 F. Supp. 2d 703, 709-10, 720-21 (E.D. Tex. 2011); TecSec, Inc. v.

Int’l Bus. Mach. Corp., 769 F. Supp. 2d 997, 1009 (E.D. Va. 2011); CIF Licensing,

LLC v. Agere Sys. Inc., 727 F. Supp. 2d 337, 353 (D. Del. 2010); Sw. Efuel Net-

work, L.L.C. v. Transaction Tracking Tech., Inc., 2010 WL 963697, at *2 (E.D.

Tex. Jan. 13, 2010). We have not located a single case in which this Court re-

versed a finding of non-infringement where a patentee asserted a patent in its com-

plaint and then failed to present any evidence of infringement at trial.

Instead, in Advance Transformer Co. v. Levinson, 837 F.2d 1081 (Fed. Cir.

1988), rev’d on other grounds, 508 U.S. 83 (1993), and Scanner Techs. Corp. v.

ICOS Vision Sys. Corp. N.V., 528 F.3d 1365 (Fed. Cir. 2008), this Court affirmed

judgments of non-infringement under similar circumstances.

First, in Advance Transformer, the patentee asserted a counterclaim of

infringement in the accused infringer’s declaratory judgment action. Years before

trial, the patentee withdrew that claim. 837 F.2d at 1084. The district court even

issued an order stating “the memoranda of both parties indicate that any claims

with regard to this patent have been abandoned.” Id. On the first day of trial,

57

however—following “inconclusive correspondence between the parties,” id.—the

patentee moved for partial summary judgment and requested that the accused

infringer “be forever estopped from asserting the invalidity or non-infringement of

[the] patent.” Advance Transformer Co. v. Levinson, 1986 WL 84365, at *22

(N.D. Ill. June 17, 1986). The court did not grant this motion. Quite the contrary,

it issued a judgment of non-infringement, because the patentee “put in no evidence

on infringement and thus could not carry his burden of proof.” Id. at *22, 27. And

this Court affirmed, noting that “[n]either party presented evidence on the question

of infringement of the [] patent” and holding that “the district court did not clearly

err in treating infringement as at issue, and in finding non-infringement.” Advance

Transformer, 837 F.2d at 1084.

Second, in Scanner, the patentee filed suit alleging infringement of all claims

of two of its patents. 528 F.3d at 1368. The defendant counterclaimed, seeking a

declaration that both patents were invalid, unenforceable, and not infringed. Id.

Prior to trial, the parties stipulated that the case “stands and falls” on one repre-

sentative claim in one of the patents. Id. at 1383. The case went to trial on only

the representative claim, and no evidence was presented on the other claims. Id. at

1383-84. After trial, the district court held that all claims of both patents were not

infringed. Id. at 1384. On appeal, the patentee charged “that the district court

failed to separately consider whether” the accused product performed the method

58

disclosed in several claims not specifically litigated at trial. Id. This Court re-

jected this argument and affirmed explaining that, quite aside from the impact of

the stipulation as to the representative claim, the district court properly determined

that those claims were not infringed because “[the patentee] failed to adduce facts

to support a finding that [the defendant] infringed” the claims not specifically liti-

gated at trial. Id.

As in Advance Transformer and Scanner, the facts of this case warrant a

judgment of non-infringement. Alcon asserted claims of infringement, failed to

remove those claims from the case before trial, and then failed to prove infringe-

ment at trial—as was its burden to do. None of that is in dispute. See A53. Like

the district courts in Advance Transformer and Scanner, the district court here

should have ruled these claims not infringed because Alcon—with the burden of

proof—failed to present any evidence to the contrary.

The district court refused to do so (A69-78), but the two cases on which it

relied are readily distinguishable. The first, 800 Adept, Inc. v. Murex Securities,

Ltd., 539 F.3d 1354 (Fed. Cir. 2008), involved invalidity, not non-infringement.

Referencing pre-litigation correspondence between the parties, Adept brought suit

seeking, among other things, a declaratory judgment of non-infringement and inva-

lidity of “asserted claims” in Targus’ patents. Id. at 1367. Targus counterclaimed,

alleging infringement of various claims in several patents. Id. at 1357. The jury

59

found non-infringement, and that both the asserted and unasserted claims of the

Targus patents were invalid. The district court then entered a judgment of invalidi-

ty on all the claims, despite the fact that Adept failed to present any evidence of in-

validity of the unasserted claims at trial. Id. at 1367. This Court reversed, explain-

ing that “a reference in the complaint is not sufficient to support a judgment that

particular claims are invalid; the specific validity of those claims must have been at

issue during the trial.” Id.

In this case, however, Barr moved at trial for a judgment of non-

infringement—not invalidity. Invalidity is different because, as 800 Adept itself

emphasized, “each claim of a patent (whether independent, dependent, or multiple

dependent form) shall be presumed valid independently of the validity of other

claims.” 800 Adept, 539 F.3d at 1368 (quoting 35 U.S.C. §282); see also Scanner

Techs., 528 F.3d at 1380 (noting that the “burden of proof never shifts to the paten-

tee to prove validity”). The challenger (Barr) has the burden of proving invalidi-

ty—and in the absence of evidence, therefore, the claims must stand. But it is the

patent-holder (Alcon) who has the burden of proving infringement, Egyptian God-

dess, Inc., 543 F.3d at 678—and in the absence of evidence, therefore, an assertion

of infringement fails. This case is thus like Advance Transformer, not 800 Adept:

Barr did not ask the district court to invalidate any claim of the ’383 and ’052 pa-

tents. It asked for, and was entitled to, a judgment of non-infringement.

60

In the second case on which the district court relied, Tol-O-Matic, Inc. v.

Proma Produkt-Und Mktg. Gesellschaft m.b.H., 945 F.2d 1546 (Fed. Cir. 1991),

the accused infringer (Tol-O-Matic) initiated litigation by seeking a declaratory

judgment of non-infringement of certain claims in one of Proma’s patents—and

Proma, the patentee, vaguely asserted in a counterclaim that Tol-O-Matic’s device

fell “within the scope” of its patent. Id. at 1554. The parties stipulated before trial

that claim 25 was representative of claims 25-31, and neither side introduced evi-

dence or requested findings on any other claim at trial. Id. The jury found that

Tol-O-Matic’s device did not infringe claims 25-31, and the trial court entered

judgment accordingly. Id. at 1547. In a post-trial motion to amend the judgment,

Tol-O-Matic sought to have the non-infringement findings applied to claims 1-24

and 32—none of which were specifically pleaded or raised at trial. Id. at 1554.

The district court denied the motion, and this Court affirmed because these claims

were “not litigated or placed fairly at issue during trial.” Id.

There are two key differences between Tol-O-Matic and this case. First, un-

like the amorphous counterclaim in Tol-O-Matic, the pleadings in this case are

concrete: Alcon specifically alleged infringement of the ’383 and ’052 patents and,

unlike the patentee in Tol-O-Matic, Alcon did not lack “sufficient and explicit no-

tice of the claims at risk.” Id. Second, as in Advance Transformer, Barr raised the

61

’383 and ’052 patents during trial. A53; A5686; contrast Tol-O-Matic, Inc., 945

F.2d at 1554-55 (appeal from “denial of its motions to amend the judgment”).

The district court also highlighted that Tol-O-Matic, 800 Adept, Advance

Transformer, and Scanner all involved counterclaims. A76. But that fact made no

difference to the analysis in any of the four cases. The district court was concerned

with notice to Alcon, given Alcon’s stated intention to abandon these claims. See

A71-72. Barr’s answer, however, expressly prayed for judgment of non-

infringement on both of those patents. A190-91. See Fed. R. Civ. P. 8(e) (“Plead-

ings must be construed so as to do justice.”). And, as discussed above, any as-

sumption on Alcon’s part that those patents were insulated from judgment was in-

correct, improper, and unworthy of protection given Rules 15 and 41.

Nor is there any notice problem or unfairness created by the parties’ recogni-

tion in the joint proposed pre-trial order that “Alcon is no longer asserting in-

fringement of [the ’383 or ’052 patents].” A2295 n.4; see also A5686 (Alcon ex-

plaining that it had advised Barr that it was “not going to be asserting those patents

at trial”). The parties simply agreed that Alcon did not intend to prove that Barr’s

ANDA product infringed the ’383 and ’052 patents. Neither party agreed that the

patents had been dismissed from the case, as they easily could have done. Barr

certainly did not agree that it was not entitled to a judgment of non-infringement.

62

Alcon’s acknowledgment that it would not even attempt to prove infringement of

the ’383 or ’052 patents does not make Barr’s request for judgment improper.

In fact, it appears that this was the situation in Advance Transformer: After

the patentee withdrew its infringement claim, the accused infringer advised the

district court that, “[i]n view of non-assertion of the ’369 patent … [it] did not

intend to introduce evidence of invalidity or non-infringement of [that patent].”

1986 WL 84365, at *21. Then, when the patentee attempted to declare victory on

this patent at trial (i.e., obtain a judgment of infringement), the district court not

only rejected that attempt; it entered judgment for the accused infringer—even

though that party, just like Barr here, put in no evidence on the issue, and the ’369

patent was no more “at issue” at trial in that case than were the ’383 and ’052 pa-

tents here. And this Court affirmed. 837 F.2d at 1084.

B. The district court should have granted Barr’s motion for judg-ment as a matter of law because Alcon is estopped from arguing infringement of the ’383 and ’052 patents.

Alcon represented to Barr and to the district court that it no longer asserted

that Barr’s ANDA product infringes the ’383 and ’052 patents, and it would seek

dismissal of those claims with prejudice. A5686. Accordingly, Alcon should not

have been heard to oppose Barr’s motion for judgment. Alcon was estopped from

asserting any such opposition, and for this reason too, Barr is entitled to a judg-

ment of non-infringement of the ’383 and ’052 patents.

63

The District of New Jersey addressed an analogous situation in Shire Labs.,

Inc. v. Corepharma, LLC, 2008 WL 4822186, at *10 (D.N.J. Nov. 3, 2008). In

that case, Shire, the patentee, asserted the ’013 patent in its complaint. Id. at *1-2.

Before Corepharma filed its initial responsive pleading, Shire amended its com-

plaint, dropping the ’013 patent from the case. Id. at *1. Corepharma, the accused

infringer, answered Shire’s amended complaint and included a counterclaim for

declaratory judgment of non-infringement of the ’013 patent. Id. Corepharma lat-

er moved for judgment on the pleadings of non-infringement of the ’013 patent.

Id. Shire represented to the Court that, in amending its complaint, Shire withdrew

all claims regarding the ’013 patent, and that it had provided Corepharma with a

covenant not to sue for infringement of the ’013 patent. Id. at *1-2. As a result of

Shire’s representations, the Court dismissed Corepharma’s declaratory judgment

counterclaim for lack of subject matter jurisdiction. Id.

Then, this Court held in Caraco Pharm. Labs., Ltd. v. Forest Labs., Ltd.,

527 F.3d 1278 (Fed. Cir. 2008), that a unilateral covenant not to sue by the paten-

tee does not divest the court of subject matter jurisdiction over a declaratory judg-

ment counterclaim for non-infringement. That led the Shire district court to vacate

its earlier order of dismissal and reinstate Corepharma’s counterclaim. Shire

Labs., Inc., 2008 WL 4822186 at *3. Corepharma then moved for summary judg-

ment, and Shire responded by arguing and presenting evidence of infringement.

64

Id. at *4. The court rejected the attempted proffer, holding that Shire was judicial-

ly estopped by its earlier representations that it was not asserting infringement of

the ’013 patent. The court ruled:

Because Shire has been judicially estopped from contesting infringe-ment of the ’013 patent, Corepharma's motion for summary judgment is unopposed. This Court finds that Shire has no evidence and cannot prove that Corepharma has infringed the ’013 patent. Corepharma is entitled to judgment of non-infringement of the ’013 patent as a mat-ter of law. The motion for summary judgment will be granted.

Id. at *10.

Here, Alcon represented to Barr and the district court that it was “no longer

asserting infringement” of the ’383 and ’052 patents (A2295), and that it intended

to enter a stipulation of dismissal with prejudice of all claims asserting infringe-

ment of the ’383 and ’052 patents. A5686; A11527-28. Alcon did not follow

through on seeking dismissal, but after those representations, Alcon should not

have been heard to oppose when Barr moved for judgment as a matter of law at tri-

al. See G-I Holdings, Inc. v. Reliance Ins. Co., 586 F.3d 247, 262 -63 (3d Cir.

2009) (explaining that judicial estoppel can apply even “where no court has ac-

cepted [a party’s] initial position,” citing by way of example, Krystal Cadillac-

Oldsmobile GMC Truck, Inc. v. General Motors Corp., 337 F.3d 314, 324-25 (3d

Cir. 2003), where, “though no court had ever relied on the debtor's initial posi-

tion[,] … creditors almost certainly had relied on it, undermining the bankruptcy

process by weakening their bargaining position”).

65

Alcon cannot now—or later—claim that Barr’s ANDA product infringes the

’383 and ’052 patents,1 which makes the decision to withhold judgment now par-

ticularly unreasonable and counter to the public interest, as discussed in the next

section below. As in Shire, Alcon should have been precluded from contesting in-

fringement, and Barr’s motion treated as unopposed.

C. In the absence of a judgment of non-infringement regarding the ’383 and ’052 patents, Alcon’s non-infringed patents may prevent Barr from entering the market.

Without a judgment of non-infringement of the ’383 and ’052 patents, as a

consequence of the procedural peculiarities of the Hatch-Waxman Act, Barr may

be prevented from launching its Travatan Z® ANDA product. That result would

deprive consumers of Barr’s lower-cost generic drug—despite the fact that every

one of Alcon’s pleaded patents has (1) expired, (2) been found invalid and not in-

fringed, or (3) cannot be asserted against Barr’s product. The manifest injustice

and harm to the public interest from such a result confirms the merit of both argu-

ments above—that Alcon should have been estopped from opposing Barr’s motion,

and that if Alcon had sought anything short of dismissal with prejudice, Barr could

have successfully opposed, or amended its own pleading to add a counterclaim.

1 Among other defenses, equitable estoppel would bar any future suit. See generally A.C. Aukerman Co. v. R.L. Chaides Const. Co., 960 F.2d 1020, 1028, 1041-42 (Fed. Cir. 1992) (en banc).

66

In this case, Par is the first ANDA filer for a generic version of Travatan

Z®. Par, therefore has 180 days of exclusivity under the statute—and “the FDA

cannot approve subsequent Paragraph IV ANDAs until the first Paragraph IV

ANDA filer’s 180-day exclusivity period expires.” Caraco, 527 F.3d at 1284.

Without FDA approval, Barr cannot launch its ANDA product and compete. And

the Hatch-Waxman Act provides that, absent other forfeiture events, Par’s exclu-

sivity period will begin only after it brings its own generic product to market or a

court decides that the ’383 and ’052 patents are not valid or not infringed. See 21

U.S.C. § 355(j); Caraco, 527 F.3d at 1284-85 & n.4. Only Par knows when it will

begin to market, and the confidential settlement agreement it entered into with Al-

con may impose restrictions on Par’s ability to launch or provide Par with an in-

centive to delay. Barr, therefore, needs a judgment to trigger Par’s exclusivity and

get to market. See 21 U.S.C. § 355(j); Caraco, 527 F.3d at 1284-85 & n.4.2

None of Alcon’s six pleaded patents should prevent Barr from launching its

generic product. Two of the patents (the ’287 and ’062—the Castor Oil Patents)

have been held to be invalid and not infringed by Barr’s ANDA product. A54.

2 As this Court has recognized, market exclusivity gives NDA holders and first-filers “a strong incentive to prevent” an event that triggers or forfeits exclusivity, like a court judgment of non-infringement or invalidity. Caraco, 527 F.3d at 1284. In fact, brand-name and generic drug companies, maneuvering alone and in concert, have for years “exploited certain aspects of the Hatch-Waxman Act” to park the first ANDA filer’s exclusivity and delay generic competition. Teva Pharm. USA, Inc. v. Novartis Pharm. Corp., 482 F.3d 1330, 1343 n.7 (Fed. Cir. 2007) (citation omitted).

67

Two other patents (the ’497 and ’253) expired in May 2012. The remaining two

patents (the ’383 and ’052) can no longer be asserted by Alcon against Barr. Being

kept off the market in these circumstances would be directly at odds with the

Hatch-Waxman Act’s policy of encouraging generic competition. See, e.g., Minn.

Mining & Mfg. v. Barr Labs., 289 F.3d 775, 786 (Fed Cir. 2002) (Garjarsa, J., con-

curring) (“allowing first ANDA filers to hold up all the generic drug manufactur-

ers for an indefinite amount of time would stifle Congress’s attempt to expedite the

marketing of generic products”). The Act does not give first filers the right to in-

definitely block other generics from entering the market.

In sum, there is no basis under the Hatch-Waxman Act or the Federal Rules

of Civil Procedure for Alcon to win by default after it put these patents at issue and

failed to prove infringement. Barr is entitled to judgment of non-infringement so

that it can bring its non-infringing product to market (after starting the clock on

Par’s exclusivity). If Par fails to use its exclusivity, that is its choice, but consum-

ers should not be held hostage by Alcon’s exploitation of the Act—particularly not

when it depends on gross disregard for Federal Procedure. Gamesmanship is bad

enough; outright failure to follow the Rules should not be countenanced.

68

CONCLUSION

For these reasons, the judgment of invalidity and non-infringement of the

Castor Oil Patents should be affirmed. In addition, the Court should remand with

instructions that the district court enter judgment of non-infringement on the ’383

and ’052 patents as well.

Dated: September 28, 2012 Respectfully submitted, /s/ William P. Ferranti

GEORGE C. LOMBARDI BRADLEY C. GRAVELINE WILLIAM P. FERRANTI Winston & Strawn LLP 35 West Wacker Drive Chicago, Illinois 60601 (312) 558-5600 [email protected] [email protected] [email protected]

Counsel for Defendant-Cross Appellant Barr Laboratories, Inc.

CERTIFICATE OF COMPLIANCE WITH TYPE-VOLUME LIMITATION, TYPEFACE

REQUIREMENTS, AND TYPE STYLE REQUIREMENTS

1. This brief complies with the type-volume limitation of Federal Rule of

Appellate Procedure 28.1(e) because it contains 16,144 words, excluding the parts

of the brief exempted by Federal Rule of Appellate Procedure 32(a)(7)(B)(iii) and

Federal Circuit Rule 32(b).

2. This brief complies with the typeface requirements of Federal Rule of

Appellate Procedure 28.1(e) and the type style requirements of Federal Rule of

Appellate Procedure 32(a)(6) because this brief has been prepared in a proportion-

ally spaced typeface using Microsoft Word 2007 in 14-point Times New Roman.

Dated: September 28, 2012 /s/ William P. Ferranti

WILLIAM P. FERRANTI Counsel for Defendant-Cross Appellant Barr Laboratories, Inc.

CERTIFICATE OF SERVICE

I hereby certify that on September 28, 2012, I caused the foregoing Brief of

Defendant-Cross Appellant Barr Laboratories, Inc. to be electronically filed with

the Clerk of Court using the CM/ECF system, and served via CM/ECF on the fol-

lowing counsel for Plaintiff-Appellant Alcon Research, Ltd.:

GLENN J. PFADENHAUER ADAM L. PERLMAN KEVIN HARDY DAVID M. KRINSKY Williams & Connolly LLP 725 Twelfth Street, N.W. Washington, DC 20005 (202) 434-5000 [email protected] [email protected] [email protected] [email protected]

Dated: September 28, 2012 /s/ William P. Ferranti WILLIAM P. FERRANTI Counsel for Defendant-Cross Appellant Barr Laboratories, Inc.